GWAS Analysis of Genetic Factors in Obsessive-Compulsive Disorder

Introduction

Anxiety is unlike fear in that it is a long-lasting response to danger signals that proceeds beyond actual risk or out of proportion to the possibility of a threat. It is an intense reaction that affects individuals tremendously, severely impacting their day-to-day lives. An example of an anxiety disorder subtype as defined by the DSM-IV is Obsessive-Compulsive Disorder (OCD). In the continued effort to understand and treat OCD, emerging research suggests a genetic and hereditary component to the disorder.

OCD is a disorder with symptoms including a complex combination of intrusive obsessions and highly ritualized compulsions. It is a relatively frequently diagnosed neuropsychiatric disorder that encompasses both compulsions and obsessions, leading to a wide range of behaviors and unique human thoughts that affect individuals (Geller et al., 2003). This report examines the genetic underpinnings of OCD using Genome-Wide Association Study (GWAS) methodology, with the aim of contributing to a more comprehensive, multi-faceted explanation for the disorder's expression.

Background on OCD Genetics and Heredity

According to Mattheisen et al. (2014), OCD is "characterized by intrusive thoughts and urges and repetitive, intentional behaviors that cause significant distress and impair functioning." These intrusive thoughts often cause individuals to obsess over otherwise seemingly trivial issues and feel the need to ritualize their lives in order to maintain a certain balance (Knopp et al., 2013). Those who suffer from OCD often express repetitive, compulsive behaviors thought to stem from an internal feeling of loss of control (Burguière et al., 2015). Many of these disruptive thoughts or ritualistic needs have been attributed to neurological abnormalities, such as the "disrupted neurotransmission of glutamate within corticostriatal-thalamocortical circuitry" (Davis et al., 2013). Several other studies have also identified abnormal neurological functioning as a primary cause of OCD expression (Ahmari et al., 2013).

Murphy et al. (2013) show that those suffering from OCD tend to exhibit symptoms of other psychological disorders as well. Over 70% of individuals with OCD have major depressive disorder, with another 10% also having bipolar disorder (Murphy et al., 2013). OCD expression is often seen alongside disorders such as Tourette Syndrome, which is also thought to have a genetic component that allows families to pass on the disorder to children (Sachdev et al., 2012). According to Murphy et al. (2013), "OCD resembles disorders such as depression, in which gene × gene interactions, gene × environment interactions, and stress elements coalesce to yield OC symptoms and, in some individuals, full-blown OCD with multiple comorbid disorders." Thus, there is a connection between other psychological stressors that have also been linked with genetic inheritance.

Research on OCD has uncovered hereditary connections regarding individuals exhibiting symptoms associated with the disorder. Studies have identified particular gene markers thought to contribute to the pathobiology of OCD (Browne et al., 2014). According to Stewart et al. (2013), in their report "Genome-Wide Association Study of Obsessive-Compulsive Disorder," OCD is "a common, debilitating neuropsychiatric illness with complex genetic etiology." A number of family-design and twin studies suggest a hereditary component associated with the etiology of OCD (Arnold et al., 2006). Specifically, genetic variance accounts for 40% of compulsive and obsessive behaviors in the sampled twin studies (Stein, Andersen, & Overo, 2007; Zohar, Greenberg, & Denys, 2012). Familial association studies have suggested that first-degree relatives of individuals with OCD are two times more likely to develop the disorder as adults compared to the rest of the population, as determined by comparison to first-degree relatives of unaffected control subjects (Pauls et al., 2014).

Additionally, Mattheisen et al. (2014) conducted a similar study on 1,065 families, totaling a sample of over 5,000 individuals. Through SNP analysis testing, the study found markers along chromosome 9, where "pre-synaptic PTPRD promotes the differentiation of glutamatergic synapses and interacts with SLITRK" (Mattheisen et al., 2014). Although no significant SNPs were exactly correlated with the presence of OCD directly, the study did show significant enrichment in certain SNP characteristics at particular chromosome markers. Thus, prior research has established clear connections between genetic heredity and the expression of OCD symptoms and behaviors within genetically linked familial groups.

A number of studies have found genetic architectural variations in key parts of the brain associated with the presence of OCD. Murphy et al. (2013) concluded that common genetic variants included issues with the serotonin transporter gene (SLC6A4), "the brain-derived neurotrophic factor (BDNF) gene, and rare variants in genes/chromosomal abnormalities." Another study conducted by Davis et al. (2013) found clear and distinct genetic architectures in the expression of OCD, mainly abnormal gene expression in the parietal cortex and cerebellum. Examining hereditary gene characteristics reveals a genetic component to OCD, but also confirms that genetics is not the only factor contributing to its condition (Moran, 2013).

In addition to genetic factors, clearly outlined environmental factors greatly increase an individual's likelihood of expressing OCD symptoms (Gorenstein et al., 2015). Heredity has been identified as one of the major causes of OCD transmitted within families; yet, genetic heredity alone does not explain the entire picture. Additional research has also linked OCD transmission within families to shared environmental factors (Ruscio et al., 2010). Beyond possible genetic heritage links, OCD behavior can be observed and learned through exposure to compulsive-obsessive symptoms within family members (Pauls et al., 2014). Given the high number of individuals with OCD who also have depressive disorders, one may infer a genetic link attached to both disorders that can be carried through family heritage and augmented by external environmental factors. Moreover, non-shared environmental factors contribute to 51% of OCD variance (Stein, Andersen, & Overo, 2007; Mattheisen et al., 2014).

However, prior research has failed to show a significant connection between genetic and environmental factors that can be reliably identified as the best predictors of risk for OCD. Stewart et al. (2013) conducted a study which found no SNPs directly significant in the presence of OCD, though there were additional findings of "enrichment of methylation QTLs and frontal lobe expression quantitative trait loci" within the highest-ranking SNPs in the control analysis. Prior research has uncovered some signs of a correlation, but more research is needed to clarify the level of significance between the hereditary nature of SNP characteristics and the later expression of OCD behaviors within familial groups.

Methodology

This current research is an attempt to develop better models that blend observations of genetic and environmental stimuli in order to provide a more multi-faceted explanation for the presence of OCD. The study focuses on the use of Genome-Wide Association Studies (GWAS) to identify potential genetic regions of interest, which can better explore the potential genetic source of OCD behaviors (Stewart and Pauls, 2010). This can then serve as a basis for studies attempting to explore both a genetic link and a combined connection between genetic and environmental factors. The objective of this lab report was to use statistical and computational tools to analyze GWAS data and deliver an effective method for understanding novel strategies to address OCD.

The study selected all subjects from European ancestry to address the problem of overrepresentation of variance that occurs across ethnic groups. If the study were to mix the racial backgrounds of cases and controls, allele frequencies might have been misrepresented as differences associated with disease risk. Accordingly, individuals of non-European ancestry were removed from the sample. Subjects with genetic gender discrepancies were also removed to avoid sample mix-up and to prevent mismatches based on differences in genetic components between individuals whose genetic and recorded sex did not correspond.

The study collected data from the Queensland Institute of Medical Research (QIMR) for the analysis of genetic factors. The data comprised an analysis of approximately 2,372,500 SNPs (single nucleotide polymorphisms) — genetic DNA variations found within a given population — which were isolated in order to compare with rates of OCD in that same population. The extracted data consisted of SNP genotypes, phenotypes, and MAF (minor allele frequencies).

Standard quality control (QC) statistics were performed on the QIMR genotype data to minimize the potential for false positive associations commonly found within poor-quality data. This enhances the validity of conclusions about the proposed relationships. The final OCD dataset was used to match the case sample size with the control sample size (n = 1,500). The control sample matched males to female subjects evenly (n = 750 each). Excessive missing values (>0.02) were dropped to remove potential errors, which consequently set cases and controls at an equal 50:50 ratio to reduce signal noise. Levels of missingness were assigned based on the details of the GWAS analysis. SNPs were removed for having a low minor allele frequency (MAF) of <0.01, as dictated by the GWAS analysis. SNPs were also dropped if they had excessive missing values greater than 0.01, or if both cases and controls showed MAFs below the sample group threshold.

Statistical analysis was carried out to test allele frequency differences between the allocated cases and controls using PLINK software. The GCTA software was used to simulate allele frequencies in cases and controls consistent with plausible SNP associations for the disorder. Data were also analyzed using the Chi-Square test and calculation of the p-value in order to identify positive correlations between allele frequencies and the presence of the disorder. The outcomes of these tests were then illustrated visually to aid in interpretation of the results.