Thesis Template

Copy Number Variation in Han Chinese Individuals with Autism Spectrum Disorder

Matthew Joseph Gazzellone

A thesis submitted in conformity with the requirements for the degree of Master of Science Department of Molecular Genetics University of Toronto

Copy Number Variation in Han Chinese Individuals with

Autism Spectrum Disorder

Matthew Joseph Gazzellone

Master of Science

Department of Molecular Genetics University of Toronto

2014 Abstract Autism spectrum disorders (ASDs) are a group of neurodevelopmental conditions with a demonstrated genetic etiology. Rare copy number variations (CNVs) account for a proportion of the genetic events involved, but their contribution in non-European ASD populations has not been well studied. This thesis examines rare CNVs detected in a cohort of Han Chinese individuals with ASD. Using the Affymetrix CytoScan HD platform, we genotyped DNA from

104 ASD trios from Harbin, China. Of the probands, 8.6% had one or more de novo CNVs.

Several candidate risk genes were also identified. A 24-kb duplication was found overlapping

YWHAE (an ASD candidate gene). This duplication is observed at a similar frequency in cases and in population controls and is likely a benign Asian-specific copy number polymorphism. Our findings help define genomic features relevant to ASD in the Han Chinese and emphasize the importance of using ancestry-matched controls in medical genetic interpretations.

Acknowledgments

First and foremost, I would like to thank my supervisor Dr. Stephen Scherer for his mentorship over the past five years during both my undergraduate and graduate studies. It has been a privilege to come to the lab each day and work with such a brilliant scientist and genuinely good person. I greatly appreciate his confidence in me over these last five years. I would also like to thank my committee members, Dr. Freda Miller and Dr. Lucy Osborne. Their scientific advice and encouragement have been invaluable during my time as a student.

I am grateful for the assistance I have received and the friendships that I have made at The Centre for Applied Genomics (TCAG). I would particularly like to thank my friend and former colleague, Dr. Anath Lionel. His advice and support have helped me become a better researcher. Many other members of the Scherer academic group and TCAG have provided assistance during the course of my time here. My profound gratitude goes to Dr. Susan Walker, Dr. Daisuke Sato, Dr. Christian Marshall, Dr. Kristiina Tammimies, Dr. Mohammed Uddin, Dr. Mehdi Zarrei, Dr. Ryan Yuen, Dr. Eric Deneault, Dr. Andy Pang, Lia D’Abate, Maggie Evans, Jenny Kaderali, Jennifer Howe, Dr. Richard Wintle, Sylvia Lamoureux, Bhooma Thiruvahindrapuram, and Sanjeev Pullenayegum.

I would also like to acknowledge my friends and family who have been supportive throughout my studies. I am especially thankful for the encouragement and understanding of Nicole Lau during this degree. I would finally like to recognize my parents and sister. I thank them for instilling within me an interest in science and for their love and support during all of my educational endeavors.

iii

Table of Contents

Acknowledgments ...... iii

Table of Contents ...... iv

List of Tables ...... vi

List of Figures ...... vii

List of Appendices ...... viii

Introduction ...... 1

1.1 Phenotypic Spectrum of ASD ...... 1

1.2 Epidemiology ...... 2

1.3 Genetic basis of ASD ...... 2

1.4 Genetic theories that may explain ASD ...... 3

1.4.1 Common Variant Model ...... 5

1.4.2 Rare Variant of High Effect Model ...... 5

1.5 Copy Number Variation ...... 7

1.6 Copy Number Variation and Human Disease ...... 9

1.7 Identifying plausible candidate genes ...... 10

1.8 Neurobiology of ASD ...... 10

1.9 Microarray Studies ...... 12

1.10 ASD in the Han Chinese population ...... 13

1.11 Rationale and Objectives of the Thesis ...... 15

1.11.1 Rationale ...... 15

1.11.2 Hypothesis ...... 15

1.11.3 Objectives ...... 15

Disclosure ...... 16

Methods ...... 17

2.1 Sample Selection ...... 17

2.2 Genotyping of Samples ...... 17

2.3 Variant Calling ...... 19

2.4 Population-based control datasets ...... 19

2.5 Experimental Validation of CNV Calls ...... 23

2.6 Expression Analysis ...... 26

Results ...... 31

3.1 De novo Variants ...... 35

3.2 Rare Inherited Variants ...... 41

3.3 Population-Specific CNV Polymorphisms ...... 43

Discussion ...... 45

Summary and Significance ...... 49

Future Directions ...... 53

Appendix ...... 57

References ...... 63

List of Tables

Table 1: Primers used for Expression Assays ...... 27

Table 2: Summary of de novo and rare inherited CNVs of interest in ASD probands...... 32

Table 3: Summary statistics of stringent CNVs larger than 20 kb ...... 33

List of Figures

Figure 1: Contribution of Genetic and Non-Genetic Factors to ASD Phenotype ...... 4

Figure 2: Ancestry Determination in ASD Cohort ...... 18

Figure 3: CNV Detection Workflow ...... 21

Figure 4: Number of calls from each algorithm per sample ...... 22

Figure 5: Agarose gel picture illustrating successful PCR amplification of FOXP2 primers ...... 24

Figure 6: qPCR assay indicating inherited loss at 10p12.33 ...... 25

Figure 7: Testing of CASKIN1 and PKD1 primer sets ...... 28

Figure 8: Confirmed amplification of CASKIN1 exons of interest ...... 29

Figure 9: Confirmed amplification of PKD1 exons of interest ...... 30

Figure 10: Number of Rare CNVs per proband ...... 34

Figure 11: Pictorial Representation of GIGYF2 deletion using ChAS software ...... 36

Figure 12: Identification of Brain-Critical Exons at 16p13.3 ...... 37

Figure 13: 16p11.2 region of interest ...... 39

Figure 14: Location of DMD Deletions ...... 40

Figure 15: Pedigree of family 694 with GRID2 deletion ...... 42

Figure 16: Genomic location of YWHAE duplications ...... 44

vii

List of Appendices

Appendix 1: List of Rare CNVs ...... 57

viii 1

Introduction 1.1 Phenotypic Spectrum of ASD

Autism Spectrum Disorder (ASD) is a common neuropsychiatric condition that presents during childhood and can have significant human health implications. Though ASD is clinically heterogeneous, impairments in social interaction and a restricted range of behaviors (including stereotyped motor movements, ritualized speech patterns, and unusually focused interests) are observed (American Psychiatric Association 2013). Language difficulties and abnormal language development are also frequently noted (Trillingsgaard et al. 2005). The condition was first recognized by Dr. Leo Kanner over seventy years ago (Kanner 1943) and the diagnostic criteria have been regularly updated as scientists and clinicians have gained a better understanding of the biological mechanisms and phenotypic characteristics that underlie the spectrum of conditions (Berg and Geschwind 2012).

Cognitive impairments are common in ASD-afflicted individuals and cognitive functioning can range from above-average intelligence to severe intellectual disability (Rutter 1978). ASD is often accompanied by other conditions including motor and sensory abnormalities, epilepsy, and other neuropsychiatric illnesses. In fact, over 70% of affected individuals present with at least one co-occurring condition (Simonoff et al. 2008; Lai et al. 2014). Typically, those concurrent conditions that manifest themselves in early childhood tend to endure through adolescence and adulthood (Simonoff et al. 2013). Importantly, individuals that present with more co-occurring conditions usually have greater cognitive impairments (Mattila et al. 2010). Altogether, the phenotypic spectrum is extremely diverse: some affected individuals are able to lead highly independent lives while others require substantial support services for day-to-day functioning (Stoddart et al. 2013). Earlier diagnosis can allow for timely targeted behavioral interventions that can improve an affected individual’s quality of life while reducing the economic burden on families and the health-care system (Stoddart et al. 2013). As a result, identifying genetic or environmental causes of ASD could have a profound positive impact in the lives of affected individuals.

1.2 Epidemiology

The prevalence of ASD has increased substantially over the past few years. The most recent statistics indicate that one in sixty-eight newborns will be diagnosed with an ASD at some point in their lives (Wingate et al. 2014). This represents between a twenty- and thirty-fold increase in diagnoses from the 1960s (Wingate et al. 2014). Though much of this increase has been attributed to changes in diagnostic criteria, other factors including earlier diagnosis, improved awareness, and an increased prevalence of risk factors should not be discounted (Devlin and Scherer 2012). Disorder onset typically occurs before age three, though signs may be visible as soon as a few months after birth (Jones and Klin 2013). However, for those individuals considered to be “higher-functioning”, only about half were diagnosed before age 21 (Stoddart et al. 2013).

One of the most striking features of ASD is a 4:1 male to female gender ratio that approaches 11:1 when considering higher-functioning manifestations of the disorder (Gillberg et al. 2006). However, when observing the 45% of individuals with who present with both an ASD and intellectual disability, the ratio of males to females is below 2:1 (Volkmar et al. 1993; Yeargin- Allsopp et al. 2003; Lai et al. 2014). These findings suggest that a greater genetic or environmental insult is required for females to meet the ASD diagnostic criteria than for males (Jacquemont et al. 2014). That said, it is also probable that higher-functioning females with an ASD may be under-represented as additional behavioral and cognitive problems might be required for diagnosis (Dworzynski et al. 2012). Overall, this highlights the importance of controlling for sex differences when attempting to uncover the underlying causes of the disorder phenotype.

1.3 Genetic basis of ASD

Previous studies of affected individuals and their families have shown that genetic factors explain a substantial proportion of the underlying ASD etiology. Longitudinal studies of ASD in monozygotic twins where one twin was previously diagnosed with autism have shown between 50-90% concordance for an ASD phenotype or subclinical cognitive problems in the second twin (Rosenberg et al. 2009; Hallmayer et al. 2011; Ozonoff et al. 2011). In dizygotic twin pairs, this concordance is about 10% (Constantino et al. 2010). Furthermore, nearly a quarter of siblings of an affected proband will also be affected (Constantino et al. 2010) and ASD concordance is

3 nearly double among full sibling than half siblings (Constantino et al. 2013). Familial clustering in other first-degree and more distant relatives has also been noted in many families (Sandin et al. 2014). Mild cognitive and behavioral problems are often seen in many relatives of ASD probands (Sucksmith et al. 2011). Individuals having these features are said to have a broader autism phenotype which is characterized by minor language or social-behavioral impairments that would not meet ASD diagnostic criteria (Losh et al. 2008).

These twin and family studies provide some of the most convincing evidence that genetics play a considerable role in the development of the conditions. The substantial elevation in ASD risk in siblings and other family members of affected probands stands in stark contrast to the 1-2% expected in the general population (Wingate et al. 2014). Though these concordance observations do not imply that environmental and other non-genetic factors play no role in the underlying etiology, they do suggest that genetic factors may explain much of the underlying architecture of the disorder.

1.4 Genetic theories that may explain ASD

Several theories have been proposed to explain the genetic contribution to ASD risk. The majority of these presuppose a polygenic contribution to phenotype as opposed to a model implying a Mendelian mode of inheritance (Berg and Geschwind 2012). Much of the evidence for this is supported by the fact that in most families, the phenotype does not segregate in a manner consistent with Mendelian inheritance (syndromic cases, which explain 5-10% of these cases, are often an exception) (Berg and Geschwind 2012; Devlin and Scherer 2012). Currently accepted genetic models involve either common or rare genetic variants or some combination of both types. The overarching hypothesis proposes that pathogenic genetic factors are additive and, once a certain threshold is reached, an ASD phenotype arises (Figure 1) (Cook and Scherer 2008). Those individuals harboring these types of genetic factors, whose cumulative effect does not exceed the threshold but instead hovers around it, are those that are frequently described as having a broader autism phenotype.

Figure 1: Contribution of Genetic and Non-Genetic Factors to ASD Phenotype

The above graphic illustrates how the combination of genetic and non-genetic factors can lead to Autism Spectrum Disorders. Each bar is representative of the susceptibility risk of ASD in an individual with a given set of genetic and non-genetic contributing factors. Certain genetic variants are highly penetrant and require few additional influences to result in an ASD. The two leftmost bars illustrate that an individual with that particular genetic variant would be highly susceptible to ASD. Conversely, an individual with genetic and non-genetic contributing factors that mirror those in the third bar would be susceptible to a broader autism phenotype but not the full-fledged manifestation of ASD. Finally, an individual with the susceptibility factors found in the fourth bar would be unlikely to have any ASD phenotype. This graphic has been adapted from a review previously published in Nature (Cook and Scherer 2008).

1.4.1 Common Variant Model

An early model that was proposed to explain this common disorder was one where either a single common genetic cause or combination of common genetic variants gives rise to ASD (Berg and Geschwind 2012). A common variant is typically classified as having a minor allele frequency exceeding 5% (Aschard et al. 2011). A genome-wide association study (GWAS) is often used to investigate whether common variants contribute to a disorder phenotype. This method utilizes an unbiased approach to identify possible disease associations with virtually every region of the genome (McCarthy et al. 2008). Since the genome can be broken into a collection of haplotypes which can be determined by genotyping a few SNPs, one can impute the rest of the sequence within the haplotype and determine whether it is associated with disease (Evangelou and Ioannidis 2013). In autism, there have been three GWASs that have had sample sizes large enough to detect common variants that have a moderate effect size (Berg and Geschwind 2012). From these studies, only two variants were significant: a SNP between CDH9 and CDH10 (Wang et al. 2009) and a variant in an intron of MACROD2 (Anney et al. 2010) (Berg and Geschwind 2012). The former variant is thought to implicate a regulatory non-coding RNA at 5p14.1 (Kerin et al. 2012) while the latter variant implicates the MACROD2 gene which, though not very well characterized, is highly expressed in brain (Anney et al. 2010). That said, the two variants found are predicted to only have a modest effect on ASD risk. Combined, these studies have shown that no common genetic lesion explains a significant amount of the genetic architecture underlying ASD.

Rare genetic variants, whose prevalence in the general population falls below 1%, have since captivated much of the focus in the research community because of the few tangible results realized in common variant studies. Studies of these types of variants have been significantly more successful at identifying plausible risk alleles. At this time, potentially causal genetic lesions have been identified in approximately 20-25% of individuals with an ASD and these genetic factors are almost exclusively rare in nature (Devlin and Scherer 2012). The take-home message from such studies is that ASDs are genetically heterogeneous.

1.4.2 Rare Variant of High Effect Model

This model is best observed in syndromic forms of autism. Studies suggest that up to 10% of individuals with an ASD have a genomic syndrome of which the phenotype has some overlap

6 with ASD, but individually, these syndromes explain no more than 1-2% of cases (Devlin and Scherer 2012). One example is Fragile X syndrome which is caused by the expansion of the trinucleotide CGG repeat region in the 5’ untranslated region of FMR1 (Santoro et al. 2012). Though this syndrome is inherited in an X-linked dominant manner, the ASD phenotype is not fully penetrant. Of the individuals with an FMR1 mutation, only 30% will receive a clinical diagnosis of ASD (Harris et al. 2008). Tuberous sclerosis (caused by mutations in TSC1 and TSC2), Rett Syndrome (caused by mutations in MECP2), and neurofibromatosis (caused by mutations in NF1) are other genetic syndromes that are most frequently accompanied by ASD (Devlin and Scherer 2012).

Rare chromosomal abnormalities have also been identified in some 5% of individuals presenting with an ASD (Devlin and Scherer 2012). These changes are easily identified via high-resolution karyotyping. It has been noted that individuals with these rare, large-scale chromosomal abnormalities are at a greater risk of presenting with a corresponding dysmorphology (Girirajan et al. 2011). The most common of these variants is a duplication of the maternally-derived 15q11-q13 region which occurs in 1-3% of ASD cases (Baker et al. 1994). Virtually every chromosome has been implicated in at least some individuals, but individually, the mutations are rare (Devlin and Scherer 2012).

Another 5-10% of cases harbor rare genic mutations that are highly penetrant (Devlin and Scherer 2012). Initially, such mutations were often identified from gene sequencing studies undertaken using Sanger sequencing of exonic regions of a candidate gene. These types of studies, while successful at identifying some potentially interesting genetic variants, suffered from several limitations. Firstly, the selection of candidate genes was done in a biased manner, usually based on the functional role of the gene product (Vourc'h et al. 2003). This is problematic as it is based on the assumption that one particular locus plays a significant role while ignoring much of the genetic heterogeneity which is innate to ASD. Secondly, single nucleotide variants that result in mostly silent or missense changes are most commonly uncovered and interpretation of these changes without any functional experiments is difficult (Wassink et al. 2005). Finally, these studies often lacked the statistical power needed to determine whether mutations arose more frequently in a cohort of cases than in population controls (Muscarella et al. 2010). These challenges have been at least partially abated by next- generation sequencing technologies. Though these technologies hold promise to one day identify

7 a wide range of genetic variants, they have thus far been used primarily to identify single nucleotide variants in the genome. In ASD, several exome sequencing studies have been undertaken to date (Iossifov et al. 2012; Neale et al. 2012; O'Roak et al. 2012; Sanders et al. 2012). These have primarily used a trio-based study design where an affected proband and both biological parents were sequenced on the same sequencing platform. The goal of these studies was to identify de novo single nucleotide variants (SNVs). Although the frequency of de novo genetic variants was similar in the case and population cohorts, a marked increase in de novo nonsynonymous and nonsense SNVs was observed in the case cohorts. Case individuals were 2- 4 times more likely to have de novo nonsense mutations than would be expected statistically (Devlin and Scherer 2012).

Another powerful study that exhibits the utility of next-generation sequencing technology is a whole-genome sequencing pilot study in autism (Jiang et al. 2013). In this study, 32 ASD trios were sequenced in order to identify potentially causal SNVs. The authors identified damaging de novo mutations in 19% of families and potentially causal autosomal or X-linked inherited changes in 31% of individuals (Jiang et al. 2013). This illustrated how both de novo and rare inherited SNVs may contribute to ASD development. These results convey the promise of whole-genome sequencing; that much of the missing heritability can be uncovered due to the coverage afforded by the technology. Though this manuscript and others examining whole genomes or exomes in ASD probands and their families have identified many interesting candidate genes and provided more support for existing ones, their utility has been focused primarily at detecting SNVs. At this time, these types of platforms are only beginning to identify deletions or duplications of DNA consistently and without a substantial number of false positives or false negatives.

1.5 Copy Number Variation

No technology has had a greater impact regarding the identification of candidate genes and loci for ASD than microarrays. Microarrays measure the intensity of probe hybridization to a particular genomic region and are able to detect duplications and deletions of segments of DNA using this information (Marshall and Scherer 2012). These deletions and duplications (as well as balanced chromosomal rearrangements including inversions and translocations) are termed Copy Number Variation (CNV). This type of variation is present in all individuals and the average

8 person possesses nearly 1,100 CNVs with a median size was 2.9 kb (Conrad et al. 2010). The genomes of any two individuals differ by about 0.1% because of CNVs (Malhotra and Sebat 2012). Approximately 12% of the genome is copy number variable and CNVs overlap about 10% of genes (Redon et al. 2006). Frequently, large deletions and duplications with significant sequence homology at the breakpoints (including many that play a role in ASD) typically arise via nonallelic homologous recombination (NAHR) (Malhotra and Sebat 2012). This process usually occurs when non-allelic low copy repeats (LCRs) align and crossing over occurs between the two regions causing a copy number change of the LCR regions and any sequence therein (Gu et al. 2008).

Though most CNV studies in ASD show that the number of inherited CNVs is similar in case and control cohorts, the number of de novo CNVs detected are consistently higher in cases. Large-scale CNV studies indicate that 5-10% of individuals with ASD have a de novo CNV while only about 1% of individuals from the general population have a de novo CNV (Sebat et al. 2007; Pinto et al. 2010; Lionel et al. 2011; Egger et al. 2014; Pinto et al. 2014). Though there is a consensus that de novo CNVs are more prevalent in cases than controls, studies of simplex families (those where only a single individual is affected) and multiplex families (those where more than one individual has an ASD) only sometimes show a greater burden of de novo CNVs in simplex than multiplex families. Two studies indicated that de novo CNVs were more common in simplex families than multiplex families (10% vs 3%) (Sebat et al. 2007) and 7% vs 2% (Marshall et al. 2008), but two more recent studies show non-significant differences between the two types of families (5.6% vs. 5.5%) (Pinto et al. 2010) and (4.9% vs. 4.2%) (Pinto et al. 2014). In affected individuals with more than one de novo CNV, the phenotype is often more severe and much more likely to be accompanied by intellectual disability (Pinto et al. 2010).

In addition to the findings surrounding de novo CNVs, additional discoveries have been made in these studies. Interestingly, rare CNVs uncovered in the ASD cohort are, on average, larger and contain more genes than those CNVs uncovered in control populations (Pinto et al. 2010; Pinto et al. 2014). These sometimes overlap loci previously implicated in genomic syndromes; however, these genomic syndromes only explain a small percentage of ASD cases. More frequently, CNVs denoted as being contributory to the ASD phenotype overlap genes that are expressed as part of the neuronal synaptic complex (Bourgeron 2009).

1.6 Copy Number Variation and Human Disease

One important fact underscored by the shortage of consistent, high impact findings from common variant studies and the greater success of rare variant studies is the genetic heterogeneity within autism. This suggests that multiple genes and loci are plausible candidates, but this doesn’t explain the whole story. Studies of inherited CNVs indicate that variable degrees of penetrance are a hallmark of many of these CNVs. This is not unexpected based on the male: female sex bias and the fact that up to 40% of the potentially causal family-specific CNVs are inherited from a parent with no detectable ASD phenotype (Devlin and Scherer 2012). One of the best examples to illustrate this phenomenon involves a four-generation family where multiple members carry a hemizygous SHANK1 deletion (Sato et al. 2012). SHANK1 is expressed at the postsynaptic membrane in excitatory synapses, acting as a scaffold for other important cellular components expressed in this region (Sheng and Kim 2000). Males carrying the deletion in this family present with a high-functioning ASD phenotype, but carrier females are unaffected (Sato et al. 2012). It has been proposed that females require a greater genetic insult than males to develop an ASD because of the presence of some protective factor (Szatmari et al. 2012). This example may explain some of the difference between males and females and ASD development.

Gene dosage changes are proposed as a possible means by which a CNV causes ASD. At the 16p11.2 locus, chromosomal gains and losses have been noted in about 1% of ASD cases (Weiss et al. 2008). While the deletions are almost always penetrant, only about half of the duplications noted are found in individuals with ASD (Fernandez et al. 2010; Devlin and Scherer 2012). In mice with deletions and duplications of the locus, changes in brain architecture and behavior are noted (Horev et al. 2011). As in humans, mice with deletions are significantly more impaired than mice with duplications (Horev et al. 2011). In many instances, potentially causal CNVs appear to contribute to the phenotype via haploinsufficiency of the genetic loci (Williams et al. 2010; Betancur and Buxbaum 2013). In such cases, if even a single copy of the gene is rendered non-functional, the individual is strongly predisposed to an ASD. However, studies of consanguineous families show that many phenotypes can arise in a recessive manner if both copies of a gene are affected (Morrow et al. 2008).

1.7 Identifying plausible candidate genes

As noted earlier, CNVs are common in the genome and even the rate of rare CNVs is equal in cases and controls. Because these studies focus on rare genetic variants, almost all of the genes that are affected by CNVs are implicated only once in a particular study. To better identify plausible candidates, there are a few different strategies that can be undertaken. Firstly, genes affected by de novo CNVs are good candidates as these mutations occur significantly more often in cases than controls (Pinto et al. 2014). This is not unexpected as mutations with severe effects are unlikely to be inherited since those with the mutation are less likely to have children. These variants would be selected against due to negative selective pressures (Julie et al. 2011; Devlin and Scherer 2012). A second means to identify putative candidates is by looking at those genes which are implicated by CNVs in unrelated individuals within a study. Rarely would one expect to see the same gene arise more than once in a rare variant study, so this finding would require further investigation. Thirdly, it is important to focus on genes with prior genetic evidence from other studies. Frequently, other groups have examined many candidate genes in model systems such as Drosophila, mouse, and iPS-derived neurons from affected individuals (Horev et al. 2011; Shcheglovitov et al. 2013; Gatto et al. 2014). In addition, other groups have examined expression and splicing patterns of ASD genes in autistic individuals (Chow et al. 2012; Uddin et al. 2014) and looked at the proteomic interactions of many of these same genes (Sakai et al. 2011). Certain gene networks have been overrepresented in cases in a number of CNV and exome sequencing studies in ASD (Iossifov et al. 2012; Pinto et al. 2014). That said, the genes arising in CNV studies frequently differ from those arising in exome and whole-genome sequencing projects though there is overlap in the pathways involved (Pinto et al. 2014). As a result, it is necessary to conduct multiple CNV studies to increase the sample sizes which can help make it more likely that the same, potentially causal genes will arise.

1.8 Neurobiology of ASD

Gene prioritization can be facilitated by utilizing knowledge of the underlying processes occurring in the brain. The brain combines and processes sensory inputs and co-ordinates motor outputs from a neuronal network to control an individual’s actions (Sudhof 2008). These neurons communicate with one another through synapses. To generate these networks though which information can be processed, neurons must progress through multiple stages which result in

11 proper synapse formation. Complications arising during any of these stages could lead to improper synapse functioning which is frequently associated with neuropsychiatric disease (Sudhof 2008).

The ASD genes that have been unearthed using the above technologies and methods fall into a range of functional classes. Owing to the significant genetic heterogeneity that exists in ASD, it is of great importance to identify the genetic networks that are believed to be involved in the disorder and to combine these results with what is known about the neurobiology to better identify causal genes to allow for more accurate diagnosis. So far, there have been a number of analyses performed that have looked at genetic and transcriptome data to classify which gene families are affected in ASD. Genes involved in neural development, cell projection and motility, and synaptic signaling were more likely to be affected by changes in copy number in the 996 individuals comprising the ASD cohort than in the population control sample (Pinto et al. 2010). One of the exome studies (Iossifov et al. 2012), showed enrichment of de novo mutations in genes involved in the FMRP pathway. The recently published Autism Genome Project--Stage 2 manuscript provided further support for these findings. In this study of nearly 2,500 ASD families, the authors identified a significant enrichment of copy number variant events in genes that target the FMRP pathway and those that play a role at the post-synaptic density (Pinto et al. 2014). An additional study attempted to combine both genomic information and gene expression data from the frontal and temporal cortexes and the cerebellum in a cohort of 19 cases and 17 controls (Voineagu et al. 2011). Using a gene-networks approach, they uncovered that genes involved in synaptic function were down-regulated in the case cohort and that many of these genes had a previous association to ASD (Voineagu et al. 2011).

The above studies and others like it have begun to explain the convergence of a genetic insult and the resulting problems in underlying neurobiology. Normal neuronal activity is often disrupted in ASD. Some of the most frequently affected genes in CNV and SNV studies are those that encode for ion channels. These channels are made up of transmembrane proteins that provide a means through which charged ions can enter or leave a cell (Neishabouri and Faisal 2014). This process is important as this movement of ions triggers action potentials, regulates cell growth, and controls the expression of some genes (Schmunk and Gargus 2013). Potassium channels, sodium channels, and calcium channels have all been identified (Schmunk and Gargus 2013). Among the most commonly implicated ion channel-encoding genes is SCN2A which has

12 been noted in several studies (Iossifov et al. 2012; Sanders et al. 2012; Jiang et al. 2013). The fact that many of these genes have also arisen in studies of epilepsy and other seizure disorders may explain some of the overlap between autism and seizure disorders.

In addition to issues with neuronal activity, impairments in transcriptional regulation at the post- synaptic density have also been observed. Genes that are a part of the (PI3K)-AKT-mTOR pathway which regulate cellular proliferation have also been previously noted in ASD (Sarbassov et al. 2005; Berg and Geschwind 2012). Additionally, genes that regulate protein metabolism at the post-synaptic density including those responsible for ubiquitination of proteins have also been found (Berg and Geschwind 2012; Tsai et al. 2012).

Another class of genes that has been repeatedly implicated in CNV and sequencing studies of ASD-affected individuals are those that play a role in establishing the structure and maintaining proper function of the synapse (Bourgeron 2009). These include genes from the NRXN, NLGN, SHANK, and DLGAP families. The protein products encoded for by the genes in these families are responsible for directing the localization of other synaptic proteins or acting as molecular scaffolds. In addition, genes that are responsible for maintaining the proper balance between excitatory and inhibitory signals have also been implicated by CNV studies of autism. The belief is that, like in situations where ion channels are affected, action potentials or neurodevelopment may be perturbed which could have a poor phenotypic outcome (Berg and Geschwind 2012).

1.9 Microarray Studies

The contribution of rare CNVs to ASD has been well described as noted above. However, due to the extreme genetic heterogeneity of the disorder, many interesting candidate genes identified have only modest support. Hence, it is important to attempt additional investigations in the same vein to identify new genes and provide more support for previously uncovered candidates. Of course, this is not meant to replace functional studies of candidate genes in ASD, but to begin to identify how frequently candidate genes and pathways are affected. So far, only a few large-scale CNV studies have been undertaken and all have been restricted by the limitations of the microarray platform used in the studies (Sebat et al. 2007; Szatmari et al. 2007; Marshall et al. 2008; Pinto et al. 2010; Levy et al. 2011; Sanders et al. 2011; Pinto et al. 2014). Specifically, these limitations refer to the inherent biases in terms of minimum detection size and probe placement which are specific to the array. Future CNV studies will need to be run on different,

13 high-resolution arrays to help identify CNVs in parts of the genome that previously may have been difficult to interrogate. The other concern pertains to the ASD populations used in the studies. Most of the investigations have focused on individuals of European ancestry. Though one would predict that most of the causal genes identified in one population would be pertinent in others, studying other populations could identify population-specific risk alleles or even variants in candidate genes that are common in other populations that may not be causal.

1.10 ASD in the Han Chinese population

ASD is thought to be under-reported outside of Western countries due to relatively poor recognition of ASD phenotypes (Elsabbagh et al. 2012). The first ASD cases in China were only first reported in the 1980s (Tao 1987) and ASD prevalence in mainland China is estimated at 10.3 per 10,000, significantly lower than recent estimates in the West (Sun et al. 2013b). Substantial differences in the prevalence have been noted even across distinct regions in China (Sun et al. 2013b). Inconsistencies between incidence in China and estimates in Western populations may be due to differences in diagnostic procedures as well as dissimilarities in the clinical ascertainment of affected individuals. Genetic studies of ASD in Han Chinese individuals have been presented primarily as either case reports or association studies of particular SNPs previously identified in studies of individuals of European ancestry. Few studies of ASD-affected individuals of Han Chinese ancestry have examined the entire genome in an unbiased manner. One notable exception is a recent manuscript that documented chromosomal abnormalities in individuals with ASD from Taiwan (Liao et al. 2013). Using karyotyping analysis in a sample of 500 probands, the authors identified four individuals with gross chromosomal abnormalities including three with sex chromosome aneuploidy (two cases having 47, XXY and a single case of 47, XYY). Sex chromosome aneuploidy has been previously noted as a risk factor underlying ASD in some individuals (Lee et al. 2012). Outside of this study, genome-wide analyses of Han Chinese ASD cohorts are rare. To date, there have been no large- scale CNV studies examining the genomes of ASD-affected individuals in the Chinese Han population.

The findings of previous CNV studies in a Han Chinese population cohort illustrate the importance of a genome-wide CNV scan in a cohort of ASD affected individuals in that population. The first is a study of the genomes of 30 individuals from three Asian populations

14 including 10 individuals of Han Chinese ethnicity (Park et al. 2010). In this study, Park and colleagues examined the genomes for copy number changes using a custom array CGH platform with 24 million probes. Their findings were significant: In total, they identified 5,177 copy number variant elements (CNV segments from different individuals were grouped together providing there was at least 50% overlap between segments). They compared these regions to CNVEs identified in European individuals (Conrad et al. 2010) and confirmed that 68.5% or 3,547 of these CNVEs were Asian-specific.

A second large study of non-European individuals examined the genomes of over 8,000 non- diseased individuals of Asian ancestry (including nearly 2,000 Han Chinese) for rare and common CNVs (Xu et al. 2011). The authors also showed substantial differences between these populations and individuals of European ancestry. Firstly, over 20% of the CNVs identified in the study were not present in the Database of Genomic Variants (http://dgv.tcag.ca/dgv/app/home) which contained CNV calls from individuals of primarily European ancestry at that time. Secondly, this study showed that most of the disease-associated common CNVs found in European populations are rare in Asians. Though the authors did find some variation within Asian populations, this diversity pales in comparison to that between European and Asian cohorts.

1.11 Rationale and Objectives of the Thesis

1.11.1 Rationale

Copy Number Variation has been shown to contribute to the risk of developing an autism spectrum disorder. There are a half-dozen large-scale CNV studies reported that examine the contribution of CNVs in ASD populations. All major studies have focused on individuals of European ancestry. While this approach has identified dozens of potentially causal genetic factors and a small number of interesting genic pathways, no study has focused on the contribution of CNVs to ASD in other populations including Han Chinese. This is especially important in light of findings from CNV studies of Han Chinese population controls which show a substantial difference in terms of the CNV architecture across different populations.

1.11.2 Hypothesis

Interrogation of the genomes of 104 parent-child trios where the proband has been diagnosed with an ASD and all family members are of Han Chinese ethnicity will identify new candidate genes, provide more support for some existing genes and pathways, and identify regions of the genome that are likely not causal based on the presence of CNVs in control individuals at these loci.

1.11.3 Objectives

1) Determine a de novo mutation rate in affected individuals and compare to the rate found in control populations.

2) Identify CNVs that overlap previously described ASD candidate loci.

3) Investigate and describe new potentially causal loci.

4) Comment on how the findings in this study can complement previous CNV studies of ASD in European populations.

Disclosure

The following manuscript was submitted to the Journal of Neurodevelopmental Disorders on June 20, 2014. It has subsequently been accepted and published after only minor revisions. This document differs slightly from the manuscript so as to stress my contributions and improve the flow of the thesis. I was responsible for analyzing and interpreting the CNV results, coordinating experiments with collaborators, performing most of the experimental validation, calculating statistical analyses shown, and drafting the manuscript.

The published manuscript describing this work can be found using the following citation:

Gazzellone, M. J., X. Zhou, et al. (2014). "Copy number variation in Han Chinese individuals with autism spectrum disorder." J Neurodev Disord 6(1): 34.

Methods 2.1 Sample Selection

ASD-affected individuals and their families were referred to the Children Development and Behavior Research Center (CDBRC) at Harbin Medical University, China by their community physician between January 2007 and June 2011. Proband diagnosis and study inclusion criteria were completed as previously described (Zhou et al. 2011). The Autism Behavior Checklist (ABC) and Childhood Autism Rating Scale (CARS) were used for diagnosis. This report examines 104 consecutive cases with an ASD diagnosis made by two psychiatrists at the CDBRC. Subsequently, DNA was also obtained from the parents of these ASD individuals. Proband participants consisted of 91 males (87.5%) and 13 females (12.5%). The mean age of the probands at enrolment was 4.31 ± 1.80 years. The study was approved by the Ethics Committee at Harbin Medical University and written consent was obtained from parents.

2.2 Genotyping of Samples

DNA samples were submitted to The Centre for Applied Genomics in Toronto and genotyping was performed on the Affymetrix CytoScan HD platform according to the manufacturer’s specifications. For arrays to be considered in the analysis, runs were required to meet or exceed the quality control thresholds specified by Affymetrix. These included a MAPD (Median Absolute Pairwise Difference) of ≤0.25, a SNP QC (SNP Quality Control) of ≥15.0, and a Waviness SD (Waviness Standard Deviation) of ≤0.12. In total, 104 probands and 206 parents met or exceeded these minimum criteria. DNA samples from two of the mothers failed to meet the appropriate thresholds and were excluded from the rest of the study. PLINK software was used to confirm the Han Chinese ethnicity of all individuals in the study from extracted SNP genotypes (Figure 2).

Figure 2: Ancestry Determination in ASD Cohort

This plot was generated using 122,368 autosomal SNPs common to both the Affymetrix 6.0 and CytoScan HD platforms. All SNPs were required to have a genotyping rate exceeding 95% and a minor allele frequency greater than 5%. The plot shows that the sample cohort clusters with individuals of Asian ethnicity while the OPGP controls cluster primarily with individuals of European ancestry. This illustrates the importance of using additional sets of Han Chinese controls when identifying rare CNVs. This plot was generated by the TCAG statistical and bioinformatics facilities.

2.3 Variant Calling

CNVs were called via the CNV detection pipeline illustrated in Figure 3. Four different CNV detection algorithms were used to limit false-positive detections and generate high-confidence calls. These algorithms included the Affymetrix Chromosome Analysis Suite (ChAS) (Affymetrix Inc., USA), iPattern (Pinto et al. 2011), Nexus (Darvishi 2010), and Partek (Downey 2006). CNVs were required to span a minimum of 5 consecutive microarray probes and 20 kb. Only those CNVs that were defined as “stringent” were subsequently analyzed. A stringent CNV implies that it was detected by one or both of ChAS or iPattern, and if detected by only one of these algorithms, also by one of Nexus or Partek. Stringent calls on the sex chromosomes required calling by both ChAS and iPattern (Figure 4). Any sample whose total number of calls exceeded the mean by three or more times the standard deviation was removed from further analysis. A total of 100 probands and 200 parents (totaling 93 complete trios) passed these quality metrics. Parentage for the 93 probands that were part of a trio was confirmed using the PLINK tool set (Purcell et al. 2007). Rare CNVs could not overlap CNVs found at a frequency of greater than 0.1% in the control cohorts and were required to be at least 50% unique by length.

2.4 Population-based control datasets

Population control datasets used to distinguish rare variants included three Han Chinese-specific cohorts and one microarray-specific set of primarily European individuals. The first included 873 DNA samples obtained as part of the Ontario Population Genomics Platform (OPGP) (Costain et al. 2013). These samples were genotyped on the Affymetrix CytoScan HD array in the same fashion as the ASD cases and analyzed for CNVs using the same methods as those obtained from the ASD cohort. Over 95% of the samples in this cohort were obtained from individuals of European ancestry. These samples were used as the primary control dataset for use in the detection of rare copy number changes and to account for calling biases inherent to this array. The other three datasets consisted of individuals of Han Chinese ethnicity and comprised a secondary set by which rare CNVs specific to the Han Chinese population could be detected. The first included 170 Han Chinese individuals from the HapMap project (Altshuler et al. 2010). The second included 147 Han Chinese controls genotyped by Lou et al (Lou et al. 2011). Both of these datasets were run on the Affymetrix 6.0 array. The third dataset contained 918 Han Chinese samples collected as part of the Singapore Genome Variation Project (Xu et al. 2011).

These samples were genotyped on the Illumina 1M Duo array. The use of these different cohorts enables the distinction of those CNVs that are truly rare in the Chinese population as opposed to rare only when compared to samples obtained from European individuals. For all case and control samples, genotyping and CNV calling were performed using identical procedures.

Figure 3: CNV Detection Workflow

Figure 4: Number of calls from each algorithm per sample

This image was generated by plotting the average number of CNV calls for female and male probands in both autosomes and sex chromosomes. This graphic illustrates how the average number of autosomal calls is considerably higher than sex chromosome calls from ChAS and iPattern, but not in Nexus and Partek (especially when compared to sex chromosome calls in males). Calls on the sex chromosomes from Nexus are substantially elevated in males compared to females as well as compared to autosomal calls, indicating poor suitability for use in generating stringent calls on the sex chromosomes. Few calls are made on the X and Y from Partek, again necessitating its exclusion when generating the set of stringent calls from the sex chromosomes.

2.5 Experimental Validation of CNV Calls

CNVs of interest were validated using SYBR-Green based real-time quantitative PCR (qPCR) where possible. Primers were designed to amplify a region 90-140 bp in size using Primer3 software v. 0.4.0 (http://bioinfo.ut.ee/primer3-0.4.0/). Control primers were designed within the FOXP2 locus which was used as a diploid locus control (Forward Primer Sequence: 5’ TGC TAG AGG AGT GGG ACA AGT A 3’; Reverse Primer Sequence: 5’ GAA GCA GGA CTC TAA GTG CAG A 3’). All primers were tested by generating a PCR fragment using genomic DNA from a HapMap sample and running on a 1% agarose gel (Figure 5). The assay utilized a standard curve was generated by preparing serial dilutions of a standard sample of known concentration. Proband and paternal DNA were interrogated using this method. In addition, two HapMap samples (NA15510 (Female) and NA10851 (Male)), were used as “normal copy number” controls. All experiments were performed in triplicate. If the ratio of signal at the locus of interest to the signal at the FOXP2 locus was below 0.7, a loss was confirmed (Figure 6). If over 1.3, a gain was called. In both cases, two-copy controls were expected to have a ratio of about 1.0, indicating no copy number change in these samples. A TaqMan Copy Number Reference Assay was used to confirm additional copy number changes found in probands and their parents. Predesigned probes were selected within the target regions of interest and a reference assay amplified a two copy region. Both NA15510 (Female) and NA10851 (Male) were used as two-copy controls for autosomal CNVs. Validation methods confirmed 96% (27/28) of the CNVs tested.

Figure 5: Agarose gel picture illustrating successful PCR amplification of FOXP2 primers

The FOXP2 primers were tested to confirm amplification of the 140 bp region using Taq2000 DNA polymerase (Agilent Technologies, USA). A 25 µL reaction was prepared including 10 µM of each of the forward and reverse primers, 2.5 µL of Taq Polymerase Buffer, 20 ng of genomic DNA, 2.5 µL of 2 mM dNTPs, and 1 U of Taq2000 DNA Polymerase. The PCR protocol was established using a 2720 Thermocycler from Applied Biosystems. It employed an initial cycle of 95°C for 1 minute, followed by 35 cycles of 95°C, 60°C, and 72°C for one minute each, and then a 10 minute elongation step at 72°C.

Figure 6: qPCR assay indicating inherited loss at 10p12.33

This image shows the ratio of copies of the 10p12.33 locus to FOXP2 in an ASD trio and in a male and female control using qPCR. No change in copy number is present in the controls or maternal sample, however, the father and proband show a loss of one copy of 10p12.33. This is illustrated by the ratio of approximately 0.5.

2.6 Expression Analysis

The SuperScript III First-Strand Synthesis SuperMix for qRT-PCR kit (Life Technologies, USA) was used to generate cDNA from RNA extracted from 11 tissues and a whole brain sample. As described in the protocol, 1 µg of RNA from each of the respective tissues was used to generate cDNA. Following cDNA production, each sample was diluted 1:10. Expression analysis and tissue distribution of CASKIN1 and PKD1 were illustrated using quantitative RT-PCR. To ensure that genomic DNA was not amplified, primers were generated in adjacent exons and tested via PCR and subsequent gel electrophoresis confirmation (Figures 7). This was corroborated by sequencing the PCR products. The sequencing products are shown in Figures 8-9. The housekeeping genes MED13 and ACTB were used to normalize the expression level. The primer sequences can be found in Table 1.

Table 1: Primers used for Expression Assays Gene Primer Sequence Forward: 5’ TAC TTG GCG TAT TCA GCC TTC 3’ CASKIN1 Reverse: 5’ CGA GAA CAT TGA TTT CAT CAC C 3’ Forward: 5’ TGC CAA ATC CTT CTC AGC AT 3’ PKD1 Reverse: 5’ CGT TTC CAT GTG GGT GTC TT 3’ Forward: 5’ ATT GCC GAC AGG ATG CAG A 3’ ACTB Reverse: 5’ GAG TAC TTG CGC TCA GGA GGA 3’ Forward: 5’ CCG CAT CCT GAT GTG TCT GA 3’ MED13 Reverse: 5’ TTG CAG GTG GAT ACG TGA CT 3’

Figure 7: Testing of CASKIN1 and PKD1 primer sets

This gel image illustrates the size of the PCR product generated using the CASKIN1 and PKD1 primer sets as tested using cDNA generated from a whole brain tissue sample. A 1 kb ladder from Agilent (Agilent Technologies, USA) is run in the first column, a non-template PCR control in the second, and two replicates using cDNA in the next two columns. The absence of PCR products in the negative control shows successful degradation of any remaining RNA following treatment with RNAse and the lack of genomic DNA contamination. The amplification of the small the 138 bp and 100 bp fragments illustrates that the primers are only amplifying the regions of interest in the cDNA.

Figure 8: Confirmed amplification of CASKIN1 exons of interest

The above image shows a screenshot generated from the UCSC Genome Browser (https://genome.ucsc.edu/index.html). The location of the primers is shown in the track towards the top of the image (indicated by solid black bars). The sequence generated from the PCR product shown in Figure 7 using the Forward primer is indicated on the second track. The presence of the coding portion of CASKIN1 is shown on the third track. The solid black line in the second track confirms that only the cDNA was amplified by these primers. This suggests that this primer set is of utility for expression experiments discussed below.

Figure 9: Confirmed amplification of PKD1 exons of interest

The above image shows a screenshot generated from the UCSC Genome Browser (https://genome.ucsc.edu/index.html). The location of the primers is shown in the track towards the top of the image (indicated by solid black bars). The sequence generated from the PCR product shown in Figure 7 using the Forward primer is indicated on the second track. The presence of the coding portion of PKD1 is shown on the third track. The solid black line in the second track confirms that only the cDNA was amplified by these primers. This suggests that this primer set is of utility for expression experiments discussed below.

Results

Using a high-resolution CNV genotyping array and a well-established CNV calling protocol, 241 rare CNVs were identified in the probands (See Appendix). Eleven of the probands from the 93 complete trios (11.8%) carried a de novo or rare inherited CNV that may contribute to ASD (Table 2). The CNV profiles of both the ASD cases and their parents indicated that there is no significant difference between the overall CNV call rate or the length of CNVs between these two groups (Table 3). After filtering for rare variants, an average of 2.41 rare CNVs was discovered per sample. A quarter of probands in the cohort had two rare CNVs and the remainder had between 0 and 7 (Figure 10)

Table 2: Summary of de novo and rare inherited CNVs of interest in ASD probands. Sample ID Cytoband Co-ordinates Type of Genes affected Inheritance (hg19) CNV 683-3 2q37.1 Chr2: 233,651,280- 22 kb GIGYF2 De novo (Female) 233,673,273 Deletion 527-3 4q28.1 Chr4: 124,063,146- 982 kb SPRY1,SPATA5 De novo (Male) 125,045,116 Duplication 517-3 16p13.3 Chr16: 843,861- 319 kb 7 genes De novo (Female) 1,162,728 Duplication 16p13.3 Chr16: 2,088,391- 327 kb 15 genes De novo 2,415,016 Duplication 503-3 16p11.2 Chr16: 28,819,029- 232 kb 9 genes De novo (Male) 29,051,191 Deletion 692-3 17p13.3, Chr17: 2,455,643- 994 kb 16 genes De novo (Male) 17p13.2 3,449,869 Duplication 567-3 Xp21.1 ChrX: 31,805,650- 154 kb DMD De novo (Male) 31,959,887 Deletion 611-3 Xp21.1 ChrX: 32,548,066- 55 kb DMD De novo (Male) 32,603,018 Deletion 552-3 Xq13.2 ChrX: 72,319,907- 33 kb NAP1L6 De novo (Male) 72,353,391 Deletion 694-3 4q22.2 Chr4: 94,144,621- 28 kb GRID2 Maternal (Male) 94,172,410 Deletion 9p21.1 Chr9: 28,491,679- 139 kb LINGO2 Paternal 28,630,598 Deletion (intronic) 511-3 9p21.1 Chr9: 28,464,218- 132 kb LINGO2 Maternal (Male) 28,596,286 Deletion 686-3 10p12.33 Chr10: 18,240,592- 73 kb SLC39A12 Paternal (Male) 18,313,842 Deletion

Table 3: Summary statistics of stringent CNVs larger than 20 kb ASD Probands Parents OPGP Controls Samples 100 200 873 Males/Females 90/10 100/100 477/396 #Stringent CNVs 1,638 3,260 15,872 Mean CNVs/sample ± 16.38±4.00 16.30±3.84 18.18±4.41 SDa Median 16 16 18 Mean CNV size (kb) ± 99.04±213.32 89.41±181.08 90.27±166.64 SDb Median CNV size (kb) 44.54 43.98 42.11 %Gain/%Loss 43.1%/56.9% 44.4%/55.6% 43.2%/56.8% #CNVs > 1Mb (%) 14 (0.85%) 21 (0.64%) 116 (0.73%) #CNVs 100kb-1Mb (%) 353 (21.55%) 658 (20.18%) 3190 (20.10%) #CNVs 20 kb-100kb (%) 1,271 (77.60%) 2,581 (79.18%) 12,566 (79.17%)

a There is no significant difference between the mean number of CNV calls in ASD probands and their parents (p=0.8669 using an unpaired two-tailed Fisher’s exact test). There is a significant difference between the number of calls noted in ASD cases and OPGP controls run on the same array (p=0.0001). This is likely due to some batch effect and is not representative of some biological difference between cases and controls as the number of calls in controls is higher. b There is no difference between the mean CNV size in ASD probands and either their parents or OPGP controls (p=0.6831 and p=0.6292, respectively).

Figure 10: Number of Rare CNVs per proband

The number of rare CNVs larger than 20 kb in each proband is plotted in this chart. The mean number of rare CNVs uncovered in probands is 2.41 and the median number of rare CNVs is 2.

3.1 De novo Variants

Nine de novo CNV events were detected in eight probands (Table 1). This represents a de novo CNV rate of 8.6% (8/93), similar to that which has been seen previously in other non-Chinese CNV studies (Marshall et al. 2008; Pinto et al. 2010; Egger et al. 2014; Pinto et al. 2014). In one female proband (683-3), a 22 kb deletion was discovered at 2q37.1 overlapping GIGYF2, which lies in a susceptibility locus for familial Parkinson’s disease (Lautier et al. 2008) (Figure 11). No ASD phenotype has been previously associated with variants affecting this gene.

A 982 kb de novo duplication was uncovered in male proband 527-3. This duplication overlaps SPRY1, a gene that regulates fibroblast growth factor signaling which plays an important role in the patterning and propagation of cells in the developing brain (Yu et al. 2011). There is no prior evidence of any link between duplications overlapping this gene or SPATA5 and ASD.

A pair of adjacent de novo duplications at 16p13.3 separated by nearly 1 Mb of two-copy intervening sequence was uncovered in a six-year-old female (517-3). To better assess which genes might be contributing to the phenotype, I referred to a recent finding showing that highly brain-expressed exons have a lower burden of rare missense variants than more ubiquitously- expressed exons and are targets for penetrant mutations in ASD (Uddin et al. 2014). Each of the exons in the genes residing within the duplications was examined to determine if any of these were characterized as a “brain-critical exon”. This was the case for at least one exon in RAB26, PKD1, E4F1, ABCA3, and CASKIN1 (Figure 12). CASKIN1 and PKD1 received special focus due to previous studies indicating that they may play some role in synaptic scaffolding and neurodevelopment (Bisbal et al. 2008; Stafford et al. 2011). In these two genes, the presence of a brain-expressed isoform whose dosage could be impacted by the duplication was confirmed (Figure 12).

Figure 11: Pictorial Representation of GIGYF2 deletion using ChAS software

This image illustrates the logR ratios of the proband and her parents. Below that is a representation of the different genes present at the locus. In the proband there is a substantial decrease in the probe hybridization intensity within the GIGYF2 gene, which is reflective of a deletion of this segment of DNA relative to a two-copy standard. Neither parent shows a similar loss (which has been verified using qPCR), confirming the de novo nature of the CNV.

Figure 12: Identification of Brain-Critical Exons at 16p13.3

The exon score was computed as described previously (Uddin et al. 2014) for each exon within the 16p13.3 duplications in the female proband to identify “brain-critical exons”. Of the five genes scoring above zero, CASKIN1 and PKD1 were selected for further investigation. The expression of these exons within these genes is higher in cerebellum and whole brain using a quantitative real-time PCR assay which looked at relative expression in these genes compared to MED13 (ACTB was used instead in a confirmatory assay).

A male proband (503-3) was found to carry a 232 kb de novo microdeletion at 16p11.2 (Figure 13). This deletion lies upstream the 600 kb ASD-implicated risk locus (Devlin and Scherer 2012). Similarly sized deletions have been previously noted in individuals with obesity and developmental delay (Bochukova et al. 2010; Egger et al. 2014). Unlike many of the individuals with similar deletions, this male has a BMI of 18.26, putting him in the normal range.

A 994 kb duplication of 17p13.3-17p13.2 was uncovered in a five-year-old autistic male (692-3). Microduplications have been previously noted at this locus and are usually associated with developmental delay and frequently with growth issues (Bi et al. 2009). This proband has typical autistic features and no evidence of any growth problems.

Two unrelated male probands harbor de novo exonic deletions of the Duchenne Muscular Dystrophy (DMD) gene (Figure 14). The first case (611-3), a five-year-old autistic male presenting with ASD, hypotonia, and progressive motor impairments including difficulty walking, has a 55 kb deletion that overlaps exons 14-17. The second proband (567-3), an autistic six-year-old male with abnormal muscular development has a deletion of 154 kb which overlaps exons 46-50. In both cases, the deletion is predicted to cause a frameshift leading to a premature stop and loss of dystrophin. In males, such mutations are predicted to result in DMD. Studies have shown a higher incidence of ASD in boys with DMD, possibly because of a secondary synaptic role for the protein (Wu et al. 2005).

A 33.5 kb deletion at Xq13.2 has also been noted in male case 552-3. This deletion overlaps the NAP1L6 nucleosome assembly protein. There are no published reports of deletions affecting this gene.

Figure 13: 16p11.2 region of interest

This image illustrates the location of the 232 kb deletion relative to the more common 600 kb 16p11.2 deletion/duplication region. The high number of segmental duplications at this locus (as noted at the top of the image) predisposes the region to copy number changes through non-allelic homologous recombination (Dittwald et al. 2013). This image was inspired by a previously published image in Genetics in Medicine (Bachmann-Gagescu et al. 2010)

Figure 14: Location of DMD Deletions

This graphic indicates the locations of the two de novo deletions overlapping DMD in the cohort. The image has been adapted from the dystrophin transcript display as illustrated by Ensembl (http://useast.ensembl.org/index.html). The transcript reads from the right to left (accounting for the fact that it is on the reverse DNA strand) and spans over 2 Mb making it the largest protein-coding gene encoded by the human genome (van Deutekom and van Ommen 2003). The two deletions occur in the major and minor hotspots for DMD deletions (Den Dunnen et al. 1989; Fletcher et al. 2012).

3.2 Rare Inherited Variants

A 27.8 kb loss of two exons of GRID2 was identified in a four-year-old male (694-3). GRID2 encodes a glutamate receptor channel subunit and mutations within GRID2 have been associated with ASD (Pinto et al. 2014). This proband inherits the mutation from his mother who has a diagnosis of intellectual disability. His maternal grandfather and a maternal cousin also have intellectual disability, but no DNA was available to test the segregation of this variant in these individuals (Figure 15). The proband does not have a diagnosis of intellectual disability.

A 73 kb deletion affecting SLC39A12 was found in a six-year-old male proband (686-3). All but the last exon of the gene was deleted in the proband and his unaffected father. A recent study suggests that this zinc transporter stimulates neurite outgrowth during neurodevelopment (Chowanadisai et al. 2013).

Finally, a male proband (511-3) with a maternally-inherited deletion of one exon of LINGO2 was identified. The gene has been previously associated with adult-onset neurodegenerative disorders and has been implicated as an ASD risk gene (Wu et al. 2011). A second male proband (694-3) in this study also harbors an intronic CNV within this gene, but its potential effect on gene expression and any possible contribution to phenotype was not possible to ascertain.

Figure 15: Pedigree of family 694 with GRID2 deletion

This figure shows how Intellectual Disability and ASD segregate through the family. If DNA can be obtained from the other members of the family, it would be interesting to see how the variant segregates with the phenotype. It would also be helpful to get more phenotype information for the uncle of the proband to see if he also has any neurocognitive issues. This figure was prepared with assistance from Dr. Xue Zhou.

3.3 Population-Specific CNV Polymorphisms

A population-specific CNV polymorphism in YWHAE, a gene previously speculated to have a role in ASD, was also identified (Bi et al. 2009). This 24 kb duplication (chr17:1,235,975- 1,259,833) overlaps the last exon of this gene in one proband and his mother. No microduplications of YWHAE were found in Caucasians in the Ontario population controls. However, similar duplications are present in two unrelated parents of other probands in this Han Chinese cohort (Figure 16). Fine-mapping of the breakpoints in the four different samples carrying the rearrangement found that both the 3’ and 5’ ends consistently map to the same regions suggesting the CNVs likely represent the same ancestral event. The precise sites of the breakpoints were not possible to characterize due to complex sequence elements located in the region. Subsequently, the frequency of this event in the Chinese population was assessed and 11/1,235 (0.9%) of the Han Chinese population controls studied were found to carry a microduplication at this locus. Using a Taqman Copy Number assay for a probe located within the breakpoints of the microduplication, duplications in 3/260 additional Han Chinese controls were found. In all, ~1% of Han Chinese individuals have this duplication, regardless of ASD status. There is no statistically significant difference between the frequency of this variant in cases versus controls (p=1.000 using a two-tailed Fisher’s exact test).

Figure 16: Genomic location of YWHAE duplications

The location of the duplications (represented by blue bars) in the four individuals tested overlapping the 3’ end of YWHAE locus are shown. Eleven similarly sized duplications are present in Chinese control samples (from samples run on different arrays). Three additional duplications in a second cohort of Chinese population controls were also found at this locus by screening using a quantitative assay (the site of the TaqMan Copy Number Assay is indicated by the vertical arrow).

Discussion

The data suggest that de novo and rare-inherited CNVs have a role in ASD in Han Chinese populations in a manner similar to what has been observed in published studies of Caucasian families with ASD. This work adds to the growing number of CNV studies in autism, all of which are critical for accurate interpretation of post- and pre-natal testing being performed or contemplated in the clinical genetics setting (Miller et al. 2010; Carter and Scherer 2013).

The prevalence of ASD in mainland China and Western countries varies considerably (Elsabbagh et al. 2012; Sun et al. 2013b; Wingate et al. 2014). This exploratory study of the role of CNVs in ASD families in China does not yet provide a genetic explanation for such differences and it is more likely that the prevalence differs due to factors regarding the ascertainment of individuals. Specifically, it is hypothesized that differences in assessment protocols and cultural expectations (especially with regards to eye contact and speech) could influence the recognition of autistic behaviors in ASD individuals with normal or mildly impaired cognitive functioning (Elsabbagh et al. 2012; Sun et al. 2013b). Moreover, parental attitudes and a stigma surrounding neuropsychiatric conditions may also contribute to a reticence towards accepting a potential autism diagnosis (Sun et al. 2013b), possibly precluding an affected child from being enrolled in any study of ASD.

The CARS and ABC screening instruments, assessment tools that are frequently used in mainland China (including this study) but not recent Western studies, may contribute to some diagnostic bias. ASD prevalence estimates in studies using the Autism Behavior Checklist are typically lower than those using other diagnostic instruments as the ABC scale preferentially identifies classical autism and often overlooks high functioning cases (Sun et al. 2013b). Similarly, the CARS diagnostic scale was created using DSM-III criteria and may not account for updates in subsequent editions of the manual (Perry et al. 2005). Some groups have argued that the CARS score may also be biased by intellectual disability in the patients, again resulting in the omission of less severely affected individuals (Perry et al. 2005).

Another potential confounder arises from the fact that many Chinese pediatricians lack training in ASD recognition and awareness (Sun et al. 2013a). This is at least partially reflected in a study showing that autism prevalence is higher in urban vs. non-urban areas (1.37% vs 0.80%) where it

46 would be less likely to encounter a physician with knowledge of ASD (Zhang and Ji 2005; Elsabbagh et al. 2012). In this manuscript, 64 affected individuals were from urban areas while 40 were from rural areas.

The de novo CNV findings in this study continue to stress the importance of synaptic and neurodevelopmental genes in the genetic architecture of ASD. CNVs overlapping previously known genetic loci contributing to ASD including variants overlapping 16p11.2, 17p13.3- 17p13.2, and DMD were identified. Interestingly, while the expected incidence of Duchenne Muscular Dystrophy is 1/3500 males, the incidence may be as high as 2/90 males in this cohort. Using a binomial distribution, the likelihood of seeing 2 or more cases in 90 males is very small (p=0.00032). The high incidence of potential Duchenne’s cases in this cohort may be reflective of some issue regarding sample ascertainment. Both of these individuals did not previously have a Duchenne’s diagnosis and were not previously assessed for muscular dystrophy. They were likely submitted to the ASD clinic due to their neurocognitive issues. As a result, testing for DMD in these individuals should be completed as soon as possible.

Perhaps the most important finding in this paper is the identification of the YWHAE CNV that appears to be a Chinese specific polymorphism and not an ASD (or developmental delay)- associated variant. The gene product, 14-3-3ε, is a member of the 14-3-3 family of genes which are highly conserved across species and play an important role in protein regulation and signal transduction (Toyo-oka et al. 2003). Changes in 14-3-3ε dosage have been shown to alter neuronal migration, thereby impairing proper neurodevelopment (Toyo-oka et al. 2003). Several studies have previously indicated that large duplications or deletions at 17p13.3 affecting this gene are associated with developmental delay or autism (Bi et al. 2009; Nagamani et al. 2009; Bruno et al. 2010; Curry et al. 2013). The small 24 kb duplication that was identified in this study has also been noted in past work (Prasad et al. 2012) and in clinical cohorts. In light of these findings, all cases with the 24 kb duplication were re-evaluated and they were found to cluster with Asian individuals when determining ancestry from extracted SNPs. Unlike the larger CNVs previously identified at this locus (Bi et al. 2009; Nagamani et al. 2009; Bruno et al. 2010; Curry et al. 2013), this small duplication is not likely to be pathogenic since the frequency of the event is similar in cases and in population controls.

The YWHAE finding provides an example of the importance of using ancestry-matched controls when characterizing the clinical relevance of rare variants in a population. As CNV studies become routine in other ASD populations, researchers will discover additional examples of population-specific CNVs, which are overlapping ASD candidate genes, but are non-pathogenic (Kamien et al. 2014).

The results of this study substantiate the extensive genetic heterogeneity that is inherent to autism. As a result, it is of great importance to examine the totality of genetic variation in order to accurately identify new, potentially causal genes. Though whole-exome and whole-genome technologies already identify interesting single nucleotide variants and show promise in detecting in/dels and copy number changes, microarrays still remain the gold-standard for CNV detection. Future CNV studies of Han Chinese individuals can build upon the foundation established by this study. Further work in this population can identify new population-specific variants that may contribute to ASD, as well as add further support for the contribution of rare CNVs to ASD.

This study serves as a pilot to provide initial insight into the genetic architecture of ASD in the Han Chinese population. Data from our lab (Jiang et al. 2013) and ongoing genome sequencing experiments (Qiu and Hayden 2008) indicate that CNVs need to be seriously considered when annotating variants. At this time, CNV testing in the Han Chinese population would be most appropriate in a confirmatory diagnostic setting, at least until larger cohorts of matched control data become available for comparisons.

In the meantime, some steps can be undertaken to improve ASD diagnosis in China. The use of modern screening instruments (i.e. ADOS and ADI-R) will allow for better identification of individuals with ASD and will allow improved comparisons between the Western and Han Chinese populations. Additionally, increased ASD research and training in Chinese medical schools could improve diagnosis of higher-functioning cases. The implementation of these suggestions will enhance the quality of future genetic studies in the Han Chinese population and will facilitate more rapid detection of ASD and quicker behavioral intervention. Finally, this study should serve as a starting point for further genetic studies of ASD in the Han Chinese population as well as other less-studied populations to identify new candidate genes, provide

48 more support for existing risk loci, and to identify non-pathogenic variants overlapping plausible candidate genes.

Summary and Significance

This study has significant immediate implications for clinical genetic testing in the Han Chinese population. China is the fastest-growing market for molecular diagnostics (Lee-Olsen 2012) and clinical microarrays are now being used for both prenatal and postnatal diagnosis of a multitude of disorders (Petrone 2012; Fu et al. 2014). In addition, outside of China and particularly in large cities like Toronto, the Han Chinese population is increasing, entailing a need for additional studies of Han Chinese individuals as more are expected to come through clinical genetics clinics in these places. As has been shown in this study, it is likely that Han Chinese-specific population variants that overlap previously discovered candidate genes will be uncovered in some individuals. Again, this stresses the importance of using a population-matched control dataset in order to properly prioritize rare genomic variants.

The results of this study also add to the growing body of knowledge showing the contribution of copy number variation to autism spectrum disorder. This study replicated risk genes/loci previously identified including 16p13.3, 16p11.2, 17p13.3-17p13.2, DMD, GRID2, and LINGO2. In addition, it identified novel candidates including GIGYF2, SPRY1, and SLC39A12. Although these genes cannot be definitively labeled as being causal, additional support may be gained from subsequent CNV/ sequencing studies. This study presents the first de novo CNV rate in a Han Chinese cohort of ASD individuals (8.6%) and has shown that this value does not differ considerably to what has been previously noted in European individuals (Marshall et al. 2008; Pinto et al. 2010; Egger et al. 2014; Pinto et al. 2014).

This study, and others like it, can contribute to better classification of copy number variants. Microarrays are now used as a first-pass technology when children present with autism spectrum disorder, developmental delay, or congenital abnormalities (Kearney et al. 2011). At present, CNVs are listed as being pathogenic, of unknown clinical significance, and benign according to American College of Medical Genetics standards (Kearney et al. 2011). These classifications are made according to previously published peer-reviewed studies. This study can help better characterize many of the CNVs found in the Han Chinese population. The YWHAE finding, among others, shows the importance of continually updating the clinical significance of the CNVs based on novel results from new studies.

Several implications for genetic counselling arise because of the work presented here. The first and most obvious consequence is that this study adds to the growing evidence that genetic factors play a role in the etiology of the disorder. Many parents of autistic children, particularly in the 1950s and 1960s, were blamed for causing their child’s autism by poorly rearing their child (Ramachandran and Oberman 2006). Though this viewpoint is no longer accepted in scientific circles, many parents still feel that they are somewhat to blame for their child’s condition (Fung et al. 2007). This self-stigmatization can result in a number of negative consequences for both the affected child and their family. In many cases, parents are more likely to suffer from additional marital tensions, parenting stress, and depression (Mak and Kwok 2010). Much of the stress is rooted in the fact that many children have not received any definitive molecular diagnosis (Narcisa et al. 2013). Genetic studies have the capacity to reduce much of this stress and put an end the diagnostic odyssey that many parents experience when searching for a cause of their child’s autism. This study ends this odyssey for several of the parents. In 11 of the 93 families, a risk variant that might explain the child’s condition was identified. More importantly, in the 8 families where the CNV of interest appears only in the child (de novo), the findings can give some consolation to those parents as the variant is unlikely to arise in any other children that they may have or may contemplate having in the future.

Another important consideration for genetic counselling is the recurrence risk in subsequent children in a family. A recent study in California showed that parents of an autistic child have a lower reproductive rate by nearly one third once the first signs of ASD are visible in a child than parents of a phenotypically “normal” child (Hoffmann et al. 2014). This effect is noted regardless of whether or not this is the couple’s first child. This reproductive stoppage could be due to several factors including an increase in parental stress, but much of this may have to do with the belief that subsequent children may be at risk of ASD (Hoffmann et al. 2014). Recurrence risk for ASD has been recently calculated in a large Danish study. In full siblings, the likelihood that the sibling of an already ASD-diagnosed child will also receive a diagnosis is 7.5% (2.4% in half-siblings with the same mother and around 1% in half-siblings with the same father) (Gronborg et al. 2013). Parents often forgo having any additional children because of these risks. However, for the parents of children carrying pathogenic de novo CNVs in this study, the results should provide them with some comfort as the recurrence risk is much lower as these mutations would not be inherited in other children (Devlin and Scherer 2012).

Another potential benefit of genetic studies is quicker detection of ASD cases using genetic biomarkers. This can have substantial benefits for affected individuals as studies have shown that earlier behavioral intervention leads to better outcomes. One of the best interventions is the Early Start Denver Model (ESDM) which is a behavioral intervention program that can begin at as early as 12 months (Dawson et al. 2010). This method has been shown to facilitate significant improvements in IQ, language acquisition, and behavior while lessening the overall level of impairment in the child (Dawson et al. 2010). Additionally, this intervention has been shown to partially normalize brain activity (Dawson et al. 2012). Earlier diagnosis via genetic methods before symptoms are even noticeable can also save the health-care system a substantial amount of money. One study showed that the societal cost of autism is $3.2 million over a lifetime with much of the costs associated with diagnosis and treatment (Ganz 2007). Earlier diagnosis can reduce the diagnostic odyssey where children undergo thousands of dollars’ worth of tests to determine their condition and its causes as well as improve functioning which can limit some of the treatment costs down the line.

Another emerging consequence of genetic studies is the discovery of potential pharmaceutical products that seek to alleviate negative ASD symptoms. Until now, the vast majority of drugs prescribed to autistic children were targeting secondary co-morbidities and none of the three dimensions (social-communication, language, interaction) specific to autism (West et al. 2009). Recent data from animal models and early-stage clinical trials has begun to address some of the core symptoms of autism, specifically social communication. One of the best examples is a drug currently in clinical trials that is supposed to alleviate some of these symptoms in individuals with Fragile X syndrome. FMR1 encodes a protein that reduces protein synthesis initiated by stimulation of metabotropic glutamate receptors (mGluRs) (Santoro et al. 2012). In mice, this syndrome is believed to alter the equilibrium of excitatory and inhibitory signals by reducing GABAergic inhibition which results in repetitive behaviors and social withdrawal (Coghlan et al. 2012; Baribeau and Anagnostou 2014). Promising research involving GABA agonists, specifically a drug called arbaclofen, showed normalized protein synthesis, improved social interactions, and fewer repetitive behaviors in mice (Henderson et al. 2012). An 8-week open label study showed improvements in a small cohort of patients with fragile X syndrome, necessitating additional placebo-controlled studies (Erickson et al. 2014). Though compounds like arbaclofen are not yet FDA-approved and much more research needs to be done to

52 determine their efficacy, they hold great promise that one day, certain pharmaceutical drugs might be prescribed based on the genetic findings of studies like this one. This is arguably the most important implication of the work presented here and one that could have substantial positive effects in the future.

Future Directions

This pilot study illustrates the importance of the continuing use of microarrays to identify copy number variation and the necessity of analysing different population cohorts. The CytoScan HD array that was used here differs from other arrays in terms of probe density and location of markers. This study is one of the first using this microarray in a cohort of autistic individuals. The analysis of different cohorts of ASD individuals on this array may identify new candidate genes or loci that were not well covered by previous microarray technologies. Screening of larger Chinese ASD cohorts also has the potential to identify Chinese-specific risk loci or population-specific variants that may not contribute to phenotype but might be rare outside of the European population previously tested. To better identify rare CNVs, the study of additional population controls (especially from Han Chinese and other less-studied ethnicities) would be helpful. Interrogating the genomes of control individuals using high-density microarrays can facilitate a more accurate identification of rare CNVs in the Chinese population.

Microarrays continue to provide important information that can improve the research community’s understanding of autism. At this time, the SFARI Gene database lists over 500 candidate genes associated with ASD, but the level of confidence for some of these genes is very low (https://gene.sfari.org). In many cases, a large number of these genes were only identified in one or two studies and additional confirmation of the variants is lacking. As a result, in genetics clinics, CNVs overlapping these loci are still labeled as being of unknown or potential pathogenicity. This is not unexpected since rare genetic variants are proposed to contribute to the etiology of the disorder. As a result, larger numbers of samples need to be run to better identify causal genetic variants. In addition, having more samples can help better implicate specific pathways affected in ASD. The early exome sequencing studies (Iossifov et al. 2012; Neale et al. 2012; O'Roak et al. 2012; Sanders et al. 2012) attempted to identify some of these pathways with varying degrees of success. Though a couple of pathways were identified including the Wnt- signalling and beta-catenin pathways as well as targets of FMR1 (Iossifov et al. 2012; O'Roak et al. 2012), the lack of samples and difficulties with interpretation of variants contributed to difficulties. Members of my lab and others have identified pathways involved in neuronal development (Hadley et al. 2014; Pinto et al. 2014), but more work needs to be done.

Another important avenue of further research is determining the phenotypic outcome of various genetic variants. Studies are now showing that many of the genes and loci implicated in one neurodevelopmental disorder are also involved in others. Attention Deficit Hyperactivity Disorder (ADHD) is co-occurring in about one third to half of ASD individuals, epilepsy in between 8-30%, and Obsessive Compulsive Disorder (OCD) in up to 25% (Lai et al. 2014). What is most interesting is that for many individuals with these and other neurodevelopmental disorders caused by variants overlapping a gene also involved in ASD, ASD symptoms are not present. Studies have begun to look into the overlap between many of these disorders and identify why some individuals with mutations overlapping a certain gene get one disorder and why some get another. I was a co-author on a study that examined ASTN2, a gene that plays an important role in neuronal migration during brain development and was implicated in individuals with neurodevelopmental disorders (Lionel et al. 2014). An enrichment of deletions was noted at the 3’ end of the gene in patients, which impacts all isoforms of ASTN2 (Lionel et al. 2014). OCD, ADHD, ASD, and speech delay were most often observed in the probands. I am also leading a project examining the contribution of rare CNVs in OCD. At this point in time, there are no published studies examining the contribution of rare CNVs to OCD (Pauls et al. 2014). In my preliminary analysis, which has not yet been reported, I have already noticed that some ASD candidate genes have already come up in this study of OCD probands. Much more research is required, but it will be important to look at both the location and size of the variants as well as which domains of the proteins are affected. Perhaps by observing this and comparing these results to what has been previously uncovered in ASD, the genotype-phenotype correlation can be better understood.

Further research is also necessary with respect to interpreting microduplications. Rare copy number losses are comparatively easier to interpret than gains as the loss of a copy of a gene is frequently reflected by a corresponding decrease in gene dosage (Chenier et al. 2014). For duplications, that is not always the case. Depending on the position of the duplicated segment of DNA, the effects vary. Should promoter sequences and enhancers get incorporated into the duplication in addition to the gene of interest, it is plausible that this additional copy could also be transcribed which would alter the dosage of the gene product. This could have some phenotypic effect. It is also possible that only a portion of the gene is duplicated and that that portion lies in tandem to the gene product or in a gene desert elsewhere in the genome. In these

55 cases, the duplication might not have any effect on dosage and may not contribute to a phenotype. Another possible outcome is that the duplicated portion is translocated into another gene. This could plausibly knock-out the function of this other gene or produce a new gene product with some new function. Some groups have attempted to develop guidelines to interpret copy number gains (Hanemaaijer et al. 2012), but they often rely on comparisons to controls or databases of previously implicated risk variants. Some strategies such as fluorescence in situ hybridization (FISH) can indicate whether the duplication occurs in tandem or whether it exists on another chromosome. Next-generation sequencing technologies can better pinpoint the location of these duplications and better infer the effect of the duplication. The development of a detailed chromosome imbalance map indicating copy number variable and copy number stable regions might also be helpful. That said, additional functional studies in cells and model organisms will be best able to determine the effects of particular duplications.

As noted earlier, only a small percentage (approximately 10-15%) of cases can be explained by copy number variation (Devlin and Scherer 2012). Firstly, much of the copy number variation below 30 kb in size is difficult to identify from normal background noise specific to the array platform (Lionel et al. 2011; Pinto et al. 2014). Next-generation sequencing holds much promise in identifying both duplications and deletions smaller than that. Already, whole genome sequencing studies in intellectual disability (Gilissen et al. 2014) and published and unpublished work by my lab group (Jiang et al. 2013) are identifying single nucleotide variants, copy number variation, and indels in autism. As more of the genome and different types of variation become available for interpretation using sequencing methods, one would expect that a clearer picture of the etiology of ASD will develop. Though many more variants are being uncovered, interpreting their clinical relevance is frequently difficult, especially for single nucleotide variants.

Animal models and cell-based models can begin to shed light on the effects of some of these mutations. Lab mice are frequently used to observe the effect of certain mutations (especially gene knockouts) due to the fact that changes in social interaction are easily noted (Kim et al. 2008; Blundell et al. 2009). As an example, loss-of-function mutations in the Nlgn1 gene in mice contribute to impairments in NMDA-receptor signalling in excitatory synapses (Kim et al. 2008). These mice displayed an increase in repetitive behaviors and deficits in spatial memory (Chubykin et al. 2007). Mice are also a good model in which to study the effects of pharmaceutical compounds and see if they lessen ASD symptoms. In the same study mentioned

56 previously, an NMDA co-agonist was given to the mice which seemed to lessen the severity of the ASD symptoms in Nlgn1 knockout mice (Chubykin et al. 2007). Other mouse models are also being developed to use for similar investigations (Baribeau and Anagnostou 2014). Although mouse models help with our understanding of the disorder, it is not possible to generate mouse models for every mutation that one would want tested due to the low-throughput and relatively high cost of generating these models.

Other model organisms are also being developed and used to test many mutations in ASD candidate genes. Drosophila embryos are a useful model for studying synapse formation as the generation of neuromuscular junctions is similar to the process by which neural circuitry develops in the human brain (Knight et al. 2011). Zebrafish are also being used to model ASD. Like mice, they are social animals and ASD-like phenotypes can be modelled in them (Stewart et al. 2014). These animals have high genetic homology to humans and their genomes are easy to manipulate (Stewart et al. 2014). Because of the low cost of the organism, quick generation time, and ease of manipulation, they are good candidates to examine some of the functional effects of various mutations.

Patient-derived induced pluripotent stem cells are another powerful way of modelling synaptic defects caused by CNVs overlapping candidate genes. Dermal fibroblasts obtained from a biopsy of an affected individual are reprogrammed so that they attain the properties of an embryonic stem cell (Kim et al. 2014). These cells are subsequently coaxed to become neurons. Experiments using this method should begin to help bridge how the genotype alters brain function and leads to the ASD phenotype. Already, models of Rett syndrome (Marchetto et al. 2010), Fragile-X syndrome (Urbach et al. 2010), and Timothy syndrome (Pasca et al. 2011) have been developed. These models can illustrate problems with the generation of synapses and also identify physiological problems of neurodevelopment and can also control for genetic background by using an unaffected relative as a control. Not only will such models provide more support for a relationship between a particular genotype and ASD, but will serve as a model with which to test specific pharmaceutical compounds that can alleviate some features of the condition. In summary, it is by the use of higher-resolution sequencing strategies and the use of both animal and cell culture models that we can gain a better understanding of the relationship of genotype to phenotype in ASD.

Appendix

Appendix 1: List of Rare CNVs

Sample chr Start end size Sex CNV genes ID 666-3 10 7,196,973 7,315,361 118,388 M Deletion SFMBT2 666-3 12 96,014,096 96,706,369 692,273 M Duplication CCDC38,HAL,ELK3,NTN4,CDK17,AMDHD1,LTA4H,SNRPF 666-3 3 55,284,137 55,403,172 119,035 M Deletion - 666-3 4 56,756,201 56,780,082 23,881 M Duplication EXOC1 666-3 Y 25,415,633 25,618,373 202,740 M Deletion DAZ3,DAZ2,DAZ4 556-3 19 17,492,673 17,683,241 190,568 M Duplication TMEM221,PGLS,COLGALT1,FAM129C,BST2,NXNL1, MVB12A,SLC27A1 556-3 2 35,142,821 35,431,860 289,039 M Deletion - 556-3 20 23,836,569 23,924,631 88,062 M Duplication CST5 556-3 4 38,798,854 38,829,628 30,774 M Deletion TLR1,TLR6 684-3 2 26,621,755 26,648,953 27,198 M Deletion DRC1 632-3 3 106,837,905 106,869,829 31,924 F Deletion - 632-3 7 1,169,619 1,192,647 23,028 F Deletion C7orf50,ZFAND2A 527-3 4 124,063,146 125,045,116 981,970 M Duplication SPRY1,SPATA5 527-3 8 8,541,144 8,566,235 25,091 M Deletion CLDN23 527-3 Y 3,951,652 4,159,244 207,592 M Deletion - 690-3 17 28,937,651 29,140,813 203,162 M Duplication CRLF3 690-3 3 19,672,864 19,795,693 122,829 M Deletion - 690-3 Y 19,797,643 19,962,926 165,283 M Duplication XKRY,XKRY2 690-3 Y 20,212,995 20,381,761 168,766 M Duplication XKRY,XKRY2 505-3 13 43,587,160 43,608,103 20,943 M Duplication DNAJC15 505-3 6 162,290,899 162,359,917 69,018 M Duplication PARK2 505-3 7 31,681,765 31,921,749 239,984 M Duplication PDE1C,PPP1R17,CCDC129 505-3 Y 6,532,597 6,577,903 45,306 M Duplication - 505-3 Y 21,231,039 21,333,110 102,071 M Duplication - 554-3 X 118,460,470 118,487,729 27,259 M Deletion - 670-3 Y 22,243,114 22,489,344 246,230 M Duplication - 681-3 14 92,422,164 92,443,907 21,743 M Duplication TRIP11 681-3 2 112,530,696 112,579,331 48,635 M Duplication ANAPC1 681-3 20 756,842 791,208 34,366 M Deletion - 681-3 3 120,614,922 120,645,266 30,344 M Deletion STXBP5L 625-3 17 20,801,835 20,893,760 91,925 M Duplication - 659-3 2 21,438,794 21,462,096 23,302 M Duplication - 659-3 7 79,927,625 79,972,710 45,085 M Deletion - 659-3 7 80,299,710 80,354,513 54,803 M Deletion CD36 678-3 16 68,284,483 68,376,819 92,336 F Duplication SLC7A6,PLA2G15,SLC7A6OS,PRMT7 517-3 16 843,861 1,162,728 318,867 F Duplication C1QTNF8,LMF1,GNG13,SOX8,SSTR5,CHTF18,PRR25 517-3 16 2,088,391 2,415,016 326,625 F Duplication BRICD5,RNPS1,RAB26,TRAF7,SLC9A3R2,PKD1,E4F1,ECI1, ABCA3,PGP,CASKIN1,NTHL1,MLST8,TSC2,DNASE1L2

517-3 18 29,638,649 29,677,400 38,751 F Duplication RNF138,RNF125 517-3 7 4,015,113 4,685,115 670,002 F Duplication SDK1 517-3 X 81,298,618 81,338,460 39,842 F Deletion - 672-3 11 129,769,926 129,815,519 45,593 M Duplication PRDM10 672-3 21 22,941,489 23,009,531 68,042 M Duplication - 672-3 4 71,227,646 71,250,769 23,123 M Deletion SMR3B,SMR3A 672-3 Y 25,415,633 25,618,373 202,740 M Deletion DAZ3,DAZ2,DAZ4 508-3 5 24,949,269 24,990,295 41,026 M Duplication - 565-3 5 96,910,556 96,935,306 24,750 M Deletion - 565-3 6 169,508,876 169,539,121 30,245 M Deletion - 565-3 X 90,817,396 90,889,532 72,136 M Deletion - 565-3 Y 16,160,929 16,189,527 28,598 M Duplication VCY,VCY1B 530-3 20 31,988,157 32,009,801 21,644 M Duplication SNTA1,CDK5RAP1 530-3 3 15,687,586 15,710,397 22,811 M Deletion ANKRD28 515-3 12 47,262,280 47,330,558 68,278 M Duplication - 515-3 15 41,585,043 41,617,637 32,594 M Duplication OIP5 537-3 19 23,895,015 23,971,507 76,492 M Deletion ZNF681 537-3 7 27,497,389 27,537,776 40,387 M Deletion - 685-3 15 42,418,261 42,447,236 28,975 M Deletion PLA2G4F 685-3 19 37,772,900 37,803,996 31,096 M Deletion - 685-3 22 47,901,616 47,922,809 21,193 M Duplication - 685-3 6 68,342,496 68,410,749 68,253 M Duplication - 685-3 X 8,289,192 8,338,653 49,461 M Deletion - 646-3 5 120,872,438 120,895,959 23,521 M Deletion - 646-3 6 169,508,876 169,539,121 30,245 M Deletion - 560-3 6 80,110,296 80,354,297 244,001 M Duplication LCA5,SH3BGRL2 560-3 6 80,884,791 80,997,902 113,111 M Duplication BCKDHB 532-3 10 133,592,188 133,652,778 60,590 M Duplication - 532-3 17 17,444,603 17,476,218 31,615 M Duplication PEMT 532-3 3 159,562,374 159,641,399 79,025 M Duplication SCHIP1,IQCJ-SCHIP1 532-3 6 169,508,876 169,539,121 30,245 M Deletion - 608-3 14 93,363,869 93,408,980 45,111 M Duplication ITPK1,CHGA 608-3 3 2,352,000 2,434,492 82,492 M Deletion CNTN4 578-3 1 155,927,397 156,002,779 75,382 F Duplication SSR2,ARHGEF2 578-3 2 68,125,250 68,176,751 51,501 F Deletion - 578-3 7 7,953,108 7,999,416 46,308 F Deletion - 578-3 7 146,329,136 146,362,505 33,369 F Deletion CNTNAP2 578-3 8 5,722,189 5,784,231 62,042 F Deletion - 548-3 15 91,145,609 91,172,887 27,278 M Deletion CRTC3 550-3 14 106,291,501 106,319,497 27,996 M Duplication - 550-3 2 209,682,592 209,703,720 21,128 M Deletion - 550-3 7 152,508,063 153,499,963 991,900 M Duplication ACTR3B

550-3 X 143,651,892 143,748,641 96,749 M Duplication - 550-3 Y 22,311,731 22,489,344 177,613 M Duplication - 540-3 16 77,191,477 77,227,551 36,074 M Duplication MON1B 540-3 6 9,965,790 10,001,901 36,111 M Deletion - 540-3 Y 27,653,382 27,809,359 155,977 M Deletion CDY1B,CDY1 664-3 X 47,850,294 48,230,012 379,718 M Duplication SPACA5,SSX5,SSX3,SSX1,ZNF630,ZNF182,SPACA5B 664-3 Y 26,513,714 26,587,395 73,681 M Deletion - 682-3 6 162,296,842 162,359,917 63,075 M Duplication PARK2 682-3 7 43,361 705,271 661,910 M Duplication PDGFA,FAM20C,PRKAR1B,LOC100288524 549-3 16 16,295,901 16,855,348 559,447 M Deletion ABCC6,NOMO3 677-3 12 99,996,714 100,080,217 83,503 M Deletion ANKS1B,FAM71C 677-3 13 43,587,160 43,608,103 20,943 M Duplication DNAJC15 677-3 X 115,862,466 116,006,048 143,582 M Deletion - 511-3 2 186,884,124 187,008,504 124,380 M Deletion - 511-3 5 123,700,908 123,883,512 182,604 M Duplication - 511-3 9 28,464,218 28,596,286 132,068 M Deletion LINGO2 511-3 X 147,489,229 147,529,551 40,322 M Duplication - 686-3 10 18,240,592 18,313,842 73,250 M Deletion SLC39A12 686-3 4 27,394,986 27,415,922 20,936 M Deletion - 686-3 9 119,547,311 119,567,789 20,478 M Deletion ASTN2 521-3 16 14,357,944 14,386,767 28,823 M Duplication MKL2 647-3 1 79,572,639 79,967,028 394,389 M Deletion - 647-3 17 80,845,676 80,969,149 123,473 M Duplication TBCD,B3GNTL1 667-3 14 83,564,570 83,593,839 29,269 M Duplication - 667-3 3 192,367,797 192,418,953 51,156 M Deletion FGF12 667-3 4 77,393,565 77,442,934 49,369 M Duplication SHROOM3 567-3 11 25,664,801 25,738,322 73,521 M Deletion - 567-3 2 53,148,333 53,290,713 142,380 M Deletion - 567-3 4 122,480,241 122,577,042 96,801 M Duplication - 567-3 X 31,805,650 31,959,887 154,237 M Deletion DMD 493-3 10 78,292,989 78,320,204 27,215 F Deletion C10orf11 493-3 X 153,822,886 153,843,460 20,574 F Deletion - 683-3 2 233,651,280 233,673,273 21,993 F Deletion GIGYF2 683-3 7 90,464,909 90,723,377 258,468 F Deletion CDK14 683-3 8 5,543,465 5,608,834 65,369 F Deletion - 534-3 15 32,206,861 32,231,263 24,402 M Duplication - 534-3 4 71,227,646 71,250,769 23,123 M Deletion SMR3B,SMR3A 502-3 2 241,104,594 241,402,775 298,181 M Duplication GPC1 502-3 X 72,872,788 72,898,554 25,766 M Duplication CHIC1 631-3 17 19,521,054 19,563,766 42,712 F Deletion ALDH3A2 631-3 2 179,752,984 179,780,855 27,871 F Deletion CCDC141 631-3 7 9,437,710 9,465,942 28,232 F Duplication -

663-3 2 68,130,508 68,179,098 48,590 M Deletion - 663-3 6 169,494,891 169,539,121 44,230 M Deletion - 663-3 8 94,866,807 94,905,198 38,391 M Deletion - 663-3 X 72,332,078 72,353,391 21,313 M Deletion - 577-3 4 71,227,646 71,250,769 23,123 M Deletion SMR3B,SMR3A 577-3 9 468,705 489,338 20,633 M Deletion KANK1 577-3 Y 9,237,095 9,304,988 67,893 M Deletion TSPY4,TSPY3,TSPY1 577-3 Y 9,308,643 9,342,828 34,185 M Deletion TSPY4 668-3 11 124,133,810 124,231,918 98,108 M Deletion OR8G5,OR8G1,OR8D1,OR8D2 668-3 18 7,079,996 7,580,481 500,485 M Duplication LRRC30,LAMA1,PTPRM 668-3 2 134,077,947 134,151,644 73,697 M Deletion NCKAP5 668-3 22 46,946,802 47,354,509 407,707 M Duplication TBC1D22A,CERK,GRAMD4 539-3 2 194,568,253 194,601,354 33,101 M Deletion - 539-3 3 83,191,785 83,232,605 40,820 M Duplication - 607-3 7 85,624,564 85,673,895 49,331 M Deletion - 688-3 20 50,582,813 50,606,329 23,516 M Deletion - 574-3 19 58,791,213 58,813,839 22,626 M Deletion ZNF8 574-3 4 54,070,949 54,091,178 20,229 M Deletion SCFD2 574-3 8 5,057,783 5,323,185 265,402 M Deletion - 545-3 6 78,214,540 78,293,398 78,858 M Deletion - 545-3 7 125,147,776 125,168,006 20,230 M Duplication - 503-3 16 28,819,029 29,051,191 232,162 M Deletion ATXN2L,ATP2A1,NFATC2IP,SPNS1,RABEP2,SH2B1,LAT, TUFM,CD19 676-3 1 246,261,903 246,324,760 62,857 M Deletion SMYD3 676-3 12 95,622,380 95,665,403 43,023 M Deletion VEZT 676-3 15 42,421,124 42,447,236 26,112 M Deletion PLA2G4F 542-3 1 65,924,510 66,031,445 106,935 M Duplication LEPR 671-3 1 192,370,110 192,394,774 24,664 M Deletion - 671-3 Y 25,415,633 25,618,373 202,740 M Deletion DAZ3,DAZ2,DAZ4 691-3 1 208,513,516 208,556,206 42,690 M Duplication - 691-3 12 86,005,217 86,139,905 134,688 M Duplication - 691-3 9 115,714,003 115,746,828 32,825 M Deletion - 528-3 16 53,747,533 53,771,635 24,102 M Deletion FTO 528-3 3 22,039,451 22,103,146 63,695 M Deletion - 528-3 4 34,116,667 34,144,022 27,355 M Deletion - 528-3 6 127,607,555 127,654,281 46,726 M Deletion ECHDC1,RNF146 692-3 1 15,276,526 15,453,728 177,202 M Duplication KAZN 692-3 17 2,455,643 3,449,869 994,226 M Duplication CLUH,OR3A3,SPATA22,PAFAH1B1,OR1E1,OR1D5,OR1E2, OR3A1,OR3A2,OR1D2,OR1A2,OR1A1,RAP1GAP2, OR1G1,TRPV3,ASPA 692-3 2 212,057,149 212,142,292 85,143 M Duplication - 692-3 6 169,508,876 169,539,121 30,245 M Deletion - 485-3 10 87,338,403 87,359,978 21,575 M Deletion GRID1 485-3 11 11,179,056 11,296,482 117,426 M Deletion GALNT18

626-3 21 22,485,241 22,526,967 41,726 M Deletion NCAM2 636-3 14 40,727,363 40,780,133 52,770 M Duplication - 538-3 1 101,104,505 101,132,410 27,905 M Deletion - 538-3 X 109,315,806 109,340,564 24,758 M Deletion TMEM164 562-3 10 135,377,075 135,400,250 23,175 F Deletion SYCE1 562-3 2 23,694,732 23,726,237 31,505 F Deletion KLHL29 562-3 3 128,340,747 128,367,730 26,983 F Duplication RPN1 562-3 3 136,633,504 136,664,731 31,227 F Duplication NCK1 562-3 6 100,423,832 100,467,829 43,997 F Duplication MCHR2 562-3 9 33,876,490 33,951,019 74,529 F Duplication UBAP2,UBE2R2 622-3 4 127,751,482 127,777,218 25,736 M Deletion - 623-3 19 55,435,082 55,754,138 319,056 M Duplication TNNT1,NLRP7,PPP6R1,NLRP2,PPP1R12C,SYT5,TNNI3, RDH13,EPS8L1,GP6,DNAAF3,TMEM86B,PTPRH 623-3 Y 27,410,703 27,607,527 196,824 M Duplication - 576-3 10 96,581,094 96,626,365 45,271 M Deletion CYP2C19 576-3 21 43,837,058 43,893,628 56,570 M Duplication UBASH3A,RSPH1 576-3 9 9,473,063 9,572,759 99,696 M Deletion PTPRD 576-3 9 109,270,732 109,310,538 39,806 M Duplication - 662-3 11 85,774,746 85,845,045 70,299 M Duplication PICALM 662-3 4 81,335,556 81,367,124 31,568 M Duplication C4orf22 552-3 11 80,696,716 80,801,906 105,190 M Duplication - 552-3 5 56,498,189 56,574,179 75,990 M Duplication GPBP1 552-3 5 178,679,462 178,790,851 111,389 M Deletion ADAMTS2 552-3 7 14,122,604 14,199,001 76,397 M Deletion DGKB 552-3 X 72,319,907 72,353,391 33,484 M Deletion NAP1L6 658-3 1 5,734,013 5,767,396 33,383 M Deletion - 658-3 5 151,292,056 151,349,089 57,033 M Duplication GLRA1 544-3 10 112,577,721 112,632,630 54,909 M Duplication PDCD4,RBM20 544-3 2 214,391,134 214,424,922 33,788 M Duplication SPAG16 520-3 7 19,980,047 20,098,212 118,165 M Deletion - 520-3 7 111,218,682 111,319,190 100,508 M Duplication - 520-3 8 53,492,860 53,549,179 56,319 M Duplication RB1CC1 519-3 2 241,544,852 241,572,469 27,617 M Deletion GPR35 519-3 X 153,408,930 153,438,781 29,851 M Duplication OPN1LW 535-3 2 35,758,580 35,795,100 36,520 M Deletion - 535-3 3 164,088,939 164,113,222 24,283 M Deletion - 514-3 11 11,116,025 11,187,898 71,873 M Duplication - 514-3 16 17,940,666 17,963,706 23,040 M Duplication - 514-3 5 96,910,556 96,935,306 24,750 M Deletion - 514-3 5 153,412,185 153,531,977 119,792 M Duplication FAM114A2,MFAP3 523-3 2 75,820,691 75,919,242 98,551 M Duplication MRPL19,GCFC2 523-3 3 285,783 318,812 33,029 M Deletion CHL1 523-3 8 89,774,917 89,993,041 218,124 M Deletion -

611-3 13 89,365,687 89,574,513 208,826 M Duplication - 611-3 7 37,913,050 38,246,663 333,613 M Duplication NME8,STARD3NL,SFRP4,EPDR1 611-3 X 32,548,066 32,603,018 54,952 M Deletion DMD 611-3 X 125,884,392 125,957,401 73,009 M Deletion CXorf64 609-3 11 10,515,751 10,537,581 21,830 M Duplication AMPD3,RNF141,MTRNR2L8 609-3 8 54,058,999 54,124,081 65,082 M Duplication - 609-3 Y 26,504,496 26,592,401 87,905 M Deletion - 687-3 5 96,910,556 96,935,306 24,750 M Deletion - 489-3 14 94,555,466 94,587,345 31,879 M Duplication IFI27,IFI27L1 489-3 17 18,692,538 18,726,389 33,851 M Deletion TVP23B 489-3 4 64,569,828 65,118,025 548,197 M Duplication - 610-3 X 82,973,565 82,997,765 24,200 M Duplication - 533-3 22 50,000,220 50,049,643 49,423 M Duplication - 693-3 2 74,892,914 74,990,885 97,971 M Duplication SEMA4F 693-3 5 164,041,749 164,131,456 89,707 M Duplication - 693-3 X 134,384,557 134,802,358 417,801 M Duplication DDX26B,ZNF75D,ZNF449 693-3 Y 9,492,812 9,522,716 29,904 M Deletion - 694-3 17 18,562,720 18,590,815 28,095 M Deletion ZNF286B 694-3 4 34,658,484 34,685,352 26,868 M Deletion - 694-3 4 48,825,457 49,093,773 268,316 M Duplication OCIAD2,CWH43,OCIAD1 694-3 4 94,144,621 94,172,410 27,789 M Deletion GRID2 694-3 5 19,409,085 19,430,309 21,224 M Deletion - 694-3 6 169,508,876 169,539,121 30,245 M Deletion - 694-3 9 28,491,679 28,630,598 138,919 M Deletion LINGO2 656-3 1 25,583,489 25,756,453 172,964 M Duplication RHCE,RHD,TMEM50A 656-3 14 87,355,480 87,446,496 91,016 M Deletion - 656-3 7 4,490,966 4,525,949 34,983 M Deletion - 628-3 20 8,913,944 8,976,141 62,197 M Duplication - 628-3 5 14,206,863 14,253,937 47,074 M Deletion TRIO 628-3 9 115,714,003 115,746,828 32,825 M Deletion - 689-3 14 62,849,708 62,960,957 111,249 M Duplication - 638-3 14 41,546,521 41,744,165 197,644 F Deletion - 638-3 15 42,421,124 42,447,236 26,112 F Deletion PLA2G4F 546-3 21 23,526,014 23,587,148 61,134 F Deletion - 546-3 5 151,263,430 152,000,379 736,949 F Duplication NMUR2,GLRA1 546-3 6 169,508,876 169,539,121 30,245 F Deletion - 546-3 9 6,518,780 6,655,056 136,276 F Deletion GLDC

References

Altshuler, D. M., R. A. Gibbs, L. Peltonen, E. Dermitzakis, S. F. Schaffner, F. Yu, P. E. Bonnen, P. I. de Bakker, P. Deloukas, S. B. Gabriel, R. Gwilliam, S. Hunt, M. Inouye, X. Jia, A. Palotie, M. Parkin, P. Whittaker, K. Chang, A. Hawes, L. R. Lewis, Y. Ren, D. Wheeler, D. M. Muzny, C. Barnes, K. Darvishi, M. Hurles, J. M. Korn, K. Kristiansson, C. Lee, S. A. McCarrol, J. Nemesh, A. Keinan, S. B. Montgomery, S. Pollack, A. L. Price, N. Soranzo, C. Gonzaga-Jauregui, V. Anttila, W. Brodeur, M. J. Daly, S. Leslie, G. McVean, L. Moutsianas, H. Nguyen, Q. Zhang, M. J. Ghori, R. McGinnis, W. McLaren, F. Takeuchi, S. R. Grossman, I. Shlyakhter, E. B. Hostetter, P. C. Sabeti, C. A. Adebamowo, M. W. Foster, D. R. Gordon, J. Licinio, M. C. Manca, P. A. Marshall, I. Matsuda, D. Ngare, V. O. Wang, D. Reddy, C. N. Rotimi, C. D. Royal, R. R. Sharp, C. Zeng, L. D. Brooks and J. E. McEwen (2010). "Integrating common and rare genetic variation in diverse human populations." Nature 467(7311): 52-58.

American Psychiatric Association, A. P. A. D. S. M. T. F. (2013). "Diagnostic and statistical manual of mental disorders : DSM-5." from http://dsm.psychiatryonline.org/book.aspx?bookid=556.

Anney, R., L. Klei, D. Pinto, R. Regan, J. Conroy, T. R. Magalhaes, C. Correia, B. S. Abrahams, N. Sykes, A. T. Pagnamenta, J. Almeida, E. Bacchelli, A. J. Bailey, G. Baird, A. Battaglia, T. Berney, N. Bolshakova, S. Bolte, P. F. Bolton, T. Bourgeron, S. Brennan, J. Brian, A. R. Carson, G. Casallo, J. Casey, S. H. Chu, L. Cochrane, C. Corsello, E. L. Crawford, A. Crossett, G. Dawson, M. de Jonge, R. Delorme, I. Drmic, E. Duketis, F. Duque, A. Estes, P. Farrar, B. A. Fernandez, S. E. Folstein, E. Fombonne, C. M. Freitag, J. Gilbert, C. Gillberg, J. T. Glessner, J. Goldberg, J. Green, S. J. Guter, H. Hakonarson, E. A. Heron, M. Hill, R. Holt, J. L. Howe, G. Hughes, V. Hus, R. Igliozzi, C. Kim, S. M. Klauck, A. Kolevzon, O. Korvatska, V. Kustanovich, C. M. Lajonchere, J. A. Lamb, M. Laskawiec, M. Leboyer, A. Le Couteur, B. L. Leventhal, A. C. Lionel, X. Q. Liu, C. Lord, L. Lotspeich, S. C. Lund, E. Maestrini, W. Mahoney, C. Mantoulan, C. R. Marshall, H. McConachie, C. J. McDougle, J. McGrath, W. M. McMahon, N. M. Melhem, A. Merikangas, O. Migita, N. J. Minshew, G. K. Mirza, J. Munson, S. F. Nelson, C. Noakes, A. Noor, G. Nygren, G. Oliveira, K. Papanikolaou, J. R. Parr, B. Parrini, T. Paton, A. Pickles, J. Piven, D. J. Posey, A. Poustka, F. Poustka, A. Prasad, J. Ragoussis, K. Renshaw, J. Rickaby, W. Roberts, K. Roeder, B. Roge, M. L. Rutter, L. J. Bierut, J. P. Rice, J. Salt, K. Sansom, D. Sato, R. Segurado, L. Senman, N. Shah, V. C. Sheffield, L. Soorya, I. Sousa, V. Stoppioni, C. Strawbridge, R. Tancredi, K. Tansey, B. Thiruvahindrapduram, A. P. Thompson, S. Thomson, A. Tryfon, J. Tsiantis, H. Van Engeland, J. B. Vincent, F. Volkmar, S. Wallace, K. Wang, Z. Wang, T. H. Wassink, K. Wing, K. Wittemeyer, S. Wood, B. L. Yaspan, D. Zurawiecki, L. Zwaigenbaum, C. Betancur, J. D. Buxbaum, R. M. Cantor, E. H. Cook, H. Coon, M. L. Cuccaro, L. Gallagher, D. H. Geschwind, M. Gill, J. L. Haines, J. Miller, A. P. Monaco, J. I. Nurnberger, Jr., A. D. Paterson, M. A. Pericak-Vance, G. D. Schellenberg, S. W. Scherer, J. S. Sutcliffe, P. Szatmari, A. M. Vicente, V. J. Vieland, E. M. Wijsman, B. Devlin, S. Ennis and J. Hallmayer (2010). "A genome-wide scan for common alleles affecting risk for autism." Hum Mol Genet 19(20): 4072-4082.

Aschard, H., W. Qiu, B. Pasaniuc, N. Zaitlen, M. H. Cho and V. Carey (2011). "Combining effects from rare and common genetic variants in an exome-wide association study of sequence data." BMC Proc 5 Suppl 9: S44.

Bachmann-Gagescu, R., H. C. Mefford, C. Cowan, G. M. Glew, A. V. Hing, S. Wallace, P. I. Bader, A. Hamati, P. J. Reitnauer, R. Smith, D. W. Stockton, H. Muhle, I. Helbig, E. E. Eichler, B. C. Ballif, J. Rosenfeld and K. D. Tsuchiya (2010). "Recurrent 200-kb deletions of 16p11.2 that include the SH2B1 gene are associated with developmental delay and obesity." Genet Med 12(10): 641-647.

Baker, P., J. Piven, S. Schwartz and S. Patil (1994). "Brief report: duplication of chromosome 15q11-13 in two individuals with autistic disorder." J Autism Dev Disord 24(4): 529-535.

Baribeau, D. A. and E. Anagnostou (2014). "Social Communication is an Emerging Target for Pharmacotherapy in Autism Spectrum Disorder - A Review of the Literature on Potential Agents." J Can Acad Child Adolesc Psychiatry 23(1): 20-30.

Berg, J. M. and D. H. Geschwind (2012). "Autism genetics: searching for specificity and convergence." Genome Biol 13(7): 247.

Betancur, C. and J. D. Buxbaum (2013). "SHANK3 haploinsufficiency: a "common" but underdiagnosed highly penetrant monogenic cause of autism spectrum disorders." Mol Autism 4(1): 17.

Bi, W., T. Sapir, O. A. Shchelochkov, F. Zhang, M. A. Withers, J. V. Hunter, T. Levy, V. Shinder, D. A. Peiffer, K. L. Gunderson, M. M. Nezarati, V. A. Shotts, S. S. Amato, S. K. Savage, D. J. Harris, D. L. Day-Salvatore, M. Horner, X. Y. Lu, T. Sahoo, Y. Yanagawa, A. L. Beaudet, S. W. Cheung, S. Martinez, J. R. Lupski and O. Reiner (2009). "Increased LIS1 expression affects human and mouse brain development." Nat Genet 41(2): 168- 177.

Bisbal, M., C. Conde, M. Donoso, F. Bollati, J. Sesma, S. Quiroga, A. Diaz Anel, V. Malhotra, M. P. Marzolo and A. Caceres (2008). "Protein kinase d regulates trafficking of dendritic membrane proteins in developing neurons." J Neurosci 28(37): 9297-9308.

Blundell, J., K. Tabuchi, M. F. Bolliger, C. A. Blaiss, N. Brose, X. Liu, T. C. Sudhof and C. M. Powell (2009). "Increased anxiety-like behavior in mice lacking the inhibitory synapse cell adhesion molecule neuroligin 2." Genes Brain Behav 8(1): 114-126.

Bochukova, E. G., N. Huang, J. Keogh, E. Henning, C. Purmann, K. Blaszczyk, S. Saeed, J. Hamilton-Shield, J. Clayton-Smith, S. O'Rahilly, M. E. Hurles and I. S. Farooqi (2010). "Large, rare chromosomal deletions associated with severe early-onset obesity." Nature 463(7281): 666-670.

Bourgeron, T. (2009). "A synaptic trek to autism." Curr Opin Neurobiol 19(2): 231-234.

Bruno, D. L., B. M. Anderlid, A. Lindstrand, C. van Ravenswaaij-Arts, D. Ganesamoorthy, J. Lundin, C. L. Martin, J. Douglas, C. Nowak, M. P. Adam, R. F. Kooy, N. Van der Aa, E. Reyniers, G. Vandeweyer, I. Stolte-Dijkstra, T. Dijkhuizen, A. Yeung, M. Delatycki, B.

Borgstrom, L. Thelin, C. Cardoso, B. van Bon, R. Pfundt, B. B. de Vries, A. Wallin, D. J. Amor, P. A. James, H. R. Slater and J. Schoumans (2010). "Further molecular and clinical delineation of co-locating 17p13.3 microdeletions and microduplications that show distinctive phenotypes." J Med Genet 47(5): 299-311.

Carter, M. T. and S. W. Scherer (2013). "Autism spectrum disorder in the genetics clinic: a review." Clin Genet 83(5): 399-407.

Chenier, S., G. Yoon, B. Argiropoulos, J. Lauzon, R. Laframboise, J. W. Ahn, C. M. Ogilvie, A. C. Lionel, C. R. Marshall, A. K. Vaags, B. Hashemi, K. Boisvert, G. Mathonnet, F. Tihy, J. So, S. W. Scherer, E. Lemyre and D. J. Stavropoulos (2014). "CHD2 haploinsufficiency is associated with developmental delay, intellectual disability, epilepsy and neurobehavioural problems." J Neurodev Disord 6(1): 9.

Chow, M. L., T. Pramparo, M. E. Winn, C. C. Barnes, H. R. Li, L. Weiss, J. B. Fan, S. Murray, C. April, H. Belinson, X. D. Fu, A. Wynshaw-Boris, N. J. Schork and E. Courchesne (2012). "Age-dependent brain gene expression and copy number anomalies in autism suggest distinct pathological processes at young versus mature ages." PLoS Genet 8(3): e1002592.

Chowanadisai, W., D. M. Graham, C. L. Keen, R. B. Rucker and M. A. Messerli (2013). "Neurulation and neurite extension require the zinc transporter ZIP12 (slc39a12)." Proc Natl Acad Sci U S A 110(24): 9903-9908.

Chubykin, A. A., D. Atasoy, M. R. Etherton, N. Brose, E. T. Kavalali, J. R. Gibson and T. C. Sudhof (2007). "Activity-dependent validation of excitatory versus inhibitory synapses by neuroligin-1 versus neuroligin-2." Neuron 54(6): 919-931.

Coghlan, S., J. Horder, B. Inkster, M. A. Mendez, D. G. Murphy and D. J. Nutt (2012). "GABA system dysfunction in autism and related disorders: from synapse to symptoms." Neurosci Biobehav Rev 36(9): 2044-2055.

Conrad, D. F., D. Pinto, R. Redon, L. Feuk, O. Gokcumen, Y. Zhang, J. Aerts, T. D. Andrews, C. Barnes, P. Campbell, T. Fitzgerald, M. Hu, C. H. Ihm, K. Kristiansson, D. G. Macarthur, J. R. Macdonald, I. Onyiah, A. W. Pang, S. Robson, K. Stirrups, A. Valsesia, K. Walter, J. Wei, C. Tyler-Smith, N. P. Carter, C. Lee, S. W. Scherer and M. E. Hurles (2010). "Origins and functional impact of copy number variation in the human genome." Nature 464(7289): 704-712.

Constantino, J. N., A. Todorov, C. Hilton, P. Law, Y. Zhang, E. Molloy, R. Fitzgerald and D. Geschwind (2013). "Autism recurrence in half siblings: strong support for genetic mechanisms of transmission in ASD." Mol Psychiatry 18(2): 137-138.

Constantino, J. N., Y. Zhang, T. Frazier, A. M. Abbacchi and P. Law (2010). "Sibling recurrence and the genetic epidemiology of autism." Am J Psychiatry 167(11): 1349-1356.

Cook, E. H., Jr. and S. W. Scherer (2008). "Copy-number variations associated with neuropsychiatric conditions." Nature 455(7215): 919-923.

Costain, G., A. C. Lionel, D. Merico, P. Forsythe, K. Russell, C. Lowther, T. Yuen, J. Husted, D. J. Stavropoulos, M. Speevak, E. W. Chow, C. R. Marshall, S. W. Scherer and A. S. Bassett (2013). "Pathogenic rare copy number variants in community-based schizophrenia suggest a potential role for clinical microarrays." Hum Mol Genet 22(22): 4485-4501.

Curry, C. J., J. A. Rosenfeld, E. Grant, K. W. Gripp, C. Anderson, A. S. Aylsworth, T. B. Saad, V. V. Chizhikov, G. Dybose, C. Fagerberg, M. Falco, C. Fels, M. Fichera, J. Graakjaer, D. Greco, J. Hair, E. Hopkins, M. Huggins, R. Ladda, C. Li, J. Moeschler, M. J. Nowaczyk, J. R. Ozmore, S. Reitano, C. Romano, L. Roos, R. E. Schnur, S. Sell, P. Suwannarat, D. Svaneby, M. Szybowska, M. Tarnopolsky, R. Tervo, A. C. Tsai, M. Tucker, S. Vallee, F. C. Wheeler, D. J. Zand, A. J. Barkovich, S. Aradhya, L. G. Shaffer and W. B. Dobyns (2013). "The duplication 17p13.3 phenotype: analysis of 21 families delineates developmental, behavioral and brain abnormalities, and rare variant phenotypes." Am J Med Genet A 161A(8): 1833-1852.

Darvishi, K. (2010). "Application of Nexus copy number software for CNV detection and analysis." Curr Protoc Hum Genet Chapter 4: Unit 4 14 11-28.

Dawson, G., E. J. Jones, K. Merkle, K. Venema, R. Lowy, S. Faja, D. Kamara, M. Murias, J. Greenson, J. Winter, M. Smith, S. J. Rogers and S. J. Webb (2012). "Early behavioral intervention is associated with normalized brain activity in young children with autism." J Am Acad Child Adolesc Psychiatry 51(11): 1150-1159.

Dawson, G., S. Rogers, J. Munson, M. Smith, J. Winter, J. Greenson, A. Donaldson and J. Varley (2010). "Randomized, controlled trial of an intervention for toddlers with autism: the Early Start Denver Model." Pediatrics 125(1): e17-23.

Den Dunnen, J. T., P. M. Grootscholten, E. Bakker, L. A. Blonden, H. B. Ginjaar, M. C. Wapenaar, H. M. van Paassen, C. van Broeckhoven, P. L. Pearson and G. J. van Ommen (1989). "Topography of the Duchenne muscular dystrophy (DMD) gene: FIGE and cDNA analysis of 194 cases reveals 115 deletions and 13 duplications." Am J Hum Genet 45(6): 835-847.

Devlin, B. and S. W. Scherer (2012). "Genetic architecture in autism spectrum disorder." Curr Opin Genet Dev 22(3): 229-237.

Dittwald, P., T. Gambin, P. Szafranski, J. Li, S. Amato, M. Y. Divon, L. X. Rodriguez Rojas, L. E. Elton, D. A. Scott, C. P. Schaaf, W. Torres-Martinez, A. K. Stevens, J. A. Rosenfeld, S. Agadi, D. Francis, S. H. Kang, A. Breman, S. R. Lalani, C. A. Bacino, W. Bi, A. Milosavljevic, A. L. Beaudet, A. Patel, C. A. Shaw, J. R. Lupski, A. Gambin, S. W. Cheung and P. Stankiewicz (2013). "NAHR-mediated copy-number variants in a clinical population: mechanistic insights into both genomic disorders and Mendelizing traits." Genome Res 23(9): 1395-1409.

Downey, T. (2006). "Analysis of a multifactor microarray study using Partek genomics solution." Methods Enzymol 411: 256-270.

Dworzynski, K., A. Ronald, P. Bolton and F. Happe (2012). "How different are girls and boys above and below the diagnostic threshold for autism spectrum disorders?" J Am Acad Child Adolesc Psychiatry 51(8): 788-797.

Egger, G., K. M. Roetzer, A. Noor, A. C. Lionel, H. Mahmood, T. Schwarzbraun, O. Boright, A. Mikhailov, C. R. Marshall, C. Windpassinger, E. Petek, S. W. Scherer, W. Kaschnitz and J. B. Vincent (2014). "Identification of risk genes for autism spectrum disorder through copy number variation analysis in Austrian families." Neurogenetics 15(2): 117-127.

Elsabbagh, M., G. Divan, Y. J. Koh, Y. S. Kim, S. Kauchali, C. Marcin, C. Montiel-Nava, V. Patel, C. S. Paula, C. Wang, M. T. Yasamy and E. Fombonne (2012). "Global prevalence of autism and other pervasive developmental disorders." Autism Res 5(3): 160-179.

Erickson, C. A., J. M. Veenstra-Vanderweele, R. D. Melmed, J. T. McCracken, L. D. Ginsberg, L. Sikich, L. Scahill, M. Cherubini, P. Zarevics, K. Walton-Bowen, R. L. Carpenter, M. F. Bear, P. P. Wang and B. H. King (2014). "STX209 (arbaclofen) for autism spectrum disorders: an 8-week open-label study." J Autism Dev Disord 44(4): 958-964.

Evangelou, E. and J. P. Ioannidis (2013). "Meta-analysis methods for genome-wide association studies and beyond." Nat Rev Genet 14(6): 379-389.

Fernandez, B. A., W. Roberts, B. Chung, R. Weksberg, S. Meyn, P. Szatmari, A. M. Joseph- George, S. Mackay, K. Whitten, B. Noble, C. Vardy, V. Crosbie, S. Luscombe, E. Tucker, L. Turner, C. R. Marshall and S. W. Scherer (2010). "Phenotypic spectrum associated with de novo and inherited deletions and duplications at 16p11.2 in individuals ascertained for diagnosis of autism spectrum disorder." J Med Genet 47(3): 195-203.

Fletcher, S., C. F. Adkin, P. Meloni, B. Wong, F. Muntoni, R. Kole, C. Fragall, K. Greer, R. Johnsen and S. D. Wilton (2012). "Targeted exon skipping to address "leaky" mutations in the dystrophin gene." Mol Ther Nucleic Acids 1: e48.

Fu, F., H. L. Liu, R. Li, J. Han, X. Yang, P. Min, L. Zhen, Y. L. Zhang, G. E. Xie, T. Y. Lei, Y. Li, J. Li, D. Z. Li and C. Liao (2014). "Prenatal diagnosis of foetuses with congenital abnormalities and duplication of the MECP2 region." Gene.

Fung, K. M., H. W. Tsang, P. W. Corrigan, C. S. Lam and W. M. Cheung (2007). "Measuring self-stigma of mental illness in China and its implications for recovery." Int J Soc Psychiatry 53(5): 408-418.

Ganz, M. L. (2007). "The lifetime distribution of the incremental societal costs of autism." Arch Pediatr Adolesc Med 161(4): 343-349.

Gatto, C. L., D. Pereira and K. Broadie (2014). "GABAergic circuit dysfunction in the Drosophila Fragile X syndrome model." Neurobiol Dis 65: 142-159.

Gilissen, C., J. Y. Hehir-Kwa, D. T. Thung, M. van de Vorst, B. W. van Bon, M. H. Willemsen, M. Kwint, I. M. Janssen, A. Hoischen, A. Schenck, R. Leach, R. Klein, R. Tearle, T. Bo, R. Pfundt, H. G. Yntema, B. B. de Vries, T. Kleefstra, H. G. Brunner, L. E. Vissers and J.

A. Veltman (2014). "Genome sequencing identifies major causes of severe intellectual disability." Nature.

Gillberg, C., M. Cederlund, K. Lamberg and L. Zeijlon (2006). "Brief report: "the autism epidemic". The registered prevalence of autism in a Swedish urban area." J Autism Dev Disord 36(3): 429-435.

Girirajan, S., Z. Brkanac, B. P. Coe, C. Baker, L. Vives, T. H. Vu, N. Shafer, R. Bernier, G. B. Ferrero, M. Silengo, S. T. Warren, C. S. Moreno, M. Fichera, C. Romano, W. H. Raskind and E. E. Eichler (2011). "Relative burden of large CNVs on a range of neurodevelopmental phenotypes." PLoS Genet 7(11): e1002334.

Gronborg, T. K., D. E. Schendel and E. T. Parner (2013). "Recurrence of autism spectrum disorders in full- and half-siblings and trends over time: a population-based cohort study." JAMA Pediatr 167(10): 947-953.

Gu, W., F. Zhang and J. R. Lupski (2008). "Mechanisms for human genomic rearrangements." Pathogenetics 1(1): 4.

Hadley, D., Z. L. Wu, C. Kao, A. Kini, A. Mohamed-Hadley, K. Thomas, L. Vazquez, H. Qiu, F. Mentch, R. Pellegrino, C. Kim, J. Connolly, J. Glessner and H. Hakonarson (2014). "The impact of the metabotropic glutamate receptor and other gene family interaction networks on autism." Nat Commun 5: 4074.

Hallmayer, J., S. Cleveland, A. Torres, J. Phillips, B. Cohen, T. Torigoe, J. Miller, A. Fedele, J. Collins, K. Smith, L. Lotspeich, L. A. Croen, S. Ozonoff, C. Lajonchere, J. K. Grether and N. Risch (2011). "Genetic heritability and shared environmental factors among twin pairs with autism." Arch Gen Psychiatry 68(11): 1095-1102.

Hanemaaijer, N. M., B. Sikkema-Raddatz, G. van der Vries, T. Dijkhuizen, R. Hordijk, A. J. van Essen, H. E. Veenstra-Knol, W. S. Kerstjens-Frederikse, J. C. Herkert, E. H. Gerkes, L. K. Leegte, K. Kok, R. J. Sinke and C. M. van Ravenswaaij-Arts (2012). "Practical guidelines for interpreting copy number gains detected by high-resolution array in routine diagnostics." Eur J Hum Genet 20(2): 161-165.

Harris, S. W., D. Hessl, B. Goodlin-Jones, J. Ferranti, S. Bacalman, I. Barbato, F. Tassone, P. J. Hagerman, H. Herman and R. J. Hagerman (2008). "Autism profiles of males with fragile X syndrome." Am J Ment Retard 113(6): 427-438.

Henderson, C., L. Wijetunge, M. N. Kinoshita, M. Shumway, R. S. Hammond, F. R. Postma, C. Brynczka, R. Rush, A. Thomas, R. Paylor, S. T. Warren, P. W. Vanderklish, P. C. Kind, R. L. Carpenter, M. F. Bear and A. M. Healy (2012). "Reversal of disease-related pathologies in the fragile X mouse model by selective activation of GABAB receptors with arbaclofen." Sci Transl Med 4(152): 152ra128.

Hoffmann, T. J., G. C. Windham, M. Anderson, L. A. Croen, J. K. Grether and N. Risch (2014). "Evidence of Reproductive Stoppage in Families With Autism Spectrum Disorder: A Large, Population-Based Cohort Study." JAMA Psychiatry.

Horev, G., J. Ellegood, J. P. Lerch, Y. E. Son, L. Muthuswamy, H. Vogel, A. M. Krieger, A. Buja, R. M. Henkelman, M. Wigler and A. A. Mills (2011). "Dosage-dependent phenotypes in models of 16p11.2 lesions found in autism." Proc Natl Acad Sci U S A 108(41): 17076-17081.

Iossifov, I., M. Ronemus, D. Levy, Z. Wang, I. Hakker, J. Rosenbaum, B. Yamrom, Y. H. Lee, G. Narzisi, A. Leotta, J. Kendall, E. Grabowska, B. Ma, S. Marks, L. Rodgers, A. Stepansky, J. Troge, P. Andrews, M. Bekritsky, K. Pradhan, E. Ghiban, M. Kramer, J. Parla, R. Demeter, L. L. Fulton, R. S. Fulton, V. J. Magrini, K. Ye, J. C. Darnell, R. B. Darnell, E. R. Mardis, R. K. Wilson, M. C. Schatz, W. R. McCombie and M. Wigler (2012). "De novo gene disruptions in children on the autistic spectrum." Neuron 74(2): 285-299.

Jacquemont, S., B. P. Coe, M. Hersch, M. H. Duyzend, N. Krumm, S. Bergmann, J. S. Beckmann, J. A. Rosenfeld and E. E. Eichler (2014). "A higher mutational burden in females supports a "female protective model" in neurodevelopmental disorders." Am J Hum Genet 94(3): 415-425.

Jiang, Y. H., R. K. Yuen, X. Jin, M. Wang, N. Chen, X. Wu, J. Ju, J. Mei, Y. Shi, M. He, G. Wang, J. Liang, Z. Wang, D. Cao, M. T. Carter, C. Chrysler, I. E. Drmic, J. L. Howe, L. Lau, C. R. Marshall, D. Merico, T. Nalpathamkalam, B. Thiruvahindrapuram, A. Thompson, M. Uddin, S. Walker, J. Luo, E. Anagnostou, L. Zwaigenbaum, R. H. Ring, J. Wang, C. Lajonchere, A. Shih, P. Szatmari, H. Yang, G. Dawson, Y. Li and S. W. Scherer (2013). "Detection of clinically relevant genetic variants in autism spectrum disorder by whole-genome sequencing." Am J Hum Genet 93(2): 249-263.

Jones, W. and A. Klin (2013). "Attention to eyes is present but in decline in 2-6-month-old infants later diagnosed with autism." Nature 504(7480): 427-431.

Julie, G., F. F. Hamdan and G. A. Rouleau (2011). "A strategy to identify de novo mutations in common disorders such as autism and schizophrenia." J Vis Exp(52).

Kamien, B., A. C. Lionel, N. Bain, S. W. Scherer and M. Hunter (2014). "Outfoxed by RBFOX1-A caution about ascertainment bias." Am J Med Genet A 164(6): 1411-1418.

Kanner, L. (1943). "Autistic disturbances of affective contact." Nervous Child 2: 217-250.

Kearney, H. M., E. C. Thorland, K. K. Brown, F. Quintero-Rivera and S. T. South (2011). "American College of Medical Genetics standards and guidelines for interpretation and reporting of postnatal constitutional copy number variants." Genet Med 13(7): 680-685.

Kerin, T., A. Ramanathan, K. Rivas, N. Grepo, G. A. Coetzee and D. B. Campbell (2012). "A noncoding RNA antisense to moesin at 5p14.1 in autism." Sci Transl Med 4(128): 128ra140.

Kim, D. S., P. J. Ross, K. Zaslavsky and J. Ellis (2014). "Optimizing neuronal differentiation from induced pluripotent stem cells to model ASD." Front Cell Neurosci 8: 109.

Kim, J., S. Y. Jung, Y. K. Lee, S. Park, J. S. Choi, C. J. Lee, H. S. Kim, Y. B. Choi, P. Scheiffele, C. H. Bailey, E. R. Kandel and J. H. Kim (2008). "Neuroligin-1 is required for normal expression of LTP and associative fear memory in the amygdala of adult animals." Proc Natl Acad Sci U S A 105(26): 9087-9092.

Knight, D., W. Xie and G. L. Boulianne (2011). "Neurexins and neuroligins: recent insights from invertebrates." Mol Neurobiol 44(3): 426-440.

Lai, M. C., M. V. Lombardo and S. Baron-Cohen (2014). "Autism." Lancet 383(9920): 896-910.

Lautier, C., S. Goldwurm, A. Durr, B. Giovannone, W. G. Tsiaras, G. Pezzoli, A. Brice and R. J. Smith (2008). "Mutations in the GIGYF2 (TNRC15) gene at the PARK11 locus in familial Parkinson disease." Am J Hum Genet 82(4): 822-833.

Lee-Olsen, M. a. F., D. (2012) "China Clears Its First Microarray Platform for in VitroDiagnostics to Accelerate Personalized Medicine." Reuters.

Lee, N. R., G. L. Wallace, E. I. Adeyemi, K. C. Lopez, J. D. Blumenthal, L. S. Clasen and J. N. Giedd (2012). "Dosage effects of X and Y chromosomes on language and social functioning in children with supernumerary sex chromosome aneuploidies: implications for idiopathic language impairment and autism spectrum disorders." J Child Psychol Psychiatry 53(10): 1072-1081.

Levy, D., M. Ronemus, B. Yamrom, Y. H. Lee, A. Leotta, J. Kendall, S. Marks, B. Lakshmi, D. Pai, K. Ye, A. Buja, A. Krieger, S. Yoon, J. Troge, L. Rodgers, I. Iossifov and M. Wigler (2011). "Rare de novo and transmitted copy-number variation in autistic spectrum disorders." Neuron 70(5): 886-897.

Liao, H. M., S. S. Gau, W. C. Tsai, J. S. Fang, Y. C. Su, M. C. Chou, S. K. Liu, W. J. Chou, Y. Y. Wu and C. H. Chen (2013). "Chromosomal abnormalities in patients with autism spectrum disorders from Taiwan." Am J Med Genet B Neuropsychiatr Genet 162B(7): 734-741.

Lionel, A. C., J. Crosbie, N. Barbosa, T. Goodale, B. Thiruvahindrapuram, J. Rickaby, M. Gazzellone, A. R. Carson, J. L. Howe, Z. Wang, J. Wei, A. F. Stewart, R. Roberts, R. McPherson, A. Fiebig, A. Franke, S. Schreiber, L. Zwaigenbaum, B. A. Fernandez, W. Roberts, P. D. Arnold, P. Szatmari, C. R. Marshall, R. Schachar and S. W. Scherer (2011). "Rare copy number variation discovery and cross-disorder comparisons identify risk genes for ADHD." Sci Transl Med 3(95): 95ra75.

Lionel, A. C., K. Tammimies, A. K. Vaags, J. A. Rosenfeld, J. W. Ahn, D. Merico, A. Noor, C. K. Runke, V. K. Pillalamarri, M. T. Carter, M. J. Gazzellone, B. Thiruvahindrapuram, C. Fagerberg, L. W. Laulund, G. Pellecchia, S. Lamoureux, C. Deshpande, J. Clayton- Smith, A. C. White, S. Leather, J. Trounce, H. Melanie Bedford, E. Hatchwell, P. S. Eis, R. K. Yuen, S. Walker, M. Uddin, M. T. Geraghty, S. M. Nikkel, E. M. Tomiak, B. A. Fernandez, N. Soreni, J. Crosbie, P. D. Arnold, R. J. Schachar, W. Roberts, A. D. Paterson, J. So, P. Szatmari, C. Chrysler, M. Woodbury-Smith, R. Brian Lowry, L. Zwaigenbaum, D. Mandyam, J. Wei, J. R. Macdonald, J. L. Howe, T. Nalpathamkalam, Z. Wang, D. Tolson, D. S. Cobb, T. M. Wilks, M. J. Sorensen, P. I. Bader, Y. An, B. L.

Wu, S. A. Musumeci, C. Romano, D. Postorivo, A. M. Nardone, M. D. Monica, G. Scarano, L. Zoccante, F. Novara, O. Zuffardi, R. Ciccone, V. Antona, M. Carella, L. Zelante, P. Cavalli, C. Poggiani, U. Cavallari, B. Argiropoulos, J. Chernos, C. Brasch- Andersen, M. Speevak, M. Fichera, C. M. Ogilvie, Y. Shen, J. C. Hodge, M. E. Talkowski, D. J. Stavropoulos, C. R. Marshall and S. W. Scherer (2014). "Disruption of the ASTN2/TRIM32 locus at 9q33.1 is a risk factor in males for autism spectrum disorders, ADHD and other neurodevelopmental phenotypes." Hum Mol Genet.

Losh, M., D. Childress, K. Lam and J. Piven (2008). "Defining key features of the broad autism phenotype: a comparison across parents of multiple- and single-incidence autism families." Am J Med Genet B Neuropsychiatr Genet 147B(4): 424-433.

Lou, H., S. Li, Y. Yang, L. Kang, X. Zhang, W. Jin, B. Wu, L. Jin and S. Xu (2011). "A map of copy number variations in Chinese populations." PLoS One 6(11): e27341.

Mak, W. W. and Y. T. Kwok (2010). "Internalization of stigma for parents of children with autism spectrum disorder in Hong Kong." Soc Sci Med 70(12): 2045-2051.

Malhotra, D. and J. Sebat (2012). "CNVs: harbingers of a rare variant revolution in psychiatric genetics." Cell 148(6): 1223-1241.

Marchetto, M. C., C. Carromeu, A. Acab, D. Yu, G. W. Yeo, Y. Mu, G. Chen, F. H. Gage and A. R. Muotri (2010). "A model for neural development and treatment of Rett syndrome using human induced pluripotent stem cells." Cell 143(4): 527-539.

Marshall, C. R., A. Noor, J. B. Vincent, A. C. Lionel, L. Feuk, J. Skaug, M. Shago, R. Moessner, D. Pinto, Y. Ren, B. Thiruvahindrapduram, A. Fiebig, S. Schreiber, J. Friedman, C. E. Ketelaars, Y. J. Vos, C. Ficicioglu, S. Kirkpatrick, R. Nicolson, L. Sloman, A. Summers, C. A. Gibbons, A. Teebi, D. Chitayat, R. Weksberg, A. Thompson, C. Vardy, V. Crosbie, S. Luscombe, R. Baatjes, L. Zwaigenbaum, W. Roberts, B. Fernandez, P. Szatmari and S. W. Scherer (2008). "Structural variation of chromosomes in autism spectrum disorder." Am J Hum Genet 82(2): 477-488.

Marshall, C. R. and S. W. Scherer (2012). "Detection and characterization of copy number variation in autism spectrum disorder." Methods Mol Biol 838: 115-135.

Mattila, M. L., T. Hurtig, H. Haapsamo, K. Jussila, S. Kuusikko-Gauffin, M. Kielinen, S. L. Linna, H. Ebeling, R. Bloigu, L. Joskitt, D. L. Pauls and I. Moilanen (2010). "Comorbid psychiatric disorders associated with Asperger syndrome/high-functioning autism: a community- and clinic-based study." J Autism Dev Disord 40(9): 1080-1093.

McCarthy, M. I., G. R. Abecasis, L. R. Cardon, D. B. Goldstein, J. Little, J. P. Ioannidis and J. N. Hirschhorn (2008). "Genome-wide association studies for complex traits: consensus, uncertainty and challenges." Nat Rev Genet 9(5): 356-369.

Miller, D. T., M. P. Adam, S. Aradhya, L. G. Biesecker, A. R. Brothman, N. P. Carter, D. M. Church, J. A. Crolla, E. E. Eichler, C. J. Epstein, W. A. Faucett, L. Feuk, J. M. Friedman, A. Hamosh, L. Jackson, E. B. Kaminsky, K. Kok, I. D. Krantz, R. M. Kuhn, C. Lee, J. M. Ostell, C. Rosenberg, S. W. Scherer, N. B. Spinner, D. J. Stavropoulos, J. H. Tepperberg,

E. C. Thorland, J. R. Vermeesch, D. J. Waggoner, M. S. Watson, C. L. Martin and D. H. Ledbetter (2010). "Consensus statement: chromosomal microarray is a first-tier clinical diagnostic test for individuals with developmental disabilities or congenital anomalies." Am J Hum Genet 86(5): 749-764.

Morrow, E. M., S. Y. Yoo, S. W. Flavell, T. K. Kim, Y. Lin, R. S. Hill, N. M. Mukaddes, S. Balkhy, G. Gascon, A. Hashmi, S. Al-Saad, J. Ware, R. M. Joseph, R. Greenblatt, D. Gleason, J. A. Ertelt, K. A. Apse, A. Bodell, J. N. Partlow, B. Barry, H. Yao, K. Markianos, R. J. Ferland, M. E. Greenberg and C. A. Walsh (2008). "Identifying autism loci and genes by tracing recent shared ancestry." Science 321(5886): 218-223.

Muscarella, L. A., V. Guarnieri, R. Sacco, P. Curatolo, B. Manzi, R. Alessandrelli, G. Giana, R. Militerni, C. Bravaccio, C. Lenti, M. Saccani, C. Schneider, R. Melmed, L. D'Agruma and A. M. Persico (2010). "Candidate gene study of HOXB1 in autism spectrum disorder." Mol Autism 1(1): 9.

Nagamani, S. C., F. Zhang, O. A. Shchelochkov, W. Bi, Z. Ou, F. Scaglia, F. J. Probst, M. Shinawi, C. Eng, J. V. Hunter, S. Sparagana, E. Lagoe, C. T. Fong, M. Pearson, M. Doco-Fenzy, E. Landais, M. Mozelle, A. C. Chinault, A. Patel, C. A. Bacino, T. Sahoo, S. H. Kang, S. W. Cheung, J. R. Lupski and P. Stankiewicz (2009). "Microdeletions including YWHAE in the Miller-Dieker syndrome region on chromosome 17p13.3 result in facial dysmorphisms, growth restriction, and cognitive impairment." J Med Genet 46(12): 825-833.

Narcisa, V., M. Discenza, E. Vaccari, B. Rosen-Sheidley, A. Y. Hardan and E. Couchon (2013). "Parental interest in a genetic risk assessment test for autism spectrum disorders." Clin Pediatr (Phila) 52(2): 139-146.

Neale, B. M., Y. Kou, L. Liu, A. Ma'ayan, K. E. Samocha, A. Sabo, C. F. Lin, C. Stevens, L. S. Wang, V. Makarov, P. Polak, S. Yoon, J. Maguire, E. L. Crawford, N. G. Campbell, E. T. Geller, O. Valladares, C. Schafer, H. Liu, T. Zhao, G. Cai, J. Lihm, R. Dannenfelser, O. Jabado, Z. Peralta, U. Nagaswamy, D. Muzny, J. G. Reid, I. Newsham, Y. Wu, L. Lewis, Y. Han, B. F. Voight, E. Lim, E. Rossin, A. Kirby, J. Flannick, M. Fromer, K. Shakir, T. Fennell, K. Garimella, E. Banks, R. Poplin, S. Gabriel, M. DePristo, J. R. Wimbish, B. E. Boone, S. E. Levy, C. Betancur, S. Sunyaev, E. Boerwinkle, J. D. Buxbaum, E. H. Cook, Jr., B. Devlin, R. A. Gibbs, K. Roeder, G. D. Schellenberg, J. S. Sutcliffe and M. J. Daly (2012). "Patterns and rates of exonic de novo mutations in autism spectrum disorders." Nature 485(7397): 242-245.

Neishabouri, A. and A. A. Faisal (2014). "Axonal noise as a source of synaptic variability." PLoS Comput Biol 10(5): e1003615.

O'Roak, B. J., L. Vives, S. Girirajan, E. Karakoc, N. Krumm, B. P. Coe, R. Levy, A. Ko, C. Lee, J. D. Smith, E. H. Turner, I. B. Stanaway, B. Vernot, M. Malig, C. Baker, B. Reilly, J. M. Akey, E. Borenstein, M. J. Rieder, D. A. Nickerson, R. Bernier, J. Shendure and E. E. Eichler (2012). "Sporadic autism exomes reveal a highly interconnected protein network of de novo mutations." Nature 485(7397): 246-250.

Ozonoff, S., G. S. Young, A. Carter, D. Messinger, N. Yirmiya, L. Zwaigenbaum, S. Bryson, L. J. Carver, J. N. Constantino, K. Dobkins, T. Hutman, J. M. Iverson, R. Landa, S. J. Rogers, M. Sigman and W. L. Stone (2011). "Recurrence risk for autism spectrum disorders: a Baby Siblings Research Consortium study." Pediatrics 128(3): e488-495.

Park, H., J. I. Kim, Y. S. Ju, O. Gokcumen, R. E. Mills, S. Kim, S. Lee, D. Suh, D. Hong, H. P. Kang, Y. J. Yoo, J. Y. Shin, H. J. Kim, M. Yavartanoo, Y. W. Chang, J. S. Ha, W. Chong, G. R. Hwang, K. Darvishi, H. Kim, S. J. Yang, K. S. Yang, M. E. Hurles, S. W. Scherer, N. P. Carter, C. Tyler-Smith, C. Lee and J. S. Seo (2010). "Discovery of common Asian copy number variants using integrated high-resolution array CGH and massively parallel DNA sequencing." Nat Genet 42(5): 400-405.

Pasca, S. P., T. Portmann, I. Voineagu, M. Yazawa, A. Shcheglovitov, A. M. Pasca, B. Cord, T. D. Palmer, S. Chikahisa, S. Nishino, J. A. Bernstein, J. Hallmayer, D. H. Geschwind and R. E. Dolmetsch (2011). "Using iPSC-derived neurons to uncover cellular phenotypes associated with Timothy syndrome." Nat Med 17(12): 1657-1662.

Pauls, D. L., A. Abramovitch, S. L. Rauch and D. A. Geller (2014). "Obsessive-compulsive disorder: an integrative genetic and neurobiological perspective." Nat Rev Neurosci 15(6): 410-424.

Perry, A., R. A. Condillac, N. L. Freeman, J. Dunn-Geier and J. Belair (2005). "Multi-site study of the Childhood Autism Rating Scale (CARS) in five clinical groups of young children." J Autism Dev Disord 35(5): 625-634.

Petrone, J. (2012) "Q&A: Shanghai Children's Medical Center's Yongguo Yu on Introducing Chromosomal Microarrays in China." GenomeWeb.

Pinto, D., K. Darvishi, X. Shi, D. Rajan, D. Rigler, T. Fitzgerald, A. C. Lionel, B. Thiruvahindrapuram, J. R. Macdonald, R. Mills, A. Prasad, K. Noonan, S. Gribble, E. Prigmore, P. K. Donahoe, R. S. Smith, J. H. Park, M. E. Hurles, N. P. Carter, C. Lee, S. W. Scherer and L. Feuk (2011). "Comprehensive assessment of array-based platforms and calling algorithms for detection of copy number variants." Nat Biotechnol 29(6): 512-520.

Pinto, D., E. Delaby, D. Merico, M. Barbosa, A. Merikangas, L. Klei, B. Thiruvahindrapuram, X. Xu, R. Ziman, Z. Wang, J. A. Vorstman, A. Thompson, R. Regan, M. Pilorge, G. Pellecchia, A. T. Pagnamenta, B. Oliveira, C. R. Marshall, T. R. Magalhaes, J. K. Lowe, J. L. Howe, A. J. Griswold, J. Gilbert, E. Duketis, B. A. Dombroski, M. V. De Jonge, M. Cuccaro, E. L. Crawford, C. T. Correia, J. Conroy, I. C. Conceicao, A. G. Chiocchetti, J. P. Casey, G. Cai, C. Cabrol, N. Bolshakova, E. Bacchelli, R. Anney, S. Gallinger, M. Cotterchio, G. Casey, L. Zwaigenbaum, K. Wittemeyer, K. Wing, S. Wallace, H. van Engeland, A. Tryfon, S. Thomson, L. Soorya, B. Roge, W. Roberts, F. Poustka, S. Mouga, N. Minshew, L. A. McInnes, S. G. McGrew, C. Lord, M. Leboyer, A. S. Le Couteur, A. Kolevzon, P. Jimenez Gonzalez, S. Jacob, R. Holt, S. Guter, J. Green, A. Green, C. Gillberg, B. A. Fernandez, F. Duque, R. Delorme, G. Dawson, P. Chaste, C. Cafe, S. Brennan, T. Bourgeron, P. F. Bolton, S. Bolte, R. Bernier, G. Baird, A. J. Bailey, E. Anagnostou, J. Almeida, E. M. Wijsman, V. J. Vieland, A. M. Vicente, G. D.

Schellenberg, M. Pericak-Vance, A. D. Paterson, J. R. Parr, G. Oliveira, J. I. Nurnberger, A. P. Monaco, E. Maestrini, S. M. Klauck, H. Hakonarson, J. L. Haines, D. H. Geschwind, C. M. Freitag, S. E. Folstein, S. Ennis, H. Coon, A. Battaglia, P. Szatmari, J. S. Sutcliffe, J. Hallmayer, M. Gill, E. H. Cook, J. D. Buxbaum, B. Devlin, L. Gallagher, C. Betancur and S. W. Scherer (2014). "Convergence of genes and cellular pathways dysregulated in autism spectrum disorders." Am J Hum Genet 94(5): 677-694.

Pinto, D., A. T. Pagnamenta, L. Klei, R. Anney, D. Merico, R. Regan, J. Conroy, T. R. Magalhaes, C. Correia, B. S. Abrahams, J. Almeida, E. Bacchelli, G. D. Bader, A. J. Bailey, G. Baird, A. Battaglia, T. Berney, N. Bolshakova, S. Bolte, P. F. Bolton, T. Bourgeron, S. Brennan, J. Brian, S. E. Bryson, A. R. Carson, G. Casallo, J. Casey, B. H. Chung, L. Cochrane, C. Corsello, E. L. Crawford, A. Crossett, C. Cytrynbaum, G. Dawson, M. de Jonge, R. Delorme, I. Drmic, E. Duketis, F. Duque, A. Estes, P. Farrar, B. A. Fernandez, S. E. Folstein, E. Fombonne, C. M. Freitag, J. Gilbert, C. Gillberg, J. T. Glessner, J. Goldberg, A. Green, J. Green, S. J. Guter, H. Hakonarson, E. A. Heron, M. Hill, R. Holt, J. L. Howe, G. Hughes, V. Hus, R. Igliozzi, C. Kim, S. M. Klauck, A. Kolevzon, O. Korvatska, V. Kustanovich, C. M. Lajonchere, J. A. Lamb, M. Laskawiec, M. Leboyer, A. Le Couteur, B. L. Leventhal, A. C. Lionel, X. Q. Liu, C. Lord, L. Lotspeich, S. C. Lund, E. Maestrini, W. Mahoney, C. Mantoulan, C. R. Marshall, H. McConachie, C. J. McDougle, J. McGrath, W. M. McMahon, A. Merikangas, O. Migita, N. J. Minshew, G. K. Mirza, J. Munson, S. F. Nelson, C. Noakes, A. Noor, G. Nygren, G. Oliveira, K. Papanikolaou, J. R. Parr, B. Parrini, T. Paton, A. Pickles, M. Pilorge, J. Piven, C. P. Ponting, D. J. Posey, A. Poustka, F. Poustka, A. Prasad, J. Ragoussis, K. Renshaw, J. Rickaby, W. Roberts, K. Roeder, B. Roge, M. L. Rutter, L. J. Bierut, J. P. Rice, J. Salt, K. Sansom, D. Sato, R. Segurado, A. F. Sequeira, L. Senman, N. Shah, V. C. Sheffield, L. Soorya, I. Sousa, O. Stein, N. Sykes, V. Stoppioni, C. Strawbridge, R. Tancredi, K. Tansey, B. Thiruvahindrapduram, A. P. Thompson, S. Thomson, A. Tryfon, J. Tsiantis, H. Van Engeland, J. B. Vincent, F. Volkmar, S. Wallace, K. Wang, Z. Wang, T. H. Wassink, C. Webber, R. Weksberg, K. Wing, K. Wittemeyer, S. Wood, J. Wu, B. L. Yaspan, D. Zurawiecki, L. Zwaigenbaum, J. D. Buxbaum, R. M. Cantor, E. H. Cook, H. Coon, M. L. Cuccaro, B. Devlin, S. Ennis, L. Gallagher, D. H. Geschwind, M. Gill, J. L. Haines, J. Hallmayer, J. Miller, A. P. Monaco, J. I. Nurnberger, Jr., A. D. Paterson, M. A. Pericak-Vance, G. D. Schellenberg, P. Szatmari, A. M. Vicente, V. J. Vieland, E. M. Wijsman, S. W. Scherer, J. S. Sutcliffe and C. Betancur (2010). "Functional impact of global rare copy number variation in autism spectrum disorders." Nature 466(7304): 368- 372.

Prasad, A., D. Merico, B. Thiruvahindrapuram, J. Wei, A. C. Lionel, D. Sato, J. Rickaby, C. Lu, P. Szatmari, W. Roberts, B. A. Fernandez, C. R. Marshall, E. Hatchwell, P. S. Eis and S. W. Scherer (2012). "A discovery resource of rare copy number variations in individuals with autism spectrum disorder." G3 (Bethesda) 2(12): 1665-1685.

Purcell, S., B. Neale, K. Todd-Brown, L. Thomas, M. A. Ferreira, D. Bender, J. Maller, P. Sklar, P. I. de Bakker, M. J. Daly and P. C. Sham (2007). "PLINK: a tool set for whole-genome association and population-based linkage analyses." Am J Hum Genet 81(3): 559-575.

Qiu, J. and E. C. Hayden (2008). "Genomics sizes up." Nature 451(7176): 234.

Ramachandran, V. S. and L. M. Oberman (2006). "Broken mirrors: a theory of autism." Sci Am 295(5): 62-69.

Redon, R., S. Ishikawa, K. R. Fitch, L. Feuk, G. H. Perry, T. D. Andrews, H. Fiegler, M. H. Shapero, A. R. Carson, W. Chen, E. K. Cho, S. Dallaire, J. L. Freeman, J. R. Gonzalez, M. Gratacos, J. Huang, D. Kalaitzopoulos, D. Komura, J. R. MacDonald, C. R. Marshall, R. Mei, L. Montgomery, K. Nishimura, K. Okamura, F. Shen, M. J. Somerville, J. Tchinda, A. Valsesia, C. Woodwark, F. Yang, J. Zhang, T. Zerjal, L. Armengol, D. F. Conrad, X. Estivill, C. Tyler-Smith, N. P. Carter, H. Aburatani, C. Lee, K. W. Jones, S. W. Scherer and M. E. Hurles (2006). "Global variation in copy number in the human genome." Nature 444(7118): 444-454.

Rosenberg, R. E., J. K. Law, G. Yenokyan, J. McGready, W. E. Kaufmann and P. A. Law (2009). "Characteristics and concordance of autism spectrum disorders among 277 twin pairs." Arch Pediatr Adolesc Med 163(10): 907-914.

Rutter, M. (1978). Language disorder and infantile autism. Autism: a reappraisal of concepts and treatment. M. Rutter, Schopler, E. New York, Plenum.

Sakai, Y., C. A. Shaw, B. C. Dawson, D. V. Dugas, Z. Al-Mohtaseb, D. E. Hill and H. Y. Zoghbi (2011). "Protein interactome reveals converging molecular pathways among autism disorders." Sci Transl Med 3(86): 86ra49.

Sanders, S. J., A. G. Ercan-Sencicek, V. Hus, R. Luo, M. T. Murtha, D. Moreno-De-Luca, S. H. Chu, M. P. Moreau, A. R. Gupta, S. A. Thomson, C. E. Mason, K. Bilguvar, P. B. Celestino-Soper, M. Choi, E. L. Crawford, L. Davis, N. R. Wright, R. M. Dhodapkar, M. DiCola, N. M. DiLullo, T. V. Fernandez, V. Fielding-Singh, D. O. Fishman, S. Frahm, R. Garagaloyan, G. S. Goh, S. Kammela, L. Klei, J. K. Lowe, S. C. Lund, A. D. McGrew, K. A. Meyer, W. J. Moffat, J. D. Murdoch, B. J. O'Roak, G. T. Ober, R. S. Pottenger, M. J. Raubeson, Y. Song, Q. Wang, B. L. Yaspan, T. W. Yu, I. R. Yurkiewicz, A. L. Beaudet, R. M. Cantor, M. Curland, D. E. Grice, M. Gunel, R. P. Lifton, S. M. Mane, D. M. Martin, C. A. Shaw, M. Sheldon, J. A. Tischfield, C. A. Walsh, E. M. Morrow, D. H. Ledbetter, E. Fombonne, C. Lord, C. L. Martin, A. I. Brooks, J. S. Sutcliffe, E. H. Cook, Jr., D. Geschwind, K. Roeder, B. Devlin and M. W. State (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.

Sanders, S. J., M. T. Murtha, A. R. Gupta, J. D. Murdoch, M. J. Raubeson, A. J. Willsey, A. G. Ercan-Sencicek, N. M. DiLullo, N. N. Parikshak, J. L. Stein, M. F. Walker, G. T. Ober, N. A. Teran, Y. Song, P. El-Fishawy, R. C. Murtha, M. Choi, J. D. Overton, R. D. Bjornson, N. J. Carriero, K. A. Meyer, K. Bilguvar, S. M. Mane, N. Sestan, R. P. Lifton, M. Gunel, K. Roeder, D. H. Geschwind, B. Devlin and M. W. State (2012). "De novo mutations revealed by whole-exome sequencing are strongly associated with autism." Nature 485(7397): 237-241.

Sandin, S., P. Lichtenstein, R. Kuja-Halkola, H. Larsson, C. M. Hultman and A. Reichenberg (2014). "The familial risk of autism." JAMA 311(17): 1770-1777.

Santoro, M. R., S. M. Bray and S. T. Warren (2012). "Molecular mechanisms of fragile X syndrome: a twenty-year perspective." Annu Rev Pathol 7: 219-245.

Sarbassov, D. D., S. M. Ali and D. M. Sabatini (2005). "Growing roles for the mTOR pathway." Curr Opin Cell Biol 17(6): 596-603.

Sato, D., A. C. Lionel, C. S. Leblond, A. Prasad, D. Pinto, S. Walker, I. O'Connor, C. Russell, I. E. Drmic, F. F. Hamdan, J. L. Michaud, V. Endris, R. Roeth, R. Delorme, G. Huguet, M. Leboyer, M. Rastam, C. Gillberg, M. Lathrop, D. J. Stavropoulos, E. Anagnostou, R. Weksberg, E. Fombonne, L. Zwaigenbaum, B. A. Fernandez, W. Roberts, G. A. Rappold, C. R. Marshall, T. Bourgeron, P. Szatmari and S. W. Scherer (2012). "SHANK1 Deletions in Males with Autism Spectrum Disorder." Am J Hum Genet 90(5): 879-887.

Schmunk, G. and J. J. Gargus (2013). "Channelopathy pathogenesis in autism spectrum disorders." Front Genet 4: 222.

Sebat, J., B. Lakshmi, D. Malhotra, J. Troge, C. Lese-Martin, T. Walsh, B. Yamrom, S. Yoon, A. Krasnitz, J. Kendall, A. Leotta, D. Pai, R. Zhang, Y. H. Lee, J. Hicks, S. J. Spence, A. T. Lee, K. Puura, T. Lehtimaki, D. Ledbetter, P. K. Gregersen, J. Bregman, J. S. Sutcliffe, V. Jobanputra, W. Chung, D. Warburton, M. C. King, D. Skuse, D. H. Geschwind, T. C. Gilliam, K. Ye and M. Wigler (2007). "Strong association of de novo copy number mutations with autism." Science 316(5823): 445-449.

Shcheglovitov, A., O. Shcheglovitova, M. Yazawa, T. Portmann, R. Shu, V. Sebastiano, A. Krawisz, W. Froehlich, J. A. Bernstein, J. F. Hallmayer and R. E. Dolmetsch (2013). "SHANK3 and IGF1 restore synaptic deficits in neurons from 22q13 deletion syndrome patients." Nature 503(7475): 267-271.

Sheng, M. and E. Kim (2000). "The Shank family of scaffold proteins." J Cell Sci 113 ( Pt 11): 1851-1856.

Simonoff, E., C. R. Jones, G. Baird, A. Pickles, F. Happe and T. Charman (2013). "The persistence and stability of psychiatric problems in adolescents with autism spectrum disorders." J Child Psychol Psychiatry 54(2): 186-194.

Simonoff, E., A. Pickles, T. Charman, S. Chandler, T. Loucas and G. Baird (2008). "Psychiatric disorders in children with autism spectrum disorders: prevalence, comorbidity, and associated factors in a population-derived sample." J Am Acad Child Adolesc Psychiatry 47(8): 921-929.

Stafford, R. L., E. Hinde, M. J. Knight, M. A. Pennella, J. Ear, M. A. Digman, E. Gratton and J. U. Bowie (2011). "Tandem SAM domain structure of human Caskin1: a presynaptic, self-assembling scaffold for CASK." Structure 19(12): 1826-1836.

Stewart, A. M., M. Nguyen, K. Wong, M. K. Poudel and A. V. Kalueff (2014). "Developing zebrafish models of autism spectrum disorder (ASD)." Prog Neuropsychopharmacol Biol Psychiatry 50: 27-36.

Stoddart, K. P., L. Burke, B. Muskat, J. Manett, S. Duhaime, C. Accardi, P. Burnham Riosa and E. Bradley (2013) "Diversity in Ontario’s Youth and Adults with Autism Spectrum Disorders: Complex Needs in Unprepared Systems.".

Sucksmith, E., I. Roth and R. A. Hoekstra (2011). "Autistic traits below the clinical threshold: re-examining the broader autism phenotype in the 21st century." Neuropsychol Rev 21(4): 360-389.

Sudhof, T. C. (2008). "Neuroligins and neurexins link synaptic function to cognitive disease." Nature 455(7215): 903-911.

Sun, X., C. Allison, B. Auyeung, S. Baron-Cohen and C. Brayne (2013a). "A review of healthcare service and education provision of Autism Spectrum Condition in mainland China." Res Dev Disabil 34(1): 469-479.

Sun, X., C. Allison, F. E. Matthews, S. J. Sharp, B. Auyeung, S. Baron-Cohen and C. Brayne (2013b). "Prevalence of autism in mainland China, Hong Kong and Taiwan: a systematic review and meta-analysis." Mol Autism 4(1): 7.

Szatmari, P., X. Q. Liu, J. Goldberg, L. Zwaigenbaum, A. D. Paterson, M. Woodbury-Smith, S. Georgiades, E. Duku and A. Thompson (2012). "Sex differences in repetitive stereotyped behaviors in autism: implications for genetic liability." Am J Med Genet B Neuropsychiatr Genet 159B(1): 5-12.

Szatmari, P., A. D. Paterson, L. Zwaigenbaum, W. Roberts, J. Brian, X. Q. Liu, J. B. Vincent, J. L. Skaug, A. P. Thompson, L. Senman, L. Feuk, C. Qian, S. E. Bryson, M. B. Jones, C. R. Marshall, S. W. Scherer, V. J. Vieland, C. Bartlett, L. V. Mangin, R. Goedken, A. Segre, M. A. Pericak-Vance, M. L. Cuccaro, J. R. Gilbert, H. H. Wright, R. K. Abramson, C. Betancur, T. Bourgeron, C. Gillberg, M. Leboyer, J. D. Buxbaum, K. L. Davis, E. Hollander, J. M. Silverman, J. Hallmayer, L. Lotspeich, J. S. Sutcliffe, J. L. Haines, S. E. Folstein, J. Piven, T. H. Wassink, V. Sheffield, D. H. Geschwind, M. Bucan, W. T. Brown, R. M. Cantor, J. N. Constantino, T. C. Gilliam, M. Herbert, C. Lajonchere, D. H. Ledbetter, C. Lese-Martin, J. Miller, S. Nelson, C. A. Samango- Sprouse, S. Spence, M. State, R. E. Tanzi, H. Coon, G. Dawson, B. Devlin, A. Estes, P. Flodman, L. Klei, W. M. McMahon, N. Minshew, J. Munson, E. Korvatska, P. M. Rodier, G. D. Schellenberg, M. Smith, M. A. Spence, C. Stodgell, P. G. Tepper, E. M. Wijsman, C. E. Yu, B. Roge, C. Mantoulan, K. Wittemeyer, A. Poustka, B. Felder, S. M. Klauck, C. Schuster, F. Poustka, S. Bolte, S. Feineis-Matthews, E. Herbrecht, G. Schmotzer, J. Tsiantis, K. Papanikolaou, E. Maestrini, E. Bacchelli, F. Blasi, S. Carone, C. Toma, H. Van Engeland, M. de Jonge, C. Kemner, F. Koop, M. Langemeijer, C. Hijmans, W. G. Staal, G. Baird, P. F. Bolton, M. L. Rutter, E. Weisblatt, J. Green, C. Aldred, J. A. Wilkinson, A. Pickles, A. Le Couteur, T. Berney, H. McConachie, A. J. Bailey, K. Francis, G. Honeyman, A. Hutchinson, J. R. Parr, S. Wallace, A. P. Monaco, G. Barnby, K. Kobayashi, J. A. Lamb, I. Sousa, N. Sykes, E. H. Cook, S. J. Guter, B. L. Leventhal, J. Salt, C. Lord, C. Corsello, V. Hus, D. E. Weeks, F. Volkmar, M. Tauber, E. Fombonne, A. Shih and K. J. Meyer (2007). "Mapping autism risk loci using genetic linkage and chromosomal rearrangements." Nat Genet 39(3): 319-328.

Tao, K. T. (1987). "Infantile autism in China." J Autism Dev Disord 17(2): 289-296.

Toyo-oka, K., A. Shionoya, M. J. Gambello, C. Cardoso, R. Leventer, H. L. Ward, R. Ayala, L. H. Tsai, W. Dobyns, D. Ledbetter, S. Hirotsune and A. Wynshaw-Boris (2003). "14-3- 3epsilon is important for neuronal migration by binding to NUDEL: a molecular explanation for Miller-Dieker syndrome." Nat Genet 34(3): 274-285.

Trillingsgaard, A., E. Ulsted Sorensen, G. Nemec and M. Jorgensen (2005). "What distinguishes autism spectrum disorders from other developmental disorders before the age of four years?" Eur Child Adolesc Psychiatry 14(2): 65-72.

Tsai, N. P., J. R. Wilkerson, W. Guo, M. A. Maksimova, G. N. DeMartino, C. W. Cowan and K. M. Huber (2012). "Multiple autism-linked genes mediate synapse elimination via proteasomal degradation of a synaptic scaffold PSD-95." Cell 151(7): 1581-1594.

Uddin, M., K. Tammimies, G. Pellecchia, B. Alipanahi, P. Hu, Z. Wang, D. Pinto, L. Lau, T. Nalpathamkalam, C. R. Marshall, B. J. Blencowe, B. J. Frey, D. Merico, R. K. Yuen and S. W. Scherer (2014). "Brain-expressed exons under purifying selection are enriched for de novo mutations in autism spectrum disorder." Nat Genet.

Urbach, A., O. Bar-Nur, G. Q. Daley and N. Benvenisty (2010). "Differential modeling of fragile X syndrome by human embryonic stem cells and induced pluripotent stem cells." Cell Stem Cell 6(5): 407-411. van Deutekom, J. C. and G. J. van Ommen (2003). "Advances in Duchenne muscular dystrophy gene therapy." Nat Rev Genet 4(10): 774-783.

Voineagu, I., X. Wang, P. Johnston, J. K. Lowe, Y. Tian, S. Horvath, J. Mill, R. M. Cantor, B. J. Blencowe and D. H. Geschwind (2011). "Transcriptomic analysis of autistic brain reveals convergent molecular pathology." Nature 474(7351): 380-384.

Volkmar, F. R., P. Szatmari and S. S. Sparrow (1993). "Sex differences in pervasive developmental disorders." J Autism Dev Disord 23(4): 579-591.

Vourc'h, P., I. Martin, F. Bonnet-Brilhault, S. Marouillat, C. Barthelemy, J. Pierre Muh and C. Andres (2003). "Mutation screening and association study of the UBE2H gene on chromosome 7q32 in autistic disorder." Psychiatr Genet 13(4): 221-225.

Wang, K., H. Zhang, D. Ma, M. Bucan, J. T. Glessner, B. S. Abrahams, D. Salyakina, M. Imielinski, J. P. Bradfield, P. M. Sleiman, C. E. Kim, C. Hou, E. Frackelton, R. Chiavacci, N. Takahashi, T. Sakurai, E. Rappaport, C. M. Lajonchere, J. Munson, A. Estes, O. Korvatska, J. Piven, L. I. Sonnenblick, A. I. Alvarez Retuerto, E. I. Herman, H. Dong, T. Hutman, M. Sigman, S. Ozonoff, A. Klin, T. Owley, J. A. Sweeney, C. W. Brune, R. M. Cantor, R. Bernier, J. R. Gilbert, M. L. Cuccaro, W. M. McMahon, J. Miller, M. W. State, T. H. Wassink, H. Coon, S. E. Levy, R. T. Schultz, J. I. Nurnberger, J. L. Haines, J. S. Sutcliffe, E. H. Cook, N. J. Minshew, J. D. Buxbaum, G. Dawson, S. F. Grant, D. H. Geschwind, M. A. Pericak-Vance, G. D. Schellenberg and H. Hakonarson (2009). "Common genetic variants on 5p14.1 associate with autism spectrum disorders." Nature 459(7246): 528-533.

Wassink, T. H., J. Piven, V. J. Vieland, L. Jenkins, R. Frantz, C. W. Bartlett, R. Goedken, D. Childress, M. A. Spence, M. Smith and V. C. Sheffield (2005). "Evaluation of the chromosome 2q37.3 gene CENTG2 as an autism susceptibility gene." Am J Med Genet B Neuropsychiatr Genet 136B(1): 36-44.

Weiss, L. A., Y. Shen, J. M. Korn, D. E. Arking, D. T. Miller, R. Fossdal, E. Saemundsen, H. Stefansson, M. A. Ferreira, T. Green, O. S. Platt, D. M. Ruderfer, C. A. Walsh, D. Altshuler, A. Chakravarti, R. E. Tanzi, K. Stefansson, S. L. Santangelo, J. F. Gusella, P. Sklar, B. L. Wu and M. J. Daly (2008). "Association between microdeletion and microduplication at 16p11.2 and autism." N Engl J Med 358(7): 667-675.

West, L., J. Waldrop and S. Brunssen (2009). "Pharmacologic treatment for the core deficits and associated symptoms of autism in children." J Pediatr Health Care 23(2): 75-89.

Williams, S. R., S. V. Mullegama, J. A. Rosenfeld, A. I. Dagli, E. Hatchwell, W. P. Allen, C. A. Williams and S. H. Elsea (2010). "Haploinsufficiency of MBD5 associated with a syndrome involving microcephaly, intellectual disabilities, severe speech impairment, and seizures." Eur J Hum Genet 18(4): 436-441.

Wingate, M., R. S. Kirby, S. Pettygrove, C. Cunniff, E. Schulz, T. Ghosh, C. Robinson, L. C. Lee, R. Landa, J. N. Constantino, R. Fitzgerald, W. Zahorodny, J. Daniels, J. Nicholas, J. Charles, W. McMahon, D. Bilder, M. Durkin, J. Baio, D. Christensen, K. V. Braun, H. Clayton, A. Goodman, N. Doernberg, M. Yeargin-Allsopp, E. Lott, K. C. Mancilla, A. Hudson, K. Kast, K. Jolly, A. Chang, R. Harrington, R. Fitzgerald, J. Shenouda, P. Bell, C. Kingsbury, A. Bakian, A. Henderson, C. Arneson, A. Washington, G. Frenkel and V. Wright (2014). "Prevalence of autism spectrum disorder among children aged 8 years - autism and developmental disabilities monitoring network, 11 sites, United States, 2010." MMWR Surveill Summ 63 Suppl 2: 1-21.

Wu, J. Y., K. C. Kuban, E. Allred, F. Shapiro and B. T. Darras (2005). "Association of Duchenne muscular dystrophy with autism spectrum disorder." J Child Neurol 20(10): 790-795.

Wu, Y. W., K. M. Prakash, T. Y. Rong, H. H. Li, Q. Xiao, L. C. Tan, W. L. Au, J. Q. Ding, S. D. Chen and E. K. Tan (2011). "Lingo2 variants associated with essential tremor and Parkinson's disease." Hum Genet 129(6): 611-615.

Xu, H., W. T. Poh, X. Sim, R. T. Ong, C. Suo, W. T. Tay, C. C. Khor, M. Seielstad, J. Liu, T. Aung, E. S. Tai, T. Y. Wong, K. S. Chia and Y. Y. Teo (2011). "SgD-CNV, a database for common and rare copy number variants in three Asian populations." Hum Mutat 32(12): 1341-1349.

Yeargin-Allsopp, M., C. Rice, T. Karapurkar, N. Doernberg, C. Boyle and C. Murphy (2003). "Prevalence of autism in a US metropolitan area." JAMA 289(1): 49-55.

Yu, T., Y. Yaguchi, D. Echevarria, S. Martinez and M. A. Basson (2011). "Sprouty genes prevent excessive FGF signalling in multiple cell types throughout development of the cerebellum." Development 138(14): 2957-2968.

Zhang, X. and C. Y. Ji (2005). "Autism and mental retardation of young children in China." Biomed Environ Sci 18(5): 334-340.

Zhou, X., Y. Xu, J. Wang, H. Zhou, X. Liu, Q. Ayub, X. Wang, C. Tyler-Smith, L. Wu and Y. Xue (2011). "Replication of the association of a MET variant with autism in a Chinese Han population." PLoS One 6(11): e27428.