The contribution of epigenetics to understanding genetic factors in autism

Abstract

Autism spectrum disorder is a grouping of neurodevelopmental disorders characterized by deficits in social communication and language, as well as by repetitive and stereotyped behaviors. While the environment is believed to play a role in the development of autism spectrum disorder, there is now strong evidence for a genetic link to autism. Despite such evidence, studies investigating a potential single-gene cause for autism, although insightful, have been highly inconclusive. A consideration of an epigenetic approach proves to be very promising in clarifying genetic factors involved in autism. The present article is intended to provide a review of key findings pertaining to epigenetics in autism in such a way that a broader audience of individuals who do not have a strong background in genetics may better understand this highly specific and scientific content. Epigenetics refers to non-permanent heritable changes that alter expression of genes without altering the DNA sequence itself and considers the role of environment in this modulation of gene expression. This review provides a brief description of epigenetic processes, highlights evidence in the literature of epigenetic dysregulation in autism, and makes use of noteworthy findings to illustrate how a consideration of epigenetic factors can deepen our understanding of the development of autism. Furthermore, this discussion will present a promising new way for moving forward in the investigation of genetic factors within autism.

Keywords

autism autism spectrum disorder epigenetic genetic

Introduction

Autism spectrum disorder (ASD) is a grouping of neurodevelopmental disorders characterized by stereotyped or repetitive behavior, as well as by impairments in social skills, language, and communication (American Psychiatric Association (APA), 2000). While a number of environmental factors have been suggested as contributors to the development of autism, there is without a doubt an effect of genetic factors, with some monozygotic twin studies revealing a heritability of 60%–90% (Veenstra-VanderWeele and Cook, 2004). More recent reports of heritability suggest that this may be an overestimation of heritability, placing estimates in the range of 37%–67% and also suggesting higher concordance rates between dizygotic twins than had been previously found. These findings imply a greater role for environmental factors (Hallmayer et al., 2011). Heritability estimates are complicated not only by factors that can decrease genetic similarity in monozygotic twins, for example, de novo mutations, copy-number variation (CNV), and asymmetrical epigenetic changes that can occur in only one twin, but are also complicated by factors that can increase genetic similarity between dizygotic twins, such as the shared gestational environment that results in similar epigenetic changes (Anderson, 2012; Bourgeron, 2012; Czyz et al., 2012). A recent review of literature estimates that 10%–20% of cases of autism have known genetic causes, attributing 6%–7% of these cases to chromosomal abnormalities, 2%–10% to CNV, and 1%–2% to single-gene disorders (Abrahams and Geschwind, 2008). Additionally, it has been reported that over 100 genomic loci have been identified as having a causal role in ASD (Betancur, 2011). To date, however, much of the research investigating the genetic etiology of idiopathic autism through linkage and association studies has been inconclusive (Grafodatskaya et al., 2010; Persico and Bourgeron, 2006; Sykes and Lamb, 2007). While confusion regarding genetic etiology supports the belief that the genetics of autism are complex and likely to involve many genes and genetic pathways, the question still remains as to what these genes and pathways may be.

Some insight into the genes and pathways involved in autism may be found by gaining a better understanding of epigenetic mechanisms and how they relate to the development of autism (Schanen, 2006). Epigenetics refers to non-permanent heritable changes that alter the expression of genes without altering the DNA sequence itself. Historically, the belief about DNA was that much of the human genome was “junk” because it did not contain genes and thus must not play a role in our growth and development. We now know that this belief was incorrect and that many of these non-gene regions play very important roles in human development, especially in the regulation of gene transcription. In part, these important roles are maintained by epigenetic regulation of the regions. The primary function of epigenetics is to regulate development through directing processes such as cell differentiation, tissue specification, and maintenance of cell lineages (Gropman and Batshaw, 2010). Changes that regulate these processes occur through alterations to the shape and formation of DNA, not to the actual nucleotide sequence, thus changing the ability of certain genes to be expressed. In other words, the strands of chromatin that contain genetic information can unwind, coil more tightly, loop, and interact with other proteins to turn specific genes on and off. Patterns of gene expression in different cells will determine their developmental fate. Epigenetic alterations result in part from the effects of the environment, thus allowing the environment to play a role in the phenotype by modulating gene expression (Van Vliet et al., 2007). It is this interaction of genetic and environmental information on gene expression and our phenotype that makes epigenetics such an attractive avenue for research in disorders such as autism, where the causes of the disorder are varied and elusive.

This article will begin with a brief discussion of some of the difficulties encountered by researchers trying to find a single gene responsible for autism. We will then give a brief discussion of epigenetics and how epigenetic studies may be more informative in the understanding of complex disorders such as autism. A review of the current body of literature regarding epigenetics as it contributes to autism will follow. Specifically, there will be a discussion of certain candidate genes under epigenetic control that have been identified as contributors to autism. Finally, we will conclude by discussing possible future directions for the field.

Single-gene findings in autism

A major area of focus research in autism research has been that of genetics, with the hope that identifying genetic links to autism will be beneficial both to the diagnosis and treatment of the disorder. The studies within this field have typically made use of one of three strategies for gene identification: linkage analysis, association studies, and cytogenetic analysis (O’Roak and State, 2008). Unfortunately, despite these varied methods and the plethora of studies that have been completed in the field, results have been difficult to replicate (Griswold et al., 2012; Sykes and Lamb, 2007; Vieland et al., 2011).

Linkage studies, which identify regions of the genome that are inherited together across generations, have yielded widely varied results, identifying multiple genes on multiple chromosomes, which are rarely identified again in other studies (Li et al., 2012; Veenstra-VanderWeele and Cook, 2004). Notably, one linkage study, which made use of 1181 families and 10,000 markers for a genome-wide scan, found significant results for linkage at the 11p12-p13 region (Szatmari et al., 2007). However, the failure of such a large and all-encompassing study to replicate other previous linkage findings is a further testament to the complexity of the genetic underpinning of autism, suggesting that many alleles that contribute to the disorder likely carry a very small individual risk (O’Roak and State, 2008).

Additionally, genetic association studies, which determine differences in the frequency of genetic variants between populations, have not seen great success in terms of consistency in their results (Abrahams and Geschwind, 2008; Li et al., 2012). Two of the greatest difficulties in producing accurate and repeatable results with this type of study are those of sample size and matching cases to controls. Specifically, association studies used to reveal common and rare variants often require very large sample sizes to produce results with enough power. Furthermore, an inability to appropriately match case groups to control groups raises questions as to whether genetic variants are identified because of a difference due to autism or a difference due to the composition of the groups. This is especially of concern with regard to differences in ethnicity. Without controlling for ethnicity through matching cases to controls, there is a strong possibility that significant gene findings could actually be indicative of gene variation that is unique to the ethnic group rather than to the autism genotype. Nonetheless, there have been some glimpses of hope regarding association studies. Most prominently, the gene CNTNAP2, which was originally discovered to play a role in autism by a linkage study, has since been suggested to have a significant role in autism via several gene-association studies (Alarcon et al., 2008; Arking et al., 2008; Li et al., 2010).

With a little more success, cytogenetic studies, which are used to identify single-gene defects (such as fragile X syndrome, Rett syndrome, and tuberous sclerosis) and chromosomal abnormalities, have suggested that such changes to the genome account for around 10% of cases of autism (Benvenuto et al., 2009; Schaefer and Mendelsohn, 2008), However, these single-gene defect studies have implicated almost every chromosome in the genome (Xu et al., 2004). One of the most commonly reported chromosomal abnormalities that has been connected to autism is a duplication of the 15q11-13 region, accounting for 1%–3% of autism cases (Depienne et al., 2009). A number of other studies have indicated the involvement of chromosomal abnormalities at a number of other loci within the genome that may play a role in autism, as well as numerous genes and gene–environment or gene–gene interactions that can influence development of autism (Benvenuto et al., 2009). However, other research groups have generally not replicated these findings.

In addition to these three traditional methods of searching for genetic links to autism, advances in microarray technology (technology for screening the genome) have made it possible to use CNVs as another avenue for genetic research in autism. By using this new technology that can detect CNVs, which are submicroscopic chromosomal abnormalities, researchers have been able to complete entire scans of the genome at a much higher resolution than has been previously possible (Kumar and Christian, 2009; LaSalle and Yasui, 2009). Although this technology is still a relatively new method used in autism research, it is already yielding intriguing results that might serve as a starting point for further research. Of note, one study that made use of CNV technology showed that de novo¹ mutations occur in 10% of simplex families (families where only one family member has been diagnosed with autism), compared to an occurrence of 3% in multiplex families and 1% in the neurotypical control group (Sebat et al., 2007). Furthermore, large-scale studies such as the Autism Genome Project and Simons Simplex Collection have used CNV technology for the detection of many new candidate genes related to autism by identifying rare but highly penetrant CNVs (Chung et al., 2013; Marshall and Scherer, 2012). Such CNVs are rare because they are found only at extremely low rates within the population, but are highly penetrant in that they will almost always produce the phenotypic traits with which they are associated — in this case, autistic traits. It is studies such as these that suggest a significant role of de novo mutations (i.e. mutations that have not been inherited from parents directly) in the genetic underpinnings of autism, a finding that would not have been possible without this new CNV technology.

While there have been some successes with basic genetic research, and while new technology proves to be promising, autism is a heterogeneous disorder and thus far the genetic findings have been highly varied (Eapen, 2011). Furthermore, although these highly varied genetic findings are in agreement with the fact that autism is an exceedingly complex disorder, the majority of this genetic research has been inconclusive. This is likely to be due in part to the difficulty of creating samples that are homogenous in terms of a particular subphenotype; using heterogeneous samples can lead to varied results. Additionally, as is common in autism research, many studies consist of small sample sizes that fail to produce significant findings due to insufficient power.

While it might be fair to attribute lack of success to issues with heterogeneity and sample size, studies that have attempted to rectify these issues still have not seen greatly improved results (e.g. Szatmari et al., 2007). The inability of studies to reveal consistent susceptibility loci or specific genes related to autism raises the question of whether alternate approaches to understanding genetic underpinning of ASD might be useful. Including consideration of epigenetic changes in autism research may aid in revealing more information about the nature and underlying mechanisms of this complex and heterogeneous disorder.

Epigenetics: mechanisms and significance to the field of autism research

Considering epigenetic theory when conducting genetic research is advantageous in that it accounts for a role of the environment in the contribution to gene expression. This is because certain environmental conditions and changes to the environment can stimulate epigenetic changes to the genome. Research with rodents has shown that limiting maternal care provided to rat pups results in changes in the expression of stress-related genes that are maintained in future generations (Meaney, 2001). Similar effects of maternal care on gene expression have been identified in nonhuman primates (Higley et al., 1991). More recently, research with human subjects has revealed that childhood abuse alters methylation patterns in the brain, resulting in epigenetic changes to gene expression (McGowan et al., 2009; Reaume and Sokolowski, 2011).

Being mindful of epigenetic processes and their interaction with environmental conditions may help us explain differences in gene expression within individuals with autism despite minimal consistent evidence of specific gene mutations. Although it is not entirely clear which environmental factors might be responsible for the epigenetic changes seen in autism, it has been suggested that they might include environmental toxins, parental age, diet, or chemicals in the womb (Chaste and Leboyer, 2012; Dolinoy et al., 2007; Jaenisch and Bird, 2003; Landrigan et al., 2012). One clear example of an environmental factor that results in epigenetic changes in autism is maternal use of valproic acid (a medication for treating seizures) during early pregnancy (Newschaffer et al., 2007). Maternal use of valproic acid has downstream effects on the production of GABAergic inhibitory neurons by inhibiting histone deacetylase, which is important for epigenetic regulation (Bourgeron, 2012).

There appear to be certain epigenetic mechanisms that might be especially relevant to the study of autism. However, before we introduce a discussion of the relevance of such mechanisms to autism, we must first gain an understanding of how these mechanisms work. Primarily, it is important to understand that the main method of epigenetic control is DNA methylation (Schanen, 2006). This is the action of adding methyl groups onto specific regions of DNA containing CpG dinucleotides. The majority of these dinucleotides are found methlyated throughout the genome (Klose and Bird, 2006). This methylation is highly important for regulating cell differentiation because of its ability to close off certain regions of DNA and subsequently alter protein transcription by influencing downstream pathways. CpGs also cluster together to form CpG islands. These islands are areas of DNA with a high concentration of the nucleotides cytosine and guanine, and they are generally found near promoters of genes, but not within the promoters themselves.² Typically these islands remain unmethylated, but occasionally they do become methylated, resulting in the methylation of a gene promoter. When the promoter region is methylated, the gene is generally turned off, or silenced, making it incapable of transcribing the proteins that it is responsible for making. Since CpG islands tend to function to maintain an unmethylated state of promoters, their methylation can be problematic in that the result is the silencing of genes whose expression is required for normal development and cellular functioning (De Leon-Guerrero et al., 2011; LaSalle, 2011). Although epigenetic mechanisms involved in autism are not well understood, there is good evidence for genome-wide methylation dysregulation in ASD (Melnyk et al., 2012), indicating that methylation is likely a significant contributor to the development of autism (Hu, 2013).

This process of methylation is responsible for two other epigenetic mechanisms that appear to be relevant to autism research: imprinting and X-inactivation (Sykes and Lamb, 2007). Imprinting refers to the process of methyl groups being used to mark genes according to the parent they are inherited from. As such, imprinting controls whether the paternal or maternal gene will be expressed in the offspring. For example, the 15q11-13 region is known as an imprinted region. When loss of function occurs in the paternally imprinted copy of 15q11-13, Prader–Willi syndrome is inherited, while loss of function in the maternally imprinted copy results in Angelman’s syndrome (Sykes and Lamb, 2007).

X-inactivation refers to a process that only occurs in females, in which one copy of the X chromosome is turned off by methylating one copy of the chromosome in order to achieve dosage compensation (to compensate for the fact that a female has two X chromosomes, while the male has only one). Either the maternal or paternal X chromosome is randomly inactivated in embryonic cells so that each copy still has expression somewhere (Li, 2002; Lyon, 1961). Problems with inactivation or removal of inactivation can result in genetic anomalies. One example of this is in the formation of tiny X rings in females. These rings prevent genes within the ring on the X chromosome from being inactivated, resulting in their overexpression, which can lead to developmental abnormalities such as Turner syndrome features, severe mental retardation, and developmental delay (Migeon, 2007).

Methylation is also responsible for the epigenetic process of long-distance chromosomal interactions. This is a process by which methylation signals interact with other chromosomes or other parts of the same chromosome that are far away to alter gene expression. An example of such interactions is found in chromatin looping, in which gene sequences are brought together via formation of a loop in the chromatin in order to facilitate long-distance gene activation (Bulger and Groudine, 1999). A review of epigenetic processes highlights the role of such looping in Rett syndrome, as well as the potential for long-distance chromosomal interaction in autism (LaSalle, 2007).

From this introduction to epigenetic theory, it is clear that the role of epigenetic processes in our genome is not a simple one. The many components of epigenetic regulation, including methylation, imprinting, X-inactivation, and long-distance chromosomal interaction, as well as others not discussed here, all work together and interact with one another to modulate a variety of different effects on different genes. Furthermore, each epigenetic process can be highly specific to a certain gene or genetic pathway, making it increasingly difficult to identify the targets and pathways involved in a specific epigenetic change. In general, there remain numerous questions within the field of epigenetics that play a critical role in our understanding of epigenetic theory; however, it is a booming field and, in light of new technological and analytical advances, we will find more answers every year (Portela and Esteller, 2010). What is clear is that epigenetic processes play a large and complex role in brain development and neurodevelopmental disorders (Gibbs et al., 2010; Houston et al., 2013; Shulha et al., 2012). Given that current genetic research in autism has been unable to provide a complete picture of the underpinnings of the disorder, combined with a rising emphasis on a role for environment in the development of autism, it follows that epigenetic mechanisms should also be studied as a potential avenue for understanding the genetics of autism.

Possible candidate genes for autism and their interactions

Research has identified a collection of genes that appear to be involved in epigenetic pathways of autism. Specifically, the GABRB3, UBE3A, and MECP2³ genes have been consistently shown to be epigenetically dysregulated in autism.

GABRB3

The GABRB3 gene creates a protein that is a subunit of the gamma-Aminobutyric acid-A (GABA_A) receptor. This receptor is important in regulating the neurotransmitter GABA and has a role in central nervous system (CNS) development, specifically synaptic function (Delahanty et al., 2011). Importantly, when the GABRB3 gene is knocked out in mice and no longer able to exert its effects, mice exhibit autistic-like behaviors including seizures, anxiety, and impaired social ability (DeLorey et al., 2008). This gene is found in the 15q11-13 region, which is an imprinted region, but the gene itself is generally not imprinted in the brains of typical controls or individuals with autism. Despite this finding, it has been recognized that the GABRB3 gene is only monoallelically expressed (only one allele of two is expressed) in autism samples, while in controls, its expression is bialellic (both alleles are expressed) (Hogart et al., 2007). This disruption in expression results in underexpression of GABA_A in autistic brains. Furthermore, duplication of the 15q11-13 region results in expression of GABRB3 that is not predicted by copy number, thus further implicating a role for epigenetic regulation (Scoles et al., 2011). It is likely that disruption in expression of GABRB3 may occur, owing to dysregulated imprinting or to issues with pairing of the two homologous alleles via disruption of long-distance chromosomal interactions (Coghlan et al., 2012; Hogart et al., 2009; Meguro-Horike et al., 2011).

UBE3A

UBE3A is best known as the Angelman’s syndrome gene because this disorder arises when there is maternal loss of function in the UBE3A gene region. Specifically, this gene appears to play a role in maintaining synaptic plasticity and is required for experience-dependent changes in the brain (Yashiro et al., 2009). Mouse models with increased UBE3A gene dosage have been shown to demonstrate autistic-like traits, suggesting a link between the gene and autism (Smith et al., 2011). Like GABRB3, UBE3A is also found in the 15q11-13 region; however, UBE3A is an imprinted gene. This gene is of specific importance because of the tendency for chromosomal abnormalities (especially duplications) in this region within autistic populations. Despite the tendency of this gene to show duplication in autism, proper imprinting patterns within the gene region are still maintained (Herzing et al., 2002), which confuses the finding that UBE3A expression is deficient in models of autism (LaSalle, 2007). In autism, although the main copy of the gene has imprinting maintained in the DNA, the complementary copy of the gene appears to lose its imprinted status, resulting in production of antisense RNA. Antisense RNA is a strand of RNA that is complementary to the mRNA produced from the same gene, and exists to inhibit the translation of this complementary mRNA into a protein by binding to it. In some cases of autism, it appears that the antisense RNA, produced via loss of imprinting, binds to the UBE3A mRNA and thus prevents protein production (Makedonski et al., 2005). Thus, like the GABRB3 gene, problems in autism relating to the UBE3A gene appear to be related to epigenetic control rather than mutation.

MECP2

MECP2 is a gene found in the X chromosome, and is responsible for making a particular kind of methyl-binding protein called Methyl-CpG-binding protein 2. Methyl-binding proteins are proteins that bind to the methylated regions of DNA to facilitate silencing genes. Thus, MECP2 is a key player in epigenetic processes, especially those critical to synaptogenesis and long-term synaptic plasticity (Kavalali et al., 2011). Like UBE3A, MECP2 is also highly associated with another disorder: Rett syndrome. Rett syndrome was until recently classified as one of the pervasive developmental disorders and is most commonly caused by mutations to MECP2 (Amir et al., 1999; Marco and Skuse, 2006). Alternately, MECP2 mutations are incredibly rare in autism (Lopez-Rangel and Lewis, 2006). Nonetheless, there have been reports that up to 79% of autistic patients experience a significant reduction in MeCP2 protein expression (Gonzales and LaSalle, 2010; Nagarajan et al., 2006). It is likely that MeCP2 disruptions in ASD are associated with decreased expression of MECP2, rather than increased expression, given that MECP2 duplications are very rare in individuals with ASD (Xi et al., 2010).

An additional interest in MECP2 in relation to Rett syndrome and autism comes from evidence showing that its protein is particularly abundant in mature neuron cells and is likely linked to proper neuron maturation (Shahbazian et al., 2002). Recent findings have revealed that MECP2 expression is heterogeneous in neuronal cells, expressing in both high (MeCP2^hi) and low (MeCP2^lo) protein levels within different protein cells. In prenatal life, there is a predominant expression of MeCP2^lo protein cells; however, after birth, expression begins to change so that MeCP2^hi-expressing neuronal cells become more predominant, which is indicative of the importance of significant MeCP2 expression in neuron maturation (Balmer et al., 2003). In Rett syndrome, the mutation of MECP2 has been shown to result in a significant decrease of MeCP2^hi protein cells compared to what is seen in typically developing brains after birth (LaSalle et al., 2001). Interestingly, the transition from low expression to high expression occurs in a developmental time frame similar to that for the onset of Rett syndrome, which is seen around 6 to 18 months, and may explain the onset of Rett syndrome as well as cases of autistic-like regression.

The decrease in MeCP2^hi cells is also seen in autism despite the inability to find evidence of MECP2 mutation in autism (Samaco et al., 2004). Instead, it has been found that this lower level of MeCP2^hi-expressing cells is linked to increased methylation of the MECP2 promoter, indicating a role for epigenetics in controlling MECP2 and in autism (Nagarajan et al., 2006). It is not clear how epigenetic factors are working in the control of MECP2; however, there has been some evidence of inability to effectively define methylation and X-inactivation boundaries (Nagarajan et al., 2008). In other words, there appear to be specific parts of the genome that should typically be methylated; however, the cellular processes responsible for ensuring methylation within these areas are not entirely accurate at recognizing the boundaries of these regions. As such, methylation may extend beyond the intended boundaries, as well as fall short of them. Results concerning a role of X-inactivation remain under debate, with some research in support of aberrant X-inactivation patterns and some research unable to produce such findings (Gong et al., 2008; Nagarajan et al., 2008; Talebizadeh et al., 2005).

In addition to evidence that MECP2 is under epigenetic control, there have also been a number of studies that point to MeCP2 as playing a role in the regulation of other genes via epigenetic processes. Specifically, MeCP2 has been shown to perform classical promoter repression by binding to methylated DNA and recruiting corepressors, as well as through facilitating chromatin looping (Chadwick and Wade, 2007; Woods and LaSalle, 2012; Yasui et al., 2007). This process of chromatin looping has been studied thoroughly with regard to Rett syndrome. Typically, MeCP2 binds at imprinted regions and brings together nucleosomes in order to form a loop of chromatin, condensing and silencing genes within the loop; however, this process breaks down in Rett syndrome (Georgel et al., 2003; Nikitina et al., 2007). Given evidence for the role of MeCP2 as an epigenetic regulator, epigenetic dysregulation of MECP2 is likely to result in a cascade of further epigenetic dysregulation of other genes (Flashner et al., 2013; Guy et al., 2011). In light of this, and with evidence of significantly decreased levels of MeCP2 in autism, it would be of interest to further investigate MeCP2 target genes and the role they might play in autism through epigenetic processes.

MECP2 as an epigenetic regulator of UBE3A and GABRB3

With evidence that these three genes are all implicated in autism and the overlapping similarities in phenotype of Rett syndrome, Angelman’s syndrome, and autism, it is possible that these genes somehow interact in autism. In fact, this appears to be the case. By both observing MECP2 knockout mice and investigating autistic brains with MeCP2 protein deficiency, it was found that there were also reductions in the expression of UBE3A and GABA_A (Kurian et al., 2007; Samaco et al., 2005). It appears that this interaction occurs through the organizing ability of MeCP2 proteins. MeCP2 proteins have been shown to directly bind to the GABRB3 gene (Hogart et al., 2007). More importantly, MeCP2 proteins have been shown to be involved in the organization of chromosomes in the brain leading to appropriate homologous pairing of the 15q11-q13 (Thatcher et al., 2005). Homologous pairing is the act of two matching chromosomes (one from the mother and one from the father) coming together, and is thought to be an important factor in gene regulation during cell differentiation. This ability of MeCP2 is based on its action in long-distance chromosomal interactions (Yasui et al., 2011). Thus, when MeCP2 is deficient, it would appear that the expression of UBE3A and GABRB3 is also epigenetically disrupted owing to their inability to come together in pairs via these long-distance interactions.

Limitations within research

While the above description provides insight into how several epigenetic mechanisms might influence the development of autism, it is important to remember that this field is new and that the body of research is still growing. Hence, replication is still needed for many findings; however, thus far replication of findings regarding the epigenetics of autism has been promising and exceeds replication abilities of other genetic research. Specifically, consistent results have been found for the involvement of the three genes discussed in detail within this article: GABRB3, UBE3A, and MECP2 (Grafodatskaya et al., 2010). There is also a growing body of literature to support the epigenetic role of other genes in autism, such as the reelin (RELN) gene (Persico et al., 2006), which plays a role in neuronal migration and synaptogenesis; the oxytocin receptor (OXTR) gene (Gregory et al., 2009), which is involved in social behavior; and the engrailed-2 gene (EN-2), which is responsible for Purkinje cell maturation (James et al., 2013). Recently, methylation profiling of lymphoblastoid cell lines has revealed differential methylation patterns between individuals with autism and their non-autistic siblings, identifying the retinoic acid-related orphan receptor alpha (RORA) gene as an epigenetically dysregulated gene in autism (Nguyen et al., 2010). As this field continues to grow, it is likely that replicated results concerning these and other genes will come forward.

In addition to replication, there are a number of methodological issues with this research, the most obvious being sample size. Samples continue to be small, thus reducing the power of the results. One of the unfortunate realities of carrying out epigenetic research is its reliance on the testing of postmortem brain tissue, which is not readily available in large amounts. Some studies have attempted to rectify this challenge by carrying out studies on mouse models. While mice have many similarities to humans and can be helpful for providing insight to both human behavior and genetic processes, we cannot be sure that results would be exactly the same in humans. Additionally, it is difficult to interpret behavior, especially autistic-like behaviors, in mice.

There are also a number of technological limits within epigenetic research, and genetic research in general. There is a constant search for more accurate and efficient methods of carrying out genetic research. The most important technological limitation in epigenetic research is the method used for detecting areas of DNA methylation. As mentioned, methylation more commonly occurs in noncoding regions of the genome than at CpG islands found in promoter regions of a gene. For example, only 5.9% of sites under epigenetic control via MeCP2 binding fall within the CpG islands (Yasui et al., 2007). Currently, the best method we have for identifying methylation sites, bisulfite sequencing, results in selective detection of the CpG islands in these promoter regions of genes, failing to reveal other regions where methylation may occur (LaSalle and Yasui, 2009).⁴ Developing methods that would allow for more complete reading of the epigenome would bring us closer to creating entire maps of DNA methylation of the human genome, thus increasing our knowledge of epigenetic control with typical and atypical populations.

Conclusion

Autism is a highly heterogeneous and complex disorder. It has been very difficult to elucidate the single-gene factors that contribute to the disorder. Much of genetic research in autism has yielded inconsistent results that suggest a wide variety of susceptibility loci and potential candidate genes, as well as a complex myriad of gene–gene and gene–environment interactions.

In light of these difficulties and inconsistencies in genetic research, recent findings in the possible epigenetic processes related to autism present a new and promising way to investigate the genetic nature of autism. Specifically, a number of genes have been revealed that may experience epigenetic dysregulation. A growing body of research strongly supports the contribution of the genes GABRB3, UBE3A, and MECP2 in the development of autism. These three genes have all been shown to have decreased expression in autistic brains, despite evidence that they remain unmutated. More so, through the use of knockout mice, there is evidence that MECP2 regulates UBE3A and GABRB3 expression. This control appears to occur by facilitating homologous pairing, which is likely to be a result of long-distance chromosomal interaction.

With the number of environmental factors that have been suggested to play a role in autism, it is not surprising that epigenetics is revealing increasing insight into the genetic regulation of autism. While it is still unclear what environmental factors might influence these epigenetic changes and how they are able to exert their effects at a cellular level, this will prove to be an interesting area of research as we gain more insight into the complex gene interactions in autism.

Research in this field has great promise not only to improve our understanding of the etiology of autism but also to make contributions to clinical aspects of autism. The ability to identify reliable markers that indicate a risk for autism would mean that we might then be able to use such markers in screening tests for individuals in the diagnostic process; such methods are already used with MECP2 in cases of Rett syndrome. Additionally, reversing epigenetic dysregulation has been shown to improve symptoms in mouse models of Rett syndrome (Percy, 2011). Elucidating epigenetic pathways would increase the likelihood of then being able to develop drugs that could target such pathways in order to treat symptoms of ASD.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

References

Abrahams

Geschwind

(2008) Advances in autism genetics: on the threshold of a new neurobiology. Nature Reviews Genetics 9(5): 341–355.

Alarcon

Abrahams

Stone

. (2008) Linkage association, and gene expression analysis identify CNTNAP2 as an autism-susceptibility gene. The American Journal of Human Genetics 82(1): 150–159.

American Psychiatric Association (APA) (2000) Diagnostic and Statistical Manual of Mental Disorders. 4th ed., text rev. Washington, DC: American Psychiatric Association.

Amir

Van den Veyver

Wan

. (1999) Rett syndrome is caused by mutations in x-linked MECP2, encoding methyl-CpG-binding protein 2. Nature Genetics 23(2): 185–188.

Anderson

(2012) Twin studies in autism: what might they say about genetic and environmental influences. Journal of Autism and Developmental Disorders 42(7): 1526–1527.

Arking

Cutler

Brune

. (2008) A common genetic variant in Neurexin superfamily member CNTNAP2 increases familial risk of autism. The American Journal of Human Genetics 82(1): 7–9.

Balmer

Goldstine

Rao

. (2003) Elevated methyl-CpG-binding protein 2 expression is acquired during postnatal human brain development and is correlated with alternative polyadenylation. Journal of Molecular Medicine 81(1): 61–68.

Benvenuto

Moavero

Alessandrelli

. (2009) Syndromic autism: causes and pathogenic pathways. World Journal of Pediatrics 5(3): 169–176.

Betancur

(2011) Etiological heterogeneity in autism spectrum disorders: more than 100 genetic and genomic disorders and still counting. Brain Research 1380: 42–77.

10.

Bourgeron

(2012) Genetics and epigenetics of autism spectrum disorders. In: Corsi

Christen

(eds) Epigenetics, Brain and Behaviour. Heidelberg: Springer, pp. 105–132.

11.

Bulger

Groudine

(1999) Looping versus linking: towards a model for long-distance gene activation. Genes & Development 13(19): 2465–2477.

12.

Chadwick

Wade

(2007) MeCP2 in Rett syndrome: transcriptional repressor or chromatin architectural protein? Current Opinion in Genetics & Development 17(2): 121–125.

13.

Chaste

Leboyer

(2012) Autism risk factors: genes, environment and gene-environment interactions. Dialogues in Clinical Neuroscience 14(3): 281–292.

14.

Chung

Tao

Tso

(2013) Copy number variation and autism: new insights and clinical implications. Journal of the Formosan Medical Association. Epub ahead of print 13 March 2013. DOI: 10.1016/j.jfma2013.01.005.

15.

Coghlan

Horder

Inkster

. (2012) GABA system dysfunction in autism and related disorders: from synapse to symptoms. Neuroscience and Biobehavioral Reviews 36(9): 2044–2055.

16.

Czyz

Morahan

Ebers

. (2012) Genetic, environmental and stochastic factors in monozygotic twin discordance with a focus on epigenetic differences. BMC Medicine 10(93): 1–12.

17.

Delahanty

Kang

Brune

. (2011) Maternal transmission of a rare GABRB3 signal peptide variant is associated with autism. Molecular Psychiatry 16(1): 86–96.

18.

De Leon-Guerrero

Padraza-Alva

Perez-Martinez

(2011) In sickness and in health: the role of methyl-CpG binding protein 2 in the central nervous system. European Journal of Neuroscience 33(9): 1563–1574.

19.

DeLorey

Sahbaei

Hashemi

. (2008) GABRB3 gene deficient mice exhibit impaired social and exploratory behaviours, deficits in non-selective attention and hypoplasia of cerebellar vermal lobules: a protein model of autism spectrum disorder. Behavioural Brain Research 187(2): 207–220.

20.

Depienne

Moreno-De-Luca

Heron

. (2009) Screening for genomic rearrangements and methylation abnormalities of the 15q11-q13 region in autism spectrum disorders. Biological Psychiatry 66(4): 349–359.

21.

Dolinoy

Weidman

Jirtle

(2007) Epigenetic gene regulation: linking early development environment to adult disease. Reproductive Toxicology 23(3): 297–307.

22.

Eapen

(2011) Genetic basis of autism: is there a way forward? Current Opinion in Psychiatry 24(3): 226–236.

23.

Feinberg

(2010) Genome-scale approached to the epigenetics of common human disease. Virchows Archiv 456(1): 13–31.

24.

Flashner

Russo

Boileau

. (2013) Epigenetic factors and autism spectrum disorders. Neuromolecular Medicine 15(2): 339–350.

25.

Georgel

Horowitz-Scherer

Adkins

. (2003) Chromatin compaction by human MeCP2. The Journal of Biological Chemistry 278(34): 32181–32188.

26.

Gibbs

van der Burg

Hernandez

. (2010) Abundant quantitative trait loci exist for DNA methylation and gene expression in human brain. PLoS Genetics 6(5): e1000952.

27.

Gong

Bacchelli

Blasi

. (2008) Analysis of X chromosome inactivation in autism spectrum disorders. American Journal of Medical Genetics 147B(6): 830–835.

28.

Gonzales

LaSalle

(2010) The role of MeCP2 in brain development and neurodevelopmental disorders. Current Psychiatry Reports 12(2): 127–134.

29.

Grafodatskaya

Chung

Szatmari

. (2010) Autism spectrum disorders and epigenetics. Journal of the American Academy of Child & Adolescent Psychiatry 49(8): 794–809.

30.

Gregory

Connelly

Towers

. (2009) Genomic and epigenetic evidence for oxytocin receptor deficiency in autism. BMC Medicine 7(1): 62–75.

31.

Griswold

Cukier

(2012) Evaluation of copy number variations reveals novel candidate genes in autism spectrum disorder-associated pathways. Human Molecular Genetics 21(15): 3513–3525.

32.

Gropman

Batshaw

(2010) Epigenetics, copy number variation and other molecular mechanisms underlying neurodevelopmental disabilities: new insights and diagnostic approaches. Journal of Developmental and Behavioral Pediatrics 31(7): 582–591.

33.

Guy

Cheval

Selfridge

. (2011) The role of MeCP2 in the brain. Annual Review of Cell and Developmental Biology 27: 631–652.

34.

Hallmayer

Cleveland

Torres

(2011) Genetic Heritability and shared environmental factors among twin pairs with autism. Archives of General Psychiatry 68(11): 1095–1102.

35.

Herzing

LBK

Cook

Jr Ledbetter

(2002) Allele-specific expression analysis by RNA-FISH demonstrate preferential maternal expression of UBE3A and imprint maintenance within 15q11-q13 duplications. Human Molecular Genetics 11(15): 1707–1718.

36.

Higley

Hasert

Suomi

. (1991) Nonhuman primate model of alcohol abuse: effects of early experience, personality and stress on alcohol consumption. Proceedings of the National Academy of Sciences of the United States of America 88(16): 7261–7265.

37.

Hogart

Leung

Wang

(2009) Chromosome 15q11-13 duplication syndrome brain reveals epigenetic alterations in gene expression not predicted from copy number. Journal of Medical Genetics 46(2): 86–93.

38.

Hogart

Nagarajan

Patzel

. (2007) 15q11-13 GABAA receptor genes are normally biallelically expressed in brain yet are subject to epigenetic dysregulation in autism-spectrum disorders. Human Molecular Genetics 16(6): 691–703.

39.

Houston

Peter

Mitchell

. (2013) Epigenetics in the human brain. Neuropsychopharmacology 38(1): 183–197.

40.

(2013) From genes to environment: using integrative genomics to build a “systems-level” understanding of autism spectrum disorders. Child Development 84(1): 89–103.

41.

Jaenisch

Bird

(2003) Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nature Genetics 33(Suppl. 3): 245–254.

42.

James

Shpyleva

Melnyk

. (2013) Complex epigenetic regulation of Engrailed-2 (EN-2) homeobox gene in the autism cerebellum. Translational Psychiatry 3: e232.

43.

Kavalali

Nelson

Monteggia

(2011) Role of MeCP2 DNA methylation and HDACs in regulating synapse function. Journal of Neurodevelopmental Disorders 3(3): 250–256.

44.

Klose

Bird

(2006) Genomic DNA methylation: the mark ad its mediators. Trends in Biochemical Sciences 31(2): 89–97.

45.

Kumar

Christian

(2009) Genetics of autism spectrum disorders. Current Neurology and Neuroscience Reports 9(3): 188–197.

46.

Kurian

Forbes-Lorman

Auger

(2007) Sex difference in MeCP2 expression during a critical period of rat brain development. Epigenetics 2(3): 173–178.

47.

Landrigan

Lambertini

Birnbaum

(2012) A research strategy to discover the environmental causes of autism and neurodevelopmental disabilities. Environmental Health Perspectives 120(7): 258–260.

48.

LaSalle

(2007) The odyssey of MeCP2 and parental imprinting. Epigenetics 2(1): 5–10.

49.

LaSalle

(2011) A genomic point-of-view on environmental factors influencing the human brain methylome. Epigenetics 6(7): 862–869.

50.

LaSalle

Yasui

(2009) Evolving role of MeCP2 in Rett syndrome and autism. Epigenomics 1(1): 119–130.

51.

LaSalle

Goldstine

Balmer

. (2001) Quantitative localization of heterogeneous methyl-CpG-binding protein 2 (MeCP2) expression phenotypes in normal and Rett syndrome brain by laser scanning cytometry. Human Molecular Genetics 10(17): 1729–1740.

52.

(2002) Chromatin modification and epigenetic reprogramming in mammalian development. Nature Reviews Genetics 3(9): 662–673.

53.

Hua

Brown

(2012) Genes associated with autism spectrum disorder. Brain Research Bulletin 88(6): 543–552.

54.

. (2010) Association analysis of CNTNAP2 polymorphisms with autism in the Chinese Han population. Psychiatric Genetics 20(3): 113–117.

55.

Lopez-Rangel

Lewis

MES

(2006) Loud and clear evidence for gene silencing by epigenetic mechanisms in autism spectrum and related neurodevelopmental disorders. Clinical Genetics 69(1): 21–25.

56.

Lyon

(1961) Gene action in the X chromosome of the mouse (Mus musculus L.). Nature 190(4773): 372–373.

57.

McGowan

Sasaki

D’Alessio

. (2009) Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse. Nature Neuroscience 12(3): 342–348.

58.

Makedonski

Abuhatzira

Kaufman

. (2005) MeCP2 deficiency in Rett syndrome causes epigenetic aberrations at the PWS/AS imprinting center that affects UBE3A expression. Human Molecular Genetics 14(8): 1049–1058.

59.

Marco

Skuse

(2006) Autism-lessons from the X chromosome. Social Cognitive and Affective Neuroscience 1(3): 183–193.

60.

Marshall

Scherer

(2012) Detection and characterization of copy number variation in autism spectrum disorder. In: Feuk

(ed.) Genomic Structural Variants: Methods and Protocols. New York: Springer, pp. 115–135.

61.

Meaney

(2001) Maternal care, gene expression and the transmission of individual difference in stress reactivity across generations. Annual Review of Neuroscience 24(1): 1161–1192.

62.

Meguro-Horike

Yasui

Powell

. (2011) Neuron-specific impairment of inter-chromosomal pairing and transcription in a novel model of human 15q duplication syndrome. Human Molecular Genetics 20(19): 3798–3810.

63.

Melnyk

Fuchs

Schulz

. (2012) Metabolic imbalance associated with methylation dysregulation and oxidative damage in children with autism. Journal of Autism and Developmental Disorders 42(3): 367–377.

64.

Migeon

(2007) Why females are mosaics, X-chromosome inactivation and sex difference in disease. Gender Medicine 4(2): 97–105.

65.

Nagarajan

Hogart

Gwye

. (2006) Reduced MECP2 expression is frequent in autism frontal cortex and correlates with aberrant MECP2 promoter methylation. Epigenetics 1(4): 1–11.

66.

Nagarajan

Patzel

Martin

. (2008) MECP2 Promoter methylation and X-chromosome inactivation in autism. Autism Research 1(3): 169–178.

67.

Newschaffer

Croen

Daniels

. (2007) The epidemiology of autism spectrum disorders. Annual Review of Public Health 28: 235–258.

68.

Nguyen

Raich

Pfeifer

. (2010) Global methylation profiling of lymphoblastoid cell lines reveals epigenetic contributions to autism spectrum disorders and a novel autism candidate gene, RORA, whose protein product is reduced in autistic brain. FASEB Journal 24(8): 3036–3051.

69.

Nikitina

Shi

Ghosh

. (2007) Multiple modes of interaction between the methylated DNA binding protein MeCP2 and chromatin. Molecular and Cellular Biology 27(3): 864–877.

70.

O’Roak

State

(2008) Autism genetics: strategies, challenges, and opportunities. Autism Research 1(1): 4–17.

71.

Percy

(2011) Rett syndrome: exploring the autism link. Archives of Neurology 68(8): 985–989.

72.

Persico

Bourgeron

(2006) Searching for ways out of the autism maze: genetic, epigenetic and environmental clues. Trends in Neurosciences 29(7): 349–358.

73.

Persico

Levitt

Pimenta

(2006) Polymorphic GGC repeat differentially regulates human reelin gene expression levels. Journal of Neural Transmission 113(10): 1373–1382.

74.

Portela

Esteller

(2010) Epigenetic modifications and human disease. Nature Biotechnology 28(10): 1057–1068.

75.

Reaume

Sokolowski

(2011) Conservation of gene function in behaviour. Philosophical Transactions of the Royal Society B: Biological Sciences 366(1574): 2100–2110.

76.

Samaco

Hogart

LaSalle

(2005) Epigenetic overlap in autism-spectrum neurodevelopmental disorders: MECP2 deficiency causes reduced expression of UBE3A and GABRB3. Human Molecular Genetics 14(4): 483–492.

77.

Samaco

Nagarajan

Braunschweig

. (2004) Multiple pathways regulate MeCP2 expression in normal brain development and exhibit defects in autism-spectrum disorders. Human Molecular Genetics 13(6): 629–639.

78.

Schaefer

Mendelsohn

(2008) Genetics evaluation for the etiologic diagnosis of autism spectrum disorders. Genetics in Medicine 10(1): 4–12.

79.

Schanen

(2006) Epigenetics of autism spectrum disorders. Human Molecular Genetics 15(2): 138–150.

80.

Scoles

Urraca

Chadwick

. (2011) Increased copy number for methylated maternal 15q duplications leads to changes in gene and protein expression in human cortical samples. Molecular Autism 2(1): 19–33.

81.

Sebat

Lakshmi

Malhotra

. (2007) Strong association of de novo copy number mutations with autism. Science 316(5823): 445–449.

82.

Shahbazian

Antalffy

Armstrong

. (2002) Insight into Rett syndrome: MeCP2 levels display tissue- and cell-specific differences and correlate with neuronal maturation. Human Molecular Genetics 11(2): 115–124.

83.

Shulha

Cheung

Whittle

. (2012) Epigenetic signatures of autism. Archives of General Psychiatry 69(3): 314–324.

84.

Smith

SEP

Zhou

Zhang

. (2011) Increased gene dosage of Ube3a results in autism traits and decreased glutamate synaptic transmission in mice. Science Translational Medicine 3(103): 103ra97.

85.

Sykes

Lamb

(2007) Autism: the quest for the genes. Expert Reviews in Molecular Medicine 9(24): 1–15.

86.

Szatmari

Paterson

Zwaigenbaum

. (2007) Mapping autism risk loci using genetic linkage and chromosomal rearrangements. Nature Genetics 39(3): 319–328.

87.

Talebizadeh

Bittel

Veatch

. (2005) Brief report: non-random X chromosome inactivation in females with autism. Journal of Autism and Developmental Disorders 35(5): 675–681.

88.

Thatcher

Peddada

Yasui

. (2005) Homologous pairing of 15q11-13 imprinted domains in brain is developmentally regulated but deficient in Rett and autism samples. Human Molecular Genetics 14(6): 785–797.

89.

Van Vliet

Oates

Whitelaw

(2007) Epigenetic mechanisms in the context of complex diseases. Cellular and Molecular Life Science 64(12): 1531–1538.

90.

Veenstra-VanderWeele

Cook

Jr (2004) Molecular genetics of autism spectrum disorder. Molecular Psychiatry 9(9): 819–832.

91.

Vieland

Hallmayer

Huang

. (2011) Novel method for combined linkage and genome-wide association analysis finds evidence of distinct genetic architecture for two subtypes of autism. Journal of Neurodevelopmental Disorders 3(2): 113–123.

92.

Woods

LaSalle

(2012) Epigenetic epidemiology of autism and other neurodevelopmental disorders. In: Michels

(ed.) Epigenetic Epidemiology. Netherlands: Springer. pp.321–342.

93.

Tan

. (2010) Analysis of MECP2 gene copy number in boys with autism. Journal of Child Neurology 26(5): 570–573.

94.

Zwaigenbaum

Szatmari

. (2004) Molecular cytogenetics of autism. Current Genomics 5(4): 347–364.

95.

Yashiro

Riday

Condon

. (2009) UBE3a is required for experience-dependent maturation of the neocortex. Nature Neuroscience 12(6): 777–785.

96.

Yasui

Peddada

Bieda

. (2007) Integrated epigenomic analyses of neuronal MeCP2 reveal a role for long-range interaction with active genes. Proceedings of the National Academy of Sciences of the United States of America 104(49): 19416–19421.

97.

Yasui

Scoles

Horike

. (2011) 15q11.2-13.3 chromatin analysis reveals epigenetic regulation of CHRNA7 with deficiencies in Rett and autism brain. Human Molecular Genetics 20(22): 4311–4323.