Analysis of a putative imprinted locus within the TRAPPC9 ......ii in PEG13 allelic expression,...
Transcript of Analysis of a putative imprinted locus within the TRAPPC9 ......ii in PEG13 allelic expression,...
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.