Analysis of a putative imprinted locus within the TRAPPC9 ......ii in PEG13 allelic expression, methylation profiles at CpG islands were determined in human extraembryonic and embryonic

Analysis of a putative imprinted locus within the TRAPPC9 intellectual disability gene

by

Rosalind Law

A thesis submitted in conformity with the requirement for the degree of Master of Science

Institute of Medical Science

University of Toronto

© by Rosalind Law 2014

i

Analysis of a putative imprinted locus within the TRAPPC9 intellectual disability gene

Rosalind Law

Master of Science

Institute of Medical Science

University of Toronto

2014

We have recently identified a subset of individuals carrying copy number variations within the

autosomal recessive intellectual disability gene TRAPPC9 (chr8q24.3) that are heterozygous and lack

apparent mutations on the second allele, some of which are inherited from unaffected parents. The

relationship of these unique deletions and duplications in autistic and ID etiology are currently

unknown; there is evidence indicating the involvement of a conserved imprinting mechanism at this

locus between mice (15qD3) and humans, specifically the maternally-expressed gene KCNK9/Kcnk9

and an intronic long non-coding RNA (ncRNA) located with an intron of Trappc9, termed

PEG13/Peg13, which is hypothesized to regulate imprinting.

I proposed that the heterozygous deletions or duplications in our probands disrupt an imprinted locus

by influencing epigenetic mechanisms, specifically DNA methylation and the expression of a

imprinted genes, PEG13 and KCNK9 according to the parent of origin.

My primary objective was to identify whether a potentially equivalent to mouse Peg13, PEG13,

located within an intron of TRAPPC9, was imprinted in various human tissues and cell lines.

To investigate imprinting at this locus, allelic expression of PEG13 was assessed using Sanger

sequencing and pyrosequencing. To identify regulatory regions that may be responsible for variations

ii

in PEG13 allelic expression, methylation profiles at CpG islands were determined in human

extraembryonic and embryonic tissues, as well as in sperm, fibroblasts and leukocytes.

The PEG13 equivalent was found to be preferentially-expressed from the paternal allele, and its

imprinted expression appears to be tissue-specific and conserved in human fetal brain. The biased

expression of PEG13 was supported by the identification of a differentially methylated region within

a CpG island associated with the long ncRNA in fetal brain, and contrasted with the hypomethylated

pattern found at the equivalent region in human sperm. Tissue expression patterns of PEG13

paralleled KCNK9, both demonstrating high expression in human whole brain and cerebellum, and

no expression in whole blood.

iii

Acknowledgements

I feel incredibly grateful for the opportunities I have experienced and the people I have encountered

in the past two years. I am also thankful for those who have guided me through the unpredictable, yet

rewarding field of medical research.

First and foremost, thank you to my supervisor Dr. John Vincent for taking a chance and allowing me

to become part of your research team. You have been incredibly supportive, patient and

understanding throughout my entire graduate experience, and had confidence in me even when I

doubted myself. Despite your busy schedule and many commitments, you were always available and

approachable. You always have best interests of your lab members and for that I could not have

asked for a better mentor.

Thank you to my committee member Dr. Rosanna Weksberg for guiding me throughout my project

from start to finish. Your insight and expertise has proven invaluable. I am also appreciative of your

support and understanding.

To the Vincent lab post-doctoral fellows – Dr. Kirti Mital, Dr. Arshad Rafiq, Dr. Nasim Vasli and

Dr. Hamid Hedari – thank you for sharing your experience, wealth of knowledge and expertise.

Thank you for guiding me when I felt lost, empathizing with my frustrations and celebrating every

milestone. To my fellow peers, Brian Degagne (MSc candidate) and Taimoor Sheikh (PhD.

Candidate), thank you for making time spent in the lab and coursework more enjoyable. Your

positivity, encouragement and words of wisdom always brightened my day.

Thank you to our research analyst Anna Mikhailov for always being there to answer my questions

and for taking time out of your schedule to help me no matter how trivial the task.

The main techniques employed in this study would not have been possible without the expertise and

hospitality of The Krembil Family Epigenetics Lab. Specifically, thank you to Dr. Tarang Khare

(PDF), Sasha Ebrahimi (PhD. Candidate) and Ricardo Harripaul (PhD. Candidate) for teaching me

the ropes of pyrosequencing and DNA methylation. Your knowledge in epigenetics continues to

astound me.

Lastly, I wish to thank my family, specifically my sister Harriet and my parents, for your unwavering

patience not only during this degree, but also in all my academic and non-academic endeavours.

Thank you for the support and encouragement during the stressful and hard times, and for equally

celebrating and rejoicing in the good. Without you, I would not be the person I am today.

iv

Contributors

This project would not have been made possible without our study collaborators Dr. Miho Ishida

(UCL, UK) and Dr. Kazuhiko Nakabayashi (NCCHD, Japan). I am grateful for the opportunity to be

a small contributor to this project. Their work and unique experimental findings were integral

components to this thesis. I would like to extend a special thank you to Dr. Ishida for allowing me to

help validate her findings and for providing me with the key component of this project, the fetal

tissues.

v

Table of Contents

Abstract i

Acknowledgements iii

Contributors iv

Table of Contents v

List of Tables viii

List of Figures ix

List of Appendices x

Abbreviations xi

Declaration of Academic Achievement xiii

CHAPTER 1: Introduction 1

1.1 Neurodevelopment 1

1.1.1 CNS organogenesis: neurulation 2

1.1.2 Regional development of the brain 3

1.1.3 Fundamental events in corticogenesis 4

1.3 Neurodevelopmental disorders 11

1.3.1 Intellectual disability 11

1.3.2 Epidemiology 13

1.3.3 Autism spectrum disorders 14

1.3.4 Comorbidities associated with ID and ASD 15

1.4 Aetiology of ID and ASD 16

1.4.1 Environmental factors 17

1.4.2 Genetic factors 17

1.4.3 Copy number variations 19

1.5 Epigenetic Mechanisms 21

1.5.1 DNA methylation and hydroxymethylation 23

1.5.2 Histone modifications 25

1.5.3 Long non-coding RNAs 27

1.5.4 CGIs and transcription 28

1.6 Genomic imprinting 29

1.6.1 Definition 29

1.6.2 Functional organization of imprinted genes 31

1.6.3 Methylation dynamics in development 32

1.6.4 Conservation of imprinted genes 33

1.6.5 Regulation of genomic imprinting: Models 35

1.6.6 Deregulation of Imprinted Genes 38

1.7 Genomic imprinting in the brain 39

1.7.1 Parent of origin effects in the brain 39

1.7.2 Identification of imprinted genes 40

1.7.3 Imprinting in neurodevelopment and brain function 41

1.8 Genomic imprinting at the TRAPPC9 locus 43

CHAPTER 2: Materials, Methods and Techniques 47

2.1 Introduction and study design 47

2.2 Statistical analysis 48

2.3. Sample quality and quantification 48

vi

2.4 Sanger sequencing 48

2.4 Isolation of nucleic acids from whole blood 49

2.5 Human fibroblasts 50

2.5.1 Culturing and growth conditions 50

2.5.2 Cryogenic Storage 51

2.6 cDNA Preparation for reverse transcription 51

2.7 Real-time quantitative PCR 51

2.7.1 Principle 51

2.7.2 Assay design and validation 55

2.7.3 Experimental design 55

2.7.4 Procedure 56

2.8 Pyrosequencing 56

2.8.1 Principle of pyrosequencing 56

2.8.2 Experimental design 57

2.8.3 Pyrosequencing procedure 59

CHAPTER 3: CNV Analysis in ID and Autism Probands 60

3.1 TRAPPC9 60

3.1.1 TRAPPC9 function in the CNS 60

3.1.2 TRAPPC9 and ID 62

3.1.3 Heterozygous CNVs at the TRAPPC9 locus 63

3.1.4 Imprinting at the Chr15qD3 domain 63

3.1.5 Imprinting at the Chr8q24.3 domain 65

3.2 Materials and methods 69

3.2.1 Sample Collection 69

3.2.2 Whole Gene Sequencing 69

3.2.3 CNV Analysis 70

3.3 Results 73

3.3.1 TRAPPC9 CNVs confer phenotypic variation 73

3.3.2 Validation of CNVs in human lymphcoytes 74

3.3.3 Relevance of characterizing PEG13 in humans 75

CHAPTER 4: Expression analysis of PEG13 80

4.1 Materials and methods: Allelic expression of PEG13 81

4.1.1 Human fetal tissues 81

4.1.2 Human fibroblasts and leukocytes 82

4.1.3 Pyrosequencing validation 83

4.2 Materials and methods: Expression analysis of PEG13 in human tissues 83

4.2.1 Sample collection 84

4.2.2 cDNA synthesis 84

4.2.3 RT-qPCR 84

4.3 Results 86

4.3.1 PEG13 is a novel paternally expressed gene in human fetal brain 86

4.3.2 Allelic expression of PEG13 is convserved in human fibroblasts 88

4.3.3 PEG13 is highly expressed in human brain 95

4.3.4 PEG13 in Autism and ID probands 95

CHAPTER 5: Methylation Analysis of CpG Islands 98

5.1 Materials and methods 99

vii

5.1.1 Sample collection 99

5.1.2 Bisulfite conversion of DNA 100

5.1.3 Quantification of DNA methylation 102

5.1.4 PCR amplification 102

5.1.5 Pyrosequencing and statistical analysis 103

5.1.6 Combined bisulfite restriction analysis 103

5.2 Results 108

5.2.1 Identification of the PEG13-DMR 108

5.2.2 PEG13-CpG island is a maternal DMR 109

5.2.3 KCNK9-associated CpG island is hypomethylated 111

5.2.4 Methylation in ID and autism probands 111

CHAPTER 6: Summary and conclusions, general discussion and future

directions

114

6.1 General discussion 114

6.2 Conclusions 125

6.3 Future directions 125

References 127

APPENDIX 1: Example of pyrograms created from a pyrosequencing program 152

APPENDIX 2: List of studies which have identified TRAPPC9 as a cause of NS-

ARID

153

APPENDIX 3: Expression analysis of Trappc9 in mice 154

APPENDIX 4: Allelic expression analysis for PEG13 157

APPENDIX 5: .Relative expression of alleles at PEG13 at SNPs rs2270409 160

APPENDIX 6: Identification of the PEG13-DMR 161

viii

List of Tables

Table 1.1 List of neurodevelopmental and neuropsychiatric disorders caused by deregulated

imprinting or demonstrate parent-of-origin effects

Table 2.1 List and description of individual fibroblast cell lines

Table 3.1 Summary of genes imprinted at human chr8q24.3 and the syntenic mouse region,

chr15qD3.

Table 3.2 List of primers pairs used to sequence TRAPPC9 exons.

Table 3.3 qPCR primers for CNV validations

Table 3.4 List of ID and autism probands with heterozygous deletions/duplications overlapping

TRAPPC9 and the CNV breakpoints according to gene assembly hg19

Table 4.1 List of tissues samples heterozygous for SNP rs4289794

Table 4.2 PCR assays for SNP loci at the putative PEG13 locus

Table 4.3 Olignonucleotides used for quantifying allelic expression of SNPs by pyrosequencing

Table 5.4 Summary of pyrosequencing results for PEG13 expression in human fetal tissues

Table 4.5 Number of informative heterozygous samples for SNP rs2270409 at PEG13

Table 4.6 Number of informative heterozygous samples for SNP rs4289794 at PEG13

Table 4.7 Number of informative heterozygous samples for SNP rs380221 at PEG13

Table 5.1 List and description of individuals from which sperm DNA was collected.

Table 5.2 List and description of control leukocyte DNA

Table 5.3 List of oligonucleotides used for bisulfite pyrosequencing and regions analyzed

across all biological samples for sperm, leukocytes and fibroblasts

ix

List of Figures

Figure 1.1 Mechanisms for deregulated imprinting in the brain

Figure 3.1 Schematic illustrating the breakpoints of CNVs spanning the TRAPPC9 locus

Figure 3.2 Schematic representation comparing the mouse and human

TRAPPC9/KCNK9/AGO2/PEG13 imprinting cluster.

Figure 3.3 Validating heterozygous CNVs in autism and ID proband by qPCR

Figure 4.1 Schematic diagram of the 17th intron of TRAPPC9 and SNPs used to quantify

allelic expression of PEG13

Figure 4.2 Electropherograms showing the imprinting status of PEG13.

Figure 4.3 Relative expressions of PEG13 alleles in human fetal tissues

Figure 4.4 Relative allelic expression of PEG13 in human fibroblasts.

Figure 4.5 Electropherograms showing the allelic bias of PEG13 in fibroblasts at

rs2270409

Figure 4.7 Relative expressions of PEG13 alleles in fibroblasts at rs2270409

Figure 4.8 Electropherograms showing the allelic bias of PEG13 at rs4289794

Figure 4.9 Relative expressions of PEG13 alleles in fibroblasts at rs4289794

Figure 4.10 Electropherograms showing the allelic bias of PEG13 at rs380221

Figure 4.11 Relative expressions of PEG13 alleles in fibroblasts at rs380221

Figure 4.12 Electropherograms showing the allelic bias of PEG13

Figure 4.13 Genotyping for PEG13 expression in Autism and ID probands with

heterozygous deletions proximal to the PEG13 locus

Figure 4.14 Relative expression of PEG13 and KCNK9 in human tissues and cells by

qPCR

Figure 5.1 Bisulfite conversion of DNA

Figure 5.2 Methylation profiles at CpG island derived from bisulfite pyrosequencing in

human peripheral blood leukocytes

Figure 5.3 Methylation profiles at CpG island derived from bisulfite pyrosequencing in

human fibroblasts

Figure 5.4 Methylation profiles at CpG island derived from bisulfite pyrosequencing in

human sperm

Figure 5.5 Schematic of the human TRAPPC9/ PEG13 locus

Figure 5.6 Methylation at the KCNK9 CpG island derived from bisulfite pyrosequencing

in human leukocytes and fibroblasts

Figure 5.7 Methylation at the PEG13 and KCNK9 CpG islands derived from bisulfite

pyrosequencing in Autism and ID proband leukocytes

Figure 6.1 Proposed model for how PEG13 instigates imprinting of KCNK9 in the brain

through higher order chromatin looping

x

List of Appendices

Appendix 1 Example of pyrograms

Appendix 2 List of studies which have identified TRAPPC9 as a cause of NS-ARID

Appendix 3 Allelic expression of TRAPPC9 in mice

Appendix 4 Electropherograms showing PEG13 expression in fetal tissues

Appendix 5 Relative expression of alleles at PEG13 at SNP rs2270409

Appendix 6 Identification of the PEG13-DMR.

xi

Abbreviations

ASD Autism Spectrum Disorder

AS Angelman Syndrome

ASE Allele-Specific Expression

ASM Allele-Specific Methylation

BWS Beckwith-Wiedmann Syndrome

cDNA Complementary DNA

CGI CpG Island

ChIP Chromatin Immunoprecipitation

CNV Copy Number Variation

COBRA Combined Bisulfite Restriction Analysis

CTCF Ccctc-Binding Factor (Zinc Finger Protein)

CpG Cytosine-phosphate-Guanine dinucleotide

CpH Cytosne-phosphate-(adenine, thymine, cytosine)

DMEM Dulbecco’s Modified Eagle Medium Nutrient Mixture

DMR Differentiatlly Methylated Region

DMSO Dimethyl sulfoxide

DSM-V The Diagnostic and Statistical Manual of Mental Disorders 5th Ed.

DNA Deoxyribonucleic Acid

DNMT DNA Methyltransferase

dNTP Dinucleotide Tri-Phosphate

EDTA Ethylenediaminetetraacetic Acid

FBS Fetal Bovine Serum

GAPDH Glyceraldehyde-3-Phosphate Dehydrogenase

HDAC1 Histone Deacetylase 1

IC Imprinting Centre

ID Intellectual Disability

IGF2 Insuline-Like Growth Factor 2

KCNK9 Potassium Channel Subfamily K Member 9

lcnRNA Long Non-coding RNA

MeCP2 Methyl-CpG Binding Protein

5mC 5-methyl Cytosine

5hmC 5-hydroxymethyl Cytosine

miRNA Micro Ribonucleic Acid

ncRNA Non-coding Ribonucleic Acid

NS-ID Non-syndromic Intellectual Disability

NDD Neurodevelopmental Disorder

NS-ARID Non-syndromic Autosomal Recessive Intellectual Disability

PCR Polymerase Chain Reaction

PEG13 Paternally-expressed Gene 13

PGC Primordial Germ Cell(s)

PTM Post-translational Modification

PWS Prader-Willi Syndrome

RT-qPCR Real-time Quantitative PCR

RGC Radial Glial Cell

RNA Ribonucleic Acid

RT-PCR Reverse Transcription PCR

xii

SAM S-Adenosyl Methionine

S-ID Syndromic Intellectual Disability

siRNA Small Interfering RNA

TRAPPC9 Trafficking Particle Protein Complex 9

TSS Transcriptional Start Site

xiii

Declaration of Academic Achievement

All work in this thesis was completed by me except the following. Tissue processing of human fetal

tissues and downstream applications, specifically nucleic acid extraction, RT-PCR and genotyping

was conducted by Dr. Miho Ishida from University College London. Identification of the PEG13-

DMR in fetal tissues was performed by Dr. Kazu Nakabayashi from the National Research Institute

for Child Health and Development (Tokyo, Japan). DNA and RNA for fibroblast trios were provided

by Dr. Rosanna Weksberg (Weksberg Lab, SickKids Hospital, Toronto). Sperm DNA was provided

by The Krembil Family Epigenetics Lab (Centre for Addiction and Mental Health, Toronto). CNV

validation and TRAPPC9 seqeuncing for Proband 1 was performed by Liana Kaufman (IMS,

University of Toronto).

1

Chapter 1: Introduction

1.1 Neurodevelopment

Temporal and spatially regulated genetic programs, along with gene-environment interactions,

underpin normal brain function and govern neurodevelopment from embryogenesis to adulthood

(Cannon et al., 2003; Vaccarino et al., 2001; Walsh et al., 2008, review). As one of the first major

organs to form, brain volume and gross brain regions – including substructures and various nuclei –

are primarily established in utero (Kretschmann et al., 1986), and mature and develop according to

unique temporal schedules (Rice & Barone, 2000). Organizational and functional changes continue

to occur during the first two years of postnatal life as the brain reaches adult size and synaptic

contacts undergo further establishishment and fine-tuning (Kretschmann et al., 1986).

Considered the defining feature of mammalian evolution, the cerebral cortex with its six-layered

neocortex is responsible for higher cognitive and executive functions including memory and learning

(Casey et al., 2005). The cerebral cortex develops in the fetal period when most neurons acquire their

spatial orientation, morphology and intercellular connectivity, where each cortical region follows

different developmental trajectories e.g. visual cortex versus prefrontal cortex (Bhardwaj et al., 2006;

Bourgeois, 1997; Huttenlocher & Dabholkar, 1997). Additional age-related changes occur well into

young adulthood, making the cortex the last brain region to complete maturation (Caviness et al.,

1996; Chugani et al., 1987; Giedd et al., 1999; Sowell et al., 2007). Animal studies, along with

human postmortem and neuroimaging studies, have unraveled the conserved developmental

programs of cortical development in mice and humans (Casey et al., 2005; Huttenlocher et al., 1982;

Shaw et al., 2008). Although these processes differ in time scales among mammals (i.e. days in

rodents versus weeks to months in humans) the sequence of events remain similar (Rice & Barone,

2000), with the foundation of neuronal architecture primarily established in fetal development

(Clancy et al., 2007; Romijin et al., 1991). A complete review of neurodevelopment and the

2

associated molecular components are not within the scope of this thesis. However, to understand the

critical periods and timing of neurodevelopment, I will briefly outline the main events.

1.1.1 CNS organogenesis: Neurulation

Neurodevelopment begins at the onset of organogenesis with the formation of the notochord. This

cellular rod derives from proliferating ectodermal cells that migrated between the hypoblast and

epiblast layers of the embryonic disc, known as the primitive streak, during gastrulation (Halacheva

et al., 2011). Through a processed termed neural induction, the notochord marks the primitive axis of

the early embryo by patterning the surface of the ectoderm (the outer of the three germ layers) with a

cadre of diffusible signaling molecules and extraembryonic interactions, which outline the antero-

posterior axis of the neuroepithelium (Imuta et al., 2014; Beddington et al., 1994). Subsequently,

multipotent stem cells between the notochord and epiblast differentiate into pseudostratified

columnar epithelium, leading to the formation of the neural plate and signifying the start of

neurulation (Tropepe et al., 2001).

The first morphological evidence of the human central nervous system (CNS) is discernible on

gestational day (GD) 16 with the genesis of the neural plate (O’Rahilly & Gardner, 1979). On GD18,

the sides of the neural plate thicken and elevate as neural folds, creating an invagination known as

the neural groove (Copp et al., 2003). Eventually the neural folds meet and fuse together at the

midline to close the neural plate (GD22) (Smith & Schoenwolf, 1997; Yamaguchi & Miura, 2013),

beginning and ending with the closure of the rostral (anterior) neuropore and caudal (posterior)

neuropore, respectively (Copp et al., 1990; O’Rahilly & Gardner, 1979). By the end of the third week

of gestation (GD 26-28), a hollow neural tube is formed (Rice & Barone, 2000). The neural tube has

apical-basal polarity, with the apical side facing inward and the basal side facing outward, which

represents the pial surface of the developing brain (Götz & Huttner, 2005). A single layer of

proliferating undifferentiated neuroepithelial cells, or neuroprogenitor cells, line the apical side of the

3

neural tube and will eventually become neurons or non-neuronal (glial) cells. Meanwhile, the cavity

of the neural tube itself forms the ventricular system of the brain and comprises the lateral (first and

second), third and fourth ventricles (Rice & Barone, 2000). Secondary neurulation also marks the

emergence and subsequent segregation of neural crest cells from the neural tube, a population of cells

lying between the neuroectoderm and the neural tube (Bronner-Fraser, 1994). Later in fetal

development, neural crest cells differentiate and give rise to most elements of the peripheral nervous

system including the ganglia, cranial nerves, Schwann cells and sensory neurons (Rice & Barone,

2000; Bronner-Fraser, 1994). The neural tube itself separates from the overlying neuroectoderm and

develops into the spinal cord and gross regions of the brain.

1.1.2 Regional development of the brain

On GD 28, three primary enlargements (or vesicles) bud from the anterior portion of the neural tube

in a caudal-to-rostral gradient, beginning with the rhombencephalon (hindbrain), followed by the

mesencephalon (midbrain), and lastly the prosencephalon (forebrain) (Rice & Barone, 2000). The

most rostral of the primary vesicles, the prosencephalon, gives rise to two secondary expansions

known as the diencephalon (the optic vesicles and hypothalamus) and the telecephalon (Müller &

O’Rahilly, 1987). The telencephalon forms the basis of the cerebral hemispheres and cortex.

Similarly, two secondary outgrowths emerge from the rhombencephalon called the metencephalon

(the pons and cerebellum) and the myelencephalon (the medulla oblongata) (Koop et al., 1986). The

mesencephalon remains as a single unit and eventually develops into the brainstem, which includes

the tectum and tegmentum. By mid-gestation, the neural tube resembles a globular-shaped brain with

discernible features, including convolutions (gyri and sulci), and enlarged brain regions such as the

cerebral hemispheres and cortex (Rice & Barone, 2000). The cerebral cortex arises from the anterior

end of the neural tube because of proliferating and differentiating neural progenitor cells. This

process, termed corticogenesis, can be defined by five successive but coordinated, overlapping

4

stages: (1) neurogenesis, characterized by neuroblasts differentiating and proliferating into both

neuronal and non-neuronal cells; (2) cortical neurons migrating to laminar position; (3)

synaptogenesis; (4) gliogenesis and myelination; and lastly (5) apoptosis, which trims and sculpts

synaptic contacts (Rice & Barone, 2000). The relative time scale and length for corticogenesis is

summarized in Figure 1.1.

1.1.3 Fundamental events in corticogenesis

Neurogenesis

Neurogenesis describes the process by which all cell types in the brain are derived from progenitor

cells (Malik et al., 2013). Regional differences exist concerning the timing of neurogenesis and the

rate of cytogenic proliferation. For the cortex, neurogenesis begins in the embryonic period at ~

human GD 43 (Rakic, 1995). All cells of the CNS, regardless of function or morphology, originate

from a layer of germinal neuroepithelial cells lining the lateral ventricles of the neural tube (Farkas &

Huttner, 2008; O’Rourke et al., 1995). These neuroepithelial cells, otherwise classified as stem cells,

exhibit self-renewing properties and are considered multipotent because of their ability to have

neurogenic or gliogenic cell fates.

The layer of neuroepithelial progenitor cells thickens as neural and glial precursor cells are

generated, creating a multilayer tissue that becomes the cerebral wall (Sidman & Rakic, 1973). A

region closest to the ventricular surface of the neural tube known as the ventricular zone (VZ)

separates from the pial surface; this region is the primary source of cortical neurons (Noctor et al.,

2002). At the VZ, two horizontal neuronal layers form. First, the preplate, which arises above the VZ

in the mid-embryonic period, contains the “pioneering” neurons (Götz & Huttner, 2005). Lastly, the

subventricular zone (SVZ) forms between the VZ and the preplate late in the third trimester (rat

GD17), and serves as another site for neuron production toward the end of fetal development (Rakic,

5

1975); the SVZ is the only germinal zone to persist postnatally (Martínez-Cerdeño et al., 2006;

Meyer et al., 2000). Completion of the VZ proliferative zone signifies the onset of neurogenesis.

The pool of progenitor cells is composed primarily of radial glial cells (RGCs) and have two roles in

neurodevelopment, serving as both progenitor cells (Malatesta et al., 2000) and as guides for

migrating neurons (Rakic, 1972). These cells are highly polarized along their apical-basal axis, and

depending on the stage in neurodevelopment, maintain contact with the ventricular surface (apical)

and the pial surface (basal) via their long ascending and descending processes (Pinto & Götz, 2007).

As cortical progenitor cells arise, their nuclei migrate along their apical-basal axis according to their

progression in the cell cycle, resulting in changes to their cell morphology (McConnell, 1995, 1991;

Sauer et al., 1935). In DNA replication, or S phase, nuclei are located in the outer third or half of the

VZ, but rapidly descend to the ventricular surface in G2 phase to complete mitosis (McConnell,

1995). With entry into G1 phase, however, nuclei move outward to the pial surface where daughter

cells can either reenter the cell cycle as a neural stem or progenitor cell or exit the VZ and

differentiate.

Prior to neurogenesis, most progenitor cells divide symmetrically and duplicate the population of

founder cells with each mitotic cycle, expanding the proliferative capacity (and size) of the

developing cortex (Noctor et al, 2004). These rapid divisions generate two daughter cells with

identical cell fates and remain in the VZ to undergo further divisions (Noctor et al., 2004). However,

as neurodevelopment continues, asymmetric divisions steadily increase until following closure of the

neural tube and later in neurogenesis, nearly all cells divide asymmetrically by two main mechanisms

(Haubensak et al., 2003). First, neurogenic divisions produce another self-renewing radial glial

progenitor that re-enters the mitotic cycle and a second, different cell type such as a non-stem

progenitor cell or a neuron (Noctor et al., 2004; Götz et al., 2002). Secondly, asymmetric progenitor

divisions can generate a self-renewing RGC and an intermediate progenitor (IP) cell that migrates to

6

the SVZ - all IP cells undergo symmetric, neurogenic divisions to either amplify the progenitor pool

or produce neurons that may adopt a neuronal or glial cell fate (Nocter et al., 2004).

Following each subsequent division, precursor cells lose their multipotency and become

progressively restricted from generating neuronal subtypes that have been born earlier in

neurodevelopment. (Quian et al., 1998; Lillien et al., 1997). The coordinated process of neuronal fate

specification is determined by an intrinsic timing mechanism at each developmental stage (Shen et

al., 2006), thus enabling each of the emerging layers to contain a unique population of neuronal cells.

Stem cell differentiation arises from the expression of transcription factors and signaling molecules

(Vaccarino et al., 2001), which regulate a combination of cellular programs (Tropepe et al., 2001)

that involve transforming growth factor Β (TGF-B) signaling, and external cues such as fibroblast

growth factors (Fgf) (Götz & Huttner, 2005). Beginning with the first asymmetric divisions,

neuroblasts leave the SVZ and VZ and differentiate as they migrate to their final cortical positions.

Neuronal migration

The cerebral cortex expands as post-mitotic neurons migrate from their germinal sites toward the pial

(outer) surface where they undergo their final differentiation and become arranged into the six

cortical layers (O’Rourke et al., 1992; Hatten, 1990). A neuron’s laminar position is determined by

its time of origin, as neocortical layers form in an “inside-out-fashion”. As such, neurons in each

layer share similar morphological and functional properties (McConnell et al., 1991). Around the 7th

week of gestation, the first wave of neurons to migrate out of the VZ form a transient layer termed

the preplate (Meyer et al., 2000; O’Rourke et al., 1995; 1992). The preplate neurons are loosely

packed and are the first to differentiate, mature and receive synaptic contacts (Supèr et al., 1998);

consequently, these neurons are considered the primitive functional cortex. Shortly thereafter, a

subsequent wave of postmitotic neurons insinuates themselves into the preplate, splitting the layer

7

into a superficial marginal zone (MZ) and a deeper subplate (SP) layer (Meyer et al., 2000). Neurons

in the MZ coalesce into a distinct monolayer, termed the cortical plate (CP), at 8-9 weeks gestation in

humans (E12-E14 in rats) (Rakic, 1972). The CP grows and thickens as newly generated neurons

migrate past earlier-generated cells to settle in progressively more superficial layers. Eventually, a

six-layered neocortex is generated, classified as layers I (most superficial, outward pial layer)

through IV (the deepest, and closest to the geometric center of the brain), and a subplate (layer VII)

(Boulder Committee, 1970; Angevine & Sidman, 1961).

Gliogenesis and myelination

In parallel to corticogenesis, myelination and gliogenesis undergo regional and temporal maturation.

Glial cells – oligodendrocytes and astrocytes – are similar to neural cells in that they arise from

multipotent cells in the VZ and SVZ (Parnavelas et al., 1999). In CNS development, gliogenesis

follows neurogenesis as progenitor cells acquire glia-like characteristics: an initial period of

asymmetrical division, followed by rapid cell division and increased motility (Quian et al., 2000).

Consequently, glioblasts are generated later in embryogenesis (Parnavelas et al., 1999), and continue

to differentiate and mature in postnatal life. Accordingly, retroviral injections in rats showed that

astrocytes are mainly formed in the third trimester and in postnatal development, while

oligodendrocyts are generated exclusively in the postnatal period (Skoff et al., 1976; Sauvegot &

Stiles, 2002). Mature oligodendrocytes appear simultaneously with myelination events, which occur

exclusively in the postnatal period (Skoff et al., 1976).

Synaptogenesis

The majority of neurons are developed in gestation, however synapse formation and reorganization

primarily occurs during the perinatal and early postnatal life (Huttenlocher, 1990). In mammals,

synaptogenesis occurs as a heterochronous event across all cortical regions (Bourgeois, 1997;

8

Huttenlocher & Dabholkar, 1997), and age-related changes have also been reported (Huttenlocher et

al., 1982; Huttenlocher et al., 1979).

The initial synapses form above and below the cortical plate within the MZ and SP shortly following

neuronal migration and coincide with neurogenesis (Huttenlocher & Dabholkar, 1997; Balslev et al.,

1996). Studies in human frontal cortex show evidence for synaptogenesis beginning as early as 6

months in gestation, which is the peak of neurogenesis and neuronal migration (Huttenlocher &

Dabholkar, 1997; Huttenlocher et al., 1982)). During this period, neurons originate from dendritic

shafts of neurons (Bourgeois, 1997). Approximately two months before birth, however, synaptic

density starts to increase primarily by the formation of dentritic spines, followed by an exponential

phase in synaptic density growth between postnatal two to four months, and plateaus at

approximately three years (Huttenlocher & Dabholkar, 1997; Hunttenlocher et al., 1982). This period

of neuron overproduction facilitates experience-dependent organization of synaptic contacts from the

third year until puberty, resulting in functional and experienced-based processes (Bourgeois, 1997;

Diamond et al., 1964). From adolescence to adulthood, synaptic density begins to decline due to a

phase of rapid neuronal apoptosis, known as synaptic pruning, which reduces 60% of the original

neuronal population and a subsequent decrease in the number of synapses (Huttenlocher &

Dabholkar, 1997; Huttenlocher et al., 1982).

Events in CNS synaptogenesis have primary been elucidated by in vitro studies using cultured

neurons. The initial synaptic contacts are formed on the dendritic shafts of neurons, where dendritic

filopodia of growth cones (Niell et al., 2004) extend toward their cellular targets as a result of

changes to cell surface adhesion molecules and dynamic changes to the cytoskeleton (Mitchison &

Kirschner, 1988). Upon reaching their synaptic contacts, axons and dendrites accrue different

synaptic and stabilizing protein complexes, depending on whether the synapses are inhibitory or

9

excitatory (Friedmanet al., 2000; Washbourne et al., 2002; Yoda & Davis, 2003). Glial cells have

also been shown to enhance synapse formation and activity in cultured RGCs (Ullian et al., 20001).

Apoptosis

Programmed cell death, or apoptosis, systematically removes large numbers of neurons in some

structures and serves to eliminate and trim synapses while strengthening others in neurodevelopment

and early life. Unlike synaptogenesis, however, cells undergoing apoptosis in the brain are targeted

synchronously as they are removed from the surrounding tissue (Huttenlocher & Dabholkar, 1997).

A 40% decline in synaptic density occurs between puberty and adolescence (Petanjek et al., 2011),

and is accompanied by decreases in neuronal density (Bourgeois & Rakic, 1993; Huttenlocher,

1978). In adult life, synaptic density stabilizes and implies the loss of neural plasticity (Huttenlocher,

1978).

10

Figure 1.1 Schematic representation of human neurodevelopment and events incorticogenesis. Corticogenesis comprises five overlapping,

but sequential stages: (1) Neurogenesis, where undifferentiated neuronal and non-neuronal precursor cells proliferate and follow cell lineages; (2)

Neuronal migration, differentiating neurons are guided towards the final laminar positions; (3) Synaptogenesis, the establishment of synaptic

contacts; (4) Gliogenesis and myelination; and (5) Apoptosis, which trims and fine-tunes synaptic contacts.

Embryonic Fetal Postnatal

Birth Adolescence

Organogenesis

Neurulation

Cell proliferation: radial glia and neurons

Neuronal migration

Cellular differentiation and synaptogenesis

Apoptosis

Gliogenesis

Myelination

Fertilization

0 1 2 3 4 5 6 7 8 9 1

month year

2 3 4 5 16

11

1.3 Neurodevelopmental disorders

The critical period between post-conception to early postnatal life has a heightened sensitivity to

genetic and environmental influences, and any deviations from strictly controlled developmental

programs underlie an increased susceptibility for neurodevelopmental disorders (NDDs). By

definition, NDDs are a group of conditions that begin in development and are characterized by

developmental deficits (American Psychiatric Association (APA), 2013). Such deficits may manifest

as impairments in personal, social, academic or occupational functioning. Autism spectrum disorders

(ASD) and intellectual disability (ID) are among the most common and prevalent NDDs.

1.3.1 Intellectual disability

Intellectual disability (intellectual developmental disorder; ID), formerly known as mental

retardation, encompasses deficits in both general mental abilities and practical, social and/or

conceptual adaptive functions (APA, 2013). Signs and symptoms for ID are apparent before

adulthood, with disorder onset characterized by the failure to reach expected milestones in early

growth and development (APA, 2013). ID, for the most part, is a non-progressive disorder. However

for certain genetic disorders there is a period of regression followed by stabilization, such as in Rett

syndrome (Chahrour et al., 2007; Amir et al., 1999), while in others there is a progressive decline in

intellectual function throughout the life course (Brereton et al., 2006; Carr, 2005). Individuals often

cannot meet personal and social responsibility in one or more aspects of daily life such as

occupational and academic pursuits as compared to peers of the same age, gender and socioeconomic

background (APA, 2013).

Historically the diagnosis of ID has relied strictly on standardized tests (e.g. Wechsler scales and

Vineland Adaptive Behaviour Scale) and placed an emphasis on intellectual ability (APA, 1994). ID

was previously defined by an intellectual quotient (IQ) score of approximately two standard

12

deviations (SD) below the population (IQ≤70), and the classification of disorder severity (i.e. mild,

moderate, severe, profound) was based solely on IQ score (APA, 1994; International Classification

of Diseases, Version 10 (ICD10); World Health Organization, 1992). However, while intellectual

deficits are central to ID, there is now recognition that this disorder requires a more flexible clinical

scope and comprehensive approach for diagnosis. Recent revisions to The Diagnostic and Statistical

Manual of Mental Disorders (5th ed.; DSM-5; APA, 2013) involve changes to ID terminology, as

well as to diagnostic criteria that places equal importance on standardized tests and individualized

clinical evaluations for measuring both intelligence and adaptive functions. When possible,

interviews from guardians and the affected individual may also be included to make more accurate

diagnoses (APA, 2013). The criteria for classifying ID severity currently focus on deficits in adaptive

functioning, which is better suited for individuals who are functionally nonverbal or have

communication impairments and cannot articulate their internal state of mind.

ID may also be subdivided according to the presence or absence of additional clinical symptoms. The

term “syndromic ID” (S-ID) describes cases with obvious patterns of physiological,

neuroradiological, and metabolic abnormalities. Common features inlcude, but are not limited to,

growth retardation; developmental anomalies of the brain (e.g. dysgenesis of the corpus callosum,

lissencephaly and microcephaly); distinctive facial dysmorphologies (Sharp et al., 2007; Andrieux et

al., 2010; Feero et al., 2012). Conversely, by definition, non-syndromic ID (NS-ID) or idiopathic ID

has no apparent physical manifestations or associated disorder symptoms, making impaired

intellectual function the sole diagnostic feature. However, because certain NS-ID individuals have

specific symptoms such as epilepsy, mild microcephaly and mild behavioural disorders, the

distinction between S-ID and NS-ID may sometimes be indistinguishable (Jamra et al., 2011).

13

1.3.2 Epidemiology

ID is a health care challenge and a socioeconomic burden in both developing and developed

countries, owing in part to limited interventional methods and treatment plans. ID affects 1-3% of the

global population (Chechlacz & Gleeson, 2003; Harris, 2006; Leonard & Wen, 2002), making it the

leading neurodevelopmental disorder worldwide. A recent meta-analysis reported a prevalence of

10.37/1000 for all severities of ID (Maulik et al., 2010), Among those affected, mild ID (49 < IQ <

70) has the highest frequency (85%), followed by moderate ID (34 < IQ <50) at 9.0%, while severe

(19 < IQ < 35) and profound (IQ < 20) ID is estimated to affect 4.0% and 2.0% of individuals,

respectively (King et al., 2009; Stevenson et al., 2000). While the prevalence of severe ID remains

relatively consistent across studies, the estimated prevalence rates for mild ID differs considerably,

with some studies reporting that ID affects as few as 1 in 10,000 people (Rauch et al., 2012) to as

many as 60 in 10, 0000 people (Roeleveld et al., 1997). Such marked differences may be attributed to

inconsistencies across study methodologies with regards to study location, sample size, method of

ascertainment and diagnostic criteria (e.g. ICD-10 vs. DSM-IV), and/or the statistical methods used

(Leonard et al., 2003; Roeleveld et al., 1997). Despite the inaccuracy in measuring true prevalence

rates, the associations between risk factors for ID are well-defined.

Ethnocultural, socio-economic and demographic factors positively correlate with susceptibility for

ID. Indeed, ID consistently affects more males than females, with male-to-female ratios ranging from

1.5 to 3.1 (Leonard et al., 2003, 2011), and is generally ascribed to X-linked genetic factors (Ropers

& Hamel, 2005; Roeleveld et al., 1997). Concerning ethnicity, African-American and Aboriginal

children have an increased risked for ID, specifically for mild-to-moderate ID as compared to

Caucasians and other major ethnic groups, which is concomitant to socioeconomic factors (Leonard

et al., 2003). Indeed, epidemiological studies indicate that poverty has a direct association with risk

for ID, as demonstrated by higher prevalence rates (i.e. two- to threefold) in developing countries as

14

compared to industrialized countries (Maulik & Darmstadt, 2007; Mercadante et al., 2009). High

levels of consanguinity, infections, prenatal malnutrition, environmental toxins, adverse prenatal

environment and limited access to interventions and preventative measures have contributed to the

increased prevalence rates of ID in low-income countries (Patel et al., 2007; Durkin, 2002).

Interestingly, while individuals with ID were once expected to have greater rates of mortality than

the general population, their overall life expectancy has increased particularly in developed countries

due to a decrease in social stigmatization, improved accessibility to mental health services and an

increased knowledge of preventative measures (Janicki et al., 1999). However, in terms of health

care conditions, the quality of life remains subpar for many affected individuals partly because an

increased risk for concurrent disorders and comorbidities, including autism, often accompanies and

aggravates ID pathophysiology, and life expectancy is significantly lower than for the non-ID

population (Bittles et al., 2002)

1.3.3 Autism Spectrum Disorders

Autism is a behavioural syndrome, although unlike previous suppositions (Kanner, 1943), autism is

not a distinct condition but rather a continuum of closely related developmental disorders collectively

referred to as the “Autism spectrum disorders” (ASD). The new diagnostic criteria by the DSM-5

(APA, 2013) reflects this notion as autistic disorder (autism), Asperger’s disorder, childhood

disintegrative disorder (CDD), pervasive developmental disorder not-otherwise-specified (PDD-

NOS) and Rett Syndrome now fall under the same broad diagnostic category as opposed to separate

entities with unique diagnostic criteria (APA, 2013). As such, ASDs are clinically defined by three

core symptomatic domains: (1) Persistent deficits in social communication and social interaction; (2)

Communication deviance; and (3) Restricted and repetitive patterns of behavior, interests, or

activities (APA, 2013). Symptoms manifest as impairments in everyday functioning such as

occupational and social domains. ASDs differ in the severity of symptoms, pattern of onset,

15

development of language, deterioration in skills once they have developed, and cognitive

development (Matson & Boisjoli, 2008). Autistic symptomatology typically emerges in the early

developmental period generally between 12 to 24 months of life when individuals experience a

regression following a normal developmental trajectory (Dawson et al., 2000; Maestro et al., 2002).

The broadening of autism nosology proves difficult for clinicians to make accurate, differential

diagnoses between ASD cases and ID cases without ASD (Matson & Shoemaker, 2009).

ASDs affect four times more males than females (Werling & Geschwind, 2013). Since the inception

of ASD nosology, the prevalence and incidence (limited) rates for all collective ASDs has greatly

increased and are now considered the second most common NDD behind ID. Recent estimates place

ASD frequency at a range that falls between 60 to 110 per 10,000 (Baird et al., 2006; Baxter et al.,

2014; Bolton et al., 2004; Chakrabarti et al., 2005). Given that ASDs appear to be increasingly

prevalent, a debate exists as to whether there is a true increase in the incidence of ASD in the general

population, or rather due to increased awareness, changes to autism inclusion criteria, or different

study methodologies (Matson & Kozlowski, 2011; Rutter et al., 2000).

1.3.4 Comorbidities associated with ID and ASD

Comorbidities and co-occurring disorders are common to ID and ASD, and include metabolic (Curtin

et al., 2010; Melville et al., 2007; Rimmer & Kiyoshi, 2006), epileptic (Bowley & Kerr, 2000;

Danielsson et al., 2005; Tuchman et al., 2002) and neuropsychiatric disorders (Leyfer et al., 2007).

Cross-sectional and longitudinal studies employing comparison groups consistently report higher

prevalence rates for comorbid psychopathology, DSM disorder, or disability in both children and

adults affected by ID or ASD than the general population (Bakken et al., 2010; Bradley et al., 2004;

Brereton, 2006; Leyfer et al., 2007; LoVullo et al., 2009; Simonoff et al., 2008; Tsankanikos et al.,

2006; Deb et al., 2001). ID and ASD share co-occurring disorders, including, but not limited to

schizophrenia (McCarthy et al., 2010), attention deficit hyperactivity disorder symptoms (Gadow et

16

al., 2004; Kenneth et al., 2006); obsessive compulsive disorders (McDougle et al., 1995), mood

disorders (Dekker & Koot, 2003; Kim et al., 2000), anxiety disorders (Gillott et al., 2001), and major

depressive disorders (Leyfer et al., 2006; Ghaziuddin et al., 2002). Most notably, ASD is also

frequently diagnosed with ID, although people with ASD fall along the entire spectrum of intellectual

ability. In addition, while not a defining feature, ID has been reported in approximately 70% of

autistic cases (Larson et al., 2001; Bradley et al., 2004; Matson & Rivet, 2008). As certain authors

note, however, recent twin studies estimate that the association between ASD and ID may be

considerably lower at 40-45% (Yeargin et al., 2003); this discrepancy has been accredited to clinical

ascertainment bias (Chakrabarti & Fombonne, 2005; Hoeskstra et al., 2009).

Dissimilarities exist concerning the rate and pattern of comorbid disorders between ID and ASD.

Indeed, the degree of severity for psychopathologies is far greater in individuals with ASD as

compared to those with ID (Matson et al., 2008; Bradley et al., 2004). Accordingly, individuals with

ASD are more likely to present multiple comorbidities, and the frequency is more pronounced in

people with ASD and severe ID than adults only affected by severe ID (Matson et al., 2008).

Estimates for a co-occurring psychopathology are as high as 37% in children (Dekker & Koot, 2003)

and 60.4% in adults with ID (Deb et al., 2001), as compared to 73% for ASD (Brereton, 2006).

Despite these differences, the considerable overlap between these developmental disorders,

particularly concerning symptomatology, comorbidities and descriptive characteristics, may partly be

attributed to the overlapping etiological factors (Kwok & Cheung, 2007).

1.4 Aetiology of ID and ASD

As previously surmised, neurodevelopment presents windows of vulnerability to environmental and

genetic perturbations, both of which have been positively associated with altered cognitive and

neurological phenotypes. Because ID and ASD arise from a complex interplay between

neurobiological, psychosocial, genetic and environmental factors (Mitchell, 2011; Heikura et al.,

17

2007; Rice & Barone, 2000; Bryan et al., 1999), the resulting clinical and genetic heterogeneity of

these disorders poses challenges when determining specific etiological causes.

1.4.1 Environmental factors

Epidemiological and clinical studies lend support to environmental contributions to ID and ASD

aetiology. These include maternal risk factors during peri- and pre-conception such as parental age

(Krakowiak et al., 2012; Leonard et al., 2006; Meyer et al., 2012; Williams et al., 2008), as well as

obstetric and perinatal complications (Gardener et al., 2009; Kolevzon et al., 2007; O’Dwyer, 1997).

Other congenital contributors include infectious diseases (Gilad et al., 2007; Odeberg et al., 2007)

and metabolic problems (Katz & Lazcano-Ponce, 2008). Prospective studies have found associations

between the development of autism, developmental delay, and intellectual impairment in infants

following intrauterine exposure to environmental neurotoxins in early prenatal life (Canfield et al.,

2003; Durkin et al., 2000; Gillberg & Soderstrom, 2003; Grandjean & Landrigan, 2006; Stanwood et

al., 2001). The impact of environmental exposures to ASD and ID remains part of ongoing research;

however, genetic susceptibility is also commonly involved in aetiology of the disorder (Edelson &

Saudino, 2009).

1.4.2 Genetic factors

The genetic contributions to ID and ASD susceptibility originate from population-based twin and

family studies (Ronald & Hoekstra, 2011, review). A seminal study of 21 twin pairs by Folstein &

Rutter (1977) observed a concordance rate of 36% for stringent autism and 82% in monozygotic

(MZ) twins and 10% in dizygotic (DZ) twins for a slightly broad autistic phenotype and other

cognitive disorders including ID. Subsequent twin studies applying broader diagnostic spectrums

with equivalent (Steffenburg et al., 1989) and larger sample sizes (Bailey et al., 1995; Mason-

Brothers et., 1985; Taniai et al., 2008; Rosenberg et al., 2009) found comparable results, where

concordance rates ranged from 88% to 96% for MZ twins. Extreme autistic traits also demonstrate a

18

heritability component as observed in various twin samples (Edelson & Saudino, 2009; Hoekstra et

al., 2007; Ronald et al., 2010) and have also been shown to be intergenerational (Constantino &

Todd, 2005). Furthermore, non-autistic relatives of individuals with ASD share similar but milder

cognitive profiles (Bailey et al., 1998, review; Bolton et al., 1994; Piven et al., 1997).

Genetics causes for ID and ASD range from straightforward to complex and heterogeneous. The

most frequently diagnosed disorders with genetic etiology arise from cytogenic anomalies such as

aneuploidies (e.g. Down’s syndrome and Turner’s syndrome), chromosomal rearrangements (e.g.

balanced or unbalanced translocations and inversions), and common microdeletion syndromes

(Rauch et al., 2006). An obvious pattern of male inheritance and the availability of family pedigrees

for linkage analysis have led to the identification of over 100 genes in X-linked ID disorders, which

accounts for the majority of all monogenic forms of ID (Lubs et al., 2012). The most common

single-gene disorders associated with ASD and ID are Rett’s Syndrome with MECP2 (Amir et al.,

1999) and Fragile X Syndrome with FMR1 (Loesch et al., 2002; Yudkin et al., 2014). Moreover,

homozygosity mapping in consanguineous ID families carrying rare homozygous mutations

(truncating, missense or nonsense) has led to the discovery of highly penetrant genes such as

MAN1B1 (Rafiq et al., 2010) and TRAPPC9 (Mir et al., 2009). To date, ~ 40 autosomal recessive

genes have been implicated in non-syndromic autosomal recessive ID (NS-ARID) (Musante &

Ropers, 2014, review).

Genomic aberrations, which are generally detectable by conventional cytogenetic techniques or

molecular karyotyping, have clear associations between the genetic factor and phenotypic profile

(Sagoo et al., 2008), and explain causation for ~40% and 20% of ID and ASD cases respectively

(Devlin & Scherer, 2012; Rauch et al., 2006). However, for most idiopathic cases, the association

between pathology and etiology proves difficult, as the contributing genetic factors may be

multigenic or not well defined. Accordingly, the advent of next-generation sequencing (NGS)

19

technology, genome-wide association studies (GWAS) and microarray platforms have led to the

identification of numerous susceptibility loci and submicroscopic genetic variants implicated in

these NDDs, among which include global and locus-specific structural variants.

1.4.3 Copy number variations

Copy number variations (CNVs) are segments of DNA ranging from 1 kilobase (kb) to several

megabases (Mb) in size that cause quantitative changes to the DNA copy number, and encompass

deletions (losses) and gains (duplications or insertional transpositions). These structural variants

follow normal Mendelian inheritance patterns (Komura et al., 2006) or can be de novo in origin

(Sebat et al., 2007), and reside within or span several genes, multiallelic regions, or complex

structural rearrangements (Fredman et al., 2004; Sebat et al., 2004; Redon et al., 2006). CNVs

account for 12% of the genome (Redon et al., 2006) and exhibit non-uniform distribution as certain

regions are more susceptible for genomic rearrangements (Hastings et al., 2009; Stankiewicz &

Luski, 2010). For example, regions harbouring repetitive sequences such as low-copy repeats (LCRs)

and microsatellites have an increased occurrence of nonallelic-homologous recombination, which

facilitate the formation of CNVs (Gu et al., 2008; Nguyen et al., 2006)). Additionally, CNVs have

enrichment at specific protein-coding genes where they affect gene dosage, and within telomeric and

centromeric regions (Nguyen et al., 2006).

Numerous microarray-based studies have found an association between CNVs and disorder

susceptibility for ID and ASD (Cooper et al., 2012; Devlin & Scherer, 2012; Iourov et al. 2012; Pinto

et al., 2010; Marshall et al., 2008; Qiao et al., 2010; Sebat et al., 2007; Ullmann et al., 2007).

Individuals with ASD have a higher frequency of CNVs as compared to the general population, and

it is estimated that rare de novo and highly penetrant CNVs contribute to 5% of ASD cases (Devlin &

Scherer, 2012). These CNVs have the tendency to overlap with genes of neurological relevance,

specifically those involved in synaptic structure and function. Family and case studies have

20

confirmed the role of candidate genes in disorder etiology, among which include synaptic scaffold

proteins (SHANK1, SHANK2, SHANK3) (Berkel et al., 2010; Durand et al., 2006), proteins from the

ubiquitin pathway (Glessner et al., 2009), postsynaptic adhesion molecules such as neuroligins

NLGN3, NLGN2 and NLGN4 (Jamain et al., 2003; Laummonnier et al., 2004) and neurexins NRXN1

and NRXN2 (Gauthier et al., 2011). It is also not coincidental that most of the genes or CNV regions

identified are implicated in ID (Berkel et al., 2010; Noor et al., 2010). Moreover, recurrent CNVs in

patients sharing common phenotypic traits are categorized as microdeletion or microduplcation

syndromes. Common syndromes associated with ASD and ID include del/dup on16p11.2 (Weiss et

al., 2008), hemizygous del on 7q11.23 which is associated with Williams-Beuren Syndrome

(Osborne et al., 2001; Sanders et al., 2011) and del on 15q13.3 (Sharp et al., 2008).

Observations that pathological CNVs can be inherited from unaffected parents (Devlin & Scherer,

2012) and those with de novo CNVs suffer from more severe ID and ASD pehnotypes as compared

to healthy individuals (Pinto et al., 2010), support the notion that different CNVs exhibit varying

degrees of penetrance. Essential to this study is the fact that the occurrence of positive or negative

sequelae depends on a) the genomic position of the CNV; and thus b) how the CNV exerts its effects

on gene transcription and/or expression, in addition to the cell or tissue type in which the gene is

expressed (Ramocki et al., 2008). Genetic syndromes with reciprocal CNVs (i.e. deletions and

duplications in the same region) can result in different, but related phenotypes of varying severities.

For instance, patients carrying hemizygous deletions (del) on 22q11.2 have developmental delay and

CNS defects (Botto et al., 2003; Ensenauer et al., 2003; Shaikh et al., 2000), while those carrying

duplications (dup) display more variable phenotypes (Wentzel et al., 2008). As indicated, CNVs can

have direct or indirect influences on gene dosage and expression (Cook & Scherer, 2008). Direct

mechanisms involve aberrations to an entire gene(s): duplications are associated with increased gene

expression, while deletions to one allele cause decreased expression (Duker et al., 2010).

21

Homozygous deletions disrupt genes or create gene fusions and result in loss of gene function

(Cerveira et al., 2006; Kleinjan & Heyningen, 2005; Potocki et al., 2007). Losses or gains outside of

coding sequences (e.g. intergenic or intragenic regions) can disrupt cis-regulatory elements (DuBose

et al., 2011). Alternatively, a CNV can alter the expression of dosage-sensitive genes by position

effect, whereby the breakpoint boundaries affect the chromatin environment in which the gene

resides or disrupts the interaction between regulatory elements controlling gene expression (Kleinjan

& van Heyningen, 1998, review). For instance, in Potocki-Shaffer syndrome (PSS), a heterozygous

deletion on chr11p11.2-p13 located downstream of the ALX4 gene was suggested to disrupt activity

of a proximal enhancer via position effect and cause haploinsufficiency (Wakui et al., 2005). For

genes that exhibit allelic expression, such as imprinted genes, downstream effects may also be

contingent on whether the CNV resides on the maternal or the paternal chromosome (Girirajan et al.,

2012; Martins-Taylor et al., 2014). However, it is important to mention that not all CNVs

overlapping gene-rich regions are pathological.

Presently, the precise downstream effects of CNVs in ID and ASD at the functional and molecular

levels remain poorly understood. Moreover, given that not all CNVs lead to direct changes in gene

expression, supports the involvement of epigenetic mechanisms

1.5 Epigenetic Mechanisms

Epigenetic regulation refers to changes in gene expression without alterations to the underlying

genetic sequence by virtue of reversible chemical modifications to chromatin. Under both

physiological and pathological conditions, epigenetic processes mediate many cellular processes

through changes to the transcriptome. Epigenetic modifications are involved in gene regulation and

gene silencing (Berger, 2007), as well as cell differentiation (Kim et al., 2010), DNA replication

(Sarraf & Stancheva, 2004), centromere inactivation (Simpson & Sullivan, 2010), and position effect

variegation (PEV) (Festenstein et al., 1996; Reute & Spierer, 1992). The epigenome is dynamic in

22

response to intrinsic and extrinsic cues, integrating signals from developmental programs, hormones,

environmental stimuli and stochastic events (Jaenisch & Bird, 2003). As such, different cell types

possess unique epigenetic signatures according to their developmental trajectory, function and

genotype (Milosavljevic, 2011).

Epigenetic mechanisms involve the interplay between chromatin modifications and chromatin

remodeling, which interact with other regulatory proteins and non-coding RNAs (ncRNAs),

specifically microRNAs (miRNAs), long ncRNAs, and small interfering RNAs (siRNAs). Different

classes of histone- and DNA-modifying enzymes tightly regulate acquisition and removal of

chromatin modifications, namely DNA methylation and posttranslational modifications (PTM) of

histone proteins, both temporally and spatially (Reik, 2007). The concerted effort of these epigenetic

modifications influences chromatin architecture.

Chromatin remodeling is the foundation of epigenetic regulation and describes the interconversion of

chromatin to active (relaxed) or repressive (condensed) states, otherwise known as euchromatin and

heterochromatin respectively (Arrighi & Hsu, 1971; Babu & Verma, 1987; Simmons, 2007). The

manner by which nucleosomes interact with DNA underlies the conformation of chromatin. Each

nucleosome contains ~147 bases of DNA wrapped twice around an octameric complex composed of

dimers of H2A, H2B, H3 and H4; neighboring nucleosomes are separated by 10-50 base pairs of

unlinked DNA (Luger et al., 1997). The epigenetic marks acquired by DNA and histone proteins

affect the positioning of nucleosomes, thereby directing the accessibility of transcriptional machinery

and protein complexes to their conjugate sequences, namely regulatory regions such as gene

promoters, insulators, enhancers, and even within gene bodies (Berger, 2007; Reik, 2007). Inherent

to the DNA-nucleosome model is the fact that nucleosomes recognize and exhibit different affinities

for particular DNA sequences in cis (Segal et al, 2006). As such, transcriptionally active start sites

are depleted of nucleosomes at the 5’end and 3’ untranslated region (UTR), enabling for the

23

assembly and disassembly of polymerase and transcription factors (Schones et al., 2008; Segal et al.,

2006). These epigenetic marks guide the recruitment of ATP-dependent remodeling complexes (e.g.

SWI/SNF2; Shen et al. 2000), methyl-CpG-binding proteins (e.g. MECP2; Fuks et al., 2003), and

nuclear scaffolding proteins to the site of interest which facilitate changes in transcriptional activity

(Berger, 2007). Taken together, epigenetic regulation requires the cross talk between DNA

methylation, histone modifications, and in certain cases, ncRNAs, all of which will be addressed in

the following sections with an emphasis on DNA methylation.

1.5.1 DNA methylation and hydroxymethylation

DNA methylation is a fundamental epigenetic modification in mammals with essential roles in

embryonic development and gene regulation (Monk et al., 1987), while conferring genomic stability

for X chromosome inactivation (Okamoto et al., 2004; Heard et al., 2001) and allele-specific

expression for imprinted genes (Li et al., 1993). Deviations from normal methylation patterns have

been associated with various human diseases, most notably in cancer (Baylin et al., 2005), imprinting

disorders (Paulsen et al., 2001), and neuropsychiatric disorders (Feng & Fan, 2009). Additionally,

global loss of methylation causes embryonic lethality in mammals (Li et al., 1992).

DNA methylation, often concentrated at repetitive regions and transposons in most eukaryotes,

involves the addition of methyl groups to the five position of carbon rings of cytosine residues (5mC)

and accounts for 1% (5x107) of total nucleotide bases in the mammalian genome (Ehrlich et al.,

1982). The post-replicative addition of a methyl group to cytosine occurs through the action of DNA

methyltransferases (DNMTs) (Bestor, 2000), which use S-adenosyl L-methionine (SAM) as a methyl

donor (Chiang et al., 1996). Under oxidative conditions, 5mC marks undergo partial conversion to 5-

hydroxymethylcytosine (5hmC) by ten-eleven translocation (TET) dioxygenases, which transfer a

hydroxyl group to 5mC (Hee al., 2011). DNA methylation patterns are mainly established by de novo

DNA methyltransferases DNMT3a and DNMT3b (Okano et al., 1999) and maintained by DNMT1 in

24

DNA replication (Bestor et al., 1998), which preferentially binds to hemimethylated DNA (Song et

al., 2011; 2012). Fidelity of DNA methylation relies not only on DNMT1 activity, but also on the

localization of different DNMT3 isoforms such as DNMT3L and other chromatin binding proteins to

methylated DNA in certain contexts (Jones & Gangning, 2009). Changes in DNA methylation are

introduced in DNA replication either due to de novo methylation or 5mC to Uracil transitions;

nonetheless, the overall methylation status (e.g. hypomethylated; hyperrmethylated) of genomic

elements essentially remains conserved following many cell divisions (Riggs et al., 2004; Ushijima et

al., 2003). Meanwhile, removal of methylated marks can occur actively or passively. Passive

demethylation takes place in the absence of maintenance DNMT1 during DNA replication.

Conversely, evidence for active demethylation pathways come from non-replicating cells where loss

of methylation was observed at specific loci in neurons and genome-wide in post-fertilized sperm in

embryogenesis (Guo et al., 2011; Lee et al., 2002). The mechanisms by which active demethylation

occurs are not well elucidated, although recent studies have proposed the involvement of TET

enzymes and 5hmC as a reaction intermediate (Kangaspeska et al., 2008; Métivier et al., 2008).

In mammals, methylation preferentially occurs at cytosine-guanine dinucleotides (CpG) (Bird, 2002).

Globally, CpGs are non-uniformly distributed in the genome and often methylated, which accounts

for ~70% of total methylation. Most CpGs, however, are predominantly concentrated within CpG

islands (CGI), genomic regions an average of 1000 bp in length distinguished by a high GC content

(>50%) and an elevated frequency of CpG sites (≥ 0.6, the ratio of the number of methylated

cytosines to the total number of methylated and unmethylated cytosine residues) (Bird et al., 1985;

Rollins et al., 2005). Over half of CGIs coincide with the promoters of annotated genes (Saxonov et

al., 2006; Illingsworth et al., 2010), mostly located upstream of transcriptional start sites (TSS) of

housekeeping genes and enriched near genes restricted to tissue-specific expression in vertebrates

(Gardiner-Garden & Frommer, 1987; Weberr et al., 2007). In general, methylation status is inversely

25

correlated to gene expression, whereby hypomethylation and hypermethylation are associated with

gene transcription and gene repression respectively (Bogadanovicc & Veenstra, 2009); however,

certain CGIs retain their unmethylated status regardless of the transcriptional activity of their

associated gene (Antequera & Bird, 1988; Saxonov et al., 2006). Morever, methylation is not limited

to CGIs. Sequences up to 2 kb outside of CGIs known as CGI shores display tissue-specific

methylation and have the ability to regulate gene expression (Doi et al., 2009; Irizarry et al., 2009).

By contrast, repetitive elements, namely LINE-1, SINE and Alu sequences, in addition to gene

bodies, are extensively methylated in human somatic tissues (Weisenberger et al., 2005). Methylation

within gene bodies has also been associated with gene repression (Bird, 2002; Wolf et al., 1984). It is

interesting to note that methylation can also exist in non-CG contexts (CpH, where H = A, T, C)

(Lister et al., 2013). Overall, non-CG methylation is a rare stochastic event in most human somatic

tissues (Lister et al., 2009), but more common in cultured pluripotent stem cells including ES cells

(Ziller et al., 2011) and the mouse germ line (Smith et al., 2012; Tomizawa et al., 2011), where it has

been associated with cell differentiation and proliferation. The brain is an exception: CpH

methylation represents a key epigenetic mark in the mammalian CNS with neuron-specific functions

(Lister et al., 2013). As previously indicated, crosstalk exists between DNA methylation and histone

modifications.

1.5.2 Histone modifications

Coordinated histone modifications bring about changes to chromatin architecture and define

chromatin domains (Cheung et al., 2000; Rice et al., 2003), otherwise known as the histone code

(Jenuwein & Allis, 2001). Each core histone, specifically H3 and H4, possesses two conserved N-

terminal amino “tails” which are highly basic due to arginine (R) and lysine (K) residues (Zheng et

al., 2003; Luger et al., 1997). Protruding from nucleosomes, histone tails undergo diverse covalent

modifications including acetylation, methylation, ubiquitylation, phosphorylation and sumoylation

26

(Peterson & Laniel, 2004) and are carried out by histone modifying enzymes (Kouzarides, 2007).

Such modifications lead to corresponding changes in chromatin packaging, thereby affecting the

accessibility of various chromatin binding proteins and RNA polymerase (RNAP) to the region of

interest (Berger, 2002; Strahl & Allis, 2000). The best characterized modifications are acetylation

and methylation of lysine and arginine residues, which are mediated by histone acetyltransferases

(HATs), histone deacetlases (HDACs), and histone methyltransferases (Garcia-Cao et al., 2003;

Peters et al., 2001; Zhang et al., 2001). Histone acetylation is restricted to lysine residues, which

relaxes chromatin conformation by affecting the electrostatic interactions between negatively-

charged DNA and positively-charged histone tails (Grunstein, 1997). By contrast, lysine residues can

be found either mono-, di- or tri-methylated (Strahl et al., 1999), while arginine residues can be

mono- or di-methylated (Kouzarides, 2007). Unlike acetylation, methylation is associated with both

open and closed chromatin conformations and does not exert its effects by charge differences to

histone tails.

Chromatin-immunoprecipitaiton (i.e. ChIP-chip) and ChiP-seq analyses has revealed different

patterns of histone modifications demarcating functional elements in the genome (Zhou et al., 2011),

and thus diverse regulatory consequences. Histone modifications present at gene promoters correlate

with the level of transcriptional activity. Indeed. promoter regions of active genes show enrichment

for tri- and monomethylation of lysine 4 of histone 3 (H3K4) surrounding TSS (Bernstein et al.,

2002; Santos-Rosa et al., 2002; Young et al., 2011) in addition to augmented histone acetylation,

which has been well-associated with gene expression and enhanced transcriptional activity (Allfrey

et al., 1966; Hebbes et al., 1988). Collectively, these marks lead to open chromatin and reduce

nucleosome occupancy. By contrast, trimethylations of H3K9, H3K27 and H3K79 are all linked to

inactive promoters and heterochromatin (Barski et al., 2007; Volpe et al., 2002). It is important to

note that while H3K27me3 is considered a repressive mark, it may be present at actively expressed

27

loci (Robertson et al., 2009). The histone profiles at enhancers (defined as cis regulatory elements

that bind transcription factors and RNAP to facilitate activation of distal gene promoters) are cell-

specific, and may display a combination of mono- or di-methylation of H3K4 and/or acetylation of

lysine 27 of H3 (H3K27ac) (Heintzman et al., 2007; Kim et al., 2010), in addition to enrichment for

the histone acetyltransferase p300 (Creyghton et al., 2010). Insulator elements, which serve as

antagonists for gene expression by blocking enhancer-promoter interactions (Phllips & Corces,

2009), invariably require the binding of CCCTC binding factor (CTCF) to mediate chromatin

looping and formation of higher order chromatin (Hou et al., 2008; Splinter et al., 2006). Special

regions in the genome, known as bivalent domains, harbour both repressive (e.g. methylation of

H3K27) and active (e.g. methylation of H3K4) histone modifications either at a single or adjacent

nucleosomes (Bradley et al., 2006). Bivalent domains are located at genes “poised” for expression,

which enables quick transitions from condensed to relaxed chromatin states for developmental genes

in ES cells, as well as for some imprinted genes (Mikkelsen et al., 2007; Sanz et al., 2008).

Moreover, histone modifications can work synergistically with long ncRNAs.

1.5.3 Long non-coding RNAs

Long ncRNAs, RNA transcripts defined as longer than 200 bp in length (Strahl et al., 1999), are

organized in functionally demarcated regions, intervening with, or interspersed between coding and

non-coding transcripts, as well as intronic and lying antisense to some imprinted and protein-coding

genes (Bertone et al., 2004; Imanishi et al., 2004; Ota et al., 2004; Kapranov et al., 2007; Mercer et

al., 2009). Many long ncRNAs are highly conserved between mice and humans and associated with

epigenetic hallmarks of regulatory regions (Guttman et al., 2009). Supporting the biological

relevance of these transcripts, intronic long ncRNAs exhibit restricted expression at specific

developmental time points (Dinger et al., 2008; Ng et al., 2011), cell type or lineage (Dinger et al.,

2007; Mercer et al., 2010), as well as subcellular compartmentalization as evidenced in mammalian

28

brain (Mercer et al., 2008; Sone et al., 2007). Most notably, long ncRNAs have temporal, regional

and sub-cellular expression patterns in adult mouse and human brains, and parallel expression to

proximal brain-specific protein-coding genes (Mercer et al., 2008; Mercer et al., 2010; Ponjavic et

al., 2009). Such stringent expression profiles may be intrinsic to the regulatory function of long

ncRNAs.

Indeed, long ncRNAs have an emerging role in transcriptional regulation of protein-coding genes by

various epigenetic mechanisms, serving as guides and scaffolds for various chromatin-binding

proteins. Many long ncRNAs have been implicated in chromatin organization by the recruitment of

chromatin remodelling complexes (Khalil et al., 2009; Bertani et al., 2011). For example, the HOXC

antisense long ncNRA HOTAIR recruits Poycomb Repressive Complex 2 (PRC2) and binds to

LSD1, a histone demethylase, to promote heterochromatin formation and repression of the HOXD

locus in trans (Rinn et al., 2007; Tasi et al., 2010). Many long ncRNAs function in large-scale

changes to chromatin architecture, namely X chromosome inactivation which includes including Xist

and Tsix (Tian et al., 2010; Wutz et al., 2002). Additionally, long ncRNAs have been shown to act as

enhancers for gene expression (Orom et al., 2010; Orom & Sheikhattar et al., 2011), as evidenced by

the expression of ncRNAs at enhancers enriched in methylation for H3K4 in mouse neurons (Kim et

al., 2010). The mechanisms by which long ncRNAs influence enhancer activity is currently

unknown. However, taken together, these findings support the role of long ncRNAs in locus-specific

and allele-specific expression (Lee, 2009).

1.5.4 CGIs and transcription

The quantity and distribution of CGIs is well-conserved between mice and humans, including those

corresponding to gene promoters, in addition to intergenic (within a gene) and intragenic (between

genes) CGIs (Illingworth et al., 2010). As previously noted, ~50% of CGIs serve as intrinsic

promoters at annotated genes and often lack methylation. Exceptions include CGIs associated with

29

imprinted genes and genes on the inactive X-chromosome, which acquire allele-specific methylation

marks during development and remain transcriptionally silent throughout the life span (Illingsworth

et al., 2008; Weber et al., 2007). However, the remaining half represent orphan CGIs, which are

defined as intragenic or intergenic CGIs of unknown function (Illingsworth et al., 2010).

Significantly, ~40% of orphan CGIs exhibit functional epigenetic marks characteristic of active gene

promoters, including TSS, trimethylation of H3K4 and a general lack of methylation (Illinsgworth et

al., 2010). ChIP studies corroborate these findings, showing that ~21% of human orphan CGIs

overlap with RNAP II peaks (Illingsworth et al., 2010), indicating the presence of an associated

transcript. Indeed, intragenic CGIs located within introns or overlapping with exons of protein-

coding genes serve as promoters for ncRNAs. Most notably, the paternally-expressed mouse long

ncRNA, Air, is expressed from its CGI-promoter located in exon 2 of the protein-coding gene, Ifg2r

(Sleutal et al., 2002). Moreover, orphan CGIs also exhibit cell- and tissue-specific methylation

(Illingsworth et al., 2010; Maunakea et al., 2010). Taken together, DNA methylation, histone

modifications and lncRNAs work synergistically to influence transient and long-term changes in

gene expression, including genomic imprinting.

1.6 Genomic imprinting

1.6.1 Definition

For autosomal genes, genomic imprinting describes the process by which one parental allele is

preferentially or predominantly expressed while the other is selectively silenced by epigenetic

mechanisms (McGrath & Solter, 1984; Surani et al., 1990). The requirement for asymmetric

functional contributions from the paternal and maternal genomes was first elucidated by nuclear

transfer experiments in uniparental diploid (UPD) mouse embryos (Barton et al., 1984; Surani et al.,

1984). Androgenetic (AG) embryos (paternal disomy) were embryonic lethal and failed to support

embryonic development, while parthenogenetic (PG) embryos (maternal disomy) did not develop

30

extraembryonic tissues. Subsequent studies in AG and PG mouse chimeras (PG and AG embryos

transfected with wild-type cells and carrying a lacZ transgene) circumvented this lethal phenotype

and found that AG cells concentrated to mesodermal derivatives, while PG cells accumulated in

ectodermal derivatives, including the epidermis and the brain (Surani et al., 1990). Taken together,

these studies pointed to the competing interests of the parental genomes, in addition to the spatial and

temporal regulation of genomic imprinting.

Imprinted genes are often tissue-specific and developmentally-regulated (Dünzinger et al., 2007;

Smit et al., 2005). For instance, the paternally imprinted (or maternally expressed) mouse zinc-finger

protein, Zim1, is highly expressed during embryogenesis, particularly in developing limb buds, but

transcribed from both alleles in neonatal and adult brains (Kim et al., 1999). Paternal imprinting of

the ubiquitin ligase, UBE3A, is restricted to the brain (Rougeulle et al., 1997). Imprinted genes also

exhibit species-specificity, whereby those imprinted in eutherian mammals and marsupials for

placental efficiency and X chromosome inactivation are not imprinted in monotremes, birds or

amphibians (Gehring et al., 2009). A specific example is Commd1 in the mouse, which is paternally

imprinted and contains a reciprocally expressed U2af1-rs1; in contrast, COMMD1 in humans shows

expression from both parental alleles and does not contain a U2af1-rs1 homologue (Zhang et al.,

2006). Additionally, genomic imprinting can be restricted to a gene isoform (Smit et al., 2005;

Arnaud et al., 2003). Most splice variants of the human GRB10 gene are paternally expressed in fetal

brain; however, one isoform is exclusively transcribed from the maternal allele in skeletal muscle and

placenta while the remaining transcripts are biallelically expressed in all other fetal somatic tissues

(Blagitko et al., 2000). Moreover, the mouse insulin-like growth factor II receptor (Igf2r) and its

antisense transcript Air are reciprocally-imprinted in cortical glial cells and fibroblasts, but not in

neurons where Igf2r is transcribed from both alleles and Air is not expressed (Yamasaki et al., 2005),

thereby illustrating that imprinting can also be restricted to a cell type. Such strict regulation may be

31

attributed to the essential biological roles of genomic imprinting and the resulting severe disorders

that arise from their dysregulation.

1.6.2 Functional organization of imprinted genes

Genomic imprints are heritable epigenetic modifications established in germ cells on parental

chromosomes (Challet et al., 1991), such that following fertilization differential marks are

maintained throughout development and propagated to somatic tissues, conferring allele specific

expression (ASE).

Imprinted genes are arranged in chromosomal clusters throughout the genome and respond to

regulatory signals in cis from imprinting centers (ICs) which can range up to several bp in size, are

CG-rich and often correspond to CGIs. How cells discriminate between parental alleles at imprinted

loci is not well understood, however differential DNA methylation at regulatory elements has been

identified as a key epigenetic regulator for differential expression. Observations for the inheritance of

parent-specific methylation patterns (Bartolomei et al., 1993; Silva et al., 1998; Stöger et al., 1993),

in conjunction with experimental evidence showing mice deficient in DNMT activity with loss of

monoallelic expression (Li et al., 1993), illustrate the functional role of DNA methylation in eliciting

ASE. Accordingly, imprinted expression is often controlled by differentially methylated regions

(DMRs) which harbour parent-dependent allele-specific methylation (ASM) in addition to other

epigenetic modifications such as histone signatures (Reik & Walter et al., 2001). ICs or DMRs

disseminate gene regulation over long distances (up to 1 Mb) and control expression of many loci

independent of the parent of origin (Reik & Walter, 2001). However, while DMRs are synonymous

with imprinted genes, it is important to mention that not all DMRs function as ICs (Rakyan et al.,

2008). For example, IGF2R is biallelically expressed in human tissues despite the presence of a

DMR, and its tissue-specific imprinted expression is dependent upon parent-specific histone

modifications (Vu et al., 2004). Moroever, evolutionarily-conserved genes such as those involved in

32

olfaction possess DMRs despite their biallelic expression. Coordinated gene expression within

imprinting clusters is therefore mainly dependent upon (1) differential epigenetic marks at regulatory

elements and (2) interactions between neighboring genes and their control sequences.

1.6.3 Methylation dynamics in development

Erasure of genomic imprints is essential for resetting parental imprints at each generation (Morgan et

al., 2005) and occurs in mammalian development, whereby the germ line undergoes two waves of

genome-wide epigenetic reprogramming. Primordial germ cell (PGCs) are subjected to

demethylation and includes erasure of parental imprinting marks, which are subsequently re-

established later in gametogenesis according to the sex of the embryo (Feng et al., 2010). The second

wave of demethylation occurs at post-fertilization; however, methylated imprinted marks are

maintained throughout embryogenesis and read in somatic tissues in postnatal life (Khatib et al.,

2007; Ueda et al., 2000). The timescale and extent of epigenetic reconfiguration differs according to

species and cell type e.g. sperm and oocytes.

Erasure and re-establishment

In the developing embryo, primordial germ cells (PGCs) are highly methylated and, because of

imprinted genes, display parent-specific epigenetic marks (Hajkova et al., 2000). Proliferating PGCs

migrate to the gonadal ridges where they undergo rapid genome-wide demethylation, which results in

the loss of DMRs at imprinted genes, and is concomitant to the loss of H3K9me2 (Lees-Murdock et

al., 2003; Lee et al., 2002; Seki et al., 2007). Once in the gonads, PGCs commit to a germ cell fate

according to whether the gonadal ridge develops into an ovary or testis, by which time oocytes and

sperm enter into meiotic and mitotic arrest, respectively. De novo methylation by DNMT1 (Jajkova

et al., 2002; Okano et al., 1999; Webster et al., 2005) occurs in the prospermatogonia stage in males

(Kato et al., 2007), and after birth in maturing oocytes prior to ovulation in females (Morgan et al.

33

2005). Here, parent-specific methylation is established according to the sex of the embryo, however

methylated marks are dependent on DNMT3A and its cofactor DNMT3L (Demars & Gicquel, 2012).

Maintenance and propagation

Following fertilization, the genome undergoes a second wave of demethylation followed by another

wave of de novo methylation in the preimplantation embryo (Santos et al., 2002); however imprinted

genes along with some other sequences retain their gametic methylation (Reik & Walter, 2001).

Hours after fertilization, most sequences in the paternal genome are actively demethylated in the

zygote just prior to DNA replication (Mayer et al., 2000; Oswald et al., 2000). In contrast, the

maternal methylome undergoes passive demethylation during cleavage (Santos et al., 2002; Oswald

et al., 2000). Genome-wide de novo methylation is believed to occur after implantation in the inner

cell mass of the blastocyst by DNMT3a and DNMT3b (Okano et al., 1999), and is essential for

establishing somatic methylation patterns (Ratnam et al., 2002). After fertilization, methylation is

maintained by DNMT1o (oocyte variant) in the zygote, and then by DNMT1s in embryonic and adult

tissues (Howell et al., 2001).

1.6.4 Conservation of imprinted genes

Genomic imprinting is limited to a small proportion of genes: ~ 100 and ~50 imprinted genes have

been identified in mice (http://www.har.mrc.ac.uk/services) and humans

(http://igc.otago.ac.nz/home.html), respectively, and account for 0.5-1.0% of 20,000 protein-coding

genes in the genome. Cross-species comparisons have revealed conservation for imprinting across

various phylogenies including higher plants and mammals (Feil & Berger, 2008), and maintenance of

syntenic clusters and paralogues amongst eutherians and marsupials (Renfree et al., 2009), as well as

between amphibians, fish and chickens (Dünzinger et al., 2007). A main school of thought is that

genomic imprinting likely evolved from duplication and translocation events of preimprinted

chromosomal regions in lower vertebrates (Dünzinger et al., 2007; Walter & Paulsen, 2003;

34

Yokomine et al., 2005). The evolutionary conservation for imprinting can partly be explained by the

parental conflict hypothesis (Moore & Haig, 1991), which corroborates observations for conflicting

interests of maternal and paternal genes in maternal resource allocation to developing offspring

(Moore & Haig, 1991): paternal genes promote growth of his offspring while maternal genes limit

growth by conserving her resources for survival while providing nourishment to her current and

subsequent litters. As such, the hypothesis postulates that the function of an imprinted gene should

correlate with the allele from which it is expressed (Moore & Haig, 1991).

In line with this school of thought is the observation that a disproportionate number (~38%) of

imprinted genes are dedicated to embryonic development in placental mammals (Coan et al., 2005;

Monk et al., 2006; Reik et al., 2004). Extraembryonic tissues are enriched in genes affecting

placental growth along with its structural tissues (Constancia et al., 2002; Li et al., 1999; Mayer et

al., 2000), and in regulating the provision of maternal nutrients to the fetus (Dao et al., 1998; Mizuno

et al., 2002). However, although mice and humans exhibit remarkable concordance for imprinted

genes themselves and their syntenic regions, as well as parental origins for the active or silenced

allele (Tycko et al., 2002), only 30% of imprinted genes overlap between mice and humans (Okazaki

et al., 2002; Monk et al., 2005; Morison et al., 2006; Su et al., 2004). The discordance can be

attributed to loss of imprinting status (e.g. biallelic in human vs. monoallelic in mouse) or absence of

orthologous genes in humans (Garg et al., 2012; Suzuki et al., 2011; Morcos et al., 2011). This has

been primarily observed for imprinted genes in the placenta (Monk et al., 2006), which is not

surprising given that most humans pregnancies are singletons as opposed to pregnancies of multiple

births, which promotes intralitter competition and places pressure on placental efficiency (Reik et al.

2003). Consequently, the limited conservation of orthologues between mice and humans in the

placenta suggests that those which remain conserved in humans are not only remaining evolutionarily

critical, but are also required in postnatal adaptation and maternal behaviour as opposed to resource

35

provision and fetal growth (Keverne, 2001; Lefebvre et al., 1998; Li et al., 1999). Accordingly, it is

estimated that ~20% of imprinted genes are involved in neurological functions (Renfree et al., 2009).

1.6.5 Regulation of genomic imprinting

Differential epigenetic marks, namely DNA methylation, of key regulatory elements associated with

imprinted genes are inherently recognized by the appropriate transcriptional machinery, thereby

resulting in differential gene expression (Hutter et al., 2006). ICRs can differ in their mode of action

according to the imprinting cluster; nonetheless, common mechanisms have emerged through studies

in mouse chimeras and transgenic mice carrying allele-specific deletions or duplications within these

regulatory elements (Fitzpatrick et al., 2002; Lin et al., 2003; Thorvaldsen et al., 1998). Such

mechanisms include sex-specific modifications of promoter sequences, silencers, enhancers,

boundary elements, and overlapping antisense transcripts at imprinted loci (Constância et al., 1998;

Reik & Walter, 2001).

Promoter methylation

Transcriptional silencing at one allele is commonly achieved by methylation at promoter regions and

TSSs which tend to be CpG rich (Weber et al., 2007). Hypermethylation of CpG dinucleotides can

inhibit gene expression by direct or indirect mechanisms. First, methylated CpG dinucleotides can

prevent transcription factors from binding directly to DNA sequences (Watt & Molloy, 1988; Iguchi-

Ariga & Schaffer, 1989). Alternatively, methylated DNA can attract transcriptional inhibitors such as

methyl-CpG binding proteins (MBDs) (Hendrich & Bird, 1998). These MBDs can recognize

methylated DNA, leading to the assembly of repression complexes comprising of transcriptional

repressors, HDACs, and DNA methyltransferases (Hendrich & Bird, 1998). In vitro, MeCP2 was

shown to localize to a transcriptional–repression domain and was associated with the co-repressor

mSin3a (Nan et al., 1999). The subsequent condensation of DNA around histone cores leads to

36

heterochromatin formation, making the region inaccessible to transcription factors and results in gene

repression (Yamashita et al., 2005).

The Insulator Model

Some imprinting clusters are regulated by insulators, which are DNA boundary elements that inhibit

the interaction between two regulatory elements, such as promoters and enhancers (Reinhart et al.,

2002). The best characterized imprinting clusters for this model is located on human 11p15.5, which

harbours two imprinted domains controlled by two ICs referred to as IC1 and IC2 (Bartolomei et al.,

1991). IC1exhibits asymmetric methylation between the parental alleles and regulates monoallelic

expression of the paternally-expressed gene IGF2 and the maternally-expressed non-coding RNA

H19, and is located 2 kb upstream of the H19 promoter (Bartolomei et al., 1991). As such, the

methylated status of this insulator influences interactions between the H19 and IGF2 promoters and

their shared enhancer elements which lie downstream of H19 (Murrell et al., 2004). Indeed, while

IC1 is methylated in the paternal allele, its unmethylated status on the maternal allele facilitates the

binding of the insulator protein CTCF to the element (Szabo et al., 2000); this prevents both de novo

methylation at the IC and enhancers from interacting with the IGF2 promoter, resulting in H19

expression and IGF2 suppression (Hark et al., 2000; Bell et al., 2000). Binding of CTCF facilitates

higher chromatin structure and creates distinct chromatin boundaries between the IGF2 and H19

genes on the maternal allele (Li et al., 2008; Murrell et al., 2004; Tabano et al., 2010). Conversely on

the paternal allele, the methylated IC inhibits CTCF-binding and enables for interactions between the

IGF2 enhancer and downstream enhancers (Li et al., 2008). Loss of the insulator results in maternal

expression of IGF2 in specific tissues (Thorvaidsen et al., 1998).

The Non-coding RNA Model

The majority of imprinting domains contain overlapping antisense transcripts, most of which encode

short and long ncRNAs. In general, these closely juxtaposed transcripts are reciprocally-imprinted

37

relative to the surrounding sense gene, such that if the ncRNA is expressed on the paternal allele, the

sense gene will be expressed from the maternal allele (Beechey et al., 2001). Most of the ncRNAs

identified thus far are themselves imprinted and paternally-expressed, and have been shown to

regulate imprinting clusters (Reik & Walter, 2001).

The promoter regions of some antisense transcripts are CpG-rich or CGIs and therefore serve as

DMRs; the methylated state of an ncRNAs promoter determines the downstream effects of the

transcript on the expression of its overlapping protein-coding genes. Indeed, ncRNAs act as long-

range cis silencing factors, demonstrating spatial control within an imprinting domain and can be

located within promoter regions of sense genes or within intronic regions (Royo & Cavaillé, 2008).

Notable imprinting clusters that express lncRNAs include the Gnas locus (Li et al., 2000), the

Dlk1/Gtl2 locus (Lin et al., 2003) and the Snrpn locus (Tsai et al., 2005; Rodriguez et al., 2005),

though the best described are the Igf2r and Kcnq1 imprinting clusters and their overlapping ncRNA

transcripts, Airn and Kcnq1ot1, which originate in the introns of their respective sense genes (.

The Igf2r imprinting cluster comprises three maternally-expressed genes, Igf2r and solute carriers,

Slc22a2 and Slc22a3, and one paternally-expressed gene, the macro long ncRNA Airn, amongst a

few other non-imprinted genes (Sleutals et al., 2003b). The IC for this imprinting cluster also serves

as the promoter for Airn (Sleutals et al., 2002a). On the paternal allele, the unmethylated IC results in

Airn expression, which overlaps the promoters of protein-coding genes in an antisense direction and

results in the subsequent suppression of surrounding genes (Lyle et al., 2000; Sleutal et al., 2003b).

Conversely, the IC is hypermethylated on the maternal allele, which prevents the transcription of

Airn and allows for the transcription of the protein-coding genes Igf2r, Slc22a2, and Slc22a3

(Sleutals et al., 2002a). Parallel to Igf2r, the Kcnq1 locus also contains several maternally-expressed

genes and one paternally-expressed gene encoding a long ncRNA, Kcnq1ot1, whose promoter also

38

acts as an IC (Macini-Dinardo et al., 2006). The IC, termed KvDMR1, is hypermethylated on the

maternal allele which suppresses the ncRNA Kcnq1ot1 and activates adjacent imprinted protein-

coding genes (Thakur et al., 2004). Kcnq1ot1 is hypothesized to target repressive chromatin

modifying complexes to imprinted genes for silencing, and is not the only mechanism thought to

regulate imprinting at the Kcnq locus (Pandey et al., 2008). Deletion of both of these DMRs leads to

loss of expression of the antisense transcripts and loss of imprinting of the sense genes (Yamasaki et

al., 2005).

1.6.6 Deregulation of imprinted genes

The balance of paternally and maternally imprinted genes is critical to normal development and any

deviations from the imprinting process in gametogenesis increases the risk lead to imprinting

disorders and abnormal developmental outcomes (Chamberlain et al., 2010). Genetic and

environmental influences determine the outcomes of epigenetic programming, and defects

encompassing either factor, in addition to epigenetic causes, have been associated with explicit

imprinting disorders (Lucifero et al., 2004). However, imprinting disorders differ from sex-linked

disorders concerning inheritance, and thus, the resulting phenotypes in offspring. Indeed, the most

striking consequence of parent-of-origin-effects (POE) is bias in the inheritance of traits, such that

maternal traits are exclusively inherited down the matriline and paternal traits are inherited

exclusively down the patriline (Reik et al., 2001). Consequently imprinted traits show non-mendelian

patterns of inheritance, result in some traits skipping generations, and effect individuals of either sex

equally (Reik & Walter, 2001). In other words, the effects of genomic imprinting on autosomal genes

is indifferent to the sex of the offspring (Davies et al., 2004), but the parental origin of the allele is

critical. This differs from sex-linked traits, whereby the trait depends on the sex of the offspring and

conseqently affects males and females differently. Thus taken together, the phenotype in offspring

resulting from an imprinting disorder or POE depends on the allele from which it is inherited.

39

Accordingly, directional effects caused by deregulation of imprinted genes have been well-

documented at human chromosome 11p15.5 and the orthologous mouse region found on distal

chromosome 7, which contains the paternally-expressed insulin-like growth factor 2 (IGF2/igf2) and

maternally-expressed genes H19/h19, CDKN1C/cdkn1c, ASCL2/ascl2, TSSC3/tssc3 and

KCNQ1/kcnq1 (Kim et al., 1999; Fowden et al., 2006). Accordingly, mutations derived from the

paternal alleles are associated with Beckwith-Wiedemann Syndrome (BWS), a prenatal overgrowth

syndrome characterized by gigantism, large organs and increased risk for childhood cancer (Ping et

al., 1989; Elliott & Maher, 1994), whereas mutations originating from the maternal allele result in

Russell-Silver Syndrome, a disorder characterized by intrauterine growth retardation, low birth

weight (LBW) and congenital hemihypertrophy (Patton, 1988). Such mutations include inherited or

sporadic uniparental disomy (Cooper et al., 2005; Slatter et al., 1994), parental isodisomy (Cooper et

al., 2005), chromosomal deletions (Niemitz et al., 2004), as well as abnormal DNA methylation

(Catchpole et al., 1997). In addition, while the significance for imprinting and the underlying

mechanisms has been well-elucidated in development and fetal processes, emerging evidence in the

CNS suggests that imprinting is a critical epigenetic process for neurodevelopment and brain

function (Keverne, 2001).

1.7 Evidence for Imprinting in the CNS

Parallel work in humans and mice provide evidence for imprinting and parent of origin effects in

neurodevelopment, cognition and behaviour (Wang et al., 2010; Wilkinson et al., 2007;

Ideraabdullah et al., 2009).

1.7.1 Parent of origin effects in the brain

The most prominent evidence for imprinted genes in brain growth comes from AG-N and PG-N

mouse chimeras. Seminal studies by Allen et al. (1995) and Keverne et al. (1996) found that brains

40

derived from uniparental chimeras have different phenotypes concerning cell localization and brain

size. Most notably, PG and AG cells displayed reciprocal patterns of distribution in specific regions

of the brain: PG cells were found concentrated in the cortex but selectively excluded from the

hypothalamus; the opposite arrangement was seen with AG cells. These sex-specific patterns

persisted throughout the lifespan for both AG and PG animals and were readily observed in early

fetal development (~E11) and maintained in adult life (Allen et al., 1995). In addition to reciprocal

spatial distributions, neurodevelopmental abnormalities were observed in the mouse chimeras. AG-N

chimeras, despite having increased body weight, had reduced brain weight, and a reduced brain/body

ratio relative to control chimeras. This phenotype became more pronounced as the degree of AG

chimerism (greater contribution from AG cells) increased. Conversely, PG-N chimeras had reduced

body weight, and when normalized for body weight, an increased brain/body ratio (specifically a

larger forebrain) Given that brain size is normally correlated to body size, this paradoxical effect

suggests the maternal and paternal genomes have functional differences in the allocation of resources

to the brain and body development. The study authors hypothesized that these observed

neuroanatomical differences were the result of parent-specific biases to gene expression and thus

brain function (Allen et al., 1995; Keverne et al., 1996).

1.7.2 Identification of imprinted genes

Genome-wide analyses have identified imprinted genes in mouse brain and mapped their respective

expression patterns (Gregg et al, 2011; Gregg et al., 2010; Wang et al., 2008). Using whole

transcriptome sequencing (RNA-seq) and allele-specific SNPs, Gregg and colleagues (2010)

identified parent-of-origin effects in over 1300 loci (both protein-coding and ncRNAs) in embryonic

and adult brains. Most notably, 204 putative imprinted clusters were identified, and 52% of these

candidate clusters containing both coding and putative noncoding loci. This is particularly

interesting, since as previously indicated, non-coding genes can regulate imprinting clusters (Pandey

41

et al., 2008; Sleutels et al., 2002). Moreover, concomitant to previous studies, these genes showed

parental bias in the expression of individual genes and of specific transcript isoforms, developmental

period, and different brain regions, revealing a dichotomy regarding the influence of the paternal

genomes in brain function where the maternal genome was the main contributor to the embryonic

brain in development (Gregg et al., 2010). However, unlike the observations made in the mouse

chimera studies (Keverne et al., 1996; Allen et al., 1995), the paternal genome had marked influences

in the adult brain, namely in the hypothalamus, cerebellum and cortex. Individual functional studies

demonstrate that imprinted genes have diverse roles in the brain, including cell cycle regulation

(Arima et al., 2005), regulatory ncRNAs (Sleutals et al., 2002), transcription factors (Keverne et al.,

1996), protein trafficking (Greer et al., 2010) and RNA processing (Runte et al., 2001), cellular

functions which underlie proper neurodevelopmental and behavioural programs. The genes identified

to date exhibit both dynamic spatial and temporal regulation (Davies et al., 2004). For instance,

maternal expression of H19 is detectable in human fetal brain (6-12 weeks gestation) where it is also

highly expressed, but becomes downregulated and restricted to the pons and globus pallidus in

adulthood (Pham et al., 1998). Conversely, mouse Igf2 undergoes monoallelic expression in the adult

hypothalamus and pons (Pham et al., 1998). Alternatively, imprinted expression of the snoRNA

SNRPN remains imprinted in adult brain from development (Runte et al., 2001). Imprinting for some

genes may also be restricted to the brain, as demonstrated by UBE3A, with expression exclusively

from the maternal allele (Vu et al., 1997).

1.7.3 Imprinting in neurodevelopment and brain function

Previous work on mouse models with targeted mutations at imprinted loci has been associated with

cognitive and behavioural perturbations (review by Isles et al., 2000). PG chimeras demonstrated

significantly more aggressive behaviour in adult life (Allen et al., 1995). Mice with maternal

deficiency for Ube3a, an ubiquitin protein ligase with monoallelic expression restricted to mitral

42

cells of the olfactory bulb, as well as in hippocampal and Purkinje neurons, had impaired foetal

growth, motor dysfunction and learning deficits (Albrecht et al., 1997; Rouguelle et al., 1997).

Additionally, the absence of Gabrb3 expression in mice results in learning and memory deficits, in

conjunction to poor motor skills and hyperactivity (Delorey et al., 1998). Lefebvre et al., (1998)

showed that female mice deficient in the maternally imprinted gene, Mest (known as Peg1), had

abnormal maternal behaviour, in addition to embryonic growth retardation and decreased life

expectancy in the postnatal period.

Syndrome/

Disorder

Parental

allele

Common chromosomal

defect/de novo or

inherited

Disrupted or

candidate genes

Clinical phenotype Reference

(first author,

year) Angelman

Syndrome

M Chr15q11-13

Maternal deletion/de novo

or inherited

Paternal UPD

Abnormal methylation at

ICR

UBE3A, ZNF127,

NDN, SNRPN, IPW,

GABAA-receptor

subtypes

Absent speech

Severe learning disability

Dysmorphic facial

features

Happy affect

Inappropriate laughter

Epileptic seizures

Buiting, 1995

Hogart, 2009

Knoll, 1989

Sutcliffe, 1994

Prader-Willi

Syndrome

P Chr15q11-13

Paternal deletion

Maternal UPD

Mutations at IC

UBE3A, ZNF127,

NDN, SNRPN, IPW,

GABAA-receptor

subtypes

Hyperphagia, obesity

Compulsive traits

ASD

Developmental delay

Infantile hypotonia

Buiting, 1995

Hogart, 2009

Sutcliffe, 1994

Beckwith-

Widemann

Syndrome

M/P Chr11p15.5

Abnormal methylation

UPD: paternal

IGF2, CDKN1C Macroglossia

Macrosomia

Organomegaly

Sparago, 2004

Russell-Silver

Syndrome

M Chr11p15 (disruption to

IC)

Hypomethylation at

H19 and IGF2

Intrauterine and postnatal

growth retardation

Relative macrocephaly

Giquel, 2005

Schizophrenia P/M 15q13-q14 Genes within this

region

Disorganized speech and

thinking

Paranoia

Auditory hallucinations

Freedman, 1997

Ingason, 2011

Autism M 15q11-q13 Genes within this

region

Developmental delay

Social and

communication deficits

Stereotypies

Cook, 1997

Schroer, 1998

Table 1.1. Imprinting and neuropsychiatric disorders in humans. Explicit imprinting disorders,

neuropsychiatric and neurodevelopmental disorders, and their associated genomic defects and clinical

43

phenotypes. The imprinting clusters and imprinting mechanisms governing their respective chromosomal

regions are well-elucidated. *M = maternal and P = paternal; UPD = Uniparental disomy

Work in humans on the patterns of inheritance for several cognitive and neurological disorders

widely acknowledges the importance of imprinted genes in the brain. Indeed, sexual dimorphisms in

human brain function are most evident in explicit imprinting syndromes, with characterization of

specific genetic factors in mouse models. Such conditions, which may be associated with disrupted

neurological function and abnormalities, manifest according to the parental origin of the genetic

aberration, as summarized in Table 1.1. For example, cytogenetic defects at chromosome 15q11-q13

lead to the neurogenetic disorders Prader-Willi (PWS) or Angelman Syndrome (AS), depending on

whether the mutation is located on the maternal or the paternal allele. PWS, characterized by infantile

hypotonia, obesity, developmental delay and ID, can arise from one of three mechanisms on the

paternal allele: mutations at the IC, maternal UPD caused by non-disjunction, and paternal deletions,

all of which involve loss of expression of paternally-derived genes (Cassidy et al., 2000; Sutcliffe et

al., 1994; Veltman et al., 2004). PWS individuals have neuroanatomical variations, including

anomalies to the anterior pituitary gland, ventriculomegaly and abnormal cortical development

(Miller et al., 1996; 2007; Yoshii et al., 2002). In contrast, AS is caused by reciprocal aberrations to

PWS, specifically maternal deletions, paternal UPD, or mutations in the genes exclusively or

preferentially expressed form the maternal allele, specifically UBE3A (Kishiro et al., 1997; Knoll et

al., 1989; Malcolm et al., 1991; Malzac et al., 1998). Individuals with AS often have ID, seizures,

motor dysfunction and a happy disposition (Dykens et al., 1992; 2000), concomitant to microcephaly

with mild cerebral atrophy marked by the loss of Purkinje and granule cells (Dorries et al., 1988; Jay

et al., 1991). Additionally, certain regions demonstrate linkage of behavioural traits with the parent-

of-origin (Francks et al., 2003; Francks et al., 2011).

44

1.8 Genomic at the TRAPPC9 locus

The trafficking protein particle complex 9 (TRAPPC9; chr8q24.3) gene has been implicated in

autosomal recessive ID (Kakar et al., 2012; Marangi et al., 2012; Jamra et al., 2011; Koifman et al.,

2010; Phillipe et al., 2009; Mir et al., 2009; Mochida et al., 2009). Homozygous frameshift or

nonsense mutations in TRAPPC9 cause nonsense-mediated mRNA decay and protein truncation; the

resulting decrease in protein expression adversely affects NF-KB signaling and vesicular trafficking,

pathways which involved in synaptogenesis and neurogenesis (Hu et al., 2005; Mattson &

Camoandola, 2001; Zong et al., 2012). We have recently identified a subset of individuals carrying

copy number variations (CNVs), some of which are from unaffected parents while others are de novo

in origin. Although the relationship of these unique deletions and duplications in autistic and ID

etiology remain unknown, there is evidence indicating the involvement of an imprinting mechanism

that may be conserved between mice and humans.

Genomic imprinting at mouse distal chromosome 15 has been examined by our group based on the

observation that the syntenic region in mouse contains the known paternally expressed gene, Peg13,

located within an intron of Trappc9 (Figure 4.2). Accordingly, allelic expression of Trappc9 was

assessed by our collaborator Dr. Kazuhiko Nakabayashi in offspring derived from reciprocally

crossed C57BL/6 x JF1 mice (to account for strain differences) in the embryonic period (15.5 dpc;

days post coitum), at postnatal day (PD) 0, and at 8 and 24 weeks. Exclusive or preferential

expression of Trappc9 from the maternal allele (paternally imprinted) was restricted to a tissue,

developmental period and isoform (Appendix 3), as would be expected for an imprinted gene. In the

CNS, Trappc9 exhibited maternal-specific expression in fetal brain, in addition to the cerebrum,

cerebellum, olfactory bulb, neurons and neural progenitor cells (NPCs); however, Trappc9 was

expressed equally from both alleles in glia, fibroblasts and leukocytes. Allelic imbalance of Trappc9

persisted until 8 weeks postnatal in mice and subsequently disappeared, at which point the gene

45

became biallelically expressed in all brain tissues analyzed.Trappc9 resides between two transcripts

with preferential maternal expression: the downstream (and proximal) two-pore potassium channel,

Kcnk9, (Ruf et al., 2007) and the upstream catalytic component of the RNA-inducing silencing

complex (RISC), Ago2 (also known as Eif2c2) (Gregg et al., 2010; Morita et al., 2007), which binds

to Dicer and is involved in miRNA biogenesis (Chendrrimada et al., 2005). A paternally-expressed

non-coding RNA (ncRNA) Peg13 on the negative strand lies in the 16th intron of Trappc9, and is

postulated to regulate imprinting and its associated CGI serves as a DMR (Smith et al., 2003; Davies

et al., 2004).

While the imprinting status of TRAPPC9 in humans has not previously been studied, EIF2C2/AGO2

is believed not to be imprinted, however, differential expression (paternal imprinting) of KCNK9

(MIM:605874) remains conserved in humans (Ruf et al., 2007; Luedi et al., 2007), in addition to the

hypomethylated profile of its promoter in brain and blood (Ruf et al., 2007). Of clinical relevance, a

mutation to the active maternal copy of KCNK9 has been linked to Birk-Barel syndrome (Barel et al.,

2008). It is currently not known what mechanism is governing KCNK9 imprinting. Two transcripts –

AK307073 and AK748239 – flanking a CGI is presumed to encode one continuous transcript on the

negative strand and lies in intron 17 of TRAPPC9. This purported ncRNA shares no sequence

conservation with mouse PEG13; however there is evidence suggesting that it, too, may be paternally

imprinted and serve as the PEG13 equivalent in humans.

I proposed that the heterozygous deletions in our probands disrupt an imprinted locus by influencing

epigenetic mechanisms, specifically DNA methylation and the expression of a regulatory long non-

coding RNA (lncRNA), in an allele-specific manner. The primary objective of this study was to

identify whether a putative long ncRNA of unknown function, PEG13, located within an intronic

region of TRAPPC9 was imprinted in human fetal brain, and in turn, determine its regulatory role on

46

the expression of proximal imprinted genes, specifically the maternally-expressed gene KCNK9. The

putative PEG13 locus was studied by:

(A) Analyzing allelic expression of PEG13 in humans

(B) Region-specific DNA methylation at CGIs at PEG13 and KCNK9

(C) To determine the relevance of the heterozygous CNVs at TRAPPC9 in ID and ASD aetiology

Figure 1.2 Mechanism of deregulated imprinting in the brain. Balanced contributions from the

paternal and maternal genomes is required for normal brain function and development. Disruptions to this

balance can lead to neurodevelopmental disorders by: (1) abnormal expression of the active allele caused

by mutations and/or (2) loss of imprinting mechanisms, such as aberrant methylation patterns at the

DMR.

Loss of ICR

Altered gene dosage

Abnormal neurodevelopment

Disruption to imprinted

genes

Uniparental disomy (UPD)

Sturctural Variation

Changes in methylation

Neurological and behavioural

disorders

47

Chapter 2: Materials, Methods and Techniques

Introduction and study design

This section will outline the shared sample collection protocols and techniques used to generate data,

as well as their purpose in the overall study design to avoid repetition. Details, including the methods

employed and data analysis, in addition to differences and deviations from the general

methodologies, will be addressed and expanded upon in each of the appropriate study chapters.

A relevant institutional ethics board approved all sample collections. Written informed consent was

obtained from all participants or their parents. Total DNA and/or RNA were collected from human

fetal tissues, fibroblasts, sperm and peripheral blood. Traditional genetic methods were applied for

identifying the putative PEG13 locus, specifically polymerase chain reaction (PCR), Sanger

sequencing and pyrosequencing technology (Chapter 5). Analysis of methylation patterns was carried

out using bisulfite pyrosequencing (Chapter 6). Lastly, real-time quantitative polymerase chain

reaction (RT-qPCR) was used to validate CNVs and for gene expression profiling (Chapter 4 and 5).

48

2.1 Statistical Analysis

This study primarily used quantitative methods. Allelic expression and methylation levels were

expressed quantitatively in percentage ranging from 0% to 100%. Pyrosequencing data for the fetal

tissues was performed in triplicate and the mean and S.D. of the technical replicates was calculated.

For data from bisulfite pyrosequencing, each CpG site was represented as a box and whisker plot to

show the distribution of methylation levels across ten samples (i.e. leukocytes, sperm and

fibroblasts). Statistical analysis (i.e. S.D., median, mean) for allele frequency and methylation was

performed by GraphPad Prism. Similarly, gene expression and DNA copy number were quantifiable

variables and normalized for variations among technical replicates using an internal control prior to

the statistical analysis.

2.2 Sample Quality and Quantification

Verification of nucleic acid integrity and concentration were assessed using the The NanoDrop®

ND-1000 spectrophotometer (Thermo Scientific) and by gel electrophoresis. An optical density (OD)

260/280 ratio of 1.8 to 2.0 and an OD 260/230 ratio greater than 2.0 was considered acceptable

nucleic acid quality. RNA was stored at -80˚C and DNA at 4˚C for long-term storage.

2.3 Sanger Sequencing

Prior to sequencing, PCR products were cleaned using an equivalent volume of microCLEAN

reagent (Microzone). Samples were diluted according to sequencing guidelines and submitted to The

Center for Applied Genomics (TCAG; Toronto, ON).

2.4 Isolation of Nucleic Acids from Whole Blood Anticoagulated blood was fractionated by centrifugation at 1000 x g for 15 min and the supernatant

was removed. Contaminants and residual erythrocytes were lysed and the remaining leukocyte-rich

fraction was pelleted by centrifugation.

49

DNA was extracted from peripheral blood leukocytes using a high salt method. Briefly, leukocytes

were re-suspended in equal volume of 2x lysis buffer (20 mM Tris-Cl pH 7.6, 20 mM KCl, 20 mM

MgCl2 and 2% Triton-X-100; 107cells/ml), incubated on ice for 5 min, and pelleted by centrifugation

at 1350 x g for 15 min at 4ºC. The supernatant was discarded and nuclei pellets were resuspended in

5ml of lysis buffer (10 mM Tris-Cl, 10 mM KCl, 10 mM MgCl2•6H2O, 2mM EDTA and 0.4 NaCl).

Protein was denatured by 10% SDS and the suspension was incubated at 55ºC for 20 min and cooled

to RT. NaCl was added to the digested product and mixed vigorously for 15 min followed by

centrifugation at 16, 000 x g for 15 min at 4ºC. The DNA was precipitated in pure EtOH and the

spooled DNA was soaked in 70% EtOH for 10 min. The DNA was air dried for 10 min and

incubated at 55ºC in 500 μl TE buffer (10 mM Tris-HCl, 0.5 mM EDTA, pH 7.6) until DNA was

dissolved. The resulting DNA was stored at 4ºC.

The RNeasy kit (Qiagen) was used for RNA isolation from peripheral blood leukocytes according to

manufacturer’s protocol. To ensure the absence of DNA contamination, RNA was treated to DNase I

per 10 μg of template (Life Technologies).

Cell Lines Peripheral leukocytes were transformed to lymphoblasts using the Epstein - Barr virus (EBV) to

create immortal cell lines for test subjects. From whole blood, isolated white blood cells were washed

with RPMI-1640 growth media, inoculated with EBV and grown at 37˚C until confluency was

reached. Lymphoblasts were fed 15% RPMI streptomycin-penicillin weekly. The transformed

lymphoblasts were transferred to a 2 mL flask, cyropreserved in an isopropanol chamber (Mr.

Frosty™) and stored in liquid nitrogen (N2).

50

2.5 Human Fibroblasts

Certain classes (5%) of autosomal genes exhibit random monoallelic or preferential allelic expression

not associated with genomic imprinting (Gimelbrant et al., 2007); moreover, few imprinted genes in

mice have been validated in human cell lines (Morcos et al., 2011). Investigating allelic expression in

fibroblasts serves as a cost-effective, yet straight-forward method for validating the imprinting status

at a novel locus. Based on an earlier study which performed a genome-wide analysis of imprinted

genes in human cells lines (Morcos et al., 2011), fibroblast cell lines (n=10) were obtained from the

NIGMS Human Genetic Cell Repository (Coriell Institute, New Jersey) on recommendation by Dr.

Tomi Pastinen (McGill University) to validate and quantify allelic expression of three SNPs found

around the PEG13 region: rs2270409, rs3802217 and rs4289794 (Table 2.1). These SNPs were

biallelic in lymphoblasts and other SNPs around TRAPPC9 were biallelic as well. Fibroblasts were

derived from individuals with rare metabolic syndromes but no neurodevelopmental disorders, and

were informative (i.e. heterozygous) for SNPs around the purported PEG13 locus. An equal number

of males (n=5) and females (n=5) were selected and ranged from 4 months to 54 years in age.

2.5.1 Culturing and Growth Conditions

Individual fibroblasts cell lines were initially sub-cultured in T25 flasks containing 5ml of

fibroblastgrowth media (Dulbecco’s Modified Eagle Medium; DMEM Gibco®; 10% fetal bovine

serum (FBS) and 5% Streptomycin-Penicillin), and incubated at 37˚C, 5% CO2 overnight. Cells were

washed in 5 ml Phosphate Buffer Saline (PBS; Sigma-Aldrich) and detached from the flask surface

by Trypsin/EDTA for 2 min (Sigma-Aldrich). Fibroblasts were subsequently passaged to T75 flasks

and cultured in 15 ml growth media at at 37˚C, 5% CO2; media wash changed every 4-5 days. When

cells reached confluency (seeding density ~ 1.0 - 1.4 x104 cells/cm2), RNA and DNA was extracted

according to manufacturer’s protocol (Machery-Nagel, Nucleospin® RNA and Cultured Cells or

51

Tissue). Cells were harvested and pelleted by centrifugation at 1000 x g for 15 min prior to nucleic

acid extraction.

2.5.2 Cryogenic storage

Cells were washed in PBS and passaged as described above (Section 3.5.1). Aliquots (5 ml each with

5.0 x 105/ml) were pelleted by centrifugation at 100 x g for 10min (10˚C). The fibroblast pellet was

re-suspended in 1 ml freeze media (5% DMSO, 1 ml; DMEM per aliquot), distributed in cryovials

and frozen for cryogenic storage in liquid N2.

2.6 cDNA preparation for reverse transcription

First-strand cDNA was synthesized from 1 μg of RNA using Superscript III™, a modified Moloney

Murine Leukemia Virus Reverse Transcriptase (M-MLV RT), and random primers (10mM) in a 20μl

reaction volume according to the manufacturer’s guidelines (Invitrogen). Reverse transcriptase and

RNA were omitted for two separate aliquots to confirm the absence of DNA contamination. The

reaction mixture was incubated at 25˚C for 5 min followed by 50˚C for 60min, and then inactivated

by heating at 70˚C for 25min. cDNA was stored at -20 ˚C.

2.7 Real-time quantitative PCR

2.7.1 Principle

Real-time quantitative PCR (RT-qPCR) involves the measurement of PCR product accumulation in

real-time and represents the standard method for quantifying gene expression and copy number

within targeted regions across multiple samples. RT-qPCR serves as an instrumental tool for

validating high throughput assays such as genome-wide microarrays, in addition to providing

complementary evidence for phenotypic studies because of its reproducibility and high sensitivity.

The molecular basis of RT-qPCR exploits the kinetics of PCR by measuring the fluorescent signal

generated by various detection chemistries, including dyes, which may intercalate with double-

52

stranded (ds) DNA (SYBR® green), hydrolysis probes (TaqMan®) or molecular beacon probes.

During the lag or linear phase, fluorescence is at background levels as the initial amplicons are

produced after 10-15 cycles (Schefe et al., 2006). Eventually, as amplicon production enters the

exponential stage, the fluorescence intensity is significantly higher than baseline and becomes

detectable. The cycle at which this occurs, otherwise known as the quantitative endpoint of the

amplification target (∆CT value), correlates to the amount of PCR product produced. The quantity of

PCR product doubles each cycle in the log-linear phase and eventually plateaus as the amount of

reagent and template diminishes, and the fluorescence signal is no longer detectable or quantifiable.

As previously indicated, experimental analysis use the CT value, defined as the PCR cycle at which

the fluorescent signal crosses an arbitrary threshold, which determines the amount of PCR product

amplified in each reaction well (Schmittgen & Livak, 2008). Therefore, the CT value is inversely

proportional to the amount of amplified product – a high ∆CT represents low expression while highly

expressed genes have low ∆CT values; the same principle is applied for quantifying genomic copy

number variation (CNVs) and for genotyping SNPs. In turn, the CT value depends on the amount of

template present at the beginning of the PCR reaction, as the more starting material, the quicker the

fluorescence reaches threshold.

53

Catalog

No Sample description

Age

(yr) Sex Origin

Passage

#* PDL rs2270409 rs4289794 rs3802217

GM00042 Hurler Syndrome 25 M Caucasian 4 0 0.3624 0.3596

GM00519

Mucopolysaccharidosis

Type VI 4 F Caucasian 3 7.87 -0.4273 -0.3762 -0.3701

GM01096

Mucopolysaccharidosis

Type IIIA NA M Caucasian 6 6 0 0.3624 0.3596

GM01376 Homocystinuria NA M Caucasian 6 -0.375 -0.3788 -0.3994

GM01391 Hurler Syndrome 9 mo F Caucasian 6 -0.3336 -0.4402 -0.4202

GM01653

apparently healthy

non=fetal tissue 37 M Caucasian 12 5.64 0 0 0

GM02175 huntington disease 55 M Caucasian 8 -0.3978 -0.3229 -0.3135

GM02455

Mucopolysaccharidosis

Type IVB 6 F Caucasian 3 -0.3434 0 -0.3641

GM03073

Adenine

monophosphate

deaminase 1 38 F Caucasian 5 -0.3183 -0.4294 -0.4267

GM09256

Von Hippel-Lindau

Syndrome 54 F Caucasian 14 0 -0.4578 -0.4186

Table 2.1 List and description of individual fibroblast cell lines. Fibroblasts (n=10) were derived from individuals with rare, metabolic

syndromes and no history of neurodevelopmental or neuropschyiatric disorders. Sample selection included an equal proportion of males

and females and age ranged from 4 months to 54 years (Coriell Insittute). Raw SNP data (Illumina IM Human Array) was provided by Dr.

Tomi Pastinen (McGill University) for SNPs within the PEG13 region for each individual. SNPs demonstrating allelic imbalance (AI) had

a ratio > 0.25 or x < - 0.25 (expression ratio of the paternal allele versus the maternal allele) and a value of 0 indicated biallelic expression.

Pyrosequencing was used to quantify AI at SNPs rs2270409, rs4289794 and rs380221 in each individual * = Passage number prior to sub-

culturing; mo = months; NA = Not available; PDL = Population doubling level.

54

Data for RT-qPCR can be obtained via two methods. The first, absolute quantification, determines

the total number of copies expressed in a system, which calibrates data against a standard curve.

Absolute quantification is not always necessary because information regarding the fold-change is

often sufficient. Alternatively, relative quantification involves measuring gene expression against a

reference gene otherwise known as an internal control. This method utilizes reference genes in order

to normalize mRNA levels between different samples, enabling for an accurate comparison of gene

expression or gene copy number across samples. The internal control is usually a housekeeping gene

expressed in all nucleated cell types and has roles in general cellular processes such as

glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in cellular respiration or Βeta-actin (β-actin)

for cytoskeleton architecture. However, unlike previous assumptions, housekeeping genes differ in

their expression levels across various biological samples in both the absence and presence of intrinsic

and extrinsic stimuli (Dheda et al., 2004; Schmittgen & Zakrajsek, 2000; Radonic et al., 2003).

Therefore, selection of a suitable reference gene(s) requires certain assumptions, mainly that the

expression of the reference gene should be uniform across samples, and if applicable, independent of

experimental treatment and developmental time points.

A number of mathematical models exist for analyzing data for relative expression across different

samples. One such method is the comparative cycle time ∆∆CT (2-∆∆CT) or Livak method (Schmittgen

& Livak, 2001), which requires the experimenter to make a few assumptions. Most notably, the PCR

efficiency must be close to 1 and uniform across all targeted genes within 5% of each other

(including the reference gene). Moreover, this method requires an equivalent amount of starting

sample for all reactions. Normalization against known reference gene(s) with stable expression

levels, in addition to good RNA quality, can circumvent most of these limitations.

55

2.7.2 Assay design and validation

Genomic positions were determined by UCSC genome browser (http://genome.ucsc.edu; hg19), from

which primer pairs (Integrated DNA Technologies) were designed according to strict parameters

using Primer3Plus (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) :

Amplicon length: 80-150 bp long

Primer length: 20/22/30 (minimum/optimum/maximum)

Primer GC content: 40%/50% (minimum/maximum)

The GC-content was greater at the 5’ end and less toward the 3’ end (max 3’ stability: 2)

Each primer had a calculated melting temperature (Tm) of ~ 60˚C and maximum Tm

difference of 0.1˚C between primer sets

To minimize primer dimer formation, primer pairs had a maximum self-complimentarity of 5

nucleotides and a max 3’ self complimentarity of 2 nucleotides

Homopolymers of more than 3 nucleotides were avoided

All primer sets were validated in silico for sequence specificity using The BLAST program from the

NCBI, as well as empirically by a temperature gradient and separation of PCR products by gel

electrophoresis. A no template control (NTC) reaction was included for each primer pair.

2.7.3 Experimental design

The ∆∆CT method was employed for all assays using SYBR® Green reagents and ROX™ reporter

dye (Life Technologies). PCR efficiency (the rate at which polymerase produces amplicons) for each

primer pair was determined by the shape of the logarithmic amplification plot and by melt-curve

analysis. Assays were designed using a single 384-well optical plate to prevent inter-plate variability

and each sample was run in quadruplicate. Internal controls were selected based on the assay type

(i.e. gene expression vs. CNV analysis) and control samples were incorporated in all runs. To test for

genomic DNA contamination, a NTC control was included for each primer set.

56

2.7.4 Procedure

For this study, each RT-qPCR assay used singleplex PCR. All wells contained a 16 μl final reaction

volume containing 2 μl template (5ng/μl), 8 μl Fast SYBR® Green Master Mix, 0.5 μl each of the

forward and reverse primer (10 μM) , and 5.0 μl of dH2O. PCR amplification was achieved using the

Applied Biosystems viiA7 TM (Life Technologies), with an initial denaturation at 95˚C for 2.5 min,

followed by 40 cycles at 95˚C for 3.0 s and 60˚C for 20 s.

2.8 Pyrosequencing

2.8.1 Principle of pyrosequencing

Pyrosequencing is a real-time, sequence-by-synthesis technique based on the detection of light

following the release of inorganic pyrophosphate (PPi) to measure allelic frequencies (SNP) and

DNA methylation patterns. The theory and methodology behind pyrosequencing is summarized in

detail by the following papers summarize the theory and methodology behind pyrosequencing

(Dupont et al., 2004; Tost & Gut, 2007) and reviews (Ronaghi, 2001; Fakhrai-Rad et al., 2002); the

biochemical basis behind the reaction will be briefly outlined below.

Following the annealing of a sequencing primer, DNA polymerase incorporates a nucleotide

complementary to the 3’ end of the immobilized DNA template. Only one nucleotide at any given

time is present in the reaction system, which contains a proportional mixture of enzymes,

deoxyribonucleotide triphosphates (dNTPs) and substrates dispensed by a cartridge according to a

predetermined program. The addition of nucleotides occurs in a stepwise manner as a consequence of

four enzymatic reactions. Once DNA polymerase I (Escherichia coli) adds a dNTP to single-stranded

(ss) DNA, PPi is released and serves as a substrate for ATP sulfurylase (Saccharomyces cerevisiae),

which converts it to ATP in the presence of adenosine 5’phosphosulfate (Tost & Gut, 2007).

Luciferase (Photinus pyralis) utilizes energy from ATP to oxidize luciferin to oxyluciferin and

57

releases light as a by-product. Because dATP is the natural substrate for luciferase, sequencing uses

alpha-thio triphosphate (dATPɑS) instead, and is recognized and efficiently used by DNA

polymerase. A charge-coupled device (CDD) camera simultaneously captures the total amount of

light produced in each well by converting photons to a quantifiable charge and are visualized as a

series of peaks in a Pyrogram® (Appendix 2). The peak heights represent the quantity of nucleotides

added to the 3’ end as the complementary DNA strand is built-up. Apyrase (Solanum tuberosum)

degrades all unincorporated dNTPs and extra ATP molecules before the addition of the next

nucleotide, after which the cartridge dispenses a new nucleotide according to the dispensation

program. Nucleotides are dispensed in alternating wells to ensure that the light measured is actually

produced by that specific well (Tost & Gut, 2007). If a dispensed nucleotide is not complementary to

the sequence, it will not release PPi and will subsequently be degraded. As such, the bioluminometric

signal is directly proportional to the amount of ATP produced, which in turn, is equimolar to the

number of PPi released or nucleotides incorporated in the sequence. For instance, if two adjacent

nucleotides in a sequence are identical, such as two thymines, the peak height will be twice as high as

compared to a peak generated by a single thymine in the sequence (Tost & Gut, 2007). Furthermore,

the ratio of nucleotides incorporated at variable positions as seen in SNPs (e.g. A/G) and methylation

(i.e. methylated vs. unmethylated) is expressed as the degree of expression or methylation

respectively. Consequently, the intrinsic kinetic properties - KM (μM) and Kcat (S-1) – of each

enzyme, in addition to the reagent concentrations, dictate the pyrogram profile (Ronaghi, 2001).

2.8.2 Experimental design

Primer and amplicon design

Three primers are required for pyrosequencing: a primer pair for PCR amplification and an internal

sequencing primer. All oligonucleotides were purchased from IDT and listed in their respective study

chapters. Online softwares OligoCalc (http://www.basic.northwestern.edu/biotools/oligocalc.html)

58

and Multiple primer analyzer by Thermo Scientific Multiprimer Tool

(http://www.thermoscientificbio.com/webtools/multipleprimer/) were used to determine melting

temperatures, possible hairpin formation and self- and cross-primer dimers. All primers had a

calculated melting temperature (TM) of ~60˚C. NCBI BLAST or UCSC BLAT validated that primer

pairs were unique to one position within the PCR product.

Amplicons between 100-300 base pairs in length were designed to facilitate amplification and

sequencing. PCR primer pairs differed in melting temperature (TM) by less than 1˚C and included the

5’ biotinylation of one amplification primer. Usually the reverse primer was biotinylated for a

forward sequencing reaction and was opposite to the direction of the sequencing primer1. The PCR

primers were designed in regions containing no variable positions and were preferably GC-rich near

the 5’ end and less toward the 3’ end to improve primer specificity. Similarly, no variable positions

were included in the design for sequencing primers, which were 18 to 25 bases (optimally 20bp) in

length and had a calculated TM between 45˚C to 50˚C.

PCR Amplification

PCR amplification of either DNA or cDNA (depending on the application) was used to generate a

sufficient amount of template and for sequence specificity. Optimization of PCR conditions

(temperature and concentration gradients) was performed using the T100™ Thermal Cycler 186-

1096 (BioRad). Each reaction system contained 2.0μl of template (> 10 ng), 1.0μl of each forward

and reverse primer (10 uM) and 12.5μl of KAPA2G Fast HotStart ReadyMix in a 25μl reaction

volume. The negative control contained no template and a positive control included already

optimized primers. PCR conditions followed conventional protocol, with an initial denaturation at

95˚C for 1 min, followed by 40 cycles at 95˚C for 10s, annealing temperature between 50˚C -65˚C

1Biotin, a B-vitamin, has high affinity for Sepharose. Streptavidin-coated sepharose beads capture

biotinylated ssDNA templates for sequencing by in the pyrosequencing assay.

59

for 10s, and extension at 72˚C for 1.0s. The high number of cycles exhausted all primers and

minimized sequence interference from potential artefacts. To ensure primers were sequence specific

and sufficient template was yielded at the appropriate size, 5.0 μl of PCR product was visualized on a

2.0% agarose gel under trans-UV light.

2.8.3 Pyrosequencing procedure

Pyrosequencing was performed using the PyroMark™ Q24 System using Q96 Gold (5x96) reagents

(Qiagen) according to manufacturer’s instructions. Each biotinylated PCR product (20 μl) was

immobilized to 2.0μl Strepavidin-coated Sepharose (GE Healthcare Life Sciences Division,

Piscataway, NJ) in 40μl 2X binding buffer and 18 μl distlled H2O. The mixture was agitated at 1400

rpm (Thermoscientific) for 10 minutes. The Vacuum Tool (Pyromark™ Q24) captured the beads

and rinsed them in 70% ethanol (70 ml), denatured in NaOH (50 ml), and neutralized in wash buffer

(70 ml). The beads, now bound to ssDNA template, were released from the vacuum into a 24-well

plate and annealed to 0.3 μM of sequencing primer at 80˚C for 2.0 min. The plate was cooled to room

temperature for 5 minutes and placed in the PyroMark™ Q24 instrument with the appropriate

quantity of dNTPs, enzyme and substrate mixtures.

Methylation patterns and SNPs were quantified using the CpG and AQ softwares, respectively

(PyroMark Q24 Software). The peak heights were compared to the expected pyrogram heights.

Pyrosequencing runs for each primer pair and sample type was optimized and included three negative

controls (no template, no sequencing primer, and no template or primer).

60

Chapter 3: CNV Analysis in ID and Autism Probands 3.1 TRAPPC9

3.1.1 TRAPPC9 function in the CNS

Several lines of evidence demonstrate that TRAPPC9 (trafficking particle protein complex 9;

chr8q24.3) is critical for neurodevelopment and brain function. TRAPPC9 has restricted expression

in cortical neurons (Mochida et al., 2009) and encodes a multi-functional protein involved in NF-κΒ

signaling (Hu et al., 2005) and vesicular trafficking (Zong et al., 2011).

In mammals, TRAPPC9 positively modulates downstream responses by NF-κΒ, a ubiquitously

expressed transcription factor composed of homo- or heterodimeric complexes of the Rel protein

family (p50, p52, p65 (RelA), RelB and c-Rel (Blank et al., 1992; Tak & Firestein 2001). In

unstimulated cells, Iκβ proteins bind to NF-κΒ and prevent its translocation to the nucleus. However

following activation by extracellular stimuli (e.g. inflammatory cytokines or brain-specific

activators), serine residues of Iκβ are phosphorylated by the signalsome complex (Traenckner et al.,

1994; 1995) which consists of protein kinases, including the Iκβ proteins Iκβɑ and Iκκβ, as well as a

non-catalytic regulatory subunit NEMO/IKKγ (Brown et al., 1993; Chen et al., 1995; Koulich et al.,

2001). As the main catalytic subunit, phosphorylation by Iκκβ targets Iκβ for ubiquitylation and its

subsequent degradation by the 26S proteasome complex in the cytoplasm (Magnani et al., 2000).

Degradation of Iκβ frees the NF-κΒ transcription factor and reveals the nuclear localization signal on

RelA (p50/p65), allowing NF-kB dimers to translocate rapidly across the nucleus to regulate

expression for genes such as cytokines, adhesion molecules and immunoreceptors in the brain

(O’Neill &Kaltschmidt, 1997). NF-κΒ transcription factors are functionally present in all neuronal

and glial cells (O’Neill & Kaltschmidt, 1997), and have critical roles in physiological and

pathological responses in the CNS, specifically for neural cell function, synaptic plasticity and

cellular immunity (Meffert & Baltimore, 2005; Mattson & Camoandola, 2001). A study by Hu et al.,

61

(2005) revealed that TRAPPC9 enhances NF-κβ activity by increasing the phosphorylation of IKKβ

and NIK, and was aptly given the alternative name, NIK (NFκβ-inducing kinase) and IKKβ binding

protein (NIBP); conversely, knockdown of TRAPPC9 by RNA interference reduced activation of the

NF-κβ pathway. In vitro, TRAPPC9 also increased neurite outgrowth following nerve growth factor-

induced neuronal differentiation, thereby suggesting a role for TRAPPC9 in neurogenesis (Hu et al.,

2005). Corroborating this finding, increased expression of the anti-apoptotic gene, Bcl-Lx, following

TRAPPC9 over expression provides evidence for TRAPPC9 as an essential protein in neuronal

survival (Hu et al., 2005). This is particularly relevant in early postnatal life when the majority of

synapses are trimmed.

TRAPPC9 also mediates vesicular trafficking of COPII vesicles in early secretory and endocytic

pathways as a component of the multi-subunit transport protein particle (TRAPP) (Zong et al., 2011).

The TRAPP complex, with the exception of one subunit, is a highly conserved guanine nucleotide-

exchange factor between yeast and mammals, serving as a tether for vesicle docking at the target

membrane for both endoplasmic reticulum to Golgi and intra-Golgi trafficking (Zong et al., 2012).

Yeast S. cerevisiae possess three different TRAPP isoforms sharing a common six-subunit core:

TRAPPI, TRAPPII and TRAPPIII (Sacher et al., 2001). For mammals, the TRAPPII complex is the

predominant isoform and primarily mediates intra-Golgi trafficking via two identified mechanisms

(Yip et al., 2010; Yamasaki et al., 2009). Within the TRAPPII complex, TRAPPC9 is a homolog to

an essential gene in yeast, Trs120 (Sacher et al., 2001). Co-immunoprecipitation experiments reveal

that TRAPPC9 interacts with TRAPPC2, a key adaptor subunit within the mammalian TRAPPII

complex, and TRAPPC10 (Zong et al., 2012). Moreover, TRAPPC9 interacts with p150Glued, a

subunit of dynactin that is required for the movement of COPII vesicles toward the ER-Golgi

intermediate compartment along microtubules (Zong et al., 2012). Individuals harbouring mutations

at TRAPPC9 illustrate the functional implications of TRAPPC9 in neurodevelopment.

62

3.1.2 TRAPPC9 and ID

To date, five homozygous frameshift or nonsense mutations in TRAPPC9 (MIM 611966) were

identified as a cause of NS-ARID in six apparently unrelated consanguineous families segregating ID

from South Asia, the Middle East and North Africa (Appendix 2). In these families, parents of

affected individuals were carriers for mutations in TRAPPC9 with an exception to a Syrian family

described by Jamra and colleagues (2011) which identified a de novo mutation at this gene locus.

Similarly, individual cases have emerged in a consanguineous family of East Asian descent and in a

non-consanguineous European family, both with no apparent history of ID (Marangi et al., 2012;

Koifman et al., 2010). The homozygous mutations comprise two submicroscopic deletions (Koifman

et al., 2010; Najmabad et al., 2009), a splice site mutation (Kakar et al., 2012), and four nonsense

mutations, three of which are located at the same genomic region (Marangi et al., 2012; Jamra et al.,

2011; Mochida et al., 2009; Mir et al., 2009). Nonsense-mediated mRNA decay and decreased

protein levels were detected in these patients via RT-qPCR and western blotting (Mir et al., 2009;

Mochida et al., 2009; Phillipe et al., 2009), and indicated loss-of-function mutations. Subsequent

functional studies by Zong and colleagues (2012) determined that truncating mutations at TRAPPC9

reduce the binding affinity of TRAPPC9 to other proteins in the TRAPPII complex.

Affected individuals carrying homozygous mutations at TRAPPC9 display a range of clinical

symptoms synonymous with an idiopathic disorder, including speech impairment, developmental

delays of varying severity (including moderate to severe ID), stereotypic behaviours and hand

flapping (Appendix 1). Other non-specific features include truncular obesity, growth retardation,

hypotonia and dysmorphic facial features. Interestingly, most individuals exhibit postnatal

microcephaly or reduced head circumference not concomitant to seizures or epilepsy. Additional

shared features include diminished cerebral white matter volume, hypoplasia of the corpus callosum

and reduced cerebellar volume.

63

3.1.3 Heterozygous CNVs at the TRAPPC9 locus

In additioin to the homozygous recessive mutations, we have recently identified a subset (n=20) of

ID and ASD individuals carrying rare CNVs that are heterozygous, some of which are from

unaffected parents while others are de novo in origin (Figure 3.1). Probands with CNVs overlapping

the TRAPPC9 locus were identified and obtained from several groups and databases, including the

Autism Genome Project (AGP) database (http://davinci.tcag.ca/agp/), a private database which we

have access to via our collaborations with the AGP, the DECIPHER database (GRCh37)

(http://decipher.sanger.ac.uk/), and the International Standards for Cytogenomic Arrays (ISCA)

consortium (https://www.iscaconsortium.org/). In addition, six independent groups contacted us with

reports of probands with TRAPPC9 CNVs. Different microarray platforms, including Roche

NimbleGen 135K array, Illumina Human 1 M array and Agilent 105 K CGH array, screened for

structural variations and estimated the breakpoints for CNVs. The relationship of these unique

deletions and duplications in autistic and ID etiology require investigation, however there is evidence

suggesting the involvement of a conserved imprinting mechanism between mice and humans (genes

summarized in Table 3.1).

3.1.4 Imprinting at the Chr15qD3 domain

Mouse distal chromosome 15 (chr15qD3), the syntenic region to chr8q24.3, is a well-characterized

imprinting cluster, comprised of three paternally imprinted protein-coding genes including Trappc9,

and a maternally imprinted intronic long ncRNA (Figure 3.2).

Our research group examined genomic imprinting at chr15qD3 based on the observation that this

region contains the known paternally expressed gene, Peg13, located within intron 16 of Trappc9

(Smith et al., 2003). Allelic expression of Trappc9 was assessed by our collaborator Dr. Kazuhiko

Nakabayashi in offspring derived from reciprocally crossed C57BL/6 x JF1 mice in the embryonic

64

period (15.5 dpc; days post coitum), at postnatal day (PD) 0, and at 8 and 24 weeks postnatal

(Appendix 3). Exclusive or preferential expression of Trappc9 from the maternal allele (paternally

imprinted) was tissue specific and restricted to development. In the CNS, Trappc9 exhibited

maternal-specific expression in fetal whole brain in addition to the cerebrum, cerebellum, olfactory

bulb, neurons and neural progenitor cells (NPCs); however, Trappc9 showed equal expression from

both alleles in glia, fibroblasts and leukocytes. Allelic imbalance of Trappc9 persisted until 8 weeks

postnatal in mice and subsequently disappeared, at which point the gene became biallelically

expressed in all brain tissues analyzed (Appendix 3). Imprinting of Trappc9 appeared to be restricted

to the CNS, as other somatic tissues such as limb, skin, placenta, spinal cord, muscle and heart

displayed biallelic expression (Appendix 3). Furthermore, RNA-seq in mouse brain found that

paternal imprinting of Trappc9 in the brain depends on the gene isoform (Gregg et al., 2010). Loss of

Trappc9 imprinting in adult life highlights its evolutionary role in mammalian neurodevelopment.

Trappc9 resides between two paternally imprinted transcripts. Lying downstream and proximal to

Trappc9 is a K2p potassium channel, Kcnk9, highly expressed in the brain (Ruf et al., 2007). Eif2c2

(argonaute 2), which lies upstream to Trappc9 and is also expressed from the maternal allele (Gregg

et al., 2010), encodes a protein involved in mRNA processing as part of the RNA-induced silencing

complex (RISC) (O’Carroll et al., 2007; Rand et al., 2004). Moreover, the chromatin-remodeling

factor 1, Chrac1, which is biallelically expressed, is located in-between Trappc9 and Eif2c2 (Kelsey,

2011). A paternally expressed long ncRNA, Peg13, on the negative strand lies in the 16th intron of

Trappc9, is presumed to regulate imprinting of these transcripts (Smith et al., 2003; Davies et al.,

2004) as the CGI associated with Peg13 was identified as a maternal DMR in the germ line and in

the brain. The promoters of neither Kcnk9 nor Eif2c2 appear to regulate their imprinting status (Ruf

et al., 2007). The CGI-promoter region of Kcnk9 is in fact hypomethylated in the germ line for both

65

sperm and oocytes. It is also worth mentioning that the transcripts exhibit restricted expression to

both neonatal and adult mouse brain (Wang et al., 2008; Davies et al., 2004; Smith et al., 2003).

Kcnk9 is highly expressed in neurons and the cerebellum (Smith et al., 2003). Similarly, expression

patterns for Peg13 in mouse brain tissues is primarily neuronal and is highly expressed in

hypothalamic regions, the hippocampus and the cerebral cortex (Davies et al., 2004), which parallels

to the expression patterns of Trappc9 in mouse brain (Mochida et al., 2009).

3.1.5 Imprinting at the Chr8q24.3 domain

The extent to which imprinting is conserved at the human syntenic region (chr8q24.3) requires

further investigation. It is currently known that EIF2C2 is not imprinted in humans, while the

imprinted status of TRAPPC9 needs to be determined. KCNK9 (MIM605874), however, remains

paternally imprinted and a mutation to the active copy (maternal allele) has been associated with

Birk-Barel syndrome (MIM612292), characterized by ID, hyperactivity, hypotonia and facial

dysmorphisms (Barel et al., 2007). Imprinting of KCNK9, in parallel to mice, is not regulated by its

promoter, which is also hypomethylated in the brain as well as in blood (Ruf et al., 2007). The

mechanism governing KCNK9 imprinting is currently unknown.

Intriguingly, intron 17 of TRAPPC9 contains numerous overlapping ESTs on the negative strand,

specifically AX307073 and AK748923, and a large, proximal CGI (Figure 3.1). ChIP-seq analysis

reveals that the CGI contains chromatin signatures indicative of an insulator element, namely

enrichment for trimethylation of H3K4 and CTCF binding. The significance of these regulatory

marks have yet to be determined in the PEG13 context. Despite the lack of sequence homology with

mouse Peg13, these ESTs appear to represent a continuous transcript, and SNPs assessed within this

region demonstrate biased expression from the paternal allele in human fibroblasts (Morcos et al.,

2011), thereby suggesting that this long ncRNA, aptly named PEG13, is the Peg13 equivalent in

66

humans. Taken together, the human PEG13 equivalent requires further investigation, along with its

relevance to the heterozygous CNVs in our ID and ASD (Figure 3.2).

Mouse

gene/Human gene

Gene

type

Classification Function Allelic

expression in

mice

Allelic

expression in

humans

Eif2c2/AGO) P Argonaute RISC

catalytic

component 2

RNA interference: may

interact with DICER

and play a role in short-

interfering-RNA-

mediated gene silencing

Maternal Biallelic

Chrac1/CHRAC1 P Chromatin

accessibility

complex protein

1

Histone fold protein;

interacts with other

histone fold proteins to

bind DNA in a

sequence-independent

manner; these dimmers

combine with larger

enzymatic complexes

for DNA transcription,

replication and

packaging

Biallelic Bialleic

Kcnk9/KCNK9 P* Potassium

channel,

subfamily K+,

member 9

Two-pore potassium

channel

Maternal Maternal

Peg13/PEG13 RNA Long ncRNA Might mediate

imprinting of

neighbouring genes

Paternal Unknown

Trappc9/TRAPPC9 P Trafficking

particle protein

complex 9

Protein involved in NF-

KB signaling and

vesicular trafficking

Maternal Biallelic

Table 3.1 Summary of genes imprinted at human chr8q24.3 and the syntenic mouse region,

chr15qD3. In mice, Trappc9, Kcnk9 and Eif2c2 are paternally-imprinted genes. Conversely in humans,

thhe imprinting status of KCNK9 remains conserved. It is hypothesized in mice and humans that the long

ncRNA, Peg13/PEG13, regulates imprinting. P = protein

67

Figure 3.1 Schematic illustrating the CNV breakpoints spanning the TRAPPC9 locus and relative to neighbouring genes (UCSC/hg19).

Blue = duplications and Red = deletions. The CNV breakpoints in ID and autism probands converge near the putative PEG13 locus, which lies

on the negative strand of the 17th intron of TRAPPC9. The CGI associated with PEG13 shows enrichment for CTCF and trimethylation and

monomethylation of H3K4, which is indicative of an insulator element and active expression respectively. As such, we hypothesize that PEG13

(AK307073 and AK748239) is a novel imprinted gene whose expression may be disrupted by the heterozygous CNVs depending on the parent

of origin in which the CNV resides. The maternally-expressed gene, KCNK9, lies downstream to TRAPPC9. CHRAC1, AGO2 and TRAPPC9

show biallelic expression in humans.

68

Figure 3.2. Schematic representation comparing the mouse Trappc9/Kcnk9/Ago2/Peg13 imprinting cluster with the human syntenic

region (Figure not drawn to scale). Allelic expression is indicated by the transcript colours: Blue = paternally-expressed gene; Red =

maternally expressed gene; Black = biallelically-expressed gene; and green rectangle = CpG island. (A) Mouse imprinting cluster at chr15D.

The ncRNA PEG13 is reciprocally imprinted to the protein-coding genes Kcnk9, Trappc9 and Eif2c2, and lies within intron 16 of Trappc9.

Peg13 is presumed to mediate imprinting. (B) Human syntenic region at chr8q24.3. According to current findings, imprinting at chr8q24.3

only remains conserved for KCNK9, a two-pore pore potassium channel highly expressed in the CNS. A maternally-inherited missense mutation

in KCNK9 has been associated with Birk-Barel Syndrome. The CpG-promoter of KCNK9 is not differentially methylated in germ cells and

brain, suggesting that imprinting of KCNK9 may occur by an alternative mechanism. While surrounding genes are biallelically-expressed and

lack DMRs, a non-coding RNA from the minus strand lies in the 17th intron of TRAPPC9, and may be the Peg13 equivalent in humans. Further

investigation is required to determine if this ncRNA is imprinted and has a regulatory role in KCNK9 imprinting.

CHRAC1

CpG 196

KCNK9

CpG 273

Intron

17

PEG13 equivalent?

CpG210 = DMR?

TRAPPC9

CpG 79

AGO2

CpG 95

Chrac1

CpG 88

Kcnk9

CpG 186

Intron

16

Peg13

CpG44 = DMR

Trappc9

CpG 38

Eif2c2

CpG 136

(A)

(B)

69

3.2 Materials and Methods I completed all work in this chapter except the following: Validation of the CNV and sequencing of

TRAPPC9 for Proband 1 was performed previously by Kaufman et al., (2011).

3.2.1 Sample Collection Blood samples from four affected individuals were provided by Suzanne Lewis, Daniel Doherty,

Edwin Kirk, and Jennifer Roggenbuck. These samples were collected in Vacutainer ACD Solution B

tubes (BD, New Jersey) and Tempus™ Blood RNA tubes (Life Technologies), and shipped to

Toronto. All genetic tests, including microarray analysis, karyotyping, Fragile X testing;

medical/blood tests (i.e. esophagram, IGF-1 and IGF-BP3; sweat chloride, thyroid levels, and Ca2+

and P levels); and behavioural assessments were performed by a genetic counselor and/or an acting

physician. Extraction of nuclei acids was performed as described in Chapter 3, using the high salt

method for DNA isolation and a commercial kit for RNA isolation (RNAeasy; Qiagen).

3.2.2 Whole gene sequencing To confirm that no secondary mutations (i.e. frameshift, missense or nonsense mutations) were

present in our probands, all exons of the TRAPPC9 gene were sequenced. PCR primers (IDT) were

designed by Liana Kaufman (2011) using UCSC genome browser (build GRCh37/hg19) and Primer3

software. Primers flanked a minimum of 70 bp from each exon boundary to ensure complete

amplification of each exon with the exception of exon 2 where two overlapping primer sets were

designed to accommodate the ~1 kb size. PCR reactions comprised 5.0 μl 2G Kapa HotStart (2X); 2

μl H2O; 1.0 μl DNA template (25 ng/μl); and 1.0 μl (10 μM) for each forward and reverse primer.

The PCR followed standard conditions with initiation at 95˚C for 1 min; and 30 cycles of

denaturation at 95˚C for 10 s; annealing at 59˚C for 1.0 s; and elongation at 72˚C for 1.0 s.

Amplification of exon 1 used hot star taq polymerase and performed according to the manufacturer’s

70

protocol (Qiagen). PCR product size was confirmed by gel electrophoresis and visualized using

ethidium bromide or Redsafe™ Nucleic acid staining solution (20, 0000X) (Frogga Bio) and the Bio

Rad Gel DocTM XR System. PCR products were Sanger sequenced (TCAG, Toronto) and sequences

were analyzed by BLAST and FinchTV.

3.2.3 CNV analysis

Validation of the CNVs was performed by RT-qPCR in test subjects using the ∆∆CT method as

previously described in Chapter 3. An internal primer pair was used as an endogenous control for all

runs and leukocyte DNA from healthy individuals was used as positive controls. Each sample was

run in quadruplicate and data was normalized against the internal control. The experimental data was

derived from the ViiA™ Software v1.1. The ∆∆CT values and statistical analysis was calculated

using the GraphPad Prism Software, with the exception of proband 1 which was derived from the

ViiA™ 7 Software.

71

Table 3.2. List of primers pairs used to sequence TRAPPC9 exons. All primer pairs had a calculated

melting temperature of ~60˚C and designed to amplify the entire exon with the exception of exon 2,

which was > 1kb in size. EXON Primer Sequence (F & R) Size (bp)

1 CGCCTTGGCACTGAACTAAC

GAGGAAAGCTTCTGCACCTG

450

2 GGGTAGGCTCTCAGCACTCA

CGGGAGTCATACAGTGTGGA

421

2-2 CCATCACAGACTGCTTCTCG

TTGCTGCAATGAAATTCTGC

421

2-3 CAAAATGAGCGTCCCTGACT

TCTTGTCTGTGGCTCTGTCCA

515

3 GAGAATTGAATAGTGCTTAGTCAGAG

GCACCCAGCCTCTCATTTTA

300

4 GCTAATGAAGGGTCACTGACTTG

TCCTAACTCAAAGGAAAAAGTAACAAA

303

5 CAAGTCAGATTTTTCTCTAGGCTATG

GGACCCACCAATCAATCATC

304

6 CATGCTCCTCTTTAACAAAGAAA

TGCATTAGAGAGCAAGACATGAA

330

7 ATGGGGGAAGTGCTGTTGTA

TTATCATGGCATGCTGAACC

344

8 CAGGTCTTTGTTTTATGAAGAGACA

CAGGGCAGGGATCTAGTCAT

447

9 GTTTGGTGGTTGAGGCATCT

CAAGCCCAGGGTTTACATGA

344

10 AATGAAAGCCAGCCACTTTG

CCGATACACAGCCAACAAAA

307

11 ATCCTGAAACCACGCTAACC

ACATGCATGAACTGGCTCAA

304

12 TGCCAGAGGAATAATTGGAAA

TCGTAAGATGTGTGCCTCAGA

265

13 TGTGCCATAGCCAATGTCTT

AACTGGTTAGGACTGGCATGA

325

14 GAAGAGGAGCCCCGTACTCT

TTACAGGCTCTTTGTGCCATT

321

15 TGGTGATTCTTCTTTGGGAAG

CTGACTTCAACTGAATCCACAAA

374

16 GGGGATTGGTAAGTCCCATA

ACAAAGTGCCCAGGAGATGT

367

17 CAAAGGTTGCAGTGAGCAGA

ACCAAACAATGCACTCAAGC

458

18 GCCTCCTGGTGGATCAAACT

ACGGGATGCATGACAAAAAC

310

19 AAGTGGTGGGAAGTCTTTGG

TGAGGACTTGGGGTGGATTA

300

20 TGAAGTCAGGGCCACCTTAG

GTGGGGCCTTTGAAGTGTTA

517

21 CCCATCTGAGGGTCTCTGTC

TTCCCGTGATGACCTTCAGT

309

22 CACACATCTGTGTCCCCAGT

CACATTTTCCTGCTTCACCA

623

23 CCTATGTCCAGGCAGATTCC

CAGTGAAGGCCTTGCTCATT

348

72

Table 3.3 qPCR assays for CNV validations. Primer pairs were designed to confirm the direction of the

CNV (i.e. deletion or duplication) in the autism and ID probands according to UCSC genome build hg19.

Source (Proband

ID)

Coordinates/CNV type Primer sequence (F/R) Size of amplicon (bp)

Proband 1 chr8:141211615-

141387894

Deletion

AGGCTTAATACCTGGGTCACAA

AGTTCAGGGCTACAAGTGAAGG

125

Proband 3 chr8:141210317-

141298087

Deletion

AGAGCACACCACGATAAATGG

TGGAGGATAAAGCATTCCAAGT

97

Proband 8 chr8:141211815-

141268818

Deletion

CACTAGCAGAAGGAGCATGAAA

GGCTGAGTATGAGGAAACCCTA

118

Proband 20 chr8:140821366-

141268818

Duplication

AGAGCACACCACGATAAATGG

TGGAGGATAAAGCATTCCAAGT

97

Internal control for CNVs (Region on

chromosome 7 not known to contain CNVs in

autism populations)

GAAGCAGGACTCTAAGTCCAGA

TGCTAGAGGAGTGGGACAAGTA

140

73

3.3 Results

3.3.1 TRAPPC9 CNVs confer phenotypic variation

Twenty probands carrying rare heterozygous CNVs overlapping TRAPPC9 exons show varying

severities of developmental delay, ID or ASD, accompanied by a range of other clinical phenotypes

(Table 3.3). All probands were from unrelated families from various geographical regions with no

known history of ID or ASD.

An autistic individual (Proband 1) carrying a 176 kb deletion was identified by a study conducted in

Vancouver (Riendeau, 2009). The CNV was confirmed by fluorescence in situ hybridization (FISH)

(Riendeau, 2009) and by RT-PCR (Kaufman et al., 2011) in both the mother and proband (Figure 3.3

a). Whole gene sequencing of TRAPPC9 in the mother and proband revealed no coding mutations

(Kaufman et al., 2011), indicating that the deletion is the only genetic variant at this gene locus.

Other autistic cases with developmental delay were found in the ISCA Database, as well as by

Sanders et al., (2011) and individual collaborators (Table 3.3). Some deletions/duplications overlap

with TRAPPC9 in addition to genes distal or proximal to TRAPPC9, including EIF2C2, KCNK9 and

CHRAC1 (Probands 4, 6 9 and 10). Mutation testing could not be performed in these probands

because DNA and RNA have not yet been obtained.

We also identified individuals harbouring deletions or duplications overlapping TRAPPC9 exons

without autism, but displaying abnormal behavioural patterns and developmental delay. An

individual (Proband 3) inherited a 129 kb deletion overlapping TRAPPC9 at exons 13-17 from a

healthy, normal father. The individual has normal intelligence and exhibits mild behavioral problems

and stereotypical behaviors. RT-qPCR confirmed the deletion in both the father and the proband and

coding sequences of TRAPPC9 did not contain any additional genetic aberrations (Figure 4.3b). A

girl 9 years at assessment (Proband 20) with ID has no history of epileptic seizures, as identified and

74

evaluated by Dr. Edwin Kirk. She has a paternally inherited duplication overlapping exons 17 to 21

in TRAPPC9; however, the father is unaffected and healthy (Figure 3.3d). The duplication overlaps

with AX748239 and it may affect TRAPPC9 function. Subsequent sequencing of TRAPPC9 in the

proband and both parents also did not reveal any additional mutations. MRI scans of the brain

revealed spaces in the cerebral spinal fluid (CSF) within the posterior cranial fossa (which contains

the brain and cerebellum), in addition to the cisterna magna. She is also microcephalic (head

circumference < 2nd percentile). There were no abnormalities in the white matter or the corpus

callosum. She has a vocabulary of 20-30 words, and although she can walk unassisted, she is ataxic.

The subject has occasional hand-flapping and limited fine-motor skills – she cannot hold a pen and

nor is she toilet trained. She also has dysmorphic facial features (Appendix 4). Methylation testing

for Angelman’s syndrome was negative. We also confirmed the patient to be negative for MECP2

mutation.

A de novo 57 kb deletion was identified in an individual with ID (Proband 8); however, the allele on

which the deletion is located is currently unknown (Figure 3.3c). This individual is short in stature

and has a distinctive facial appearance. The individual is also noted to have borderline

hypothyroidism. Her hair is described as sparse and brittle. Fragile X testing and screening for

Noonan syndrome were both negative. We also confirmed the patient to be negative for MECP2

mutation. Furthermore, TRAPPC9 sequencing in this proband did not identify other coding

mutations.

3.3.2 Validation of CNVs in human leukocytes

As described above, whole genome sequencing of TRAPPC9 did not reveal any coding mutations in

our probands and RT-qPCR confirmed the heterozygous nature of the CNVs. The fact that the CNVs

are heterozygous and no other coding mutation were found indicates that these individuals are not

75

compound heterozygotes for TRAPPC9 mutations, which is clearly a recessive disease gene.

Therefore, the disease mechanism, if related to the CNVs, must be independent of TRAPPC9 if

imprinting of TRAPPC9 is ruled out.

3.3.3 Relevance for characterizing PEG13 in humans

The CNV deletion in Proband 1 was previously shown to have no effect on TRAPPC9 expression in

blood as compared to the healthy parents, including the mother who carries the deletion (Kaufman et

al., 2011). Moreover, our collaborator (Dr. Miho Ishida, UCL, UK) recently showed that TRAPPC9

is not imprinted in humans using the SNPs rs3735801 and rs3735802 within coding regions of

TRAPPC9. Of the 58 fetal DNA samples that were sequenced, 22 cases were heterozygous for both

SNPs, 17 of which were also heterozygous for rs3735803 and 8 additional cases were heterozygous

for SNP rs3735803. As such, five heterozygous cases with brain and other tissues were available for

assessment of TRAPPC9 allelic expression. In all tissues tested, including brain, TRAPPC9 exhibited

biallelic expression, indicating that TRAPPC9 is not imprinted in human foetus.

Further evidence for the involvement of disrupted imprinting in disorder etiology is the analysis of a

maternally inherited truncating mutation at TRAPPC9 in unaffected carriers from a consanguineous

family segregating ID (Mir et al., 2009). Kaufman et al. (2011) found that unaffected carriers with

the maternally inherited mutations were phenotypically normal, while paternally inherited mutations

displayed delayed cognitive development and poor performance at school. All family members who

were heterozygous for the mutations coming from the maternal allele also showed very high

performance at school. The cognitive performance of individuals carrying the mutation on the

paternal allele indicates parent-of-origin effects (POE) at chr8q4.3. However, it should be noted that

evidence for POE is weak because of the small sample size. Also, there may be a culturally related

gender bias, whereby greater efforts would be made to educate males in this family. Taken together,

the absence of additional coding mutations at TRAPPC9 supports the hypothesis that the proband

76

CNVs may exert their effects via an alternative mechanism, such as imprinting of another proximal

gene.

77

Table 3.3 CNVs identified in TRAPPC9 and summary of clinical phenotypes. Different CHIP platforms identified the estimated breakpoints of

CNVs overlapping TRAPPC9 and, in some cases, neighboring genes including KCNK9 and EIF2C2. N/O = no overlap

PROBAND

ID

DATA SOURCE MICROARRAY

USED FOR

ANALYSIS

CNV

SIZE

AND

TYPE

CNV

COORDINATES

(HG19)

AFFECTED

EXONS

OTHER GENES

OVERLAPPED

INHERITA

NCE

PHENOTY

PE

SEQUE

NCING

OF

TRAPP

C9

CNV

VALID

ATION

1 Suzanne Lewis Lab,

UBC

Agilent 105K CGH

array

176 kb

deletion

chr8:141211615-

141387894

8-17 None Maternal Autism Yes Yes

2 Marvin Natowicz,

Cleveland Clinic

Unknown 129 kb

deletion

chr8:141142433-

141318712

18-19 None Unknown/

De Novo

Profound

global

development

al delays

and epilepsy

No No

3 Dan Doherty,

University of

Washington

Roche NimbleGen

135K array

(SignatureChipOS

version 2, hg18

assembly)

88 kb

deletion

chr8:141210317-

141298087

13-17 None Paternal Behavioral

problems

and

stereotypical

behaviors

Yes Yes

4 ISCA Database Unknown 660.5 kb

deletion

chr8:141070844-

141731423

1-17 CHRAC1,

EIF2C2, PTK2,

AK130220,

AX748239

Maternal ASD/DD No No

5 ISCA Database Unknown 106 kb

deletion

chr8:141297551-

141403714

8-13 None Unknown ASD/DD No No

6 Bridget Fernandez,

Memorial

Illumina Human 1M

array

1 Mb

duplicatio

n

chr8:140016461-

141168970

18-23 KCNK9, C8orf17,

AX748239

Maternal Autism and

speech

delays

Yes Yes

7 TCAG Autism

Chromosome

Rearrangement

Database

Unknown 281 kb

duplicatio

n

chr8:140789805-

141070993

18-21 None Maternal Autism/PD

D

No No

8 Jennifer

Roggenbuck

Unknown 57 kb

deletion

chr8:141211815-

141268818

16-17 None De Novo ID/short

stature,

facies

Yes Yes

Proband

9=Proband

6?

DECIPHER

#250624

Unknown 1.15Mb chr8:140016461-

141168970

18-23 KCNK9, C8orf17,

AX748239

Unknown No No

78

10 ISCA Database

nssv576297

Unknown 1,165 Dup chr8:140876636-

142041711

1-19 CHRAC1,

EIF2C2, PTK2,

AX748239

Unknown Autism; ID No No

11 ISCA Database

nssv582488

Unknown 246Kb del chr8:140980052-

141226564

15-17 AX748239 Unknown Developmen

tal delay

No No

12 ISCA Database

nssv579107

Unknown 5.7Mb

dup

chr8:140459470-

146280161

1-21 KCNK9,

CHRAC1,

EIF2C2, PTK2,

AX748239

Many genes

Unknown Not reported No No

13 Sanders et al, 2011

#11194

Unknown 1.291Kb

Del

chr8:141120454-

141121745

N/O AK130220 Unknown mother No No

14 Sanders et al, 2011

#11290 Unknown 21.143Kb

Del

chr8:141203709-

141224852

N/O AK130220 Unknown father No No

15 Sanders et al, 2011

#11290

Unknown 21.143Kb chr8:141203709-

141224852

N/O AK130220 Unknown Autism

proband

No No

16 Sanders et al, 2011

#11644

Unknown 2.881Kb

dup

chr8:141117789-

141120670

N/O none Unknown mother No No

17 Sanders et al, 2011

#11825

Unknown 216bp

Dup

chr8:141120454-

141120670

N/O none Unknown Autism

proband

No No

18 Sanders et al, 2011

#11846

Unknown 216bp

Dup

chr8:141120454-

141120670

N/O none Unknown Autism

proband

No No

19 Sanders et al, 2011

#12224

Unknown 216bp

Dup

chr8:141120454-

141120670

N/O none Unknown father No No

20 Edwin Kirk CGH array ~500Kb

Dup

chr8:140821366-

141268818

17-21 AX748239 Paternal intellectual

disability; microcephali

c (HC <2nd

centile)

Yes Yes

79

Figure 3.3 Validating heterozygous CNVs in autism and ID probands by RT-qPCR. Gene copy number within the TRAPPC9 locus was

measured in probands relative to three normal, healthy controls. (A) Confirmation of a maternally-inherited deletion in Proband 1 (B)

Confirmation of a paternally-inherited deletion from an unaffected father in Proband 3. (C) Confirmation of a de novo deletion in Proband 8. (D)

Confirmation of a parentally inherited duplication in Proband 20.

Co

ntr

ol 1

Co

ntr

ol 2

Co

ntr

ol 3

Pro

ban

d 8

0 .0

0 .5

1 .0

1 .5

Re

lati

ve

Qu

an

tity

(R

Q)

Co

ntr

ol 1

Co

ntr

ol 2

Co

ntr

ol 3

Pro

ban

d 2

0

0 .0

0 .5

1 .0

1 .5

2 .0

Re

lati

ve

Qu

an

tity

(R

Q)

Co

ntr

ol 1

Co

ntr

ol 2

Co

ntr

ol 3

Fath

er

Pro

ban

d 3

0 .0

0 .5

1 .0

1 .5

Re

lati

ve

Qu

an

tity

(R

Q)

C)

A) B) A)

C) D)

80

Chapter 4: Expression analysis of PEG13

All work in this chapter was completed by me except the following: Acquisition and processing of

human fetal tissues and downstream applications, specifically nucleic acid extraction, RT-PCR and

genotyping was conducted by Dr. Miho Ishida from University College London (UK).

Current data supports the evolutionary conservation of mouse Peg13 and its exclusive expression

from the paternal allele in humans. Given that the deletions/duplications of the ID and autism

probands lie proximal to or overlap with the putative PEG13 locus and no other mutations were

found at TRAPPC9, indicates that the heterozygous CNVs in our probands may affect the expression

of PEG13 in an allele-specific manner. Based on the hypothesis that PEG13 is maternally imprinted

in humans, deletions/duplications proximal to the PEG13 locus on the paternal allele may affect

PEG13 expression and have downstream effects on KCNK9 expression. In contrast, disruptions on

the maternal allele should have no effect on PEG13 function or expression, or on KCNK9 expression.

Allelic expression of PEG13 (AK307073; chr8:141105709-141107269; hg19) (Figure 4.1) was

assessed using human fibroblasts, whole blood and fetal tissues. DNA was genotyped at SNP loci

within the transcripts of interest to identify heterozygous individuals and the relative abundance of

each allele was quantified by pyrosequencing. Expression profiles of PEG13 and KCNK9 were also

studied in various somatic and germ line tissues to determine if these genes share similar patterns of

expression. Informative SNPs of cDNA sequence suggests that PEG13 is preferentially expressed

from the paternal allele in human fetal brain and kidneys, in addition to fibroblasts, but not in skin,

placenta, and whole blood. Moreover, both PEG13 and KCNK9 exhibit high expression in human

cerebellum and whole brain.

81

4.1 Materials and Methods: Allelic expression of PEG13

4.1.1 Human fetal tissues

Sample Collection

Human fetal tissues, brain, skin, placenta and kidney, were obtained from the Moore Tissue bank at

the UCL Institute of Child Health (UK). The Hammersmith, Queen Charlotte’s and Chelsea and

Acton Hospital Research Ethics Committee approved the samples (Project Registration 2001 /6029

and 2001/6028).

RNA Isolation and cDNA Synthesis

Tissue samples were homogenized in 1ml TRIzol (Invitrogen) per 100 mg of tissue. TRIzol®

Reagent (Invitrogen) extraction of RNA was performed on tissues according to the manufacturer’s

instructions. RNA samples were treated with TURBO™ DNase (Ambion®) for residual DNA. First

strand cDNA synthesis was performed using 1μg of RNA volume with Promega M-MLV RT in a

25μl reaction according to manufacturer’s protocol. Reverse Transcriptase was omitted in a separate

reaction for each sample to screen for DNA contamination. The cDNA samples were shipped to

Toronto on dry ice and stored at -20˚C.

Tissue

Type

# of

heterozygotes

# of DNA/cDNA samples

available for genotyping

# of cDNA samples

available for

pyrosequencing

Brain 10 10 8

Kidney 1 1 2

Liver 1 1 0

Placenta 3 3 2

Skin 1 1 2

Blood 10 10 0

Total 26 26 14

Table 4.1. List of human fetal tissues samples heterozygous for SNP rs4289794. A

total of 10 individuals had brain samples in addition to other somatic tissues that were

informative for rs4289794. Maternal DNA was available for four individuals. Fourteen

samples, including eight brain samples, were available for pyrosequencing.

82

PCR amplification and genotyping

PCR was performed using BIOTAQ DNA polymerase (Bioline). Cycling conditions began with an

incubation at 94°C for 5 min, followed by 35 cycles of 94°C for 30 sec, 58°C for 30 sec and 72°C for

30 sec, and final extension at 72°C for 2 min. DNA and cDNA was amplified by primer sequences in

5’ to 3’ direction for SNP genotyping (Table 4.1). Prior to sequencing, PCR products were purified

with microCLEAN (Microzone) following manufacture’s instruction. BigDye® Terminator v1.1

Cycle Sequencing Kit (Life Technologies) carried out the sequencing reaction. Allelic expression

was analyzed by comparing the peak heights of the parental alleles in the electropherograms.

4.1.2 Human Fibroblasts and leukocytes

RNA Isolation and cDNA Synthesis

Total RNA was extracted from human fibroblasts and leukocytes using the NucleoSpin® RNA kit

(Machery-Nagel) following the manufacturer’s guidelines. First-strand cDNA was synthesized from

1μg of RNA using Superscript III™ Reverse Transcriptase. Reverse Transcriptase was omitted in a

parallel reaction for each sample to ensure the absence of DNA.

Genotyping: Identification of SNPs

PCR primers (IDT) are listed in Table 4.2. PCR was carried out using KAPA2G™ Fast HotStart

ReadyMx (2X) (Kapa Biosystems) in a 10μl reaction volume containing 1.0μl (30 ng/μl) template.

Q-Solution (Qiagen) in place of H2O facilitated the amplification of GC-rich sequences. Each PCR

reaction followed an initial denaturation at 95˚C and 35 cycles of 95˚C for 10 sec, 60˚C for 10 sec

and 72˚C for 1 sec. PCR products were checked for DNA contamination and integrity by 2% gel

electrophoresis and Sanger sequenced by TCAG (Toronto).

83

SNPs Genotyped PCR Primers (annealing temperature) Amplicon Size

rs4289794

GCTTGGTTTGTGCATTCCTT

TCATTTTGGGTCTGCTTTCC (58˚C)

203

rs4289794 &

rs2270409

TTCTGCAGTGGCTCCCTA

CCAGACAGGACCCAAAGAAA (61˚C)

794

rs3802217

CGTCTGGGTCCGTAGTGTG

GAGTGTTCTCCGCTGTGGAT (61˚C)

677

rs4455807 CCTTCCACAGGGTAGGAACA

AGATTGCGGCACTCGTTC (61˚C)

323

Table 4.2. PCR assays for SNP loci at the putative PEG13 locus. This table outlines the primer sets

used to sequence SNPs within the PEG13 region and includes of the amplicon size and annealing

temperatures used for each PCR program.

4.1.3 Pyrosequencing validation

PCR primer pairs flanked the SNP loci within PEG13 and featured 5’ biotin modification of the

reverse primer. Table 4.3 lists all the oligonucleotides used, including sequencing primers. To

optimize PCR conditions, cDNA was diluted and aliquoted into tubes for 10 μl reaction as previously

described in Chapter 2. PCR for pyrosequencing was carried out in 25 μl reaction volumes

comprising 2.0 μl of cDNA template, 12.5 μl of KAPA 2G and 1.0μl (10 μM) of each forward and

reverse primer. Amplification conditions were as described in Chapter 2. PCR products were

pyrosequenced using the PyroMark™ Q24 System (Qiagen) according to the manufacturer’s

instructions and results were analyzed using the AQ software included with the instrument (Chapter

3). Pyrosequencing assays for human fetal tissues were performed in triplicate and we calculated the

mean and S.D. for each sample.

4.2 Materials and methods: Tissue expression of PEG13

The functional role of PEG13 in humans is currently unknown. An important aspect in characterizing

PEG13 is a description of its tissue expression profile. As previously mentioned, intronic long

ncRNAs with regulatory functions tend to share similar expression patterns as their protein-coding

counterparts. Therefore, the comparative CT method for qPCR was used to show the relative

expression of PEG13 and KCNK9 across various human tissues. Here, we show that concomitant to

KCNK9, PEG13 expression is restricted to the brain and cerebellum.

84

4.2.1 Sample collection

Human RNA was obtained from the Human Total RNA Master Panel II (BD Biosciences Clontech).

Fibroblast and leukocyte RNA were from normal, healthy individuals.

4.2.2 cDNA synthesis

First-strand cDNA was synthesized according to standard procedures from 1.0 μg RNA using

Superscript Reverse® Transcriptase III (Life Technologies) (Chapter 3). Prior to use, RNA was

diluted 1:1 with nuclease-free H2O (Ultrapure™ Invitrogen™). Reverse Transcriptase was omitted in

parallel reactions for each sample to ensure the absence of DNA contamination.

4.2.3 RT-qPCR

All oligonucleotides were purchased from IDT and designed using the primer express software

(Applied Biosystems Inc., Foster city, CA). With the exception of PEG13, primers spanned intron

junctions between exons to limit amplification to the coding sequence (CDS), as well as to prevent

amplification of genomic DNA. All cDNA samples and primers were optimized by standard PCR

and visualized under UV light following gel electrophoresis. Because the ∆∆CT method assumes

equal amplification efficiency for all samples, a serial dilution for each primer set was created. qPCR

was carried out using SYBR green reagents and performed in quadruplicate in a 384-well plate. A

NTC was included for each primer set. Cerebellum was used as a reference gene and Human Β-Actin

was used as an internal control. Data was normalized against B-actin prior to statistical analysis using

GraphPad Prism Software and values were represented as a fold-change in expression compared to

cerebellar tissue. An unpaired two-tailed Student’s t-test was used to assess differences in gene

expression across tissue samples.

85

AK307073

CpG 210

rs2270409 rs4289794 rs3802217 rs4455807

AX748239

Intron 17

Cen Tel

SNP and region Amplification Primers (annealing

temperature ˚C)

Amplicon

Size

Sequencing Primer Sequence

length

Nucleotide

dispensation

rs4289794

chr8:141,106,145-

141,106,258

Forward 1

GGCGCACATTCCACCTAGT

Reverse

(B)-GGCTTGGTTTGTGCATTCCT

(62˚C)

114 Primer 1

GGGACTCATAAATATGAC

T

Primer 2

GGACTCATAAATATGACT

G

36

27

11

4

rs2270409

chr8:141,105,656-

141,105,838

Forward

CCAGACAGGACCCAAAGAAA

Reverse

(B) –

CCTCTCTCCCAAGACCCTCT

(52˚C)

183 Primer 1

GAGGGGGACTCATAAAT

AT

16 8

rs3802217

chr8:141,107,907-

141,108,043

Forward

CCCAGGGTCATGACGCCTA

Reverse

(B)-

GGTGGGCATCCTTATCCTTCA

(58˚C)

139 Primer 1

CAGTCCTACCGGGAAAG

23 6

Table 4.3 Oigonucleotides used for quantifying allelic expression of SNPs by pyrosequencing. This

table outlines the PCR primer sets used and the annealing temperatures for the PCR program, in addition

to the sequencing primer used for pyrosequencing. *(B) denotes 5’ nucleotide biotin modification

Figure 4.1 Schematic of the PEG13 locus within intron 17 of TRAPPC9 (not drawn to scale). The blue

rectangular boxes represent the PEG13 transcript, which encompasses AK307073 (chr8:141,105,709-

141,107,269) and AX748239. (chr8:141,111,257-141,113,775). The SNPS rs2270409, rs4289794,

rs3802217, rs4455807 and rs7828256 were used to assess allelic expression of PEG13. The green

rectangular box is the CpG island embedded within PEG13 and the black rectangular boxes show the

centromere (Cen) and telomere (Tel).

rs7828256

86

4.3 Results

4.3.1 PEG13 is a novel paternally expressed gene in human fetal brain

Fifty-five DNA samples from human fetal tissues were genotyped for SNP rs4289794 to identify

heterozygous individuals (Figure 4.1). Among the twenty-nine samples that were informative, ten

had brain samples available in addition to other somatic tissues for RNA extraction (Appendix 5.1).

PEG13 showed predominant or preferential expression from one allele in seven brain samples, and

had biased expression in fetal kidney. Three brain samples, however, expressed both gene copies

although this may be due to the heterogeneous nature of the brain tissues such that imprinting of

PEG13 may be restricted to a specific brain region such as cerebellum. Biallelic expression was

observed in all other tissues tested i.e. placenta and skin (Table 4.4; Appendix 4). The parental

genotype was available for one case and confirmed that PEG13 is paternally-expressed, as the

mother was homozygous (C/C, homozygous) for the non-expressed allele (Figure 4.2). Relative

allelic expression was quantified for the seven brain samples by pyrosequencing and is presented in

Figure 4.3 as a ratio of allelic expression: 0:100 represents imprinted expression (silencing of one

allele) and 50:50 represents biallelic expression. Since these exact ratios are not always represented

biologically, transcripts were considered to be preferentially or exclusively expressed (imprinted) if

the difference in allelic bias between the alleles was > 60% for one allele, whereas expression

between 40-60% for an allele indicated biallelic expression. Out of the seven heterozygous samples,

only six samples confirmed biased expression in fetal brain (including the individual with the

maternal genotype) (Figure 4.3; Appendix 4.2). The discrepancy between the Sanger sequencing data

and pyrosequencing data may be attributed to bias in the PCR reaction or sequence heterogeneity.

87

Figure 4.2 Electropherograms showing the imprinting status of PEG13. SNP rs4289794 was used to

assess the parental origin of PEG13 expression in human fetal brain. A) Genomic DNA sequence of case

1 (C/T heterozygous). B) Brain cDNA sequence of case 1 (Monoallelic, T expression). C) DNA sequence

of case 1’s mother, who is homozygous for the non-expressed allele (C/C homozygous).

Case Tissue Active allele Active allele

expression (%)

Maternal

genotype

GA (weeks)

1 Brain T 87% C/C 18

Kidney Pref T 76%

Liver C/T N/A

Skin C/T 58%

Placenta C/T N/A

2 Brain C 77% C/T 14

Skin C/T 54%

3 Brain C 45% C/T 10

4 Brain C 62% C/T 10

5 Brain T 89% NA 16

Placenta C/T N/A

6 Brain T 90% NA 10

Placenta C/T 67%

7 Brain T 90% NA 11

Placenta C/T 53%

8 Brain C/T N/A NA 14

9 Brain C/T 57% NA 15

10 Brain C/T 67% NA 10

Table 4.4 Number of informative heterozygous samples for SNP rs4289794 at PEG13. Seven out of

ten brain samples showed predominant expression from an allele and was confirmed in six brain samples

by pyrosequencing. The paternal origin of PEG13 expression was confirmed by case 1’s maternal

genotype who was homozygous (C/C) for the non-expressed allele. For sample 3 (red), the

pyrosequencing assay did not validate the Sanger Sequencing data. N/A = sample cDNA was not

available; Pref = preferential expression from one allele (both alleles were expressed; however the peak

height for one allele was greater than the other allele in the electropherogram).

88

4.3.2 Allelic expression of PEG13 is conserved in human fibroblasts

Fibroblasts were analyzed in parallel to fetal tissues to confirm imprinting of PEG13 in humans.

Allelic expression was quantified by pyrosequencing in ten unrelated fibroblast cell lines

heterozygous for the SNPs rs2270409; rs4289794; and rs3802217 (refer to Table 3.2 for list and SNP

array data). Based on the assumption that the SNPs tested belong to the same non-coding transcript,

it would be expected that all SNPs have preferential expression. All individuals, even those that did

not exhibit preferential expression from the data provided by Dr. Tomi Pastinen, had predominant to

exclusive allelic expression at all SNPs (Figure 4.4; Appendix 4). The discrepancy between the

findings may be due to experimental error, either with the inaccuracy of reads from the Illumina

platform or PCR amplification bias for the pyrosequencing reaction. The results, however, are

consistent with predictions for a novel imprinted locus. Indeed, a distinguishing feature for

imprinted genes is concordant allelic or biased expression at neighbouring SNPs within the region of

interest (DeVeale et al., 2012). Furthermore, the degree of allele-specific expression (ASE) appeared

to occur independent of age (5 months to 54 years), and no correlation was observed between age

and the degree of biased expression. To determine the parental transmission of the expressed allele,

four fibroblast trios (fibroblast RNA from proband and leukocyte DNA from parents) were assessed

for the abovementioned SNPs, in addition to rs4455807. The probands were screened for and did not

have any developmental or neuropsychiatric disorders. Among the probands who were informative,

manual inspection of electropherogram peaks showed preferential or exclusive expression at each

SNP with the exception of two probands (Proband 1 and 3) at rs2270409. The parental genotypes that

were informative were available (Figures 4.5, 5.7, 4.9 and 4.11) and validated the paternal expression

of PEG13 where the maternal genotype was homozygous for the non-expressed allele. Tables 4.5-4.8

summarize the relative expression at each SNP. Pyrosequencing data was in agreement with the

Sanger sequencing data (Figures 4.6, 4.8, and 4.10). Therefore, we confirmed that PEG13 is

preferentially-expressed from the paternal allele in humans.

89

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Placenta

Kidney

Cerebellum

Whole Brain

Fetal Brain

10 Brain

9 Brain

7 Placenta

7 Brain

6 Placenta

6 Brain

5 Brain

4 Placenta

4 Brain

3 Brain

2 Skin

2 Brain

1 Skin

1 Kidney

1 Brain

PEG13 rs4289794 % expression from each allele

Sam

ple

tis

sue

T allele

C allele

Figure 4.3 Relative expression of PEG13 alleles in human fetal tissues. The coloured bars represent the proportion of expression (%) at PEG13 rs4189794. PEG13 shows biased expression in human fetal brain and kidney, and biallelic expression in skin and placenta. Pooled commercial RNA was

run in parallel, indicating that PEG13 shows biased expression in human cerebellum and kidney. C = Cytosine; T = Thymine

90

0% 20% 40% 60% 80% 100%

GM00042

GM01096

GM01376

GM01391

GM01653

GM02175

GM02455

GM03073

GM09256

GM00519

PEG13 rs38022177 % expression from each allele

Sam

ple

G

A

0% 20% 40% 60% 80% 100%

GM00042

GM01096

GM01376

GM01391

GM01653

GM02175

GM02455

GM03073

GM09256

GM00519

PEG13 rs4289794 % expression from each allele

Sam

ple

C

T

Figure 4.4 Relative allelic expression of PEG13 in

human fibroblasts. The coloured bars represent the

ratio of allelic expression (%) at PEG13 rs4189794

and rs3802217. PEG13 shows preferential

expression in most individuals, however the parent-

of-origin could not be determined as parental DNA

was not available. A = adenine; C = cytosine; G =

guanine; T = thymine

91

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 Leukocyte

PEG

13

rs2

27

04

09

%

exp

ress

ion

fro

m e

ach

alle

le

Fibroblast probands (RT-PCR)

A

G

a) Fibroblast gDNA b) Fibroblast cDNA c) Maternal DNA d) Paternal DNA

Figure 4.5 Electropherograms showing the allelic bias of PEG13 at rs2270409. SNP rs2270409 was

used to assess the parental origin of PEG13 allelic expression in fibroblast trios. A) Genomic DNA

sequence of proband 2 (G/A heterozygous). B) Fibroblast cDNA sequence of proband 2 (preferential

expression from the A allele). C) DNA sequence of proband 2’s mother, who is homozygous for the non-

expressed allele (G/G). D) Genomic DNA sequence of proband 2’s father who carries the preferentially

expressed allele (G/A, heterozygous)

Family Active allele Active allele

expression (%)

Maternal

Genotype

Proband

Genotype

Age (years)

1 G 99% G/G G/A 21

2 Pref. A 60% G/G G/A 20

3 G 77% G/A G/A 7

4 G/A 60% G/A G/A 18

Table 4.5 Number of informative heterozygous samples for SNP rs2270409 at PEG13. Two out of

four probands showed predominant expression from an allele, and was confirmed by pyrosequencing. The

paternal origin of PEG13 expression was confirmed by proband 2’s maternal genotype, who was

homozygous (G/G) for the non-expressed allele. Pref.A = preferential expression from adenine (the

electropherogram showed expression for both alleles; however A had a higher peak).

Figure 4.6 Relative expressions of PEG13 alleles in fibroblasts at rs2279409. The coloured bars

represent the proportion of expression (%) at PEG13 rs2270409. PEG13 shows preferential expression in

two probands. DNA sample from a control leukocyte that was heterozygous for the SNP was used as a

positive control.

* *

*

*

92

a) Fibroblast gDNA b) Fibroblast cDNA c) Maternal DNA d) Paternal DNA

Figure 4.7 Electropherograms showing the allelic bias of PEG13 at rs4289794. SNP rs4289794 was

used to assess the parental origin of PEG13 allelic expression in fibroblast trios. A) Genomic DNA

sequence of proband 1 (C/T heterozygous). B) Fibroblast cDNA sequence of proband 1 (Preferential A

expression). C) DNA sequence of proband 1’s mother, who is homozygous for the non-expressed allele

(C/C). D) Genomic DNA sequence of case 2 father who carries the expressed alleles (C/T, heterozygous).

Family Active allele Active allele

expression (%)

Maternal

Genotype

Proband

Genotype

Age (years)

1 T 86% C/C C/T 21

2 C 81% C//T C/T 20

3 C 71% C/T C/T 7

4 C 94% C/T C/C 18

Table 4.6 Number of informative heterozygous samples for SNP rs4289794 at PEG13. All four

probands showed predominant to exclusive allelic expression and was confirmed by pyrosequencing. The

paternal origin of PEG13 expression was confirmed by case 1’s maternal genotype, who was

homozygous (C/C) for the non-expressed allele.

Figure 4.8 Relative expressions of PEG13 alleles in fibroblasts. The coloured bars represent the

proportion of expression (%) at PEG13 rs4289794. PEG13 shows preferential expression in all probands.

DNA sample from a control leukocyte that was heterozygous for the SNP was used as a positive control.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 Leukocyte

PEG

13

rs4

28

97

94

%

exp

ress

ion

fro

m e

ach

alle

le

Fibroblast probands (RT-PCR)

C

T

* *

* *

93

a) Fibroblast gDNA b) Fibroblast cDNA c) Maternal DNA d) Paternal DNA

Figure 4.9 Electropherograms showing the allelic bias of PEG13 at rs380221. SNP rs380221 was

used to assess the parental origin of PEG13 allelic expression in fibroblast trios. A) Genomic DNA

sequence of proband 2 (C/T heterozygous). B) Fibroblast cDNA sequence of proband 1 (Preferential A

expression). C) DNA sequence of proband 2’s mother, who is homozygous for the non-expressed allele

(C/C). D) Genomic DNA sequence of case 2 father who carries the expressed alleles (C/T, heterozygous).

Family Active allele Active allele

expression (%)

Maternal

Genotype

Proband

Genotype

Age (years)

1 C 99% C/C C/T 21

2 Pref. C 81% C/T C/T 20

3 C 84% C/T C/T 7

4 C 99% C/C C.T 18

Table 4.7 Number of informative heterozygous samples for SNP rs380221 at PEG13. All probands

showed predominant or exclusive allelic expression and was confirmed by pyrosequencing. The paternal

origin of PEG13 could not be confirmed in any of the fibroblasts Pref .C = the electropherogram showed

expression for both alleles, however C had a higher peak.

Figure 4.10 Relative expressions of PEG13 alleles in fibroblasts. The coloured bars represent the

proportion of expression (%) at PEG13 rs380221. PEG13 SHOWS exclusive expression in all probands.

DNA sample from a control leukocyte who was heterozygous for the SNP was used as a positive control.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1 2 3 4 Leukocyte

PEG

13

rs3

38

02

21

7

% e

xpre

ssio

n f

rom

eac

h a

llele

Fibroblast probands

C

T

* * *

*

94

a) Fibroblast gDNA b) Fibroblast cDNA c) Maternal DNA d) Paternal DNA

Figure 4.11 Electropherograms showing the allelic bias of PEG13 at rs4455807. SNP rs4455807 was

used to assess the parental origin of PEG13 allelic expression in fibroblast trios. A) Genomic DNA

sequence of proband 3 (C/T heterozygous). B) Fibroblast cDNA sequence of proband 3 (Preferential A

expression). C) DNA sequence of proband 3’s mother, who is homozygous for the non-expressed allele

(G/G). D) Genomic DNA sequence of case 2 father who carries the expressed alleles (G/T, heterozygous).

Family Active allele Active allele

expression (%)

Maternal

Genotype

Proband

Genotype

Age (years)

1 G N/A G/G G/G 21

2 Pref. T N/A G/T G/T 20

3 T N/A G/G G/T 7

4 G N/A G/G G/G 18

Table 4.8 Number of informative heterozygous samples for SNP rs4455807 at PEG13. All probands

showed predominant expression from an allele; however this could not be confirmed by pyrosequencing.

The paternal origin of PEG13 expression was confirmed by proband 3’s maternal genotype, who was

homozygous (G/G) for the non-expressed allele Pref. T = the electropherogram showed expression for

both alleles, however C had a higher peak; N/A = not available.

* *

*

*

95

4.3.3 PEG13 is highly expressed in human brain PEG13 and KCNK9 displayed similar expression profiles in both germ and somatic tissues (Figure

5.14 a and b). Both genes were prominently expressed in the brain with highest expression in the

cerebellum p<0.5), indicating that PEG13 expression and its function may also be region-specific in

the CNS. PEG13 exhibited minimal expression in skeletal muscle, kidney and placenta, while

KCNK9 expression was non-existent in other tissues except for kidney and adrenal gland. The

absence of PEG13 and KCNK9 in leukocytes supports earlier findings that PEG13 is biallelically-

expressed in this cell. Consequently, we could not address the hypothesis that CNVs in the ID and

Autism probands may have downstream effects on PEG13 and KCNK9 expression.

4.3.4 PEG13 in Autism and ID Probands

We postulated that CNVs overlapping TRAPPC9 and adjacent to the putative PEG13 locus may alter

the expression of KCNK9 and PEG13 according to the parent of origin. RNA and DNA from

leukocytes extracted from whole blood were available for Proband 3 (including father) and Proband

8, while only RNA was available for Proband 1. However, given that leukocytes do not express

PEG13 or KCNK9 (Figure 5.14b), RT-qPCR could not accurately measure gene expression in our

probands. Instead, DNA was genotyped for informative SNPs within the PEG13 locus to determine if

PEG13 is imprinted in human leukocytes. Proband 8 and the father of Proband 3 were heterozygous

for PEG13 rs7828256, while Proband 3 did not have any informative SNPs (Figure 4.15). The cDNA

of Proband 1 was also genotyped and showed expression for both gene copies at rs4289793,

indicating that PEG13 is biallelically expressed in human blood. Similarly, both alleles were

expressed in the cDNA sequences in Proband 8 and the father of Proband 3. Although we could not

elucidate any changes in PEG13 expression, based on the fact that PEG13 imprinting and expression

is restricted to the brain, suggests that the CNVs may manifest specifically in the brain, and the

genetic aberrations observed in our probands are not detectable with our current materials.

96

Figure 4.12 Relative expression of PEG13 and KCNK9 in human tissues and cells by RT-qPCR. A)

KCNK9 expression is restricted to the brain, particularly the cerebellum. B) PEG13 has a similar expression

pattern as KCNK9, but also shows minimal expression in other somatic tissues, including skeletal muscle,

kidney, adrenal gland and placenta. Both genes are not expressed in leukocytes. */ ** = p <0.05

Cere

bllu

m

Wh

ole

Bra

in

Ad

ren

al G

lan

d

Kid

ney

Skele

tal M

uscle

Pla

cen

ta

Th

yro

id G

lan

d

Test i

s

Lu

ng

Sto

mach

Sp

leen

Fib

rob

last

Ute

rus

Liv

er

Hear t

Lym

ph

ocyte

0 .0

0 .5

1 .0

1 .5

T is s u e

Re

lati

ve

Qu

an

tity

(R

Q)

of

KC

NK

9

Cere

bellu

m

Wh

ole

Bra

in

Skele

tal M

uscle

Ad

ren

al G

lan

d

Kid

ney

Pla

cen

ta

Th

yro

id G

lan

d

Test i

s

Lu

ng

Sto

mach

Sp

leen

Fib

rob

last

Ute

rus

Liv

er

Hear t

Lym

ph

ocyte

0 .0

0 .5

1 .0

1 .5

T is s u e

Re

lati

ve

Qu

an

tity

(R

Q)

of

PE

G1

3

a)

b)

*

**

*

**

97

a) cDNA Proband 1 b) gDNA Proband 3 Father c) cDNA Proband 3 Father

d) gDNA Proband 8 e) cDNA Proband 8

Figure 4.13 Electropherograms in ID and autism probands with deletions proximal to PEG13.

Deletions in probands had no effect on PEG13 expression levels in leukocytes. A) Proband 1 showed

biallelic expression at rs4289794. b) Leukocyte cDNA sequence of proband 3’s father who is

heterozygous for rs7828256 (C/G). c) Father’s cDNA sequence was heterozygous (C/G), indicating that

PEG13 is not imprinted in leukocytes. Proband 3 did not have any informative SNPs at the PEG13 locus;

therefore the paternal DNA was used instead. d) Proband 8 DNA sequence at rs7828256 (C/G,

heterozygous). E) Proband 8 cDNA sequence was heterozygous ((C/G, heterozygous), which also

confirmed biallelic expression of PEG13 in human leukocytes.

* * *

* *

98

Chapter 5: Methylation analysis of CpG islands

All work in this chapter was completed by me except the following: Identification of the PEG13-

DMR in fetal tissues was performed by Dr. Kazu Nakabayashi from the National Research Institute

for Child Health and Development (Tokyo, Japan). Sperm DNA was provided by The Krembil

Family Epigenetics Lab (CAMH, Toronto).

Differentially methylated CGIs are predictive features of imprinted genes, and to date, the ICs for a

number of long ncRNAs have been identified as maternally-methylated DMRs. Here, we assessed

methylation patterns at the PEG13-associated CGI and the KCNK9-associated CGI by bisulfite

pyrosequencing in fibroblasts and leukocytes, in conjunction to combined bisulfite restriction

analysis (COBRA) in fetal tissues. Methylation was also studied in Probands 3, 8 and 20 to identify

whether the CNVs affect methylation at the PEG13-DMR or KCNK9 CGI. As PEG13 is a paternally

expressed gene, we hypothesized that methylation would be maternally derived and therefore, the

DMR would be hypomethylated in sperm DNA.

We identified the PEG13-DMR in human brain, whole blood and fibroblasts, and maternal-specific

methylation in placenta. Low methylation levels were indicative of hypomethylation in sperm and a

germ line-derived DMR, corroborating our findings that PEG13 is a paternally expressed gene.

5.1 Materials and Methods

5.1.1 Sample collection

Isolation of DNA from Human Sperm

Sperm DNA from 10 healthy males (Table 5.1) was extracted using a standard phenol-chloroform

method. Briefly, semen was washed twice with 150mM NaCl and 10mM EDTA solution (pH 8.0)

and centrifuged at 1000 x g for 10min. The pellet was vortexed and resuspended in lysis buffer (6.0

99

guanidine thiocyanate, 5 M NaCl, 30% Sarkosyl, 1M DTT, 20, 000 mg/ml, 240μl H2O) and

incubated at 56˚C for 2 h, after which 2.4 ml of isopropanol was added for DNA precipitation. The

spooled DNA was removed from solution and transferred to a tube containing 2.0ml 0.1M sodium

citrate in 10% EtOH. After 30 min at RT, the buffer was removed and DNA was washed with 70%

EtOH and air-dried. The DNA was rehydrated in 10 mM Tris-HCl (pH 8.0), and incubated at 65˚C

for 1h until DNA was in solution.

Isolation of DNA from whole blood

The sample set (n=10) was taken from 440 unrelated individuals of European Caucasian ancestry

recruited at The Center for Addiction and Mental Health (CAMH, Toronto) and when possible, age-

matched to the sperm DNA sample set (Table 5.2). Individuals were screened and negative for

psychiatric conditions and serious medical conditions as described by Xu et al., (2014). DNA was

isolated from whole blood using a high-salt method as described in Chapter 3. Individual fibroblast

cell lines were obtained from the Coriell Institute (Camden, New Jersey) (Table 3.2). DNA was

extracted from fibroblasts using the Nucleospin® RNA kit following manufacturer’s protocol as

previously described in Chapter 3.

Sample Age Medical problem Smoker Rx Meds Supplements Ethnicity

C37 21 N/I N - - Caucasian

C117 24 CN N - CN-VitD Caucasian

C10 25 Eczema Y Hydrocortis - Caucasian

C46 29 HBP Y - Caucasian

C16 32 N/I N - VitD Caucasian

C77 33 CN N CN CN Caucasian

C28 39 Cleft palate N - - Caucasian

C104 40 CN Y Cephalexin CN Mixed

C44 41 N/I N - Multivitamin Asian

C125 52 CN N CN CN Caucasian

Table 5.1 List and description of individuals from which sperm DNA was collected. DNA from ten

sperm samples was used to assess methylation at PEG13-DMR. CN = ; HBP = high blood pressure; N/I =

not identified; N = No Y = Yes

100

Sample Age Ethnicity

GBP 7208 25 Caucasian

GBP 7014 26 Caucasian

GBP 7148 26 Caucasian

GBP 7031 29 Caucasian

GBP 7012 33 Caucasian

GBP 7113 33 Caucasian

GBP 7288 39 Caucasian

GBP 7058 40 Caucasian

GBP 7047 41 Caucasian

GBP 7207 52 Caucasian

Table 5.2 List and description of control leukocyte DNA. DNA isolated from whole blood for 10

healthy males that were age-matched to DNA from sperm samples.

Isolation of DNA from Fibroblasts

5.1.2 Bisulfite Conversion of DNA

Bisulfite conversion for DNA samples from fibroblasts, leukocytes and sperm was carried out with

the EZ DNA Methylation-Lightning Kit™ (Zymo Research) using 1.0μg of DNA. Converted DNA

was eluted in 20μl of TE buffer (pH 8.0). A brief summary of the reaction protocol is outlined below

in Figure 6.1.

Figure 5.1 Bisulfite conversion involves three main steps following DNA strand separation (98ºC for 8

min). Steps (1) and (2), the actual bisulfite conversion, occurs at 54ºC for 1 h, and step (3) occurs at RT

for 20 min. (1) Sulphonation: the addition of bisulfite (HSO3-) to the 5-6 bond of cytosine; (2) Hydrolic

deamination: the removal of the amine group (NH2) from the cytosine-bisulfite intermediate, resulting in

the loss of ammonia and the creation of a cytosine-uracil derivative; (3) Alkali desulphonation: the

removal of sulphonate from the uracil ring by alkali treatment. 5-mC is protected from sodium bisulfite

and retains its integrity as a cytosine residue.

Bisulfite conversion involves the deamination of unmethylated cytosine residues to uracil by sodium

bisulfite while leaving methylated cytosines intact (Figure 3.3) (Frommer et al., 1992; Clark et al.,

101

1994). Because DNA polymerases cannot discriminate between uracil and thymine, subsequent PCR

amplifications replace uracil with thymine while amplifying 5mC (and 5hmC) residues as cytosines

(Warnecke et al., 2002). Downstream applications for bisulfite-treated DNA generally involve assays

that quantitatively determine the ratio of thymine “T” versus cytosine “C” at specific CpG sites. Both

strands can be studied individually for DNA methylation due to a loss of DNA complementarity

following bisulfite conversion.

Bisulfite treatment is advantageous primarily because of its efficiency in converting ~99% of all

cytosine residues to uracil (Holmes et al., 2014). However, despite the range of bisulfite-based

applications, there remain certain limitations and caveats. First, bisulfite conversion leads to

considerable DNA degradation. Earlier bisulfite conversion methods involved the use of harsh

conditions (i.e. long incubation periods, strong chemical agents and high temperatures), resulting in

fragmented single stranded DNA and very low DNA concentration (Raizis et al., 1995). However,

while conditions have improved with the introduction of ready-made kits and decreased incubation

times, up to 90% of DNA undergoes degradation following bisulfite conversion. Therefore both

DNA quality and quantity prior to bisulfite treatment is important before commencing downstream

applications, particularly for PCR, which depends on the number of DNA molecules Secondly, the

formation of artifacts such as incomplete conversion and strand reannealing can lead to artificial

results (Warnecke et al., 2002). Furthermore, errors in PCR amplification (e.g. PCR bias and hybrid

products) can hinder sequencing results due to background interference (Warnecke et al., 2002).

Optimization of bisulfite PCR conditions – annealing temperature, DNA polymerase and Nested

PCR –can circumvent some amplification-based problems (Li & Tollefsbol, 2011).

5.1.3 Quantification of DNA methylation

CGI analysis and primer design

102

CpG islands were identified using EMBOSS CpGplot (EMBL-EBI) and Methyl Primer Express v1.0

software (Life Technologies: http://www.appliedbiosystems.com). Primers for PCR amplification

were designed using Microsoft Word and followed recommended guidelines for bisulfite treated

DNA (Patterson et al., 2011):

1. Primers were 25-30 bp in length with Primers a melting temperature of ~ 60˚C, and did not

differ by more than 1˚C for each pair

2. Primers had ~20% T-content (including both converted and non-converted T)

3. The 3’ end of primers (when possible) were a converted thymine (T) to ensure amplification

of bisulfite converted DNA

4. Primers were not designed within repetitive regions, palindromes, self-chain alignments or

polymorphic regions

5. Primer sequences were devoid of CpG nucleotides to prevent biased PCR amplification

6. Amplicon length did not exceed 300 bp to ensure maximum yield of the PCR product and to

facilitate amplicon capturing during the pyrosequencing reaction

PCR assays were designed to amplify a part of the CGIs associated with PEG13 and KCNK9 and

encompassed 5 to 7 CpG sites. All reverse primers featured a 5’ biotin modification to create a

forward sequencing assay.

5.1.4 PCR amplification

Each PCR reaction contained 2.0-3.0μl of bisulfite-treated DNA in a 25μl reaction volume and was

carried out using 12.5μl of KAPA2G™ Fast HotStart ReadyMx (2X) (Kapa Biosystems), 7.5-8.5μl

H2O and 1.0μl of each forward and reverse primer (10μM). DNA or polymerase was omitted in a

separate reaction for each sample to check for DNA contamination. Cycling conditions began with an

initial incubation at 95˚C for 1 min, followed by 40 cycles of 95˚C for 10 sec, 50˚C to 63˚C for

10sec, and 72˚C for 1 sec. Table 6.3 lists the specific annealing temperatures and oligonucleotides

103

used for each amplicon (IDT). The quality and quantity of PCR product in addition to DNA

contamination was verified by 2% gel electrophoresis.

5.1.5 Pyrosequencing and statistical analysis

PCR products were pyrosequenced using the PyroMark™ Q24 System according to the

manufacturer’s instructions (Qiagen) and each CpG site was analyzed using the pyrogram generated

by the CpG software provided by the manufacturer where methylation at each CpG site was

represented as a percentage (%) ranging from 0% to 100% (Chapter 3). Methylation profiles were

visualized as a box and whisker plot showing the median and interquartile ranges at a single CpG site

for all samples (n=10), except for methylation analysis in the ID and autism probands. Statistical

analysis was performed using GraphPad Software and involved calculations of the median, S.D., and

interquartile values.

5.1.6 Combined-bisulfite restriction analysis

Combined-bisulfite restriction analysis (COBRA) was performed to identify the PEG13-DMR and

allele-specific methylation (ASM) in DNA samples from human placenta and fetal brain. DNA

samples were treated to bisulfite by The EZ DNA Methylation-Gold Kit™ (Zymo Research ) using

1.0 μg of DNA. COBRA-primers used to investigate the PEG13-DMR were: forward (KS185)

5’TTATAGTTATTTGTAGGGTAGGGAA-3’; and reverse (KS186)

5’CTTAAAAACCTAACTAAATATCCTCCC-3’. The primers were complementary to bisulfite

treated DNA and did not contain CpG sites. Following PCR amplification, amplicons were digested

by BstUI restriction enzyme (New England BioLabs) and fragments were separated by

polyacrylamide gel electrophoresis (PAGE). Quantification of methylation was represented as a ratio

of all digested fragments to the total number of undigested and digested fragments.

104

Table 5.3 DNA methylation assays. PCR conditions and oligonucleotides used for bisulfite pyrosequencing and regions analyzed across all

biological samples for sperm, leukocytes and fibroblasts. (B)- represents 5’ biotin modification of the primer. Region Amplification Primers and hybridization

temperature

Product

length

(bp)

Sequencing Primers No.

of

CpGs

Sequence

length

No. of

nucleotide

dispensation

PEG13 DMR

chr8:141,110,441-

141,110,641

Forward

GAGTYGTAGTGGTATATGGAAGATT

Reverse

(B)- AATACRACACCCAACCAAATCTCAA

201 Primer 1

GTGGTATATGGAAGATTTGG

7 50 49

PEG13, DMR

chr8:141,108,632-

1411087555

Forward

GATGTTGAAGTTATGTATTGAGATT

Reverse

(B)-ACACCACCTTCCACAATATAATAA

122 Primer 1

GGTGTTTTTATTTAGTGGA

7 46 40

PEG13, DMR

chr8:141,108,731-

141,108,8807

Forward

GTTATTATTTTGTGTATAGTAGGAATGT

Reverse

(B)-AATCTCAATACATAACTTCAACATCC

140 Primer 1

TATTGAGATTTAGTTGGTGT

7 67 60

PEG13, DMR

chr8:141,107,853-

141,108,138

Forward

TAGGATAAGAATGTTTAGTTGGATTTTG

Reverse

(B) –ACAACTACTCATTAAAAACATAAACTAAC

(60.5)

287 Primer 1

GTTTATTTAGGTTTTTAGTG

7 61 61

PEG13, DMR

chr8:141,107,731-

141,108,007

Forward

TTTGTTAATAAGGAGGTTTTAATGGTGT

Reverse

(B) –TCAAAAATACCTCAATTCACAAAATACTA

277 Primer 1

GTATATGGAAGATTTGGG

5 54 47

KCNK9 CGI

chr8:140,714,342-

140,714,644

Forward

TTAGGGGTTTGGGAAGTTTAAGTTT

Reverse

(B) – AAATAATCCCATAACAAACACCCAC

333 Primer 1

GAGTTTAGAGAAGTTTGGA

Primer 2

GTAAGTGTAAGGTTTAGAGTT

4

5

51

26

40

27

105

Figure 5.2 Plots of methylation at the PEG13-CGI in human leukocyte DNA. Box and whisker plots show median and interquartile ranges at

single CpG sites for10 healthy individuals (Table 6.2). The error bars indicate the S.D across samples. Methylation patterns were assessed at four

regions. (A) and (B) Region 1 (chr8:141,107,853-141,108,138) and region 2 (chr8:141,107,731-141,108,007) demonstrate high levels of

methylation, synonymous to hypermethylation. (C) and (D) Methylation at region 3 (chr8:141108632-1411087555) and region 4

(chr8:141,108,731-141,108,8807) have differential methylation, indicating a DMR

1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S I t e

%

Me

thy

lati

on

1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e%

M

eth

yla

tio

n

1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4 5

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

A) B)

C) D)

106

Figure 5.3 Plots of methylation at the PEG13-CGI in human fibroblasts DNA. Box and whisker plots show median and interquartile ranges at

single CpG sites for 10 healthy individuals (Table 3.1). The error bars indicate the S.D. of methylation in all samples. (A) Methylation profiles at

region 1 (chr8:141,107,853-141,108,138), similar to leukocytes, is highly methylated akin to hypermethylation. (B) and (C) Methylation profiles

at region 2 (chr8:141108632-1411087555) and region 3 (chr8:141,108,731-141,108,8807) showed differential methylation, indicating a DMR

1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

% M

eth

yla

tio

n1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

% M

eth

yla

tio

n

C) B)

A)

107

Figure 5.4 Plots of methylation at the PEG13-CGI in human sperm DNA. Box and whisker plots show median and interquartile ranges at

single CpG sites for 10 healthy individuals. The error bars indicate the S.D. across samples. (A) Methylation profiles at Region 1

(chr8:141,108,731-141, 108, 8807) demonstrate low methylation levels (i.e. hypomethylation) in sperm as compared to (B) differential

methylation in leukocytes. (C) and (D) Methylation profiles at Region 2 (chr8:141,110,441-141,110,641) and Region 3 (chr8:141108632-

1411087555) illustrate a hypomethylated pattern.

1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4 5

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4 5

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4 5 6 7 8

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

A) B)

C) D)

108

5.2 Results

To determine if DNA methylation levels reflect allelic expression of PEG13, individual CpG sites

were profiled at various regions within the CpG island overlapping PEG13 (chr8: 141107838-

141110984; GRCh37/hg19) and the KCNK9 CGI (chr8:140,714,586-140,718,259; GRCh37/hg19) by

bisulfite pyrosequencing. For pyrosequencing, methylation levels were measured as a ratio of

cytosine:thymine residues (%) at each CpG site and were defined as a range where 0:100 to 30:70

was considered lowly methylated (hypomethylation), 70-30 to 100:0 was highly methylated

(hypermethylation) and a ratio between 30:70 to 60:40 was indicative of a DMR. The regions

selected spanned the entire PEG13 CGI for an accurate representation of methylation patterns.

5.2.1 Identification of the PEG13-DMR

Methylation was assessed at four regions within the PEG13-CGI for human leukocyte DNA (n=10).

Figure 5.2 illustrates methylation patterns at single CpG sites for each region. Regions 1

(chr8:141,107,853-141,108,138) comprised seven CpG sites and region 2 (141,107,731-141,108,007)

had five CpG sites. Both regions were highly methylated and consistent with a hypermethylated

pattern, where region 1 and region 2 had mean methylation of 93.0 ± 4.14% and 90.0 ± 5.55%,

respectively. Region 3 (chr8:141108632-1411087555) encompassed seven CpG sites with a mean

methylation level of 40.75 ± 8.34%, indicative of a DMR. Similarly, region 4 (chr8:141,108,731-141,

108, 8807) spanned seven CpG sites and had an average methylation level of 39.69 ± 3.96%.

Fibroblasts (n=10) shared similar methylation patterns as leukocytes (Figure 5.3). Region 1

(141,107,853-141,108,138) in fibroblasts was also hypermethylated (89.49 ± 5.54%), and was

similar to leukocytes. Region 2 (chr8:141108632-1411087555) had a mean methylation level of

38.06 ± 4.53% and region 3 (chr8:141,108,731-141, 108, 8807) had an average methylation level of

36.40 ±6.66%). We identified the PEG13-DMR in human leukocytes and fibroblasts, and supports

imprinting of PEG13 in fibroblasts.

109

5.2.2 PEG13-CpG island is a maternal DMR

Methylation patterns at three different regions within the PEG13-CGI (chr8: 141107838-141110984;

GRCh37/hg19) were investigated using human placental and fetal brain DNA by combined bisulfite

restriction analysis (COBRA) (Figure 5.5; Appendix 6). Region 1 (chr8:141107809-141108250) and

region 3 (chr8:141109944-141110307) exhibited hypermethylated and hypomethylated patterns,

respectively. Region 2 (chr8:141108200-141108523) was found to be differentially methylated,

however the parental origin could not be determined due to the absence of informative SNPs within

this region. Intriguingly, region 1 showed predominant maternal methylation in placental DNA,

suggesting that region 2 acquires methylation on the maternal allele.

Accordingly, methylation patterns were assessed at three regions in sperm DNA (Figure 5.4). As

expected, the average methylation at each region was lowly methylated and indicative of

hypomethylation. Region 1 (chr8:141,108,731-141, 108, 8807) encompassed five CpG sites within

the PEG13-DMR, had a mean methylation of 1.80% ± 1.8% in sperm as compared to 42.2 ± 1.8% in

whole blood. The hypomethylated profile observed in sperm suggests that the PEG13-DMR is a

maternally-derived DMR.

110

Figure 5.5. Schematic diagram of the human TRAPPC9/PEG13 locus and identification of the PEG13 DMR in human fetal brain and

placenta (Figure not drawn to scale). One of the blue rectangular boxes represents the PEG13 transcript (AK307073) and lies within intron 17 of

TRAPPC9. The black arrows indicate the transcriptional direction and the black rectangular boxes depict the centromere (cent) and telomere (tel).

Methylation at the PEG13-associated CpG island (green rectangular box) was assessed at three regions. Each circle denotes a single CpG site,

where methylated and unmethylated are black and white circles, respectively. Allelic methylation at region 1 (chr8:141107809-141108250) was

evaluated using placental DNA at rs3802217, where G was the paternal allele and A was the maternal allele, which was methylated. The overall

methylation pattern at region 1 indicates hypermethylation. Region 2 (chr8:141108200-141108523) and region 3 (chr8:141109944-141110307)

were assessed using fetal brain DNA. Region 2 exhibited differential methylation, indicating the PEG13-DMR. Conversely, region 3 showed

methylation levels reflecting hypomethylation.

111

5.2.3 KCNK9-associated CpG island is hypomethylated

To support our hypothesis that PEG13 influences KCNK9 expression, we measured DNA

methylation levels at the KCNK9 CGI (chr8: 140714586-140718259) (Figure 5.6). The region

analyzed (chr8:140,714,585-140,714,615) contained 5 CpG sites and had low levels of methylation

in both leukocytes (9.3 ± 3.39%) and fibroblasts (6.65 ± 4.38%), similar to sperm. Methylation was

assessed at four CpG sites just outside the promoter region (chr8:140,714,444-140,714,512) as a

comparison, where the overall methylation levels varied (26.2± 4.98%). The overall hypomethylated

pattern at the KCNK9 CGI in leukocytes and fibroblasts indicates that a mechanism other than

promoter methylation regulates KCNK9 imprinting.

5.2.4 Methylation in ID and autism probands Methylation patterns at the PEG13-CCGI were analyzed using DNA from leukocytes of ID and

autism probands to determine if the CNVs proximal or distal to the PEG13-DMR affect methylation.

Similar methylation patterns at the PEG13-DMR and CGI of KCNK9 were observed in all probands

as compared to leukocyte DNA from healthy controls (Figure 5.7). Region 1 (chr8: 141,110,441-

141,110,641) encompassed seven CpG sites. For the deletion probands, a mean methylation level of

37.9 ± 2.54% and 38.1 ± 2.12% was observed for Proband 3 and 8 respectively. Proband 20, who has

a paternally-inherited duplication, had a mean methylation level of 32.6 ± 2.51 %. All probands had

low methylation levels at the KCNK9 CGI similar to control leukocytes, where Proband 3, 8 and 20

had mean methylation levels of 6.40 ± 1.52%, 8.40 ± 2.70% and 10.4 ± 3.78%. Taken together, the

CNVs do not appear to affect methylation at the PEG13-DMR or KCNK9 CGI, suggesting that either

the disorder mechanism is tissue specific (i.e. not in leukocytes), or the CNVs may exert their

pathological effect by alternative method.

112

Figure 5.6 Plots of methylation at the PEG13-CGI and KCNK9 CGI in human leukocytes and fibroblasts. Box and whisker plots show

median and interquartile ranges at single CpG sites for 10 healthy individuals. The error bars indicate the S.D. across samples. (A) Methylation

profiles at individual CpG sites outside the KCNK9 CpG island (chr8:140,714,444-140,714,512) were highly variable and differed significantly (p

< 0.5). (B) and (C) Methylation at region 1 (chr8:140,714,585-140,714,615) in leukocytes and fibroblasts were lowly methylated, indicative of

hypomethylation.

1 2 3 4 5

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

thy

lati

on

1 2 3 4 5

0

2 0

4 0

6 0

8 0

1 0 0

C p G S i t e

%

Me

th

yla

tio

n

A)

B) C)

113

Figure 5.7 Methylation at the PEG13 and KCNK9 CpG islands derived from bisulfite

pyrosequencing in Autism and ID proband leukocytes. (A) Methylation profiles at Region 1 were not

different between probands and controls, (C) Methylation profiles at region 2 the KCNK9 CpG island was

lowly methylated in all probands, similar to control leukocytes. Circles represent the methylation at a

single CpG site. Black circles = proband 3; Red circles = Proband 8; White circles = Proband 20

1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 6 7

0

2 0

4 0

6 0

8 0

1 0 0

C p G S I t e

% M

eth

yla

tio

n

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

0

2 0

4 0

6 0

8 0

1 0 0

C p G S I t e

% M

eth

yla

tio

n

P ro b a n d 3 P ro b a n d 8 P ro b a n d 2 0

A)

B)

114

Chapter 6 Summary and conclusions, general discussion and future directions

The data presented here highlights the association between genomic imprinting and

neurodevelopment by providing evidence for a novel imprinted gene, PEG13, showing preferential

expression towards the paternal allele in human fetal brain. Further hypotheses require investigation

in brain samples from both normal, healthy individuals and our deletion probands in conjunction with

functional in vitro experiments to a) determine the specific role of PEG13 on KCNK9 imprinting in

the brain; and to b) elucidate any neurophysiological implications for ID and ASD.

6.1 General Discussion

PEG13 is a paternally expressed ncRNA in the brain

We identified an intronic long ncRNA running antisense to the intellectual disability gene TRAPPC9

and illustrate partial conservation of the imprinting that was identified for mouse Peg13 at the human

syntenic region (chr8q24.3). Human PEG13 was found to be preferentially expressed from the

paternal allele like its mouse equivalent (Wang et al., 2008; Davies et al., 2004; Smith et al., 2003),

specifically in fetal brain, as well as fibroblasts and kidney. Interestingly, PEG13 is biallelically

expressed in placenta. There is currently no data indicating the evolutionary loss of imprinting for

PEG13 in the placenta before species divergence between mice and humans. Moreover, whether

knockout of Peg13 or deregulated imprinting confers embryonic lethality or abnormal behavioural

and/or growth phenotypes in mice requires investigation. The fact that PEG13 is biallelically

expressed in placenta supports the possibility that PEG13, along with its brain-specific expression,

may have arisen in a lineage specific to ethuerian mammals, or was recently acquired in mice and

maintained in humans. Indeed, unlike most imprinted long ncRNAs, and dissimilar to mouse Peg13,

human PEG13 lies within a non-imprinted gene (i.e. TRAPPC9), an arrangement analogous to a

microimprinted domaim. Microimprinted domains are hypothesized to have eutherian-specific

115

lineages acquired through late retrotransposition events. For example, neuronatin (NNAT), a

paternally-expressed gene, resides within the single intron of BLCAP on chr20q11.2 and is highly

expressed in the CNS from mid-gestation to early postnatal life (Evans et al., 2001). A subsequent

study revealed that imprinting of NNAT is restricted to placental mammals, with no traceable lineage

in monotremes or marsupials (Evans et al., 2005). As susch, phylogenetic analysis across species for

PEG13 may help to address the evolutionary impact of this gene and its role in the brain.

Nonetheless, various lines of evidence support the importance of PEG13’s imprinted expression in

brain function and development. First, PEG13 remains imprinted from fetal (Law, 2014; this thesis)

to adult life (Court et al., 2014). Secondly, PEG13 is the only gene along with KCNK9 who’s

imprinting status remains conserved from mice. The fact that only seven out of ten fetal brain

samples displayed preferential expression may be attributed to the heterogeneity of the tissue

samplesLastly, PEG13 has a similar expression profiles as the mouse gene: we showed that PEG13

is highly expressed in the brain (Davies et al., 2004), as well as in adrenal gland, and has minimal

expression in kidney, heart and placenta (Smith et al., 2003). Davies et al., (2004) found that mouse

Peg13, similar to other paternally-expressed genes in the brain, has a neuron-specific localization in

olfactory bulbs, cerebral cortex, hippocampus and thalamus. In situ hybridization of PEG13 in

human brain tissues (or other techniques such as gene expression microarray, SAGE (serial analysis

of gene expression) or RNA-seq) are required to determine if this brain-specific expression pattern is

conserved. Furthermore, we identified a DMR in the CGI associated with PEG13 in all cells and

tissues examined, including fetal brain. Analysis of the PEG13-DMR showed that the maternal allele

was methylated in extra-embryonic tissue, and hypomethylated in sperm, indicative of a germ line-

derived maternal DMR.

Previous studies have identified imprinted long ncRNAs in mouse brain, although few have been

characterized in humans. A recent survey identified 34 imprinted ncRNAs in mouse brain which

116

corresponded to the Allen Brain Atlas (ABA) (Nikaido et al., 2003), and included Air (Sleutals et al.,

2005), Copg2as (Lee et al., 2000), Gtl2 (Wylie et al., 2000), Rian (Cavaillé et al., 2002), and Mirg

(Gregg et al., 2010). Subsequent analysis of ABA genes in mouse brain illustrated that long ncRNAs

have regional expression patterns with enrichment in the hippocampus, cerebellum, olfactory bulb

and cerebral cortex, in addition to subcellular compartmentalization (Mercer et al., 2008). Some cis-

antisense and intronic transcripts show expression patterns corresponding to their protein-coding

genes of neurological relevance (Mercer et al., 2008), indicating that unlike previous suppositions;

some long ncRNAs are not merely genomic artefacts. A proportion of long ncRNAs been

characterized as imprinted genes in mouse brain and impart regulation in brain development and

function (Davies et al., 2004). The ncRNA Air is paternally expressed in glial cells and regulates

maternal expression of Igf2r, and loss of Air imprinting in neurons results in biallelic expression of

Igf2r (Yamasaki et al., 2008). Mercer et al., (2008) also identified a novel antisense transcript,

AK045070, which shows parallel expression patterns to the nuclear receptor, Coup-Tfll, in mouse

brain. Another imprinted long ncRNA, Kcnq1ot1, which shows region-specific expression in tissues,

may also confer silencing in specific brain regions in addition to the placenta (Korostowski et al.,

2012; Robins et al., 2012). Based on these findings, further analysis is required to determine how

PEG13 plays a role in KCNK9 imprinting and if it acts similar to other antisense transcripts in the

brain.

Accordingly, a recent study by Court et al., (2014) confirmed that PEG13 is maternally imprinted in

human adult frontal cortex, and demonstrated that the mechanism by which PEG13 regulates KCNK9

expression parallels the reciprocal expression of H19 and Igf2 through higher-order chromatin

remodeling (Kurukuti et al., 2003; Murrell et al., 2004) (Figure 8.1). Analogous to our findings, the

study authors found that the imprinted expression of PEG13 was associated with methylation at the

PEG13-DMR on the maternal allele. Furthermore, the non-methylated state of the PEG13-DMR on

117

the paternal allele was found to preferentially bind to CTCF-cohesin in the brain (Court et al., 2014).

Most notably, the study identified a novel brain-specific enhancer element distal to PEG13 and

located in the 17th intron of TRAPPC9. Subsequent chromatin-conformation capture (3C)

experiments revealed that in human cerebellum, the distal enhancer interacts with the KCNK9

promoter (which also has CTCF enrichment) and the PEG13-DMR through chromatin looping

according to the parental allele (Court et al., 2014) (Figure 7.1). In other words, the enhancer

interacts with PEG13-DMR depending on its methylated state: on the paternal allele, the

unmethylated PEG13-DMR possesses enhancer-blocker activity (or acts as an insulator) as it can

bind to CTCF-cohesin to induce higher order chromatin folding, thereby preventing the enhancer

from interacting with the promoter of KCNK9. Moreover, enrichment of the epigenetic marks

H3Kme1, H3K27ac and p300 in at the PEG13-DMR was restricted to the brain and not found in

leukocytes (Court et al., 2014), thus supporting our conclusion that imprinting of PEG13 is primarily

restricted to the brain. The higher-order chromatin silencing model has been previously described in

Saccharomyces cerevisiae where chromatin looping has been observed at the processed promoter

(5’end) and terminator (3’end) regions of genes transcribed by RNAP II and is required for

transcriptional activation (O’Sullivan et al., 2004). As indicated above, chromatin looping also

separates promoters and enhancers in the Igf2-H19 imprinting cluster according to the methylation

status of the IC (Murrell et al., 2004). This enables Igf2 to reside within open chromatin for gene

expression or within repressive chromatin depending on the allele, wherebinding of CTCF to the

Igf2/H19 IC facilitates chromatin packaging on the maternal allele for silencing of Igf2 (Kurukuti et

al., 2006). The Kcnq1ot1 long ncRNA has also has a role in the imprinted silencing of Kcnq1 by

affecting chromatin conformation, which in turn, influences interactions between a local enhancer

and the Kcnq1 promoter (Korostowski et al., 2012).

118

We found no significant differences in methylation levels between our probands and normal

individuals at the PEG13-DMR or KCNK9 CGI. Our results here do not show any association

between the CNVs and disorder aetiology in our probands. However, we cannot disregard the

possibility for a mechanism in the imprinting context. First, the epigenetic changes may be far more

robust in tissues that are actually affected by PEG13 and KCNK9 activity, namely the brain.

Secondly, other epigenetic mechanisms, such as chromatin conformation and histone modifications

may constitute the primary epigenetic marks regulating PEG13-KCNK9 imprinting. Downstream

effects of deletions/duplications around PEG13 may manifest by modulations to chromatin

architecture. Indeed, the juxtaposition of euchromatic and heterochromatic domains by

nucleoproteins and epigenetic marks define chromosomal regions and dictate the positioning of

regulatory elements and the accessibility of interacting proteins (Kleinjan & van Heyningen, 1998).

Structural rearrangements, translocations, inversions, deletions, duplications, can influence gene

expression through “position effects” by altering the genomic position or chromatin environment in

which a gene or regulatory element reside (Feuk et al., 2006). Position effect refers to the stochastic

spreading of heterochromatin across a euchromatic region (which assumes the conformation of

heterochromatin) due to chromosomal rearrangements, resulting in the inactivation of expressed

genes (Kleinjan & van Heyningen, 1998). This phenomenon was originally observed in Drosophila

for eye pigmentation and referred to as position effect variegation (PEV) (Wallrath & Reuter, 1995).

Genes encoding chromatin-modifying proteins, either as enhancers or silencers of gene expression

including Su (var) proteins, were identified via direct mutagenesis and associated to genes silenced

by PEV (Fodor et al., 2010; Tschiersch et al., 1994). Furthermore, yeast also exhibit gene

transcriptional silencing by position effect (Gottcschling et al., 1990) and is a natural occurrence in

119

Figure 6.1 Proposed model for PEG13-mediated imprinting of KCNK9 in the brain through higher order chromatin looping (based on

diagram from Court et al., 2014) (Not drawn to scale). A brain-specific enhancer is located in intron 17 of TRAPPC9 and has enrichment for

H3K4me1, H3K27ac and p300. (A) On the paternal allele, the unmethylated PEG13-DMR binds to a CTCF-cohesin complex to facilitate

chromatin looping, and blocks the enhancer from interacting with the promoter of KCNK9, resulting in silencing of KCNK9. (B) Conversely on

the maternal allele, the methylated PEG13-DMR cannot interact with the enhancer element, allowing the enhancer to associate with the KCNK9

promoter region to facilitate KCNK9 transcription on the maternal allele. The specific role of the long ncRNA PEG13 in this mechanism is

currently unknown. Grey circles = unmethylated and black circles = methylated; Green rectangular boxes = CGI

A) B)

120

mammals (Capel et al., 1993; Koopman et al., 1991). Chromatin-mediated inactivation has been

connected to human diseases such as aniridia (Fantes et al., 1994), Potocki-Shaffer syndrome (Feuk

et al., 2006; Wakui et al., 2008), and ɑ-Thalassemia (Barbour et al., 2000). Position effect has also

been inferred in developmental disorders where mutations to gene coding sequences have not be

identified in affected individuals (Lee et al., 2006; Velagaleti et al., 2005). Indeed. breakpoints of

chromosomal rearrangements are often distal to coding sequences and open reading frames, which

are free of mutations but concomitant to gene downregulation through various mechanisms. Position

effects manifest by altering chromatin structure around a regulatory element, dissociating

transcriptional units and regulatory elements, and/or may perturb methylation at gene promoters

(Barbour et al., 2000; Kleinjan & van Heyningen, 1998). Indeed, insulator elements help demarcate

chromatin boundaries through chromatin looping and higher order chromatin structure (Labrador &

Corces, 2002). It is therefore plausible that CNVs may affect how chromatin is organized within and

surrounding the PEG13 locus. Intriguingly, and in support of this hypothesis, the chromatin-

remodelling complex CHRAC1 resides proximal and upstream to TRAPPC9. CHRAC1, which

comprises five subunits including the ATPase ISWI and topoisomerase II, increases the accessibility

to chromatin by facilitating nucleosome assembly and disassembly (Varga-Weisz et al., 1997).

Whether the genomic position of CHRAC1 is merely coincidental and if its expression is required for

cis-regulation of chr8q24.3 remains unexplored.

PEG13: Potential implications in ID and ASD etiology

Given that PEG13 and KCNK9 are not highly expressed in leukocytes, we could not identify any

changes in gene expression that might result from deletions proximal or distal to the PEG13-DMR.

Additionally, we could not determine if loss of imprinting at PEG13 contributes to the neurological

profiles observed in our ID patients. Based on the assumption that PEG13 regulates imprinting at

KCNK9, it is possible to expect some phenotypic similarities between our affected ID and autistic

121

individuals and those with Birk-Barel Syndrome. Barel et al., (2007) described an imprinting

syndrome caused by a missense mutation replacing glycine at position 236 by arginine (G236R) in

exon 2 on the maternal copy of KCNK9. The affected individuals displayed mild to severe ID,

generalized hypotonia in early life which progressed to muscle weakness in major muscle groups, as

well as unique facial dysmorphisms including elongated faces. Subsequent whole cell

electrophysiology in oocytes found that the K2p9.1 (G236R) channel produced no measurable

currents. Moreover, oocytes expressing the wild-type and mutant channel in a heterologous system

elicited a decrease in K+ currents as compared to wild-type cells; the same phenotype was observed

in oocytes co-expressing the mutant channel along with TASK-1, a channel which under normal

physiological conditions heterodimerizes with TASK-3 in the human and mouse CNS (Barel et al.,

2007). Interestingly, the study authors did not report any behavioural abnormalities other than ID in

the affected individuals, and since this study, no other disorders associated with KCNK9 have been

reported. Intriguingly, one of our study subjects (proband 20) has ID, ataxia, as well as characteristic

dysmorphisms (Appendix 4). While these comparisons are not conclusive, the overlap of certain

clinical features suggests that it would be worthwile to obtain more clinical information from our

probands for more in-depth analyses.

Role of KCNK9/TASK-3 in the brain

PEG13 expression was mostly restricted to the brain with a tissue-specific pattern similar to other

identified brain-specific ncRNAs (Qureshi et al., 2010). Moreover, PEG13’s high expression in the

cerebellum suggests region-specific and sub-comparmental localization in the brain. The fact that

KCNK9 and PEG13 exhibit overlapping expression profiles suggests some functional importance,

specifically with regards to how PEG13 regulates KCNK9 expression. Moreover, both genes are

highly expressed in the cerebellum, which influences both motor and non-motor functions, and

122

ncRNAs have subcellular localization in cerebellar Purkinje cells (i.e. diffused in the soma, nuclei, or

foci) (Mercer et al., 2008).

The KCNK9 gene encodes the K2p9.1 potassium channel (also known as TASK-3) from the two pore

domain (K2p) subfamily. K2p channels are among the largest class of K+ channels with over 50

related genes identified in Caenorhabiditis elegans (Bargmann, 1998); comparatively in mammals,

14 genes have been identified, of which 11 are functional and 8 are highly expressed in the CNS

(Talley et al., 2003). Structurally, K2p channels are characterized by their four membrane spanning

domains and two pore-forming domains (2P/4TM) which dimerize to form a conducting K+ pore

(Lesage & Lazdunski, 2000). These background or “leaky” channels are both voltage and time-

independent (Duprat et al., 1997), producing basal outward rectifying potassium currents that not

only maintain resting membrane potential (RMP), but also modulate cell excitability by affecting the

duration, frequency and amplitude of action potentials during hyperpolarization and depolarization

(Talley et al., 2001). This is particularly important for neuronal excitability, heart rate, muscle

contractions, as well as hormone secretion. Furthermore, K2p channels are responsive to a variety of

other chemical and physical agents, including volatile anesthetics (Lesage, 2003; Veale et al., 2007),

oxygen (Mulkey et al., 2007), neurotransmitters, mechanical stretch, lipids, and G-protein coupled

receptors (Duprat et al., 2007). However, despite their structural similarities, K2p channels share

relatively low sequence similarity and consequently, exhibit diverse cellular functions and expression

patterns in mammals.

Cloning of the K2p channels in heterologous systems has enabled for their classification into five

subfamilies based on their electrophysiological and pharmacological properties: the acid-sensitive K+

(TASK), the weakly inward rectifying K+ channel (TWIK), the TWIK-related K+ channel (TREK), the

TWIK-related arachidonic acid stimulated K+ channel (TRAAK) and tandem-pore domain halothane-

123

inhibited K+ (THIK) (review by Goldstein et al., 2001; 2005). The mammalian TASK family

comprises TASK-1 (Duprat, 1997), TASK-2 (Medhurst et al., 2001; Lesage et al., 2000), TASK-3

(Kim et al., 2000) and the non-functional TASK-5 (Ashmole et al., 2001; Duprat et al., 1997), which

are all expressed in rodents and humans. Only TASK-1 and TASK-3 share sequence homology (Rajan

et al., 2002), and as their nomenclature implies with the exception for TASK-5, respond to slight

variations in extracellular pH (Lesage & Lazdunski, 2000; Rajan et al., 2000; Kim et al., 2000).

K2p9.1 activity may also be modulated by local volatile anesthetics such as halothane and isoflurane

(Czirják & Enyedi, 2003), and serve as a therapeutic target.

Previous expression studies illustrate that TASK channels are widely expressed in mammals and

exhibit different tissue expression profiles in the CNS and peripheral tissues (Chapman et al., 2000;

Medhurst et al., 2001); our current work is in agreement with these reports, specifically for TASK-3,

and poses some interesting questions with regards to the functional implications of PEG13 and its

affect on KCNK9 expression. Indeed, TASK-3 is prominently expressed in the human brain where it

is restricted to the cerebellum, and has little to no expression in other somatic tissues (Chapman et

al., 2000; Medhurst et al., 2001). Interestingly, TASK-3 expression is more regionally distributed in

the rodent CNS, showing moderate to high expression in select hypothalamic nuclei, the amygdala,

the neocortex and spinal cord, yet its marked expression in the cerebellum, specifically the granule

cell layer, remains conserved (Karschin et al., 2001; Talley et al., 2001; Vega-Saenz et al., 2001).

TASK-3 expression overlaps with some other KCNK transcripts, particularly with TASK-1, in both

neuronal and non-neuronal tissues (Karschin et al., 2001; Talley et al., 20001). The complementary

expression patterns in the brain between TASK-1 and TASK-3 are not surprising given that these

channels share 54% sequence homology and have been shown to heterodimerize in these regions

(Aller et al., 20005; Millar et al., 2000; Talley et al., 2003).

124

At the physiological level, knockout TASK-3 in cortical granular neurons fire at higher frequencies

during depolarization, as expected (Brickley et al., 2007), indicating a role in controlling neuronal

output and excitability. Accordingly, loss of TASK-3 expression is assoociated with impaired

memory and altered cognitive function (Linden et al., 2007). Therefore, dysregulation of PEG13

expression may lead to a similar neurophysiological phenotype.

Hydroxymethylation: another epigenetic signature in the brain

Bisulfite sequencing cannot distinguish between 5-methylcytosine (5mC) and 5-

hydroxymethylcytosine (5hmC) (Hahn et al., 2013). Consequently, 5mC may not account for all the

methylated marks we observed at the PEG13-DMR, namely in brain tissues. Previous studies showed

that the mammalian brain is enriched in 5hmC marks (Guo et al., 2011; Szulwach et al., 2011) and

accounts for 0.20% of cytosine bases in fetal cortex and 0.87% inadult cortex (Lister et al., 2013).

However, although 5hmC does not represent a significant fraction of cytosine methylation, it is

widely hypothesized to serve as a functional intermediate in regulating gene expression as the

majority of 5hmC marks are found at active genomic sequences i.e. promoters, intergenic regions and

gene bodies (Jin et al., 2010; Robertson et al., 2011). Moreover, genes required for neuronal function

such as ion channels and scaffold proteins show enrichment for 5hmC marks (Robertson et al.,

2011). Recent whole-genome studies illustrate dynamic changes to the brain methylome that occur in

development and with aging, and recognize 5hmC as a key regulatory mark (Guo et al., 2014; Lister

et al., 2013; Xie et al., 2012). Lister and colleagues (2013) found that while fetal genomes are

enriched in 5mC marks, early postnatal genomes were marked by the accruement of conserved non-

CG methylation, including 5hmC (Szulwach et al., 2011); this trend, which continues into

adolescence, coincides with the primary period of neurogenesis (Lister el., 2013). Furthermore, the

increase of 5hmC marks from fetal life to adulthood is restricted to neuronal cells and not glial cells

(Hahn et al., 2013; Lister et al., 2013). 5hmC profiles are also region-specific, thus adding another

125

layer of gene regulation. For instance, 5hmC marks in human cerebellum were shown to significantly

increase from fetal to adult life, and the change was correlated to changes in gene expression (Wang

et al., 2012). Presently, the role of 5hmC in the context of imprinting has yet to be addressed and the

purpose of 5hmC versus 5mC methylation is not well-understood, although certain studies postulate

that 5hmC may be a stronger silencing mark than 5mC (Robertson et al., 2011). Therefore,

measuring hmC marks at the PEG13 DMR, as well as in relation to KCNK9, may provide more

insight into the function of this conserved imprinted locus in neurodevelopment and adulthood.

6.2 Conclusions

This study identified a novel imprinted gene PEG13 in human fetal brain and provides evidence that

the long ncRNA has a role in the imprinting of KCNK9. The tissue- and temporal-specific imprinting

of mouse Trappc9 is not conserved in humans. The role of the heterozygous deletions/duplications in

ID and ASD etiology require further investigation, however based on our current knowledge, a

disease mechanism including PEG13 cannot be ruled out. There is the possibility that these structural

variations may affect chromatin architecture surrounding the PEG13 locus, though future studies will

need to address this hypothesis.

6.3 Future Directions

Studies using induced pluripotent stem cell (iPS) generated from our ID and autistic probands and

differentiated to neurons will enable us to properly study the effects of the heterozygous deletions in

the brain given that the expression of PEG13 and KCNK9 are restricted to the CNS. Moreover, other

epigenetic mechanisms, such as chromatin remodeling and 5hmC marks will need to be studied at the

PEG13 locus by chromatin immunoprecipitation (ChIP) assays and protein assays such as western

blotting in neurons. Additionally, if deletions/duplications affect KCNK9 expression,

electrophysiology studies measuring K+ currents will provide more insight into the neurophysiologic

effects in ID and ASD etiology for our probands. In fact, electrophysiological studies, as well as

126

facial dysmorphology, and neurological testing would be helpful in order to compare the clinical

presentation of these CNV individuals to reported cases of Birk-Barel syndrome. Knockdown of the

long ncRNA by small interfering RNAs (siRNAs) or transfection of PEG13 in cells may also help

determine the role of PEG13 in KCNK9 imprinting.

127

References

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th

ed., text rev.). doi:10.1176/appi.books.9780890423349.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders:

DSM-5. Neurodevelopmental Disorders (pp. 31-86). Washington, D.C: American Psychiatric

Association.

Albrecht, U. et al., (1997). Imprinted expression of the murine Angelman syndrome gene, Ube3a, in

hippocampal and Purkinje neurons. Nature genetics, 17(1), 75-78.

Allfrey, V. G., Faulkner, R., & Mirsky, A. E. (1964). Acetylation and methylation of histones and

their possible role in the regulation of RNA synthesis. Proceedings of the National Academy

of Sciences of the United States of America, 51(5), 786.

Allen, N.D. Asperger, H. (1944). Die “Autistichen Psychopathen” im Kindesalter. Arch Psyciatr

Nervenkr. 117, 76-136.

Amir, R.E. et al., (1999). Rett syndrome is caused by mutations in X-linked MECP2, encoding

methyl-CpG-binding protein 2. Nature Genetics, 23 (1999), 185–188.

Antequera, F., & Bird, A. P. (1988). Unmethylated CpG islands associated with genes in higher plant

DNA. The EMBO journal, 7(8), 2295.

Antequera, F., & Bird, A. (1993). Number of CpG islands and genes in human and mouse.

Proceedings of the National Academy of Sciences, 90(24), 11995-11999.

Arima, T. et al., (2005). ZAC, LIT1 (KCNQ1OT1) and p57KIP2 (CDKN1C) are in an imprinted

gene network that may play a role in Beckwith–Wiedemann syndrome. Nucleic acids

research, 33(8), 2650-2660.

Arney, K. L., Bao, S., Bannister, A. J., Kouzarides, T., & Surani, M. A. (2002). Histone methylation

defines epigenetic asymmetry in the mouse zygote.International Journal of Developmental

Biology, 46(3), 317-320.

Ashmole, I., Goodwin, P. A., & Stanfield, P. R. (2001). TASK-5, a novel member of the tandem pore

K+ channel family. Pflügers Archiv, 442(6), 828-833.

Babu, A., & Verma, R. S. (1987). Chromosome structure: euchromatin and heterochromatin. Int.

Rev. Cytol, 108, 1-60.

Bailey, A., Palferman, S., Heavey, L., & Le Couteur, A. (1998). Autism: the phenotype in relatives.

Journal of autism and developmental disorders, 28(5), 369-392.

Bailey, A. et al., (1995). Autism as a strongly genetic disorder: evidence from a British twin

study. Psychological medicine, 25(1), 63-78.

Baird, G. et al., (2000). A screening instrument for autism at 18 months of age: a 6-year follow-up

study. Journal of the American Academy of Child & Adolescent Psychiatry, 39(6), 694-702.

Bakken, T. L. et al., (2010). Psychiatric disorders in adolescents and adults with autism and

intellectual disability: A representative study in one county in Norway. Research in

Developmental Disabilities, 31(6), 1669-1677.

Bargmann, C. I. (1998). Neurobiology of the Caenorhabditis elegans genome.Science, 282(5396),

2028-2033.

Bartolomei, M. S., Webber, A. L., Brunkow, M. E., & Tilghman, S. M. (1993). Epigenetic

mechanisms underlying the imprinting of the mouse H19 gene. Genes & development, 7(9),

1663-1673.

Bartolomei, M. S., Zemel, S., & Tilghman, S. M. (1991). Parental imprinting of the mouse H19 gene.

Nature, 351(6322), 153-155.

Barton, S.C. et al. (1984). Role of paternal and maternal genomes in mouse development. Nature.

311, 374-376.

128

Baylin, S. B. (2005). DNA methylation and gene silencing in cancer. Nature Clinical Practice

Oncology, 2, S4-S11.

Baxter, A. J. et al., (2014). The epidemiology and global burden of autism spectrum disorders.

Psychological medicine, 1-13.

Beddington, R.S. (1994). Induction of a second neural axis by the node. Development. 120(3), 613-

620.

Bell, A.C. & Felsenfeld, G. (2000). Methylation of a CTCF-dependent boundary controls imprinted

expression of the Igf2 gene. Nature. 405(6785), 462-485.

Bestor, T. H. (2000). The DNA methyltransferases of mammals. Human molecular genetics, 9(16),

2395-2402.

Bestor, T. H., & Ingram, V. M. (1983). Two DNA methyltransferases from murine erythroleukemia

cells: purification, sequence specificity, and mode of interaction with DNA. Proceedings of

the National Academy of Sciences, 80(18), 5559-5563.

Bestor, T., Laudano, A., Mattaliano, R., & Ingram, V. (1988). Cloning and sequencing of a cDNA

encoding DNA methyltransferase of mouse cells: the carboxyl-terminal domain of the

mammalian enzymes is related to bacterial restriction methyltransferases. Journal of

molecular biology, 203(4), 971-983.

Berger, S.L. (2007). The complex language of chromatin regulation during transcription. Nature.

447, 407-412.

Berkel, S. et al., (2010). Mutations in the SHANK2 synaptic scaffolding gene in autism spectrum

disorder and mental retardation. Nature genetics, 42(6), 489-491.

Bernstein, B. E. et al., (2002). Methylation of histone H3 Lys 4 in coding regions of active genes.

Proceedings of the National Academy of Sciences, 99(13), 8695-8700.

Bertani, S., Sauer, S., Bolotin, E., & Sauer, F. (2011). The Noncoding RNA< i> Mistral</i>

Activates< i> Hoxa6</i> and< i> Hoxa7</i> Expression and Stem Cell Differentiation by

Recruiting MLL1 to Chromatin. Molecular cell, 43(6), 1040-1046.

Bertone, P. et al., (2004). Global identification of human transcribed sequences with genome tiling

arrays. Science, 306(5705), 2242-2246.

Bertrand, J. et al., (2001). Prevalence of autism in a United States population: the Brick Township,

New Jersey, investigation. Pediatrics, 108(5), 1155-1161.

Bhardwaj, R. et al., (2006). Neocortical neurogenesis in humans is restricted to development. PNAS.

103(33), 12564-12568.

Bird, A. P. (2002). DNA methylation patterns and epigenetic memory. Genes Dev. 16(1), 6–21.

Bird, A.P. & Wolffe, A.P. (1999). Methylation-induced repression – belts, braces, and chromatin.

Cell. 99, 451-454.

Bittles, A. H. et al., (2002). The influence of intellectual disability on life expectancy. The Journals

of Gerontology Series A: Biological Sciences and Medical Sciences, 57(7), M470-M472.

Blagitko, N. et al., (2000). Human GRB10 is imprinted and expressed from the paternal and maternal

allele in a highly tissue-and isoform-specific fashion. Human Molecular Genetics, 9(11),

1587-1595.

Blank, V., Kourilsky, P., & Israël, A. (1992). NF-κB and related proteins: Rel/dorsal homologies

meet ankyrin-like repeats. Trends in biochemical sciences, 17(4), 135-140.

Bogdanović, O., & Veenstra, G. J. C. (2009). DNA methylation and methyl-CpG binding proteins:

developmental requirements and function. Chromosoma, 118(5), 549-565.

Bolton, P. et al., (1994). A case‐control family history study of autism. Journal of Child Psychology

and Psychiatry, 35(5), 877-900.

Botto, L. D. et al., (2003). A population-based study of the 22q11. 2 deletion: phenotype, incidence,

and contribution to major birth defects in the population. Pediatrics, 112(1), 101-107.

129

Bourch’his, D et al., (2001). Dnmt3L and the establishment of maternal genomic imprints. Science.

294(5551), 2536-2539.

Bourgeois, J.P. & Rakic, P. (1993). Changes of synaptic density in the primary visual cortex of the

macaque monkey from fetal to adult stage. The Journal of Neuroscience. 13(7), 2801-2820.

Bourgeois, J.P. (1997). Synaptogenesis, heterochrony and epigenesist in the mammalian neocortex.

86(S422), 27-33.

Bowley, C., & Kerr, M. (2000). Epilepsy and intellectual disability. Journal of Intellectual Disability

Research, 44(5), 529-543.

Bradley, E. A., Summers, J. A., Wood, H. L., & Bryson, S. E. (2004). Comparing rates of psychiatric

and behavior disorders in adolescents and young adults with severe intellectual disability

with and without autism. Journal of autism and developmental disorders, 34(2), 151-161.

Branco, M. R., Ficz, G., & Reik, W. (2011). Uncovering the role of 5-hydroxymethylcytosine in the

epigenome. Nature Reviews Genetics, 13(1), 7-13.

Brandeis, M. et al., (1993). The ontogeny of allele-specific methylation associated with imprinted

genes in the mouse. The EMBO Journal. 12(9), 3669-3677.

Brereton, A.V., Tonge, B.J. (2006). Psychopathology in children and adolescents with autism

compared to young people with intellectual disabiltiy. J Autism Dev. Discord. 36(7), 863-

870.

Brickley, S. G. et al., (2007). TASK-3 two-pore domain potassium channels enable sustained high-

frequency firing in cerebellar granule neurons. The Journal of Neuroscience, 27(35), 9329-

9340.

Bronner-Fraser, M. (1994). Neural crest cell formation and migration in the developing embryo. The

FASER Journal. 8(10), 699-706.

Brown, K., Park, S., Kanno, T., Franzoso, G., & Siebenlist, U. (1993). Mutual regulation of the

transcriptional activator NF-kappa B and its inhibitor, I kappa B-alpha. Proceedings of the

National Academy of Sciences, 90(6), 2532-2536.

Browning, S.R. & Browning, B.L. (2010). High-resolution detection of identity by descent in

unrelated individuals. The Am J Hum Gen. 86(4), 526-539.

Cameron. E.E. et al., (1999). Synergy of demethylation and histone deacetylase inhibition in the re-

expression of genes silenced in cancer. Nature Genetics. 21(1), 103-107.

Canfield, R. L. et al., (2003). Intellectual impairment in children with blood lead concentrations

below 10 μg per deciliter. New England Journal of Medicine,348(16), 1517-1526.

Cannon, T. D. et al., (2003). Early and late neurodevelopmental influences in the prodrome to

schizophrenia: contributions of genes, environment, and their interactions. Schizophrenia

Bulletin, 29(4), 653-669.

Capel, B. et al., (1993). Deletion of Y chromosome sequences located outside the testis determining

region can cause XY female sex reversal. Nature genetics, 5(3), 301-307.

Casey, B.J., Giedd, J.N., & Thomas, K.M. (2000). Structural and functional brain development and

its relation to cognitive development. Biologicla Psychology. 54(1), 241-257.

Casey, B. J., Tottenham, N., Liston, C., & Durston, S. (2005). Imaging the developing brain: what

have we learned about cognitive development?. Trends in cognitive sciences, 9(3), 104-110.

Carr, J. (2005). Stability and change in cognitive ability oRver the life span: a comparison of

populations with and without Down’s syndrome. 49(12), 915-928.

Catchpoole, D. et al., (1997). Epigenetic modification and uniparental inheritance of H19 in

Beckwith-Wiedemann syndrome. Journal of medical genetics, 34(5), 353-359.

Cattanach, B.M., Beechey, C.V., & Peters, J. (2004). Interactions between imprinting effects in the

mouse. Genetics Society of America. 168(1), 397-413.

130

Cavaillé, J. et al., (2002). Identification of tandemly-repeated C/D snoRNA genes at the imprinted

human 14q32 domain reminiscent of those at the Prader–Willi/Angelman syndrome region.

Human molecular genetics, 11(13), 1527-1538.

Caviness, V.S., Kennedy, D.N., Richelme, C., Rademacher, J., Filipek, P.A., (1996). The human

brain age 7–11 years: a volumetric analysis based on magnetic resonance images. Cerebral

Cortex 6(5), 726–736.

Chahrour, M. & Zoghibi, H.Y. (2007). The story of rett syndrome: from clinica to neurobiology.

Neuron. 56(3), 422-437.

Chaillet, J. R., Vogt, T. F., Beier, D. R., & Leder, P. (1991). Parental-specific methylation of an

imprinted transgene is established during gametogenesis and progressively changes during

embryogenesis. Cell, 66(1), 77-83.

Chamberlain, S.J. & Lalande, M. (2010). Neurodevelopmental disorders involving genomic

imprinting at human chromosome 15q11-q13. 39(1), 13-20.

Chakrabarti, S., & Fombonne, E. (2001). Pervasive developmental disorders in preschool

children. Jama, 285(24), 3093-3099.

Chakrabarti, S., & Fombonne, E. (2005). Pervasive developmental disorders in preschool children:

confirmation of high prevalence. American Journal of Psychiatry, 162(6), 1133-1141.

Chen, Z. et al., (1995). Signal-induced site-specific phosphorylation targets I kappa B alpha to the

ubiquitin-proteasome pathway. Genes & development, 9(13), 1586-1597.

Chen, P. et al., (2011). A comparative analysis of DNA methylation across human embryonic stem

cell lines. Genome Biology. 12, 1-12.

Chendrimada, T. P. et al., (2005). TRBP recruits the Dicer complex to Ago2 for microRNA

processing and gene silencing. Nature, 436(7051), 740-744.

Cheung, P., Allis, C. D., & Sassone-Corsi, P. (2000). Signaling to chromatin through histone

modifications. Cell, 103(2), 263-271.

Choufani, S. et al., (2011). A novel approach identifies new differentially methylated regions

(DMRs) associated with imprinted genes. Genome Research. 21(3), 465-476.

Chugani, H.T., Phelps, M.E., & Mazziotta, J.C. (1987). Positron emission tomopgraphy study of

human brain functional development. Annals of Neurology. 22(4), 487-497.

Clancy, B., Finlay, B. L., Darlington, R. B., & Anand, K. J. S. (2007). Extrapolating brain

development from experimental species to humans. Neurotoxicology, 28(5), 931-937.

Constância, M. et al., (2002). Placental-specific IGF-II is a major modulator of placental and fetal

growth. Nature, 417(6892), 945-948.

Constância, M et al., (2000). Deletion of a silencer element in Igf2 results in loss of imprinting

independent of H19. Nature. 26(2), 203-206.

Constância, M., Pickard, B., Kelsey, G., & Reik, W. (1998). Imprinting mechanisms. Genome

Research, 8(9), 881-900.

Constantino, J. N., & Todd, R. D. (2005). Intergenerational transmission of subthreshold autistic

traits in the general population. Biological psychiatry, 57(6), 655-660.

Cook Jr, E. H., & Scherer, S. W. (2008). Copy-number variations associated with neuropsychiatric

conditions. Nature, 455(7215), 919-923.

Coop, A.J., Greene, N.D.E., & Murdoch, J.N. (2003). The genetic basis of mammalian neurulation.

Nature Genetics. 4(10), 784-794.

Cooper, W. N. et al., (2005). Molecular subtypes and phenotypic expression of Beckwith–

Wiedemann syndrome. European journal of human genetics, 13(9), 1025-1032.

Courchesne, E., Carper, R., & Akshoomoff, N. (2003). Evidence of brain overgrowth in the first year

of life in autism. Jama, 290(3), 337-344.

131

Crackower, M. A. et al., (1996). Characterization of the split hand/split foot malformation locus

SHFM1 at 7q21. 3–q22. 1 and analysis of a candidate gene for its expression during limb

development. Human molecular genetics, 5(5), 571-579.

Creyghton, M. P. et al., (2010). Histone H3K27ac separates active from poised enhancers and

predicts developmental state. Proceedings of the National Academy of Sciences, 107(50),

21931-21936.

Curtin, C., Anderson, S. E., Must, A., & Bandini, L. (2010). The prevalence of obesity in children

with autism: a secondary data analysis using nationally representative data from the National

Survey of Children's Health. BMC pediatrics, 10(1), 11.

Dao, D. et al., (1998). IMPT1, an imprinted gene similar to polyspecific transporter and multi-drug

resistance genes. Human molecular genetics, 7(4), 597-608.

Danielsson, S. et al., (2005). Epilepsy in Young Adults with Autism: A Prospective Population‐based

Follow‐up Study of 120 Individuals Diagnosed in Childhood. Epilepsia, 46(6), 918-923.

Davies, W., Isles, A.R., & Wilkinson, L.S.(2001). Imprinted genes and mental dysfunction. Trends in

Molecular Medicine. 33(6), 428-436.

a) Davies, W. et al. (2004). Expression patterns of the novel imprinted genes Nap15 and PEG13 and

their non-imprinted host genes in the adult mouse brain. GEP. 4(6), 741-747.

b) Davies, W.et al. (2004). Imprinted gene expression in the brain. Neuroscience and Biobehavioural

Reviews. 29(3), 421-430.

Dawson, G., Osterling, J., Meltzoff, A. N., & Kuhl, P. (2000). Case study of the development of an

Infant with autism from birth to two years of age. Journal of Applied Developmental

Psychology, 21(3), 299-313.

Deb, S., Thomas, M., & Bright, C. (2001). Mental disorder in adults with intellectual disability. I:

prevalence of functional psychiatric illness among a community-based population aged

between 16 and 64 years. J Intel Disab Research. 45(6), 495-505.

Dekker, M. C., & Koot, H. M. (2003). < i> DSM-IV</i> Disorders in Children With Borderline to

Moderate Intellectual Disability. II: Child and Family Predictors. Journal of the American

Academy of Child & Adolescent Psychiatry, 42(8), 923-931.

DeLorey, T. M. et al., (1998). Mice lacking the β3 subunit of the GABAA receptor have the epilepsy

phenotype and many of the behavioral characteristics of Angelman syndrome. The Journal of

neuroscience, 18(20), 8505-8514.

DeVeale, B., Van Der Kooy, D., & Babak, T. (2012). Critical evaluation of imprinted gene

expression by RNA–Seq: a new perspective. PLoS genetics, 8(3), e1002600.

Dheda, K. et al., (2004). Validation of housekeeping genes for normalizing RNA expression in real-

time PCR. Biotechniques, 37, 112-119.

Diamond, M. C., Krech, D., & Rosenzweig, M. R. (1964). The effects of an enriched environment on

the histology of the rat cerebral cortex. Journal of Comparative Neurology, 123(1), 111-119.

Dinger, M. E. et al., (2008). Long noncoding RNAs in mouse embryonic stem cell pluripotency and

differentiation. Genome research, 18(9), 1433-1445.

Dörries, A., Spohr, H. L., & Kunze, J. (1988). Angelman (“happy puppet”) syndrome—seven new

cases documented by cerebral computed tomography: review of the literature. European

journal of pediatrics, 148(3), 270-273.

DuBose, A. J., et al., (2011). A new deletion refines the boundaries of the murine Prader–Willi

syndrome imprinting center. Human molecular genetics, 20(17), 3461-3466.

Duker, A. L. et al., (2010). Paternally inherited microdeletion at 15q11. 2 confirms a significant role

for the SNORD116 C/D box snoRNA cluster in Prader–Willi syndrome. European Journal

of Human Genetics, 18(11), 1196-1201.

132

Dupont, J. M., Tost, J., Jammes, H., & Gut, I. G. (2004). De novo quantitative bisulfite sequencing

using the pyrosequencing technology. Analytical biochemistry, 333(1), 119-127.

Duprat, F. et al., (1997). TASK, a human background K+ channel to sense external pH variations

near physiological pH. The EMBO Journal, 16(17), 5464-5471.

Duprat, F., Lauritzen, I., Patel, A., & Honoré, E. (2007). The TASK background K< sub> 2P</sub>

channels: chemo-and nutrient sensors. Trends in neurosciences, 30(11), 573-580.

Durand, C. M. et al., (2006). Mutations in the gene encoding the synaptic scaffolding protein

SHANK3 are associated with autism spectrum disorders. Nature genetics, 39(1), 25-27.

Durkin, M. S. et al., (2000). Prenatal and postnatal risk factors for mental retardation among children

in Bangladesh. American journal of epidemiology, 152(11), 1024-1033.

Dykens, E.M. (2000). Annotation: psychopathology in children with intellectual disability. J Child

Psychol Psychiat. 41(4), 407-417.

Dykens, E. M., Hodapp, R. M., Walsh, K., & Nash, L. J. (1992). Profiles, correlates, and trajectories

of intelligence in Prader-Willi syndrome. Journal of the American Academy of Child &

Adolescent Psychiatry, 31(6), 1125-1130.

Edelson, L. R., & Saudino, K. J. (2009). Genetic and environmental influences on autistic-like

behaviors in 2-year-old twins. Behavior genetics, 39(3), 255-264.

Egger, G, Liang, G., Aparicio, A. & Jones, P.A. (2004). Epigenetics in human disease and prospects

for epigenetic therapy. Nature. 429(6990), 457-463.

Einfeld, S. L., et al. (2011). Comorbidity of intellectual disability and mental disorder in children and

adolescents: A systematic review. J of Intellectual & Dev Disability. 36, 137-143.

Elliott, M., & Maher, E. R. (1994). Beckwith-Wiedemann syndrome. Journal of medical genetics,

31(7), 560.

Emerson, E. (2003). Prevalence of psychiatric disorders in children and adolescents with and without

intellectual disability. Journal of Intellectual Disability Research. 47(1), 51-58.

Emerson, E. (2004). Poverty and children with intellectual disabilities in the world’s richer countries.

J Intel Dev Disab. 29(4), 319-338.

Ensenauer, R. E. et al., (2003). Microduplication 22q11. 2, an emerging syndrome: clinical,

cytogenetic, and molecular analysis of thirteen patients. The American Journal of Human

Genetics, 73(5), 1027-1040.

Evans, H. K., Weidman, J. R., Cowley, D. O., & Jirtle, R. L. (2005). Comparative phylogenetic

analysis of Blcap/Nnat reveals eutherian-specific imprinted gene. Molecular Biology and

Evolution, 22(8), 1740-1748.

Evans, H. K., Wylie, A. A., Murphy, S. K., & Jirtle, R. L. (2001). The neuronatin gene resides in a

“micro-imprinted” domain on human chromosome 20q11. 2. Genomics, 77(1), 99-104.

Fakhrai‐Rad, H., Pourmand, N., & Ronaghi, M. (2002). Pyrosequencing™: an accurate detection

platform for single nucleotide polymorphisms. Human mutation, 19(5), 479-485.

Fantes, J. et al., (1995). Aniridia-associated cytogenetic rearrangements suggest that a position effect

may cause the mutant phenotype. Human molecular genetics, 4(3), 415-422.

Farkas, L.M. & Huttner, W.B. (2008). The cell biology of neural stem and progenitor cells and its

significance for their proliferation versus differentiation during mammalian brain

development. Curr Op Cell Biol. 20(6), 707-715.

Feng, S., Jacobsen, S. E., & Reik, W. (2010). Epigenetic reprogramming in plant and animal

development. Science, 330(6004), 622-627.

Feng, J., & Fan, G. (2009). The role of DNA methylation in the central nervous system and

neuropsychiatric disorders. International review of neurobiology, 89, 67-84.

Festenstein, R. et al., (1996). Locus control region function and heterochromatin-induced position

effect variegation. Science, 271(5252), 1123-1125.

133

Feuk, L., Marshall, C. R., Wintle, R. F., & Scherer, S. W. (2006). Structural variants: changing the

landscape of chromosomes and design of disease studies. Human molecular genetics,

15(suppl 1), R57-R66.

Finlay, B. L., & Darlington, R. B. (1995). Linked regularities in the development and evolution of

mammalian brains. Science, 268(5217), 1578-1584.

Fitzpatrick, G.V., Soloway, P.D., & Higgins, M.J. (2002). Regional loss of imprinting and growth

deficiency of mice with a targeted deletion of KvDMR1. Nature. 32, 426-431.

Fodor, B. D., Shukeir, N., Reuter, G., & Jenuwein, T. (2010). Mammalian Su (var) genes in

chromatin control. Annual review of cell and developmental biology, 26, 471-501.

Folstein, S., & Rutter, M. (1977). Infantile autism: a genetic study of 21 twin pairs. Journal of Child

psychology and Psychiatry, 18(4), 297-321.

Fournier, C. et al., (2002). Allele-specific histone lysine methylation marks regulatory regions at

imprinted mouse genes. EMBO. 21(23), 6560-6570.

Francks, C. et al., (2003). Parent-of-origin effects on handedness and schizophrenia susceptibility on

chromosome 2p12–q11. Human Molecular Genetics, 12(24), 3225-3230.

Francks, C. et al., (2007). LRRTM1 on chromosome 2p12 is a maternally suppressed gene that is

associated paternally with handedness and schizophrenia. Molecular psychiatry, 12(12),

1129-1139.

Fredman, D. et al., (2004). Complex SNP-related sequence variation in segmental genome

duplications. Nature genetics, 36(8), 861-866.

Freedman, R. et al. (1997). Linkage of a neurophysiological deficit in schizophrenia to a

chromosome 15 locus. Proc. Natl. Acad. Sci. U. S. A.94(2), 587–592.

Friedman, H. V., Bresler, T., Garner, C. C., & Ziv, N. E. (2000). Assembly of new individual

excitatory synapses: time course and temporal order of synaptic molecule recruitment.

Neuron, 27(1), 57-69.

Fuks, F. et al.,. (2003). The methyl-CpG-binding protein MeCP2 links DNA methylation to histone

methylation. Journal of Biological Chemistry, 278(6), 4035-4040.

Fuks, F. (2005). DNA methylation and histone modifications: teaming up to silence genes. Curr Op

Gen & Dev. 15(5), 490-495.

Gadow, K.D., DeVincent, C.J., Pomeroy, J. & Azizian, A. (2004). Psychiatric symptoms in

preschools children with PDD and clinic comparison samples. Journal of autism and

developmental disorders. 34(4), 379-393.

Gama-Sosa, M. A., Midgett, R. M., Slagel, V. A., Githens, S., Kuo, K. C., Gehrke, C. W., & Ehrlich,

M. (1983). Tissue-specific differences in DNA methylation in various mammals. Biochimica

et Biophysica Acta (BBA)-Gene Structure and Expression, 740(2), 212-219.

García-Cao, M., O'Sullivan, R., Peters, A. H., Jenuwein, T., & Blasco, M. A. (2003). Epigenetic

regulation of telomere length in mammalian cells by the Suv39h1 and Suv39h2 histone

methyltransferases. Nature genetics, 36(1), 94-99.

Gauthier, J. et al., (2011). Truncating mutations in NRXN2 and NRXN1 in autism spectrum

disorders and schizophrenia. Human genetics, 130(4), 563-573.

Ghaziuddin, M., Ghaziuddin, N., & Greden, J. (2002). Depression in persons with autism:

Implications for research and clinical care. Journal of autism and developmental disorders.

32(4), 299-306.

Gécz, J., Shoubridge, C., Corbett, M. (2009). The genetic landscape of intellectual disability arising

from chromosome X. Cell. 25(7), 308-316.

Gehring, M., Bubb, K. L., & Henikoff, S. (2009). Extensive demethylation of repetitive elements

during seed development underlies gene imprinting. Science, 324(5933), 1447-1451.

Georgiades, P. et al., (2001). Roles for genomic imprinting and the zygotic genome in placental

development. PNAS. 98(8), 4522-4527.

134

Gertz, J. et al., (2011). Analysis of DNA methylation in a three-generation family reveals widespread

genetic influence on epigenetic regulation. PLoS Genetics. 7(8), e1002228.

Giedd, J.N. et al., (1999). Brain Development during Childhood and Adolescence: A Longitudinal

MRI Study. Nat. Neurosci. 2(10), 861–863.

Gilad, R. Lampl, Y., Blumstein, G., Dan, M. (2007). Neurosyphilis: the reemergence of an historical

disease. Isr Med Assoc J. 9(2), 117-118.

Gillott, A., Furniss, F. & Walter, A. (2001). Anxiety of high-functioning children with Autism.

Autism. 5(3), 277-286.

Gimelbrant, A., Hutchinson, J. N., Thompson, B. R., & Chess, A. (2007). Widespread monoallelic

expression on human autosomes. Science, 318(5853), 1136-1140.

Ginsburg, M. A. L. K. A., Snow, M. H., & McLaren, A. N. N. E. (1990). Primordial germ cells in the

mouse embryo during gastrulation. Development,110(2), 521-528.

Girirajan, S. et al., (2012). Phenotypic heterogeneity of genomic disorders and rare copy-number

variants. New England Journal of Medicine, 367(14), 1321-1331.

Glessner, J. T. et al., (2009). Autism genome-wide copy number variation reveals ubiquitin and

neuronal genes. Nature, 459(7246), 569-573.

Goldstein, S. A., Bockenhauer, D., O'Kelly, I., & Zilberberg, N. (2001). Potassium leak channels and

the KCNK family of two-P-domain subunits.Nature Reviews Neuroscience, 2(3), 175-184.

Goldstein, S. A. et al., (2005). International Union of Pharmacology. LV. Nomenclature and

molecular relationships of two-P potassium channels. Pharmacological reviews, 57(4), 527-

540.

Goldzik, D. et al., (2013). Inference of identity by descent in population isolates and optimal

sequencing studies. European Journal of Human Genetics. 1-6.

Gottschling, D. E., Aparicio, O. M., Billington, B. L., & Zakian, V. A. (1990). Position effect at S.

cerevisiae telomeres: reversible repression of Pol II transcription. Cell, 63(4), 751-762.

Götz, M., Hartfuss, E., & Malatesta, P. (2002). Radial gila cells as neuronal precursors: A new

perspective on the correlation of morphology and lineage restriction in the developing

cerebral cortex of mice. Brain Res Bulletin. 57(6), 777-788.

Götz, M. & Huttner, W.B. (2005). The cell biology of neurogenesis. Nature Reviews: Molecular Cell

Biology. 6(10), 777-788.

Grandjean, P. & Landrigan, P.J. (2006). Developmental neurotoxicity of industrial chemicals. The

Lancet. DOI:10.1016/S0140-6736(06)69665-7.

Gregg, C. (2010a). High resoluation analysis of parent-of-origin allelic expression in the mouse

brain. 329(5992), 643-648.

Gregg, C. (2010b). Sex-specific parent-of-origin allelic expression in the mouse brain. Science.

329(5992), 682-685.

Greer, P. L. et al., (2010). The Angelman Syndrome protein Ube3A regulates synapse development

by ubiquitinating arc. Cell, 140(5), 704-716.

Grunstein, M. (1997). Histone acetylation in chromatin structure and transcription. Nature,

389(6649), 349-352.

Guo, J. U. et al., (2011). Hydroxylation of 5-methylcytosine by TET1 promotes active DNA

demethylation in the adult brain. Cell, 145(3), 423-434.

Guttman, M. et al., (2009). Chromatin signature reveals over a thousand highly conserved large non-

coding RNAs in mammals. Nature, 458(7235), 223-227.

Halacheva, V. et al., (2011). Planar cell movements and oriented cell division during early primitive

streak formation in the mammalian embryo. Developmental Dynamics, 240(8), 1905-1916.

Harris, J. C. (2006). Intellectual disability: Understanding its development, causes, classification,

evaluation, and treatment (pp. 11-41). New York: Oxford University Press.

135

Hastings, P. J., Lupski, J. R., Rosenberg, S. M., & Ira, G. (2009). Mechanisms of change in gene

copy number. Nature Reviews Genetics, 10(8), 551-564.

Hata, K., Okano, M, Lei, H. & Li, En. (2002). Dnmt3L cooperates with the Dnmt3 family of de novo

DNA methyltransferases to establish maternal imprints in mice. 129(8), 1983-1993.

Hatten, M.E. (1990). Riding the glial monorail: a common mechanism for glial-guided neuronal

migration in different regions of the developing mammalian brain. TINS. 13(5), 179-184.

Hebbes, T. R., Thorne, A. W., & Crane-Robinson, C. (1988). A direct linkbetween core histone

acetylation and transcriptionally active chromatin. The EMBO Journal, 7(5), 1395-1402.

Heard, E., Rougeulle, C., Arnaud, D., Avner, P., Allis, C.D. & Spector, D.L. (2001). Methylation of

histone H3 at Lys-9 is an early mark on the X chromosome during X inactivation. 107(6),

727-738.

Heikura, U. et al., (2003).Temporal changes in incidence and prevalence of intellectual disability

between two birth cohorts in Northern Finland. American Journal of Mental Retardation.

108(1), 19–31.

Heintzman, N. D. et al., (2009). Histone modifications at human enhancers reflect global cell-type-

specific gene expression. Nature, 459(7243), 108-112.

Heintzman, N. D. et al., (2007). Distinct and predictive chromatin signatures of transcriptional

promoters and enhancers in the human genome. Nature genetics, 39(3), 311-318.

Henckel, A. & Arnaud, P. (2010). Genome-wide identification of new imprinted genes. Briefings in

Functional Genomics. 9, 304-314.

Hendrich, B., & Bird, A. (1998). Identification and characterization of a family of mammalian

methyl-CpG binding proteins. Molecular and cellular biology, 18(11), 6538-6547.

Hoekstra, R. A., Bartels, M., Verweij, C. J., & Boomsma, D. I. (2007). Heritability of autistic traits in

the general population. Archives of Pediatrics & Adolescent Medicine, 161(4), 372-377.

Hoekstra, R.A., Happé, F., Baron-Cohen, S., & Ronald, A. (2009). Association between extreme

autistic traits and intellectual disability: insights from a general population twin study. BJP.

195(6), 531-536.

Hogart, A. et al., (2009). Chromosome 15q11-13 duplication syndrome brain reveals epigenetic

alterations in gene expression not predicted from copy number. 46(2), 86-93.

Holmes, E. E. et al., (2014). Performance Evaluation of Kits for Bisulfite-Conversion of DNA from

Tissues, Cell Lines, FFPE Tissues, Aspirates, Lavages, Effusions, Plasma, Serum, and Urine.

PloS one, 9(4), e93933.

Hou, J.-W., Wang, T.-R., &, Chuang, S.-M. (1998). An epidemiological and aetiological study of

children with intellectual disability in Taiwan. J Intel Disab Res. 42(2), 137-143.

Hou, C., Zhao, H., Tanimoto, K., & Dean, A. (2008). CTCF-dependent enhancer-blocking by

alternative chromatin loop formation. Proceedings of the National Academy of Sciences,

105(51), 20398-20403.

Howell, C. Y., Bestor, T. H., Ding, F., Latham, K. E., Mertineit, C., Trasler, J. M., & Chaillet, J. R.

(2001). Genomic Imprinting Disrupted by a Maternal Effect Mutation in the< i> Dnmt1</i>

Gene. Cell, 104(6), 829-838.

Hu, W.-H. et al., (2005). NIBP, a novel NIK and IKKB-binding protein that functions in NF-kB

activation. The J Biol Chem. 280(32), 29233-29241.

Hultman, C. M., Sparén, P., & Cnattingius, S. (2002). Perinatal risk factors for infantile

autism. Epidemiology, 13(4), 417-423.

Huttenlocher, P.R. (1979). Synaptic density in human frontal cortex. Developmental changes and

effects of aging. Brain Res. 163(2), 195-205.

Huttenlocher, P. R. (1990). Morphometric study of human cerebral cortex development.

Neuropsychologia, 28(6), 517-527.

136

Huttenlocher, P.R., de Courten, C., Garey, L.J., & van der Loos, H. (1982). Synaptogenesis in human

visual cortex – evidence for synapse elimination during normal development. Neuroscience

Letters. 33(3), 247-252.

Huttenlocher, P.R. & de Courten, C. (1987). The development of synapses in striate cortex in man.

Human Neurobiology. 6(1), 1-9.

Huttenlocher, P. R., & Dabholkar, A. S. (1997). Regional differences in synaptogenesis in human

cerebral cortex. Journal of comparative Neurology, 387(2), 167-178.

Hutter, B., Helms, V., & Paulse, M. (2006). Tandem repeats in the CpG islands of imprinted genes.

88(3), 323-332.

Illingworth, R. S. et al., (2010). Orphan CpG islands identify numerous conserved promoters in the

mammalian genome. PLoS genetics, 6(9), e1001134.

Illingworth, R. et al., (2008). A novel CpG island set identifies tissue-specific methylation at

developmental gene loci. PLoS biology, 6(1), e22.

Imanishi, T. et al., (2004). Integrative annotation of 21,037 human genes validated by full-length

cDNA clones. PLoS biology, 2(6), e162.

Imuta, Y. et al., (2014). Mechanical control of notochord morphogenesis by extra-embryonic tissues

in mouse embryos. Mechanisms of development, 132, 44-58.

Ingason, A. et al., (2011). Maternally derived microduplications at 15q11-q13: Implication of

imprinted genes in psychotic illness. Am J Psychiatry. 168(4), 408-417.

Iourov, I. Y. et al., (2012). Molecular karyotyping by array CGH in a Russian cohort of children with

intellectual disability, autism, epilepsy and congenital anomalies. Mol Cytogenet, 5, 46.

Isles, A.R. & Humby, T. (2006). Modes of imprinted gene action in learning disability. J of

Intellectual Disability Research. 50(5), 318-325.

Jaenisch, R. & Bird, A. (2003). Epigenetic regulation of gene expression: how the genome integrates

intrinsic and environmental signals. Nature Genetics. 33, 245-254.

Jamain, S. et al., (2003). Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4

are associated with autism. Nature genetics,34(1), 27-29.

Jamra, R.A. et al., (2011). Homozygosity mapping in 64 Syrian consanguineous families with non-

specific intellectual disability reveals 11 novel loci and high heterogeneity. 1-6,

doi:10.1038/ejhg.2011.98.

Janicki, P., Arthur J. Dalton, C. Michael Henderson, Philip W. Davidson, M. (1999). Mortality and

morbidity among older adults with intellectual disability: health services considerations.

Disability & Rehabilitation, 21(5-6), 284-294.

Jay, V., Becker, L. E., Chan, F. W., & Perry, T. L. (1991). Puppet‐like syndrome of Angelman A

pathologic and neurochemical study. Neurology, 41(3), 416-416.

Jenuwein, T., & Allis, C. D. (2001). Translating the histone code. Science, 293(5532), 1074-1080.

Jones, P. A., & Liang, G. (2009). Rethinking how DNA methylation patterns are maintained. Nature

Reviews Genetics, 10(11), 805-811.

Jones, P. L. et al., (1998). Methylated DNA and MeCP2 recruit histone deacetylase to repress

transcription. Nature genetics, 19(2), 187-191.

Kakar, N. et al., (2012). A homozygous splice site mutation in TRAPPC9 causes intellectual

disability and microcephaly. Eur J Med Gen. 55, 727-731.

Kaneda, M., Okano, M., Hata, K., Sado, T., Tsujimoto, N., Li, E., & Sasaki, H. (2004). Essential role

for de novo DNA methyltransferase Dnmt3a in paternal and maternal

imprinting. Nature, 429(6994), 900-903.

Kato, Y., Kaneda, M., Hata, K., Kumaki, K., Hisano, M., Kohara, Y., ... & Sasaki, H. (2007). Role of

the Dnmt3 family in de novo methylation of imprinted and repetitive sequences during male

germ cell development in the mouse.Human molecular genetics, 16(19), 2272-2280.

137

Katz, G., & Lazcano-Ponce, E. (2008). Intellectual disability: definition, etiological factors,

classification, diagnosis, treatment and prognosis. Salud Pública de México, 50, s132-s141.

Karschin, C. et al., (2001). Expression Pattern in Brain of TASK-1, TASK-3, and a Tandem Pore

Domain K< sup>+</sup> Channel Subunit, TASK-5, Associated with the Central Auditory

Nervous System.Molecular and Cellular Neuroscience, 18(6), 632-648.

Kaufman, L. et al. (2011, October). Evidence of imprinting at the autosomal recessive intellectual

disability gene, TRAPPC9, and a possible role for CNVs at this locus in autism. Poster

presented at the meeting of The American Society of Human Genetics, Montreal, Canada.

Kenneth, D., Gadow, C.J. & DeVincent, J.P. (2006). ADHD symptom subtypes in children with

pervasive developmental disorder. Journal of Autism and Developmental Disorders. 36(2),

271-283.

Keverne, E. B. (2001). Genomic imprinting, maternal care, and brain evolution. Hormones and

behavior, 40(2), 146-155.

Khalil, A. M. et al., (2009). Many human large intergenic noncoding RNAs associate with

chromatin-modifying complexes and affect gene expression. Proceedings of the National

Academy of Sciences, 106(28), 11667-11672.

Kim, Y., Bang, H., & Kim, D. (2000). TASK-3, a new member of the tandem pore K+ channel

family. Journal of Biological Chemistry, 275(13), 9340-9347.

Kim, J., Szatmari, P., Bryson, S., Streiner, D. & Wilson, F. (2000). The prevalence of anxiety of

mood problems among Autism and Asperger Syndrome. Autism. 4(2), 117-132.

Kim, K., Doi, A., Wen, B., Ng, K., Zhao, R. et al., (2010). Epigenetic memory in induced pluripotent

stem cells. Nature. 467(7313), 285-290.

Kim, T.H. et al., (2007). Analysis of the vertebrate insulator protein CTCF binding sites in the human

genome. Cell. 128(6), 1231-1245.

Kim, T. K. et al., (2010). Widespread transcription at neuronal activity-regulated enhancers. Nature,

465(7295), 182-187.

Kim, J., Lu, X., & Stubbs, L. (1999). Zim1, a maternally expressed mouse Kruppel-type zinc-finger

gene located in proximal chromosome 7. Human molecular genetics, 8(5), 847-854.

Kitsberg, D. et al., (1993). Allele-specific replication timing of imprinted gene regions. Nature.

364(6436), 459-463.

Kleinjan, D. J., & Van Heyningen, V. (1998). Position effect in human genetic disease. Human

molecular genetics, 7(10), 1611-1618.

Kleinjan, D. A., & Van Heyningen, V. (2005). Long-range control of gene expression: emerging

mechanisms and disruption in disease. The American Journal of Human Genetics, 76(1), 8-

32.

Knoll, J. H. M. et al., (1989). Angelman and Prader‐Willi syndromes share a common chromosome

15 deletion but differ in parental origin of the deletion. American journal of medical genetics,

32(2), 285-290.

Koifman, A. et al., (2009). A homozygous deletion of 8q24.3 including the NIBP gene associated

with severe developmental delay, dysgenesis of the corpus callosum, and dysmorphic facial

features. Am J Med Gen. 152(5), 1268-1272.

Komura, D. et al., (2006). Genome-wide detection of human copy number variations using high-

density DNA oligonucleotide arrays. Genome research, 16(12), 1575-1584.

Koop, M., Rilling, G., Herrmann, A., & Kretschmann, H. J. (1986). Volumetric development of the

fetal telencephalon, cerebral cortex, diencephalon, and rhombencephalon including the

cerebellum in man. Bibl Anat, 28, 53-78.

Koopman, P. et al., (1991). Male development of chromosomally female mice transgenic for Sry.

Nature, 351(6322), 117-121.

138

Kotzot, D. et al., (1995). Uniparental disomy 7 in Silver—Russell syndrome and primordial growth

retardation. Human molecular genetics, 4(4), 583-587.

Korostowski, L., Sedlak, N., & Engel, N. (2012). The Kcnq1ot1 long non-coding RNA affects

chromatin conformation and expression of Kcnq1, but does not regulate its imprinting in the

developing heart. PLoS genetics, 8(9), e1002956.

Koulich, E. et al., (2001). NF‐κB is involved in the survival of cerebellar granule neurons:

association of Iκβ phosphorylation with cell survival. Journal of neurochemistry, 76(4),

1188-1198.

Kouzarides, T. (2007). Chromatin modifications and their function. Cell, 128(4), 693-705.

Krakowiak, P. et al., (2012). Maternal metabolic conditions and risk for autism and other

neurodevelopmental disorders. Pediatrics, 129(5), e1121-e1128.

Kretschmann, H.J., Kammradt, G., Krauthausen, I., Sauer, B., Wingert, F., 1986. Brain growth in

man. Biblio. Anat. 28, 1–26.

Kurukuti, S. et al., (2006). CTCF binding at the H19 imprinting control region mediates maternally

inherited higher-order chromatin conformation to restrict enhancer access to Igf2.

Proceedings of the National Academy of Sciences, 103(28), 10684-10689.

La Malfa, G. et al., (2004). Autism and intellectual disability: a prevalence on a sample of the Italian

population. J Intel Disab Res. 48(3), 262-267.

Labrador, M., & Corces, V. G. (2002). Setting the boundaries of chromatin domains and nuclear

organization. Cell, 111(2), 151-154.

Langer, E.S. & Botstein, D. (1987). Homozygosity mapping: a way to map human recessive traits

with the DNA of inbred children. Science. 236(4808), 1567-1570.

Larsen, F., Solheim, J., & Prydz, H. (1993). A methylated CpG island 3'in the apolipoprotein-E gene

does not repress its transcription. Human molecular genetics, 2(6), 775-780.

Laumonnier, F. et al., (2004). X-Linked Mental Retardation and Autism Are Associated with a

Mutation in the< i> NLGN4</i> Gene, a Member of the Neuroligin Family. The American

Journal of Human Genetics, 74(3), 552-557.

Lee, J. T. (2009). Lessons from X-chromosome inactivation: long ncRNA as guides and tethers to the

epigenome. Genes & development, 23(16), 1831-1842.

Lee, J. et al., (2002). Erasing genomic imprinting memory in mouse clone embryos produced from

day 11.5 primordial germ cells. Development, 129(8), 1807-1817.

Lee, J. A. et al., (2006). Spastic paraplegia type 2 associated with axonal neuropathy and apparent

PLP1 position effect. Annals of neurology, 59(2), 398-403.

Lees-Murdock, D. J., De Felici, M., & Walsh, C. P. (2003). Methylation dynamics of repetitive DNA

elements in the mouse germ cell lineage.Genomics, 82(2), 230-237.

Lefebvre, L. et al., (1998). Abnormal maternal behaviour and growth retardation associated with loss

of the imprinted gene Mest. Nature genetics, 20(2), 163-169.

Lehnertz, B. et al., (2003). Suv39h-mediated histone H3 lysine 9 methylation directs DNA

methylation to major satellite repeats at pericentric heterochromatin. 13(14), 1192-1200.

Leonard, H.&X. Wen. (2002). The epidemiology of mental retardation: challenges and opportunities

in the new millennium. Ment Retard Dev Disabil Res Rev. 8(3), 117-34.

Leonard, H. et al., (2011). Autism and intellectual disability are differentially related to

sociodemographic background at birth. PLoS ONE. 6(3), e17875.

Leonard, H., de Klerk, N., Bourke, J., & Bower, C. (2006). Maternal health in pregnancy and

intellectual disability in the offspring: a population-based study.Annals of

epidemiology, 16(6), 448-454.

Lepage, J., et al., (2012). Genomic imprinting effects on cognitive and social abilities in prepubertal

girls with turner syndrome. J Clin Endocrinol Metab. 97(3), e460-e464.

139

Lesage, F., & Lazdunski, M. (2000). Molecular and functional properties of two-pore-domain

potassium channels. American Journal of Physiology-Renal Physiology, 279(5), F793-F801.

Lesage, F. (2003). Pharmacology of neuronal background potassium

channels.Neuropharmacology, 44(1), 1-7.

Leventer, R.J., Phelan, E.M., & L.T. Coleman. (1999). Clinical and imaging features of cortical

malformations in childhood. Neurology. 53(4), 709-722.

Leyfer, O.T. et al., (2006). Comorbid psychiatric disorders in children with autism: interview

development and rates of disorders. Journal of Autism and Developmental Disorders. 36(7),

849-861.

Li, E., Bestor, T. H., & Jaenisch, R. (1992). Targeted mutation of the DNA methyltransferase gene

results in embryonic lethality. Cell, 69(6), 915-926.

Li, E., Beard, C., & Jaenisch, R. (1993). Role for DNA methylation in genomic

imprinting. Nature, 366(6453), 362-365.

Li, T. et al., (2008). CTCF regulates allelic expression of Igf2 by orchestrating a promoter-polycomb

repressive complex 2 intrachromosomal loop. Mol Cell Bio, 28(20), 6473-6482.

Li, L. L. et al., (1999). Regulation of maternal behavior and offspring growth by paternally expressed

Peg3. Science, 284(5412), 330-334.

Li, Y., & Tollefsbol, T. O. (2011). DNA methylation detection: bisulfite genomic sequencing

analysis. In Epigenetics Protocols (pp. 11-21). Humana Press.

Li, T. et al., (2000). Tissue-specific expression of antisense and sense transcripts at the imprinted

Gnas locus. Genomics, 69(3), 295-304.

Lillien, L. (1997). Neural development: instructions for neural diversity 7.3. Curr Biol. 7(3), R168-

R171.

Lin, S. P. et al., (2003). Asymmetric regulation of imprinting on the maternal and paternal

chromosomes at the Dlk1-Gtl2 imprinted cluster on mouse chromosome 12. Nature

genetics, 35(1), 97-102.

Lister, R. et al., (2013). Global epigenomic reconfiguration during mammalian brain development.

Science, 341(6146).

Lister, R., Pelizzola, M., Dowen, R. H., Hawkins, R. D., Hon, G., Tonti-Filippini, J., ... & Ecker, J.

R. (2009). Human DNA methylomes at base resolution show widespread epigenomic

differences. nature, 462(7271), 315-322.

Livak, K. J., & Schmittgen, T. D. (2001). Analysis of Relative Gene Expression Data Using Real-

Time Quantitative PCR and the 2< sup>− ΔΔCT</sup> Method. methods, 25(4), 402-408.

Loesch, D. Z., Huggins, R. M., Bui, Q. M., Epstein, J. L., Taylor, A. K., & Hagerman, R. J. (2002).

Effect of the deficits of fragile X mental retardation protein on cognitive status of fragile X

males and females assessed by robust pedigree analysis. Journal of Developmental &

Behavioral Pediatrics, 23(6), 416-423.

LoVullo, S. V., & Matson, J. L. (2009). Comorbid psychopathology in adults with autism spectrum

disorders and intellectual disabilities. Research in Developmental Disabilities, 30(6), 1288-

1296.

Lubs, H. A., Stevenson, R. E., & Schwartz, C. E. (2012). Fragile X and X-linked intellectual

disability: four decades of discovery. AJHG, 90(4), 579-590.

Luger, K., Mäder, A. W., Richmond, R. K., Sargent, D. F., & Richmond, T. J. (1997). Crystal

structure of the nucleosome core particle at 2.8 Å resolution. Nature, 389(6648), 251-260.

Lyle, R. et al., (2000). The imprinted antisense RNA at the Igf2r locus overlaps but does not imprint

Mas1. Nature genetics, 25(1), 19-21.

Mancini-DiNardo, D. et al., (2006). Elongation of the Kcnq1ot1 transcript is required for genomic

imprinting of neighboring genes. Genes & Development, 20(10), 1268-1282.

140

Mackarehtschian, K., Lau, C. K., Caras, I., & McConnell, S. K. (1999). Regional differences in the

developing cerebral cortex revealed by ephrin-A5 expression. Cerebral Cortex, 9(6), 601-

610.

Magnani, M., Crinelli, R., Bianchi, M., & Antonelli, A. (2000). The ubiquitin-dependent proteolytic

system and other potential targets for the modulation of nuclear factor-kB (NF-kB). Current

drug targets, 1(4), 387-399.

Malatesta, P., Harfuss, E., & Magdalena, G. (2000). Isolation of radial glial cells by fluorescent-

activated cell soring reveals a neuronal lineage. Development. 127(24), 5253-5263.

Malik, S. et al., (2013). Neurogenesis continues in the third trimester of pregnancy and is suppressed

by premature birth. The Journal of Neuroscience, 33(2), 411-423.

Malzac, P. et al., (1998). Mutation analysis of UBE3A in Angelman syndrome patients. The

American Journal of Human Genetics, 62(6), 1353-1360.

Maestro, S. et al., (2002). Attentional skills during the first 6 months of age in autism spectrum

disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 41(10), 1239-

1245.

Matson, J. L., & Kozlowski, A. M. (2011). The increasing prevalence of autism spectrum disorders.

Research in Autism Spectrum Disorders, 5(1), 418-425.

Matson, J.L. & Shoemaker, M. (2009). Intellectual disability and its relationship to autism spectrum

disorders (2009). Research in Dev Disabilities. 30(6), 1107-1114.

Mattson, M.P. & Camandola, S. (2001). NF-kB in neuronal plasticity and neurodegenerative

disorders. J Clin Inves. 107(3), 247-254.

Marangi, G. et al., (2012). TRAPPC9-related autosomal recessive intellectual disability: report of a

new mutation and clinical phenotype. Eur J Hum Gen. 21(2), 1-4.

Maulik, P. K., & Darmstadt, G. L. (2007). Childhood disability in low-and middle-income countries:

overview of screening, prevention, services, legislation, and epidemiology. Pediatrics,

120(Supplement 1), S1-S55.

Maulik, P.K. et al., (2010). Prevalence of intellectual disability: A meta-analysis of population-based

studies. Research in Dev Disabilities. doi:10.1016/j.ridd.2010.12.018.

Maunakea, A. K. et al., (2010). Conserved role of intragenic DNA methylation in regulating

alternative promoters. Nature, 466(7303), 253-257.

Martins-Taylor, K. et al., (2013). Imprinted expression of UBE3A in non-neuronal cells from a

Prader-Willi syndrome patient with an atypical deletion. Human molecular genetics, ddt628.

Mayer, W. et al., (2000). Embryogenesis:demethylation of the zygotic paternal genome. Nature,

403(6769), 501-502.

Mayer, W. et al., (2000). Expression of the imprinted genes MEST/Mest in human and murine

placenta suggests a role in angiogenesis. Developmental dynamics, 217(1), 1-10.

McCarthy, J., Hemmings, C., Kravariti, E., Dworzynski, K., Holt, G., Bouras, N., & Tsakanikos, E.

(2010). Challenging behavior and co-morbid psychopathology in adults with intellectual

disability and autism spectrum disorders. Research in Developmental Disabilities, 31(2),

362-366.

McConnell, S.K. & Kaznowski, C.E. (1991). Cell cycle dependence of laminar determination in

developing neocortex. Science. 254(5029), 282-285.

McConnell, S.K. (1995). Constructing the cerebral cortex: neurogenesis and fate determination.

Neuron. 15(4), 761-768.

McDougle, C.J., Kresch, K.E., Goodman, W.K, Naylor, S.T. et al., (1995). A case-controlled study

of repetitive thoughts and behavior in adults with autistic disorder and obsessive-compulsive

disorder. The American Journal of Psychiatry. 152(5), 772-777.

McGrath, J. & Solter, D. (1984). Completion of mouse embryogenesis requires both the maternal and

paternal genomes. Cell. 37(1), 179-183.

141

Meffert, M.K. et al., (2003). NF-kB functions in synaptic signaling and behaviour. Nature. 6(10),

1072-1329.

Mefford, H.C. et al., (2012). Genomics, intellectual disability, and autism. N Engl J Med. 366(8),

733-743.

Melville, C. A., Hamilton, S., Hankey, C. R., Miller, S., & Boyle, S. (2007). The prevalence and

determinants of obesity in adults with intellectual disabilities.Obesity Reviews, 8(3), 223-230.

Melom, J.E. & Littleton, J.T. (2011). Synapse development in health and disease. Curr Op Gen &

Dev. 21(3), 256-261.

Mercadante, M. T., Evans-Lacko, S., & Paula, C. S. (2009). Perspectives of intellectual disability in

Latin American countries: epidemiology, policy, and services for children and adults.

Current Opinion in Psychiatry, 22(5), 469-474.

Mercer, T. R., Dinger, M. E., Sunkin, S. M., Mehler, M. F., & Mattick, J. S. (2008). Specific

expression of long noncoding RNAs in the mouse brain. Proceedings of the National

Academy of Sciences, 105(2), 716-721.

Mercer, T. R., Qureshi, I. A., Gokhan, S., Dinger, M. E., Li, G., Mattick, J. S., & Mehler, M. F.

(2010). Long noncoding RNAs in neuronal-glial fate specification and oligodendrocyte

lineage maturation. BMC neuroscience, 11(1), 14.

Mikkelsen, T. S. et al., (2007). Genome-wide maps of chromatin state in pluripotent and lineage-

committed cells. Nature, 448(7153), 553-560.

Mir, A. et al., (2009). Identification of mutations in TRAPPC9, which encodes the NIK- and IKK-B-

Binding protein, in nonsyndromic autosomal-recessive mental retardation. Am J Hum Gen.

85(6), 909-915.

Millar, J. A. et al., (2000). A functional role for the two-pore domain potassium channel TASK-1 in

cerebellar granule neurons. Proceedings of the National Academy of Sciences, 97(7), 3614-

3618.

Miller, L., Angulo, M., Price, D., & Taneja, S. (1996). MR of the pituitary in patients with Prader-

Willi syndrome: size determination and imaging findings. Pediatric radiology, 26(1), 43-47.

Miller, J. L. et al., (2007). Intracranial abnormalities detected by three‐dimensional magnetic

resonance imaging in Prader–Willi syndrome. American Journal of Medical Genetics Part A,

143(5), 476-483.

Milosavljevic, A. (2011). Emerging patterns of epigenomic variation. Trends in Genetics, 27(6), 242-

250.

Mitchell, K.J. (2011). The genetics of neurodevelopmental disease. Curr Op Neurobiol. 21(1), 197-

203.

Mitchison, T., & Kirschner, M. (1988). Cytoskeletal dynamics and nerve growth. Neuron, 1(9), 761-

772.

Mizuno, Y. et al., (2002). Asb4, Ata3, and Dcn are novel imprinted genes identified by high-

throughput screening using RIKEN cDNA microarray. Biochemical and biophysical research

communications, 290(5), 1499-1505.

Mochida, G. H. et al., (2009). A truncating mutation of TRAPPC9 is associated with autosomal-

recessive intellectual disability and postnatal microcephaly. Am J Hum Gen. 85(6), 897-902.

Mohammad, F., Mondal, T., & Kanduri, C. (2009). Epigenetics of imprinted long noncoding RNAs.

Epigenetics. 4(5), 277-286.

Monk, D. et al., (2006). Limited evolutionary conservation of imprinting in the human placenta.

Proceedings of the National Academy of Sciences, 103(17), 6623-6628.

Monk, M., Boubelik, M., & Lehnert, S. (1987). Temporal and regional changes in DNA methylation

in the embryonic, extraembryonic and germ cell lineages during mouse embryo development.

Development. 99(3), 371-382.

142

Moore, T. et al., (1997). Multiple imprinted sense and antisense transcripts, differentia methylation

and tandem repeats in a putative imprinting control region upstream of mouse Igf2. PNAS.

94(23), 12509-12514.

Moore, T., & Haig, D. (1991). Genomic imprinting in mammalian development: a parental tug-of-

war. Trends in Genetics, 7(2), 45-49.

Morcos, L. (2011). Genome-wide assessment of imprinted expression in human cells. Genome

Biology. 12(3), R25.

Morison, I.M., Ramsay, J.P., & Spencer, H.G. (2005). A census of mammalian imprinting. TRENDS

in Genetics. 21(8), 457-465.

Morita, S. et al. (2007). One argonaute family member, Eif2c2 (Ago2), is essential for development

and appears not to be involved in DNA methylation. Genomics. 89(6), 687-696.

Mulkey, D. K. et al., (2007). TASK channels determine pH sensitivity in select respiratory neurons

but do not contribute to central respiratory chemosensitivity. The Journal of

Neuroscience, 27(51), 14049-14058.

Müller, F. & O’Rahilly, R. (1987). The development of the human brain, the closure of the caudal

neuropore, and the beginning of the secondary neurulation at stage 12. Anat Embryol. 176(4),

413-430.

Müller, F. & O’Rahilly, R. (1988). The development of the human brain from a closed neural tube at

stage 13. Anat Embryol. 177(3), 203-224.

Murrell, A., Heeson, S., & Reisk, W. (2004). Interaction between differentially methylated regions

partition the imprinted genesIgf2 and H19 into parent-specific chromatin loops. Nature.

36(8), 889-893.

Nan, X. et al., (1999). Transcriptional repression by the methyl-CpG-binding protein MeCP2

involves a histone deacetylase complex. Nature. 393(6683), 386-389.

Najmabadi, H. et al., (2007). Homozygosity mapping in consanguineous families reveals extreme

heterogeneity of non-syndromic autosomal recessive mental retardation and identifies 8 novel

gene loci. Hum Genet. 121(1), 43-48.

Nicholls, R.D. (2000). The impact of genomic imprinting for neurobehavioral and developmental

disorders. The Journal of Clinical Investigation, 105(4), 413-418.

Niell, C. M., Meyer, M. P., & Smith, S. J. (2004). In vivo imaging of synapse formation on a

growing dendritic arbor. Nature neuroscience, 7(3), 254-260.

Niemitz, E. L. et al., (2004). Microdeletion of< i> LIT1</i> in Familial Beckwith-Wiedemann

Syndrome. The American Journal of Human Genetics, 75(5), 844-849.

Noctor, S. C. et al., (2001). Neurons derived from radial glial cells establish radial units in neocortex.

Nature, 409(6821), 714-720.

Noctor, S. C. et al., (2002). Dividing precursor cells of the embryonic cortical ventricular zone have

morphological and molecular characteristics of radial glia. The Journal of Neuroscience,

22(8), 3161-3173.

Noor, A. et al., (2010). Disruption at the PTCHD1 Locus on Xp22. 11 in Autism spectrum disorder

and intellectual disability. Science translational medicine, 2(49), 49ra68-49ra68.

Odeberg, J., et al., (2007). Late human cytomegalovirus (HCMV) proteins inhibit differentiation of

human neural precursor cells into astrocytes. J Neurosci Res. 85(3), 583-593

O'Dwyer, J. M. (1997). Schizophrenia in people with intellectual disability: The role of pregnancy

and birth complications. Journal of Intellectual Disability Research, 41(3), 238-251.

Okamoto, I., Otte, A.P., Allis, C.D, Reinberg, D. & Heard, E. (2004). Epigenetic dynamics of

imprinted X inactivation during early mouse development. Science. 303(5658), 644-649.

Okano, M., Bell, D. W., Haber, D. A., & Li, E. (1999). DNA methyltransferases Dnmt3a and

Dnmt3b are essential for de novo methylation and mammalian development. Cell, 99(3),

247-257.

143

Okano, M., Xie, S., & Li, E. (1998). Cloning and characterization of a family of novel mammalian

DNA (cytosine-5) methyltransferases. Nature genetics, 19(3), 219-220.

Okita, C. et al., (2003). A new imprinted cluster on the human chromosome 7q21-q31, identified by

human-mouse monochromosomal hybrids. Genomics, 81(6), 556-559.

O'Neill, L. A., & Kaltschmidt, C. (1997). NF-kB: a crucial transcription factor for glial and neuronal

cell function. Trends in neurosciences, 20(6), 252-258.

O’Rahilly, R. & Gardner, E. (1979). The initial development of the human brain. Acta anat. 104(2),

123-133.

Ørom, U. A., et al., (2010). Long noncoding RNAs with enhancer-like function in human cells. Cell,

143(1), 46-58.

Osborne, L. R. et al., (2001). A 1.5 million–base pair inversion polymorphism in families with

Williams-Beuren syndrome. Nature genetics, 29(3), 321-325.

Oswald, J. et al., (2000). Active demethylation of the paternal genome in the mouse zygote. Current

Biology, 10(8), 475-478.

O'Sullivan, J. M. et al., (2004). Gene loops juxtapose promoters and terminators in yeast. Nature

genetics, 36(9), 1014-1018.

Ota, T. et al., (2003). Complete sequencing and characterization of 21,243 full-length human cDNAs.

Nature genetics, 36(1), 40-45.

Ørom, U. A., & Shiekhattar, R. (2011). Long non-coding RNAs and enhancers. Current opinion in

genetics & development, 21(2), 194-198.

Pàldi, A., Gyapay, G., & Jami, J. (1995). Imprinted chromosomal regions of the human genome

display sex-specific meiotic recombination frequencies. Current Biology. 5(9), 1030-1035.

Pandey, R. R. et al., (2008). < i> Kcnq1ot1</i> Antisense Noncoding RNA Mediates Lineage-

Specific Transcriptional Silencing through Chromatin-Level Regulation. Molecular

cell, 32(2), 232-246.

Patel, V. et al., (2007). Treatment and prevention of mental disorders in low-income and middle-

income countries. The Lancet, 370(9591), 991-1005.

Paulsen, M., & Ferguson‐Smith, A. C. (2001). DNA methylation in genomic imprinting,

development, and disease. The Journal of pathology, 195(1), 97-110.

Petanjek, Z. et al., (2011). Extraordinary neoteny of synaptic spines in the human prefrontal cortex.

PNAS. 108(32), 13281-13286.

Peters, A. H. et al., (2001). Loss of the< i> Suv39h</i> Histone Methyltransferases Impairs

Mammalian Heterochromatin and Genome Stability. Cell, 107(3), 323-337.

Peterson, C. L., & Laniel, M. A. (2004). Histones and histone modifications. Current Biology,

14(14), R546-R551.

Pfeifer, K. (2000). Mechanisms for genomic imprinting. Am J Hum Gen. 67(4), 777-787.

Pham, N. V. et al., (1998). Dissociation of< i> IGF2</i> and< i> H19</i> imprinting in human brain.

Brain research, 810(1), 1-8.

Phillipe et al., (2009). Combination of linkage mapping and microarray-expression analysis identifies

NF-KB signaling defect as a cause of autosomal-recessive mental retardation. Am J Hum

Gen. 85(6), 903-908.

Phillips, J. E., & Corces, V. G. (2009). CTCF: master weaver of the genome. Cell, 137(7), 1194-

1211.

Ping, A. J. et al., (1989). Genetic linkage of Beckwith-Wiedemann syndrome to 11p15. American

journal of human genetics, 44(5), 720.

Pinto, L., & Götz, M. (2007). Radial glial cell heterogeneity—the source of diverse progeny in the

CNS. Progress in neurobiology, 83(1), 2-23.

144

Piven, J., Palmer, P., Jacobi, D., Childress, D., & Arndt, S. (1997). Broader autism phenotype:

evidence from a family history study of multiple-incidence autism families. American

Journal of Psychiatry, 154(2), 185-190.

Ponjavic, J., Oliver, P. L., Lunter, G., & Ponting, C. P. (2009). Genomic and transcriptional co-

localization of protein-coding and long non-coding RNA pairs in the developing brain. PLoS

genetics, 5(8), e1000617.

Potocki, L. et al., (2007). Characterization of Potocki-Lupski syndrome (dup (17)(p11. 2p11. 2)) and

delineation of a dosage-sensitive critical interval that can convey an autism phenotype. The

American Journal of Human Genetics, 80(4), 633-649.

Qian, X. et al., (2000). Timing of CNS cell generation: a programmed sequence of neuron and glial

cell production from isolated murine cortical stem cells. Neuron, 28(1), 69-80.

Qian, X. et al., (1998). Intrinsic programs of patterned cell lineages in isolated vertebrate CNS

ventricular zone cells. Development. 125(16), 3143-3152.

Qureshi, I. A., Mattick, J. S., & Mehler, M. F. (2010). Long non-coding RNAs in nervous system

function and disease. Brain research, 1338, 20-35.

Radonić, A. et al., (2004). Guideline to reference gene selection for quantitative real-time PCR.

Biochemical and biophysical research communications, 313(4), 856-862.

Rafiq, M. A. et al., (2011). Mutations in the Alpha 1, 2-Mannosidase Gene,< i> MAN1B1</i>, Cause

Autosomal-Recessive Intellectual Disability. The American Journal of Human Genetics,

89(1), 176-182.

Raizis, A. M., Schmitt, F., & Jost, J. P. (1995). A bisulfite method of 5-methylcytosine mapping that

minimizes template degradation. Analytical biochemistry, 226(1), 161-166.

Rajan, S. et al., (2002). Interaction with 14‐3‐3 proteins promotes functional expression of the

potassium channels TASK‐1 and TASK‐3. The Journal of physiology, 545(1), 13-26.

Rajan, S. et al., (2000). TASK-3, a Novel Tandem Pore Domain Acid-sensitive K+ Channel AN

EXTRACELLULAR HISTIDINE AS pH SENSOR. Journal of Biological

Chemistry, 275(22), 16650-16657.

Ramocki, M. B., & Zoghbi, H. Y. (2008). Failure of neuronal homeostasis results in common

neuropsychiatric phenotypes. Nature, 455(7215), 912-918.

Rakic, P. (1974, December). Timing of major ontogenic events in the visual cortex of the rhesus

monkey. In UCLA Forum Med Sci. (No. 18, pp. 3-40.

Rakic, P. A small step for the cell, a giant leap for mankind: a hypothesis of neocortical expansion

during evolution. Trends Neurosci. 18(9), 383-388.

Rauch, A. et al., (2006). Diagnostic yield of various genetic approaches in patients with unexplained

developmental delay or mental retardation. American journal of medical genetics Part A,

140(19), 2063-2074.

Redon, R. et al., (2006). Global variation in copy number in the human genome. nature, 444(7118),

444-454.

Reik, W. (2007). Stability and flexibility of epigenetic gene regulation in mammalian development.

Nature. 447(7143), 425-432.

Reik, W. & Walter, J. (2001). Genomic imprinting: parental influence on the genome. Nature

Reviews. 2(1), 21-32.

Reik, W. et al., (2003). Regulation of supply and demand for maternal nutrient transfer in mammals

in imprinted genes. The J of Physiol. 547(1), 35-44.

Renfree, M. B. et al., (2009). Evolution of genomic imprinting: insights from marsupials and

monotremes. Annual review of genomics and human genetics, 10, 241-262.

Reute, G., & Spierer, P. (1992). Position effect variegation and chromatin proteins. Bioessays, 14(9),

605-612.

145

Rice, D. & Barone S. jr. (2000). Critical periods of vulnerability for the developing nervous system:

evidence from humans and animal models. Environmental Health Perspectives. 108(S3),

511-533.

Rice, J. C. et al., (2003). Histone methyltransferases direct different degrees of methylation to define

distinct chromatin domains. Molecular cell, 12(6), 1591-1598.

Rideout, W.M., Eggan, K., & Jaenisch, K. (2001). Nuclear cloning and epigenetic reprogramming of

the genome. Science. 293(5532), 1093-1098.

Riggs, A. D., & Xiong, Z. (2004). Methylation and epigenetic fidelity. Proceedings of the National

Academy of Sciences, 101(1), 4-5.

Rimmer, J. H., & Yamaki, K. (2006). Obesity and intellectual disability. Mental retardation and

developmental disabilities research reviews, 12(1), 22-27.

Rinn, J. L. et al., (2007). Functional Demarcation of Active and Silent Chromatin Domains in

Human< i> HOX</i> Loci by Noncoding RNAs. Cell, 129(7), 1311-1323.

Robertson, A.G. et al., (2008). Genome-wide relationship between histone H3 lysine 4 mono-and tri-

methylation and transcription factor binding. Genome Res. 18(12), 1906-1917.

Robertson, J., Robertson, A. B., & Klungland, A. (2011). The presence of 5-hydroxymethylcytosine

at the gene promoter and not in the gene body negatively regulates gene expression.

Biochemical and biophysical research communications, 411(1), 40-43.

Rodriguez-Jato, S. et al., (2005). Characterization of cis-and trans-acting elements in the imprinted

human SNURF-SNRPN locus. Nucleic acids research, 33(15), 4740-4753.

Rollins, R. A., et al., (2006). Large-scale structure of genomic methylation patterns. Genome

research, 16(2), 157-163.

Romijin, H.J., Hofman, M.A., Gramsbergen, A. (1991). At what age is the developing cerebral cortex

of the rate comparable to that of the full-term newborn human baby? Early Human

Development. 26, 61-67.

Ronaghi, M. (2001). Pyrosequencing sheds light on DNA sequencing. Genome research, 11(1), 3-11.

Ronald, A., & Hoekstra, R. A. (2011). Autism spectrum disorders and autistic traits: a decade of new

twin studies. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics,

156(3), 255-274.

Ronald, A., Butcher, L. M., Docherty, S., Davis, O. S., Schalkwyk, L. C., Craig, I. W., & Plomin, R.

(2010). A genome-wide association study of social and non-social autistic-like traits in the

general population using pooled DNA, 500 K SNP microarrays and both community and

diagnosed autism replication samples. Behavior genetics, 40(1), 31-45.

Rosenberg, R. E., Law, J. K., Yenokyan, G., McGready, J., Kaufmann, W. E., & Law, P. A. (2009).

Characteristics and concordance of autism spectrum disorders among 277 twin pairs. Archives

of pediatrics & adolescent medicine, 163(10), 907-914.

Rougeulle, C. et al. (1997) The Angelman syndrome candidate gene, UBE3A/E6-AP, is imprinted in

brain. Nat. Genet. 17(1), 14–15.

Royo, H. & Cavaillé, J. (2008). Non-coding RNAs in imprinted gene clusters. Biol Cell. 100(3), 149-

166.

Runte, M., Hüttenhofer, A., Groß, S., Kiefmann, M., Horsthemke, B., & Buiting, K. (2001). The IC-

SNURF–SNRPN transcript serves as a host for multiple small nucleolar RNA species and as

an antisense RNA for UBE3A. Human Molecular Genetics, 10(23), 2687-2700.

Rutter, M. (2005). Incidence of autism spectrum disorders: Changes over time and their

meaning. Acta Paediatrica, 94(1), 2-15.

Sacher, M. et al., (2001). TRAPP I implicated in the specificity of tethering in ER-to-Golgi transport.

Molecular cell, 7(2), 433-442.

146

Sagoo, G. S. et al., (2009). Array CGH in patients with learning disability (mental retardation) and

congenital anomalies: updated systematic review and meta-analysis of 19 studies and 13,926

subjects. Genetics in Medicine, 11(3), 139-146.

Sanders, S. J. et al., (2011). Multiple recurrent de novo CNVs, including duplications of the 7q11. 23

Williams syndrome region, are strongly associated with autism. Neuron, 70(5), 863-885.

Santos, F., Hendrich, B., & Reik, W. (2002). Dynamic reprogramming of DNA methylation in the

early mouse embryo. Dev Biol. 241(1), 172-182.

Santos-Rosa, H. et al., (2002). Active genes are tri-methylated at K4 of histone H3. Nature.

419(6905), 407-411.

Sanz, L. A., et al., (2008). A mono‐allelic bivalent chromatin domain controls tissue‐specific

imprinting at Grb10. The EMBO journal, 27(19), 2523-2532.

Sarraf, S. A., & Stancheva, I. (2004). Methyl-CpG binding protein MBD1 couples histone H3

methylation at lysine 9 by SETDB1 to DNA replication and chromatin assembly. Molecular

cell, 15(4), 595-605.

Saxonov, S., Berg, P., & Brutlag, D. L. (2006). A genome-wide analysis of CpG dinucleotides in the

human genome distinguishes two distinct classes of promoters. Proceedings of the National

Academy of Sciences of the United States of America, 103(5), 1412-1417.

Schefe, J. H., Lehmann, K. E., Buschmann, I. R., Unger, T., & Funke-Kaiser, H. (2006). Quantitative

real-time RT-PCR data analysis: current concepts and the novel “gene expression’s C T

difference” formula. Journal of molecular medicine, 84(11), 901-910.

Schmittgen, T. D., & Livak, K. J. (2008). Analyzing real-time PCR data by the comparative CT

method. Nature protocols, 3(6), 1101-1108.

Schmittgen, T. D., & Zakrajsek, B. A. (2000). Effect of experimental treatment on housekeeping

gene expression: validation by real-time, quantitative RT-PCR. Journal of biochemical and

biophysical methods, 46(1), 69-81.

Schroer, R.J. et al. (1998) Autism and maternally derived aberrations of chromosome 15q. Am. J.

Med. Genet. 76(4), 327–336.

Schones, D. E. et al., (2008). Dynamic regulation of nucleosome positioning in the human genome.

Cell, 132(5), 887-898.

Sharp, A. J. et al., (2008). A recurrent 15q13. 3 microdeletion syndrome associated with mental

retardation and seizures. Nature genetics, 40(3), 322-328.

Shaw, P. et al., (2008). Neurodevelopmental trajectories of the human cerebral cortex. The Journal of

Neuroscience, 28(14), 3586-3594.

Seki, Y. et al., (2007). Cellular dynamics associated with the genome-wide epigenetic

reprogramming in migrating primordial germ cells in mice.Development, 134(14), 2627-

2638.

Shaikh, T. H. et al., (2000). Chromosome 22-specific low copy repeats and the 22q11. 2 deletion

syndrome: genomic organization and deletion endpoint analysis. Human Molecular

Genetics, 9(4), 489-501.

Shen, X., Mizuguchi, G., Hamiche, A., & Wu, C. (2000). A chromatin remodelling complex

involved in transcription and DNA processing. Nature,406(6795), 541-544.

Shen, Q. et al., (2006). The timing of cortical neurogenesis is encoded within lineages of individual

progenitor cells. Nature Neuroscience. 9(6), 743-751.

Sidman, R.L. & Rakic, P. (1992) Neuronal migration with special reference to developing human

brain: a review. Science. 62(1), 419–23.

Sidorov, M. S., Krueger, D. D., Taylor, M., Gisin, E., Osterweil, E. K., & Bear, M. F. (2014).

Extinction of an instrumental response: a cognitive behavioral assay in Fmr1 knockout mice.

Genes, Brain and Behavior.

147

Silva, A. J., & White, R. (1988). Inheritance of allelic blueprints for methylation patterns. Cell, 54(2),

145-152.

Simonoff, E. et al., (2008). Psychiatric disorders in children with autism spectrum disorders:

prevalence, comorbidity, and associated factors in a population-derived sample.Journal of the

American Academy of Child & Adolescent Psychiatry, 47(8), 921-929.

Simpson, K. M., & Sullivan, B. A. (2010). Epigenomics of centromere assembly and

function. Current opinion in cell biology, 22(6), 772-780.

Singh, P. et al., (2011). Chromosome-wide analysis of parental allele-specific chromatin and DNA

methylation. Mol Cell Biol. 31(8), 1757-1770.

Sion, I. et al., (1995). Asyncrhonous replication of imprinted genes is established in the gametes and

maintained during development. Nature. 40(6756), 929-932.

Sleutals, F., Zwart, R. & Barlow, D.P. (2002). The non-coding Air RNA is required for silencing

autosomal imprinted genes. Nature. 415(6873), 810-813.

Sleutals, F., Tjon, G., Ludwig, T., Barlow, D.P. (2003). Imprinted silencing of Slcc22a2 and Slc22a3

does not need transcriptional overlap between Ifg2r and Air. EMBO. 22(14), 3696-3704.

Smith, R.J., Dean, W., Konfortova, G., Kelsey, G. (2003). Identification of novel imprinteind genes

in a genome-wide screen for maternal methylation. Genome Research. 13(4), 558-569.

Smith, Z. D., Chan, M. M., Mikkelsen, T. S., Gu, H., Gnirke, A., Regev, A., & Meissner, A. (2012).

A unique regulatory phase of DNA methylation in the early mammalian embryo. Nature,

484(7394), 339-344.

Smith, J. L., & Schoenwolf, G. C. (1997). Neurulation: coming to closure. Trends in neurosciences,

20(11), 510-517.

Sone, M. et al., (2007). The mRNA-like noncoding RNA Gomafu constitutes a novel nuclear domain

in a subset of neurons. Journal of cell science, 120(15), 2498-2506.

Song, J., Rechkoblit, O., Bestor, T. H., & Patel, D. J. (2011). Structure of DNMT1-DNA complex

reveals a role for autoinhibition in maintenance DNA methylation. Science, 331(6020), 1036-

1040.

Song, J., Teplova, M., Ishibe-Murakami, S., & Patel, D. J. (2012). Structure-based mechanistic

insights into DNMT1-mediated maintenance DNA methylation. Science, 335(6069), 709-

712.

Sowell, E. R. et al., (2002). Development of cortical and subcortical brain structures in childhood and

adolescence: a structural MRI study. Developmental Medicine & Child Neurology, 44(1), 4-

16.

Sparago, A. et al., (2004). Microdeletions in the human H19 DMR result in loss of IGF2 imprinting

and Beckwith-Widemann syndrome. 36(9), 958-960.

Sparks, B. F. et al., (2002). Brain structural abnormalities in young children with autism spectrum

disorder. Neurology, 59(2), 184-192.

Splinter, E. et al., (2006). CTCF mediates long-range chromatin looping and local histone

modification in the β-globulin locus. Genes & Dev. 20(17), 2349-2354.

Stadtfeld, M. et al., (2010). Aberrant silencing of imprinted genes on chromosome 12qF1 in mouse

induced pluripotent stem cells. Nature, 465(7295), 175-181.

Stanwood, G.D., Washington, R.A., Shumsky, J.S., & Levitt, P. (2001). Prenatal cocaine exposure

produces consistent developmental alterations in dopamine- rich regions of the cerebral

cortex. Neuroscience. 106(1), 5-14.

Steffenburg, S., Gillberg, C., Hellgren, L., Andersson, L., Gillberg, I. C., Jakobsson, G., & Bohman,

M. (1989). A twin study of autism in Denmark, Finland, Iceland, Norway and

Sweden. Journal of Child Psychology and Psychiatry, 30(3), 405-416.

Stevenson, R. E., Schwartz, C. E., & Schroer, R. J. (2000). X-linked mental retardation (pp. 385-8).

New York: Oxford University Press.

148

Stöger, R. et al., (1993). Maternal-specific methylation of the imprinted mouse Igf2r locus identifies

the expressed locus as carrying the imprinting signal. Cell. 73(1), 61-71.

Strahl, B. D., & Allis, C. D. (2000). The language of covalent histone modifications. Nature,

403(6765), 41-45.

Strahl, B. D., Ohba, R., Cook, R. G., & Allis, C. D. (1999). Methylation of histone H3 at lysine 4 is

highly conserved and correlates with transcriptionally active nuclei in Tetrahymena.

Proceedings of the National Academy of Sciences, 96(26), 14967-14972.

Sturrock, R.R. (1974). Histogeneiss of the anterior limb of the anterior commissure of the mouse

brain. J Anat. 117(Pt 1), 37-53.

Sun, F. et al., (2001). Transactivation of Igf2 in a mouse model of Beckwith-Wiedemann syndrome.

Nature. 389(6653), 809-815.

Supèr, H., Soriano, E., & Uylings, H.B. (1998). The functions of the preplate in development and

evolution of the neocortex and hippocampus. Brain Res Reviews. 27(1), 40-64.

Surani, M.A.H., Barton, S.C., & Nords, M.L. (1984). Development of reconstituted mouse eggs

suggests imprinting of the genome during gametogenesis. Nature. 308(5959), 548-550.

Suzuki, S. et al., (2011). The evolution of mammalian genomic imprinting was accompanied by the

acquisition of novel CpG islands. 3, 1276-1283.

Szabó, P.E. & Mann, J.R. (1995). Biallelic expression of imprinted genes in the mouse germ line:

implications for erasure, establishment, and mechanisms of genomic imprinting. Genes &

Development. 9, 1857-1868.

Szabó, P. E., Tang, S. H. E., Rentsendorj, A., Pfeifer, G. P., & Mann, J. R. (2000). Maternal-specific

footprints at putative CTCF sites in the< i> H19</i> imprinting control region give evidence

for insulator function. Current Biology, 10(10), 607-610.

Tada, M. et al., (1997). Embryonic germ cells induce epigenetic reprogramming of somatic nucleus

in hybrid cells. The EMBO Journal. 16(21), 6510-6520.

Tak, P. P., & Firestein, G. S. (2001). NF-κB: a key role in inflammatory diseases. Journal of Clinical

Investigation, 107(1), 7-11.

Talley, E. M., Sirois, J. E., Lei, Q., & Bayliss, D. A. (2003). Two-pore-Domain (KCNK) potassium

channels: dynamic roles in neuronal function. The Neuroscientist, 9(1), 46-56.

Talley, E. M., Solórzano, G., Lei, Q., Kim, D., & Bayliss, D. A. (2001). CNS distribution of

members of the two-pore-domain (KCNK) potassium channel family. The Journal of

Neuroscience, 21(19), 7491-7505.

Taniai, H., Nishiyama, T., Miyachi, T., Imaeda, M., & Sumi, S. (2008). Genetic influences on the

broad spectrum of autism: Study of proband‐ascertained twins. American Journal of Medical

Genetics Part B: Neuropsychiatric Genetics, 147(6), 844-849.

Thorvaldsen, J.L., Duran, K.L., & Marisa, S.B. (1998). Deletion of the H19 differentially methylated

domain results in loss of imprinted expression of H19 and Igf2. Genes and Development.

12(23), 3693-3702.

Tabano, S. et al., (2010). Epigenetic modulation of the IGF2/H19 imprinted domain in human

embryonic and extra-embryonic compartments and its possible role in fetal growth

restriction. Epigenetics, 5(4), 313-324.

Thakur, N. et al., (2004). An antisense RNA regulates the bidirectional silencing property of the

Kcnq1 imprinting control region. Molecular and cellular biology, 24(18), 7855-7862.

Tomizawa, S. I. et al., (2011). Dynamic stage-specific changes in imprinted differentially methylated

regions during early mammalian development and prevalence of non-CpG methylation in

oocytes. Development, 138(5), 811-820.

149

Tonge, B. J., & Einfeld, S. L. (2003). Psychopathology and intellectual disability: The Australian

child to adult longitudinal study. International review of research in mental retardation, 26,

61-91.

Tost, J., & Gut, I. G. (2007). DNA methylation analysis by pyrosequencing. Nature protocols, 2(9),

2265-2275.

Traenckner, E. B. et al., (1995). Phosphorylation of human I kappa B-alpha on serines 32 and 36

controls I kappa B-alpha proteolysis and NF-kappa B activation in response to diverse

stimuli. The EMBO journal, 14(12), 2876.

Traenckner, E. B. et al., (1995). Phosphorylation of human I kappa B-alpha on serines 32 and 36

controls I kappa B-alpha proteolysis and NF-kappa B activation in response to diverse

stimuli. The EMBO journal, 14(12), 2876.

Tropepe, V., Hitoshi, S., Sirard, C., & Mak, T.W. (2001). Direct neural fate specification from

embryonic stem cells: a primitive mammalian neural stem cell stage acquired through a

default mechanism. Neuron. 30(1), 65-78.

Tsakanikos, E., Cotello, H., Holt, G., Bouras, N., STurmey, P., & Newton, T. (2006).

Psychopathology in adults with autism and intellectual disability. Journal of Autism and

Developmental Disorders, 36(8), 1123-1129.

Tschiersch, B., et al., (1994). The protein encoded by the Drosophila position-effect variegation

suppressor gene Su (var) 3-9 combines domains of antagonistic regulators of homeotic gene

complexes. The EMBO Journal, 13(16), 3822.

Tsai, T. F. et al., (2002). Evidence for translational regulation of the imprinted Snurf–Snrpn locus in

mice.Human molecular genetics, 11(14), 1659-1668.

Tse, C., Sera, T., Wolffe, A. P., & Hansen, J. C. (1998). Disruption of higher-order folding by core

histone acetylation dramatically enhances transcription of nucleosomal arrays by RNA

polymerase III. Molecular and cellular biology, 18(8), 4629-4638.

Tuchman, R., & Rapin, I. (2002). Epilepsy in autism. The Lancet Neurology, 1(6), 352-358.

Ullian, E. M., Sapperstein, S. K., Christopherson, K. S., & Barres, B. A. (2001). Control of synapse

number by glia. Science, 291(5504), 657-661.

Ushijima, T. et al., (2003). Fidelity of the methylation pattern and its variation in the genome.

Genome research, 13(5), 868-874.

Vaccarino, F. M., Ganat, Y., Zhang, Y., & Zheng, W. (2001). Stem cells in neurodevelopment and

plasticity. Neuropsychopharmacology, 25(6), 805-815.

Varga-Weisz, P. D., Wilm, M., Bonte, E., Dumas, K., Mann, M., & Becker, P. B. (1997). Chromatin-

remodelling factor CHRAC contains the ATPases ISWI and topoisomerase

II. Nature, 388(6642), 598-602.

Veale, E. L., Buswell, R., Clarke, C. E., & Mathie, A. (2007). Identification of a region in the

TASK3 two pore domain potassium channel that is critical for its blockade by

methanandamide. British journal of pharmacology, 152(5), 778-786.

Velagaleti, G. V. et al., (2005). Position Effects Due to Chromosome Breakpoints that Map∼ 900 Kb

Upstream and∼ 1.3 Mb Downstream of< i> SOX9</i> in Two Patients with Campomelic

Dysplasia. The American Journal of Human Genetics, 76(4), 652-662.

Veltman, M. W., Thompson, R. J., Roberts, S. E., Thomas, N. S., Whittington, J., & Bolton, P. F.

(2004). Prader-Willi syndrome. European child & adolescent psychiatry, 13(1), 42-50.

Verpelli, C. & Sala, C. (2012). Molecular and synaptic defects in intellectual disability syndromes.

Curr Op in Neurbiol. 22(3), 530-546.

Volpe, T. A., Kidner, C., Hall, I. M., Teng, G., Grewal, S. I., & Martienssen, R. A. (2002).

Regulation of heterochromatic silencing and histone H3 lysine-9 methylation by RNAi.

science, 297(5588), 1833-1837.

150

Vu, T. H., & Hoffman, A. R. (1997). Imprinting of the Angelman syndrome gene, UBE3A, is

restricted to brain. Nature genetics, 17(1), 12-13.

Vu, T. H., Li, T., & Hoffman, A. R. (2004). Promoter-restricted histone code, not the differentially

methylated DNA regions or antisense transcripts, marks the imprinting status of IGF2R in

human and mouse. Human molecular genetics, 13(19), 2233-2245.

Wakui, K. et al., (2005). Construction of a natural panel of 11p11. 2 deletions and further delineation

of the critical region involved in Potocki–Shaffer syndrome. European journal of human

genetics, 13(5), 528-540.

Wallrath, L. L., & Elgin, S. C. (1995). Position effect variegation in Drosophila is associated with an

altered chromatin structure. Genes & Development, 9(10), 1263-1277.

Wang, X. et al., (2008). Trancriptome-wide identification of novel imprinted genes in neonatal

mouse brain. PLoS ONE. 3(12), e3839.

Wang, J. et al., (2010). Dual DNA methylation pattersn in the CNS reveals developmentally poised

chromatin and monoallelic expression of critical genes. PLoS ONE. 5, e13843.

doi:10.1371/journal.pone.0013843.

Warnecke, P. M., et al., (2002). Identification and resolution of artifacts in bisulfite

sequencing.Methods, 27(2), 101-107.

Washbourne, P., Bennett, J. E., & McAllister, A. K. (2002). Rapid recruitment of NMDA receptor

transport packets to nascent synapses. Nature neuroscience, 5(8), 751-759.

Weber, M., Hellmann, I., Stadler, M. B., Ramos, L., Pääbo, S., Rebhan, M., & Schübeler, D. (2007).

Distribution, silencing potential and evolutionary impact of promoter DNA methylation in

the human genome. Nature genetics, 39(4), 457-466.

Webster, K. E. et al., (2005). Meiotic and epigenetic defects in Dnmt3L-knockout mouse

spermatogenesis. Proceedings of the National Academy of Sciences of the United States of

America, 102(11), 4068-4073.

Weisenberger, D. J. et al., (2005). Analysis of repetitive element DNA methylation by MethyLight.

Nucleic acids research, 33(21), 6823-6836.

Weiss, L. A. et al., (2008). Association between microdeletion and microduplication at 16p11. 2 and

autism. New England Journal of Medicine, 358(7), 667-675.

Weksberg, R., Smith, A. C., Squire, J., & Sadowski, P. (2003). Beckwith–Wiedemann syndrome

demonstrates a role for epigenetic control of normal development. Human molecular

genetics, 12(suppl 1), R61-R68.

Wentzel, C., Fernström, M., Öhrner, Y., Annerén, G., & Thuresson, A. C. (2008). Clinical variability

of the 22q11. 2 duplication syndrome. European journal of medical genetics, 51(6), 501-510.

Werling, D. M., & Geschwind, D. H. (2013). Sex differences in autism spectrum disorders. Current

opinion in neurology, 26(2), 146-153.

Wilkinson, L.S., Davies, W., & Isles, A.R. (2007). Genomic imprinting effects on brain development

and function. Nature. 8, 832-843.

Williams, K., Helmer, M., Duncan, G. W., Peat, J. K., & Mellis, C. M. (2008). Perinatal and maternal

risk factors for autism spectrum disorders in New South Wales, Australia. Child: care, health

and development, 34(2), 249-256.

Wolf, S. F., Jolly, D. J., Lunnen, K. D., Friedmann, T., & Migeon, B. R. (1984). Methylation of the

hypoxanthine phosphoribosyltransferase locus on the human X chromosome: implications for

X-chromosome inactivation. Proceedings of the National Academy of Sciences, 81(9), 2806-

2810.

Woods, C.G., Valente, E.M., Bond, J., Roberts, E. (2004). A new method for autozygosity mapping

using single nucleotide polymoprhisms (SNPs) and EXCLUDEAR. J Med Gen. 41(8), 1-4.

World Health Organization. (1992). The International Classification of Diseases–Tenthrevision

(ICD10). Geneva:World Health Organization.

151

Wylie, A. A., Murphy, S. K., Orton, T. C., & Jirtle, R. L. (2000). Novel imprinted DLK1/GTL2

domain on human chromosome 14 contains motifs that mimic those implicated in IGF2/H19

regulation. Genome research, 10(11), 1711-1718.

Wutz, A., Rasmussen, T. P., & Jaenisch, R. (2002). Chromosomal silencing and localization are

mediated by different domains of Xist RNA. Nature genetics, 30(2), 167-174.

Xie, W. et al., (2012). Base-resolution analyses of sequence and parent-of-origin dependent DNA.

Cell. 148(4), 816-831.

Xin, Z., Allis, C.D., Allis, & Wagstaff, J. (2001). Parent-specific omcplementary patterns of histone

3 H3 lysine 9 and H3 lysine 4 methylation at the Prader-Willi Syndrome. Am J Hum Genet.

69(6), 1389-1394.

Xu, W. et al., (2014). Genome-wide association study of bipolar disorder in Canadian and UK

populations corroborates disease loci including SYNE1 and CSMD1. BMC medical genetics,

15(1), 2.

Xu, Y. Q., Goodyer, C. G., Deal, C., & Polychronakos, C. (1993). Functional Polymorphism in the

Parental Imprinting of the Human< i> IGF2R</i> Gene. Biochemical and biophysical

research communications, 197(2), 747-754.

Yamaguchi, Y., & Miura, M. (2013). How to form and close the brain: insight into the mechanism of

cranial neural tube closure in mammals. Cellular and Molecular Life Sciences, 70(17), 3171-

3186.

Yamasaki, Y. et al., (2005). Neuron-specific relaxation of Igf2r imprinting is associated with neuron-

specific histone modifications and lack of its antisense transcript Air. Hum Mol Gen. 14(17),

2511-2520.

Yeargin-Allsopp, M., Rice, C., Karapurkar, T., Doernberg, N., Boyle, C., & Murphy, C. (2003).

Prevalence of autism in a US metropolitan area. Jama, 289(1), 49-55.

Yip, C.K., Berscheminski, J., & Waltz, T. (2010). Molecular architecture of the TRAPPII complex

and implications for vesicle tethering.. Nat Struct Mol Biol. 17(11), 1298-1304.

Yokomine, T. et al., (2005). Structural and functional analysis of a 0.5-Mb chicken region

orthologous to the imprinted mammalian Ascl2/Mash2–Igf2–H19 region. Genome research,

15(1), 154-165.

Yoshii, A., Krishnamoorthy, K. S., & Grant, P. E. (2002). Abnormal cortical development shown by

3D MRI in Prader–Willi syndrome. Neurology, 59(4), 644-644.

Yotova, I. Y. et al., (2008). Identification of the human homolog of the imprinted mouse< i> Air</i>

non-coding RNA. Genomics, 92(6), 464-473.

Yudkin, D. et al., (2014). Chromosome fragility and the abnormal replication of the FMR1 locus in

fragile X syndrome. Human molecular genetics, ddu006.

Zheng, C., & Hayes, J. J. (2003). Structures and interactions of the core histone tail domains.

Biopolymers, 68(4), 539-546.

Ziller, M. J. et al., (2011). Genomic distribution and inter-sample variation of non-CpG methylation

across human cell types. PLoS genetics, 7(12), e1002389.

Zoghbi, H. Y., & Bear, M. F. (2012). Synaptic dysfunction in neurodevelopmental disorders

associated with autism and intellectual disabilities. Cold Spring Harbor perspectives in

biology, 4(3), a009886.

Zong, M. et al., (2011). The adaptor function of TRAPPC2 in mammalian TRAPPs explains

TRAPPC2-associated SEDT and TRAPPC9-associated congenital intellectual disability.

PLoS ONE. 6(8), e23350.

Zong, M. et al., (2012). TRAPPC9 mediates the interaction between p150Glued and COPII vesicles at

the target membrane. PLoS ONE..7(1), e29995.

152

APPENDIX 1

Figure A.1 Pyrosequencing traces analyzed to validate allelic expression and methylation. The x-

axis shows the order of nucleotide dispensation and the y-axis measures the amount of nucleotide added

for each reaction well (the amount of light produced). The ascending slope of a peak is determined by

DNA polymerase and sulfurylase activity, while the peak height depends on the oxidative capacity of

luciferase, and the rate of nucleotide degradation by apyrase determines the descending slope (Rongahi

et al., 2010). (A) Pyrogram quantifying allelic expression at a SNP (B) Pyrogram measuring DNA

methylation.

A)

B)

153

APPENDIX 2

First

author,

year

Family; Affected

members

Mutation CNV

(size)

Inheritance pattern Main clinical

phenotypes

Marangi,

2012

2 Italian (Caucasian,

Western European)

sisters born to healthy,

non-consanguineous

parents

Truncating c.2851-

2A>C [p.T951Y

fsXD17]

N/I* Parents heterozygous

for shortened gene

variant

Cerebellar aphasia

Microcephaly

Kakar, 2012

Consainguineous

Pakistani family: 3

affected girls

Homozygous

transversion donor

splice site of Exon 3:

c.1024G>T

FrameshiftTruncatin

g

N/I Parents were

heterozygous

Microcephaly

Jamra, 2011 Syrian consanguineous

families: 3 affected boys

Truncating, nonsense

c.1423C>T

[p.R475X]**

N/I De novo Microcephaly

Stereotypies

Koifman,

2010

1 female proband born to

normal, healthy parents

of Filipino-descent;

healthy sister; no

consainguinity (but

great-grandmothers were

first cousins)

Deletion Chr8:

140,879,937

141,021,392

x0 mat pat

(141.46 kb)

Parents were

heterozygous

Hypoplasia of the

corpus callosum

Developmental

delay

Dysmorphic facial

features

Mochida,

2009

Israeli Arab

consanguineous family;

3 affected girls

Truncating

c.1423C>T[p.R475X]

(variant 1)

c.1129C>T [p.R475X]

(variant 2)

N/I Parents heterozygous

carriers for missense

mutation

Moderate to

severe ID

↓volume of

cerebral white

matter

Mir, 2009 Consanguineous

Pakistani family; 7

affected (6F;1M)

Truncating frameshift

mutation: c.1422C>T

[p.R475X7]

Nonsense-mediated

RNA decay

Deletion;

includes

KCNK9

Heterozygous carriers Microcephaly

↓volume of

cerebral white

matter

Najmabad,

2009 (from

Mir, 2009)

Consanguineous Iranian

family

c.2311-2314

[p.L772RfsX7]

deletion N/I N/A

Phillipe,

2009

Consanguineous

Tunsinian family; 3

affected boys

Truncating c.1708C>T

[p.R570X9]

Nonsense-mediated

decay

N/I Healthy parents

Microcephaly

Truncular obesity

Myelination

defect

Dysmorphic

features

Table A.2 List of studies which have identified TRAPPC9 as a cause of NS-ARID. All studies with

the exception of the Koifman et al., (2011) study used homozygosity mapping to determine segregation of

disease-causing mutations in TRAPPC9. N/A = Not available; N/I = Not identified; ** This mutations

was identified in three families (Jamra et al., 2011, Mochida et al., 2009, Mir et al., 2009)

154

APPENDIX 3

A.3 Expression analysis of Trappc9 in mice. Allelic expression of Trappc9 was assessed in B6 males

crossed with JF1 females, and JF1 males crossed with B6 females to account for strain variation. Trappc9

showed preferential or exclusive expression from the maternal allele in fetal brain, neuroprogenitor cells

and neural cells. Allelic imbalance disappeared at 8 weeks postanatal. These findings were confirmed by

pyrosequencing.

155

BJ15.5_whole brain

JB15.5_whole brain

BJ P0♀_brain1

JB P0♀-brain1

BJ P0♀_brain2

JB P0♀_brain2

BJ P0♀_brain3

JB P0♀-brain3

BJ P0♀_brain5

JB P0♀-brain5

JB 24W♀_brain1

JB 24W♀_brain3

JB 24W ♀_brain5

BJ 24W ♀_brain1

BJ 24W ♀_brain3

BJ 24W ♀_brain5

BJ 8W ♂ _brain1

BJ 8W ♀ _brain1

JB 5W ♂ _brain1

JB 5W ♀ _brain1

SNP#1: C/T (B6/JF1)

1810044A24Rik (mouse TRAPPC9)_allelic expression analysis_2010/4/16

brain1：cerebrum brain2：olfactory bulb brain3：cerebellum brain4：midbrain region brain5：thalamic/subthalamic region NPCs: neuronal progenitor cells prepared from 16.5dpc cerebral cortices Neuron and glia cells: In vitro-differentiated neuronal and glial cells from NPCs

★

★ ★

?

★ ★

★ ★

★ ★

★ ★

★ ★

★ ★

★ ★

★ ★

★

★ ★

★

★ ★

★

★

★ ★

★

★

★ ★

★ ★

★ ★

★ ★

BJ_neuron1

BJ_neuron2

BJ_glia1

JB_neuron1

JB_glia1

BJ_neuron3

BJ_glia1

BJ_NPCs

JB_NPCs

JB_fibroblast

BJ_fibroblast

failed failed

★ ★

★ ★

★ ★

★ ★

★ ★

★ ★

★ ★

★

★ ★

★

★ ★

SNP#2: T / C (B6/JF1)

SNP#1: C/T (B6/JF1)

SNP#2: T / C (B6/JF1)

PCR and sequencing primers - product size 973 bp - F-WY1593 5’-GAAACACGTTGGGGACTTGT-3’ - R-WY1594 5’-AATGGGCAACTTGGTAAACG-3’ - PCR cycle numbers: x 30 – 33

★ exclusive maternal expression (paternal peak barely seen)

★ biallelic expression (the maternal peak looks less than twice higher than the paternal peak)

★ preferentiall maternal expression (the maternal peak looks more than twice higher than the paternal peak)

BJ: born from the mating of B6♀ x JF1♂ JB: born from the mating of JF1♀ x B6♂ 15.5: 15.5 dpc (day post coitum) P0: postnatal day0 W: weeks

SNP#1 nt 1285 of NM_180662 aggccgtgcg[C/T]gtcctagcga

SNP#2 nt 1612 of NM_180662 ttcagatgcg[T/C]ctgctgcatg

156

157

APPENDIX 4

A.4. Allelic expression analysis for PEG13. Electropherograms showing allelic expression of PEG13 in

fetal brain, kidney, skin and placenta using SNP rs4289794 . Among the ten individuals who were

informative, seven showed preferential expression at PEG13. Three samples showed biallelic expression

for PEG13 in fetal brain.

158

159

160

APPENDIX 5

A.5. Relative expression of alleles at PEG13 in ten fibroblasts samples at SNPs rs2270409. G =

guanine; A= adenine. All individuals showed preferential or exclusive expression from an allele; however

the parent-of-origin could not be determined because we did not have parental DNA.

0% 20% 40% 60% 80% 100%

GM00042

GM01096

GM01376

GM01391

GM01653

GM02175

GM02455

GM03073

GM09256

GM00519

PEG13 rs2270409 % expression from each allele

Sam

ple

G

A

161

APPENDIX 6

A.6. Identification of the PEG13-DMR. Three regions of the CpG island investigated for allelic

methylation are shown as region 1 (chr8:141107809-141108250), region 2 (chr8:141108200-141108523),

and region 3 (chr8:141109944-141110307). Allelic methylation of region 1 was assessed in placental and

cord DNA using rs3802217. Methylation in cord blood was highly methylated, while methylation in

placenta varied, but was overall highly methylated. Methylation profiles at region 2in placenta, cord

blood and fetal brain shows differential methylation, indicating the PEG13-DMR. Conversely,

methylation at region 3 was lowly methylated in all tissues tested.