Chloe Wong

A data-driven approach to preprocessing Illumina 450K methylation array data

BackgroundAs the most stable and experimentally accessible epigenetic mark, DNA methylation is of great interest to the research community. The landscape of DNA methylation across tissues, through development and in disease pathogenesis is not yet well characterized. Thus there is a need for rapid and cost effective methods for assessing genome-wide levels of DNA methylation. The Illumina Infinium HumanMethylation450 (450K) BeadChip is a very useful addition to the available methods for DNA methylation analysis but its complex design, incorporating two different assay methods, requires careful consideration. Accordingly, several normalization schemes have been published. We have taken advantage of known DNA methylation patterns associated with genomic imprinting and X-chromosome inactivation (XCI), in addition to the performance of SNP genotyping assays present on the array, to derive three independent metrics which we use to test alternative schemes of correction and normalization. These metrics also have potential utility as quality scores for datasets.ResultsThe standard index of DNA methylation at any specific CpG site is β = M/(M + U + 100) where M and U are methylated and unmethylated signal intensities, respectively. Betas (βs) calculated from raw signal intensities (the default GenomeStudio behavior) perform well, but using 11 methylomic datasets we demonstrate that quantile normalization methods produce marked improvement, even in highly consistent data, by all three metrics. The commonly used procedure of normalizing betas is inferior to the separate normalization of M and U, and it is also advantageous to normalize Type I and Type II assays separately. More elaborate manipulation of quantiles proves to be counterproductive.ConclusionsCareful selection of preprocessing steps can minimize variance and thus improve statistical power, especially for the detection of the small absolute DNA methylation changes likely associated with complex disease phenotypes. For the convenience of the research community we have created a user-friendly R software package called wateRmelon, downloadable from bioConductor, compatible with the existing methylumi, minfi and IMA packages, that allows others to utilize the same normalization methods and data quality tests on 450K data.

A longitudinal study of epigenetic variation in twins.

DNA methylation is a key epigenetic mechanism involved in the developmental regulation of gene expression. Alterations in DNA methylation are established contributors to inter-individual phenotypic variation and have been associated with disease susceptibility. The degree to which changes in loci-specific DNA methylation are under the influence of heritable and environmental factors is largely unknown. In this study, we quantitatively measured DNA methylation across the promoter regions of the dopamine receptor 4 gene (DRD4), the serotonin transporter gene (SLC6A4/SERT) and the X-linked monoamine oxidase A gene (MAOA) using DNA sampled at both ages 5 and 10 years in 46 MZ twin-pairs and 45 DZ twin-pairs (total n=182). Our data suggest that DNA methylation differences are apparent already in early childhood, even between genetically identical individuals, and that individual differences in methylation are not stable over time. Our longitudinal-developmental study suggests that environmental influences are important factors accounting for interindividual DNA methylation differences, and that these influences differ across the genome. The observation of dynamic changes in DNA methylation over time highlights the importance of longitudinal research designs for epigenetic research.

Methylomic analysis of monozygotic twins discordant for autism spectrum disorder and related behavioural traits

Autism spectrum disorder (ASD) defines a group of common, complex neurodevelopmental disorders. Although the aetiology of ASD has a strong genetic component, there is considerable monozygotic (MZ) twin discordance indicating a role for non-genetic factors. Because MZ twins share an identical DNA sequence, disease-discordant MZ twin pairs provide an ideal model for examining the contribution of environmentally driven epigenetic factors in disease. We performed a genome-wide analysis of DNA methylation in a sample of 50 MZ twin pairs (100 individuals) sampled from a representative population cohort that included twins discordant and concordant for ASD, ASD-associated traits and no autistic phenotype. Within-twin and between-group analyses identified numerous differentially methylated regions associated with ASD. In addition, we report significant correlations between DNA methylation and quantitatively measured autistic trait scores across our sample cohort. This study represents the first systematic epigenomic analyses of MZ twins discordant for ASD and implicates a role for altered DNA methylation in autism.

Drugs and addiction: an introduction to epigenetics

Addiction is a debilitating psychiatric disorder, with a complex aetiology involving the interaction of inherited predispositions and environmental factors. Emerging evidence suggests that epigenetic alterations to the genome, including DNA methylation and histone modifications, are important mechanisms underlying addiction and the neurobiological response to addictive substances. In this review, we introduce the reader to epigenetic mechanisms and describe a potential role for dynamic epigenetic changes in mediating addictive behaviours via long-lasting changes in gene expression. We summarize recent findings from both molecular and behavioural experiments elucidating the role of epigenetic changes in mediating the addictive potential of various drugs of abuse, including cocaine, amphetamine and alcohol. The implications of these findings for molecular studies of addiction and the future development of novel therapeutic interventions are also discussed.

Genetic and epigenetic associations of MAOA and NR3C1 with depression and childhood adversities

Monoamine oxidase A (MAOA) harbours a polymorphic upstream variable-number tandem repeat (u-VNTR). The MAOA-L allele of the u-VNTR leads to decreased gene expression levels in vitro and has been found to increase the risk of conduct disorder in males with childhood adversities. Early-life adversities have been associated with hypermethylation of the glucocorticoid receptor (NR3C1). In this study, we first performed a genetic association analysis of the MAOA u-VNTR using individuals with depression (n = 392) and controls (n = 1276). Next, DNA methylation analyses of MAOA and NR3C1 were performed using saliva samples of depressed and control subgroups. Adult MAOA-L females with childhood adversities were found to have a higher risk of developing depression (p = 0.006) and overall MAOA methylation levels were decreased in depressed females compared to controls (mean depressed, 42% vs. mean controls, 44%; p = 0.04). One specific childhood adversity [early parental death (EPD)] was associated with hypermethylation of NR3C1 close to an NGFI-A binding site (mean EPD, 19% vs. mean non-EPD, 14%; p = 0.005). Regression analysis indicated that this association may be mediated by the MAOA-L allele (adjusted R² = 0.24, ANOVA: F = 23.48, p < 0.001). Conclusively: (1) depression in females may result from a gene × childhood-adversity interaction and/or a dysregulated epigenetic programming of MAOA; (2) childhood-adversity subtypes may differentially impact DNA methylation at NR3C1; (3) baseline MAOA-genotypic variations may affect the extent of NR3C1 methylation.

Increased serotonin transporter gene (SERT) DNA methylation is associated with bullying victimization and blunted cortisol response to stress in childhood: a longitudinal study of discordant monozygotic twins.

BACKGROUND Childhood adverse experiences are known to induce persistent changes in the hypothalamic-pituitary-adrenal (HPA) axis reactivity to stress. However, the mechanisms by which these experiences shape the neuroendocrine response to stress remain unclear. Method We tested whether bullying victimization influenced serotonin transporter gene (SERT) DNA methylation using a discordant monozygotic (MZ) twin design. A subsample of 28 MZ twin pairs discordant for bullying victimization, with data on cortisol and DNA methylation, were identified in the Environmental Risk (E-Risk) Longitudinal Twin Study, a nationally representative 1994-1995 cohort of families with twins. RESULTS Bullied twins had higher SERT DNA methylation at the age of 10 years compared with their non-bullied MZ co-twins. This group difference cannot be attributed to the childrens genetic makeup or their shared familial environments because of the study design. Bullied twins also showed increasing methylation levels between the age of 5 years, prior to bullying victimization, and the age of 10 years whereas no such increase was detected in non-bullied twins across time. Moreover, children with higher SERT methylation levels had blunted cortisol responses to stress. CONCLUSIONS Our study extends findings drawn from animal models, supports the hypothesis that early-life stress modifies DNA methylation at a specific cytosine-phosphate-guanine (CpG) site in the SERT promoter and HPA functioning and suggests that these two systems may be functionally associated.

The PsychENCODE project

Recent research on disparate psychiatric disorders has implicated rare variants in genes involved in global gene regulation and chromatin modification, as well as many common variants located primarily in regulatory regions of the genome. Understanding precisely how these variants contribute to disease will require a deeper appreciation for the mechanisms of gene regulation in the developing and adult human brain. The PsychENCODE project aims to produce a public resource of multidimensional genomic data using tissue- and cell type–specific samples from approximately 1,000 phenotypically well-characterized, high-quality healthy and disease-affected human post-mortem brains, as well as functionally characterize disease-associated regulatory elements and variants in model systems. We are beginning with a focus on autism spectrum disorder, bipolar disorder and schizophrenia, and expect that this knowledge will apply to a wide variety of psychiatric disorders. This paper outlines the motivation and design of PsychENCODE.

Sleep quality and diurnal preference in a sample of young adults: Associations with 5HTTLPR, PER3, and CLOCK 3111†

Research investigating associations between specific genes and individual differences with regards to the quality and timing of sleep has primarily focussed on serotonin‐related and clock genes. However, there are only a few studies of this type and most of those to date have not considered the possibility of gene–environment interaction. Here, we describe associations between sleep quality and diurnal preference and three functional polymorphisms: 5HTTLPR, PERIOD3, and CLOCK 3111. Furthermore, we assessed whether associations between genotypes and sleep phenotypes were moderated by negative life events—a test of gene–environment interaction. DNA from buccal swabs was collected from 947 individuals [mean age = 20.3 years (SD = 1.77), age range = 18–27 years; 61.8% female] and genotyped for the three polymorphisms. Participants completed the Pittsburgh Sleep Quality Index and the Morningness‐Eveningness Questionnaire. There was a significant main effect of 5HTTLPR on sleep quality, indicating that “long–long” homozygotes experienced significantly poorer sleep quality (mean = 6.35, SD = 3.36) than carriers of at least one “short” allele (mean = 5.67, SD = 2.96; β = −0.34, P = 0.005). There were no main effects of 5HTTLPR on diurnal preference; no main effects of PERIOD3 or CLOCK on sleep quality or diurnal preference; and no significant interactions with negative life events. The main effect of the “long” 5HTTLPR allele contradicts previous research, suggesting that perhaps the effects of this gene are heterogeneous in different populations. Failure to replicate previous research in relation to PERIOD3 and CLOCK concurs with previous research suggesting that the effects of these genes are small and may be related to population composition.

Genome-wide Methylomic Analysis of Monozygotic Twins Discordant for Adolescent Depression

Background Adolescent depression is a common neuropsychiatric disorder that often continues into adulthood and is associated with a wide range of poor outcomes including suicide. Although numerous studies have looked at genetic markers associated with depression, the role of epigenetic variation remains relatively unexplored. Methods Monozygotic (MZ) twins were selected from an adolescent twin study designed to investigate the interplay of genetic and environmental factors in the development of emotional and behavioral difficulties. There were 18 pairs of MZ twins identified in which one member scored consistently higher (group mean within the clinically significant range) on self-rated depression than the other. We assessed genome-wide patterns of DNA methylation in twin buccal cell DNA using the Infinium HumanMethylation450 BeadChip from Illumina. Quality control and data preprocessing was undertaken using the wateRmelon package. Differentially methylated probes (DMPs) were identified using an analysis strategy taking into account both the significance and the magnitude of DNA methylation differences. The top differentially methylated DMP was successfully validated by bisulfite-pyrosequencing, and identified DMPs were tested in postmortem brain samples obtained from patients with major depressive disorder (n = 14) and matched control subjects (n = 15). Results Two reproducible depression-associated DMPs were identified, including the top-ranked DMP that was located within STK32C, which encodes a serine/threonine kinase, of unknown function. Conclusions Our data indicate that DNA methylation differences are apparent in MZ twins discordant for adolescent depression and that some of the disease-associated variation observed in buccal cell DNA is mirrored in adult brain tissue obtained from individuals with clinical depression.

Variation in 5-hydroxymethylcytosine across human cortex and cerebellum

Background The most widely utilized approaches for quantifying DNA methylation involve the treatment of genomic DNA with sodium bisulfite; however, this method cannot distinguish between 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC). Previous studies have shown that 5hmC is enriched in the brain, although little is known about its genomic distribution and how it differs between anatomical regions and individuals. In this study, we combine oxidative bisulfite (oxBS) treatment with the Illumina Infinium 450K BeadArray to quantify genome-wide patterns of 5hmC in two distinct anatomical regions of the brain from multiple individuals. Results We identify 37,145 and 65,563 sites passing our threshold for detectable 5hmC in the prefrontal cortex and cerebellum respectively, with 23,445 loci common across both brain regions. Distinct patterns of 5hmC are identified in each brain region, with notable differences in the genomic location of the most hydroxymethylated loci between these brain regions. Tissue-specific patterns of 5hmC are subsequently confirmed in an independent set of prefrontal cortex and cerebellum samples. Conclusions This study represents the first systematic analysis of 5hmC in the human brain, identifying tissue-specific hydroxymethylated positions and genomic regions characterized by inter-individual variation in DNA hydroxymethylation. This study demonstrates the utility of combining oxBS-treatment with the Illumina 450k methylation array to systematically quantify 5hmC across the genome and the potential utility of this approach for epigenomic studies of brain disorders. Electronic supplementary material The online version of this article (doi:10.1186/s13059-016-0871-x) contains supplementary material, which is available to authorized users.BackgroundThe most widely utilized approaches for quantifying DNA methylation involve the treatment of genomic DNA with sodium bisulfite; however, this method cannot distinguish between 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC). Previous studies have shown that 5hmC is enriched in the brain, although little is known about its genomic distribution and how it differs between anatomical regions and individuals. In this study, we combine oxidative bisulfite (oxBS) treatment with the Illumina Infinium 450K BeadArray to quantify genome-wide patterns of 5hmC in two distinct anatomical regions of the brain from multiple individuals.ResultsWe identify 37,145 and 65,563 sites passing our threshold for detectable 5hmC in the prefrontal cortex and cerebellum respectively, with 23,445 loci common across both brain regions. Distinct patterns of 5hmC are identified in each brain region, with notable differences in the genomic location of the most hydroxymethylated loci between these brain regions. Tissue-specific patterns of 5hmC are subsequently confirmed in an independent set of prefrontal cortex and cerebellum samples.ConclusionsThis study represents the first systematic analysis of 5hmC in the human brain, identifying tissue-specific hydroxymethylated positions and genomic regions characterized by inter-individual variation in DNA hydroxymethylation. This study demonstrates the utility of combining oxBS-treatment with the Illumina 450k methylation array to systematically quantify 5hmC across the genome and the potential utility of this approach for epigenomic studies of brain disorders.