Exploring the moderating effects of dopaminergic polymorphisms and childhood adversity on brain morphology in schizophrenia-spectrum disorders

Highlights

• Genetic and environmental factors contribute to the neurobiology of schizophrenia.

• Dopamine is a key neurotransmitter in schizophrenia pathophysiology.

• Childhood adversity and dopaminergic polymorphisms regulate the dopamine system.

• The effect of these factors on dopamine-related brain structures was explored.

• Left putamen volume was moderated by these factors in a dose-dependent fashion.

Abstract

Genetic and environmental etiologies may contribute to schizophrenia and its associated neurobiological profile. We examined the interaction between dopaminergic polymorphisms, childhood adversity and diagnosis (schizophrenia/schizoaffective disorder) on dopamine-related brain structures. Childhood adversity histories and structural MRI data were obtained from 249 (153 schizophrenia/schizoaffective, 96 controls) participants registered in the Australian Schizophrenia Research Bank. Polymorphisms in DRD2 and COMT were genotyped and a dopaminergic risk allelic load (RAL) was calculated. Regression analysis was used to test the main and interaction effects of RAL, childhood adversity and diagnosis on volumes of dopamine-related brain structures (caudate, putamen, nucleus accumbens, dorsolateral prefrontal cortex and hippocampus). A schizophrenia/schizoaffective diagnosis showed significant main effects on bilateral hippocampus, left dorsolateral prefrontal cortex and bilateral putamen volumes. RAL showed a significant main effect on left putamen volumes. Furthermore, across the whole sample, a significant two-way interaction between dopaminergic RAL and childhood adversity was found for left putamen volumes. No brain structure volumes were predicted by a three-way interaction that included diagnosis. Our finding suggests the left putamen may be particularly sensitive to dopaminergic gene-environment interactions regardless of diagnosis. However, larger studies are needed to assess whether these interactions are more or less pronounced in those with schizophrenia/schizoaffective disorders.

Introduction

Schizophrenia is a psychiatric illness that typically emerges during late adolescence, and often results in a lifetime of disability (Lewis and Lieberman, 2000). Although the disorder appears to lack a discretely causal aetiology, dopaminergic abnormalities remain a central hypothesis of the pathophysiology of schizophrenia and are the primary target of current pharmacological treatments (Howes and Kapur, 2009). The dopamine hypothesis of schizophrenia suggests that characteristic symptoms such as hallucinations, delusions and abnormal cognitive functioning are caused by a synergistic imbalance of dopamine neurotransmission in cortical and subcortical brain regions (Howes and Kapur, 2009). Thus, genetic variants and brain regions implicated in the function of the dopaminergic system may contribute to the disorder's aetiology.

A number of variants within genes in the dopaminergic system have been studied, and among them, the dopamine receptor D2 (DRD2) and catechol-o-methytransferase (COMT) genes have arguably been interrogated the most. DRD2 was recently identified as one of 108 loci associated with schizophrenia in the largest schizophrenia genome-wide association study to date (Psychiatric Genomics Consortium, 2014). One highly researched DRD2 variant that occurs at rs1076560 (G > T) determines whether mRNA splices into long or short isoforms (Zheng et al., 2012). Past literature has identified this substitution as a risk-conferring variant for schizophrenia, most likely due to the resultant decrease of dopamine transmission in the frontal cortex (Tallerico et al., 2001, Zheng et al., 2012). Another relevant DRD2 mutation is located at rs12364283 (T > C). This polymorphism results in enhanced total D2 mRNA expression, which may exacerbate already elevated striatal dopamine transmission in patients with schizophrenia (Bertolino et al., 2009a). A third DRD2 variant relevant to schizophrenia occurs at rs1801028 (C > G), where cysteine production replaces serine production. A meta-analysis has suggested that there is a link between this polymorphism and an increased susceptibility to schizophrenia through alteration of D2 receptor physiology and functioning (Glatt et al., 2003). Some of these DRD2 variants have also been shown to effect schizophrenia-associated intermediate phenotypes such as disrupted prefrontal-striatal activity (Bertolino et al., 2008) and morphological changes including smaller caudate volumes (Bertolino et al., 2009b). Lastly, a mutation in the COMT gene at rs4680 (A > G) enhances coding of valine instead of methionine, resulting in a higher enzymatic catabolism of dopamine in the prefrontal cortex (Egan et al., 2001). Whilst some studies have identified COMT as a possible candidate gene for schizophrenia (Kunugi et al., 1997, Li et al., 2000, Wonodi et al., 2003), others have revealed no association between the val/met polymorphism and the disorder (de Chaldée et al., 2001, Okochi et al., 2009). As opposed to the manifested clinical outcome, the COMT polymorphism is much more strongly associated with schizophrenia-related intermediate phenotypes such as brain morphology; for example, volumetric changes in the hippocampus and dorsolateral prefrontal cortex (DLPFC) (Cerasa et al., 2008, Honea et al., 2009, Kates et al., 2006). It has been widely accepted that these observed gene-brain associations are likely due to the interactive and cumulative effect of molecular mechanisms downstream from genotype (Harrison and Weinberger, 2005), which are heavily affected by environmental influences (van Os et al., 2008a).

One potent and established environmental risk factor for schizophrenia and other psychiatric disorders is childhood adversity (Matheson et al., 2013). Childhood adversity has been defined as any form of emotional or physical ill-treatment, sexual abuse, exploitation or neglect during childhood or teen years (Rosenman and Rodgers, 2004). Perhaps the most frequently reported types of early trauma associated with psychosis are sexual and physical abuse, which are often examined together (Davies-Netzley et al., 1996, Read et al., 2003, Read and Argyle, 1999). Since attributing salience to threatening or adverse environmental stimuli (Kapur, 2003) and stress-mediated responses (Laruelle, 2000) both implicate the dopamine system, childhood adversity may contribute to pathological dopamine neurotransmission and, in turn, alter neurobiology (Read et al., 2005, Van Winkel et al., 2008, Walker et al., 2008). In those who have experienced early maltreatment, atrophy has been especially noted in the hippocampus (Bremner et al., 1997, Hoy et al., 2012, Rao et al., 2010, Woon and Hedges, 2008) and may occur in other dopamine-related regions such as the prefrontal cortex and striatal structures (Cohen et al., 2006, Frodl et al., 2010, Tomoda et al., 2009). It has been proposed that early physical and emotional adversities each have specific neurobiological targets, as different neural mechanisms and pathways are employed to cope with different types of traumatic experience (Edmiston et al., 2011, Teicher and Samson, 2016). As such, exposure to physical abuse, emotional abuse or emotional neglect may have regionally-specific effects on the brain.

Notably, not all individuals who carry ‘risk’ polymorphisms or experience childhood adversity develop schizophrenia, suggesting genetic and environmental factors likely interact rather than act alone. As such, the current study used a gene  ×  environment (G  ×  E) framework to investigate the interaction between genetic variation in dopaminergic genes (DRD2 and COMT) and childhood adversity in determining schizophrenia-associated brain morphology. Although there is a wealth of G  ×  E literature examining schizophrenia (Tienari et al., 2004, van Os et al., 2008a, Wahlberg et al., 1997), few studies have utilized this framework to investigate intermediate phenotypes associated with the disorder. The present study focused on brain structures which have shown to be both atrophied in schizophrenia patients and influenced by genetic and environmental factors, thus fulfilling the two defining targets of an intermediate phenotype (Meyer-Lindenberg and Weinberger, 2006). These structures were the hippocampus, DLPFC, nucleus accumbens, caudate nucleus and putamen. It was hypothesised that individuals with a high proportion of dopaminergic ‘risk’ polymorphisms, elevated levels of childhood adversity and a diagnosis of schizophrenia-spectrum disorders would have the lowest volumes in these brain regions.