White matter integrity in major depressive disorder: Implications of childhood trauma, 5-HTTLPR and BDNF polymorphisms

Highlights

• Examined impact of neglect, 5-HTTLPR and BDNF polymorphisms on white matter in MDD.

• Trauma history, DTI, genotyping collected on 55 unmedicated MDD patients; 18 controls.

• MDD subgroups: 5-HTTLPR heterozygotes and BDNF-Met had reduced white matter integrity.

• In highly depressed neglect was associated with FA in the Uncinate and SLF.

• Neglect and genetic risk factors and depression affect frontal/limbic connectivity.

Abstract

This study examined the impact of childhood neglect, serotonin transporter (5-HTTLPR) and brain derived neurotrophic factor (BDNF) polymorphisms on white matter (WM) integrity in major depressive disorder (MDD) using diffusion tensor imaging (DTI). Fifty-five medication-free MDD patients and 18 controls underwent diffusion tensor imaging scanning, genotyping and completed the Childhood Trauma Questionnaire. Tract based spatial statistics (TBSS) findings revealed reduced fractional anisotropy (FA) in the MDD group in the anterior internal capsule. 5-HTTLPR–S′L′ heterozygotes in the MDD group exhibited reduced FA in the internal capsule relative to S′S′ and reduced FA in corona radiata compared to L′L′. Probabilistic tractography revealed higher FA in the uncinate fasciculus (UF) for BDNF val/val genotype relative to met-carriers, particularly in individuals with high depression severity. High depression severity and experiences of childhood physical or emotional neglect predicted higher FA in the UF and superior longitudinal fasciculus. Reductions in FA were identified for subgroups of MDD patients who were 5-HTTLPR heterozygotes and BDNF-met carriers. An association between emotional/physical neglect and FA was observed in subjects with high depressive symptoms. Our findings suggest that WM connectivity within frontal and limbic regions are affected by depression and influenced by experiences of neglect and genetic risk factors.

Introduction

Growing evidence suggests that major depressive disorder (MDD) is a complex brain disorder that results from structural and functional abnormalities of multiple limbic and cortical brain regions involved in mood regulation. In order to understand the structural connectivity of brain regions implicated in MDD, much research has focused on the abnormalities of white matter integrity and fiber tracts between brain regions. Diffusion tensor imaging (DTI) provides a non-invasive MR-method of assessing myelin integrity (and thus neuronal connectivity between associated brain regions). Fractional anisotropy (FA) is a DTI parameter that reflects the ratio of directional to non-directional water movement in a single imaging voxel and provides information on relative axonal size, myelination, axon connections and orientation (Sexton et al., 2009, Smith et al., 2006). In regions with reduced white matter structure integrity and axon connectivity, FA values are low. The potential use of diffusion tensor imaging as a tool to guide treatment has promoted investigations to determine whether FA can predict depression severity, illness course/duration and treatment outcomes.

A number of DTI studies on MDD have reported white matter abnormalities in tracts and regions including the superior longitudinal fasciculus (SLF) (Dalby et al., 2010, Wu et al., 2011, Zuo et al., 2012; see Murphy and Frodl (2011) for meta-analysis), the fornix (Zou et al., 2008), internal capsule (Zuo et al., 2012), external capsule (Guo et al., 2012), corpus callosum (Guo et al., 2012, Kieseppä et al., 2010, Korgaonkar et al., 2011), sagittal striatum (Kieseppä et al., 2010), hippocampus (Zhou et al., 2011, Zhu et al., 2011), cingulum (Keedwell et al., 2012), uncinate fasciculus (Cullen et al., 2010, Dalby et al., 2010) and projection fibers associated with the thalamus (Korgaonkar et al., 2012). However, these findings have been variable in terms of location and FA changes. Among several genetic and environmental risk factors that impact on depression onset, course and outcome, particular attention has been payed to early life stress and serotonin transporter (5-HTTLPR) and brain-derived neurotrophic factor (BDNF) genes. Experiences of early life stress are predictive of depression with earlier age of onset, greater chronicity (Heim and Nemeroff, 2001), depression later in life (Karg, 2011) and are associated with a higher number of psychiatric co-morbidities (Bernet and Stein, 1999). Only a few studies have examined connectivity and integrity of white matter tracts in subjects who have experienced childhood adversity. Alterations have been detected in the uncinate fasciculus (connecting the orbitofrontal cortex to the amygdala) (Eluvathingal, 2006), arcuate fasciculus, fornix and cingulum (Choi et al., 2009) and consistent volume reductions have been reported in the corpus callosum ( De Bellis et al., 2002, De Bellis and Thomas, 2003, Teicher et al., 2004; see Brem and Walitza (2002) for review). However, it remains unclear whether depression associated with early life stress could be a subtype of depression and therefore associated with different brain mechanisms.

Among several candidate genes, 5-HTTLPR and BDNF have been widely investigated in MDD, as these two genes confer the largest risk for depression and moderate the effect of childhood adversity on depression development (Bukh et al., 2009). The human 5-HT transporter (5-HTT) is encoded by the SLC6A4 gene on chromosome 17p12, and constitutes a transmembrane protein that is important in presynaptic re-uptake of serotonin following neurotransmitter release (Lesch et al., 1995). The short allele has significantly reduced transcription efficiency relative to the long allele (Lesch et al., 1996). Furthermore, research has determined that the long allele consists of two distinct variants, L_A and L_G, with the L_G variants having a transcription efficiency that is significantly less than the L_A variant, and similar in efficiency to the short S_A allele (Hu et al., 2005). The short allele of the 5-HTTLPR is strongly associated with depression and anxiety symptomology (Collier et al., 1996, Bellivier et al., 1998, Furlong et al., 1998, Lesch et al., 1995, Lesch et al., 1996), stress related MDD (Caspi et al., 2003) and poor response to selective serotonin reuptake inhibitors (SSRIs) (Serretti et al., 2007).

Neuro-structural and functional effects have been reported in depressed individuals with the S allele. Similar to depressed populations smaller grey matter volumes in the perigenual cingulate and amygdala, altered frontal-limbic white matter connectivity and functional uncoupling of the frontal-limbic circuit have been found in healthy S allele carriers (Pacheco et al., 2009, Pezawas et al., 2005). An uncoupling of this circuit may reduce top-down regulation, which further substantiates findings of increased amygdala activity to aversive stimuli in S allele carriers (Hariri et al., 2005, Heinz et al., 2004). However, the role of 5-HTTLPR polymorphisms on WM tracts of other corticolimbic regions in MDD remains largely unknown.

Chen et al. (2006), have tied the more rare met-BDNF variant of the val66met polymorphism to diminished BDNF secretion, smaller hippocampal volume, less dendritic arborization and 30% reductions in activity dependent secretion of BDNF (Chen et al., 2006).

Previous research has associated gene and environmental factors with depression onset and severity (Caspi et al., 2003). Furthermore, depression development has been consistently associated with changes in neural connectivity (see review Murphy and Frodl (2011)). Moreover, changes in neural connectivity have predicted performance on executive function tasks in individuals with depression (Murphy et al., 2007). Understanding how measures of environmental adversity and genetic risk factors influence and moderate the effect of depression on neural connectivity will allow clinicians to understand and predict potential neurocognitive and emotion regulation outcomes in individuals with multiple depression risk factors.

The objective of the present study was to draw together a number of the above findings, by assessing the influence of childhood neglect, 5-HTTLPR and BDNF polymorphisms on myelin integrity in the brain in MDD. Myelin integrity was assessed by two analysis methods: tract based spatial statistics (TBSS) and probabilistic tractography. TBSS allows whole-brain group-level analysis of functional connectivity and suffers from less user biases (Smith et al., 2006). In contrast, probablistic tractography calculates the probability that two voxels are connected and is useful to examine specific white matter tracts by using seed or target regions (Mori and Van Zijl, 2002). We examined myelin integrity within white matter fiber tracts within the fronto-limbic circuit including the uncinate (UF) and superior longitudinal (SLF) fasciculus, as these tracts have been implicated in depression (Murphy and Frodl, 2011). The UF connects the amygdala to the medial and orbito frontal cortex and the SLF connects the lateral prefrontal cortex to the temporal and occipital lobe. These tracts are predominantly involved in bottom up and top down emotional processing (Pacheco et al., 2009; Phillips et al., 2003). We anticipated that childhood neglect and genetic risk factors would influence abnormalities of WM integrity and pathways, specifically in the UF and SLF tracts, as they are implicated in both normal emotional regulation and MDD. We predicted voxel-wise differences in the corpus callosum, anterior internal capsule, hippocampus and corona radiata. Additionally, we predicted that experiences of childhood trauma and genetic risk factors (low transcriptionally efficient alleles of 5-HTTLPR and the met allele of the BDNF gene) would influence myelin integrity, and moderate the influence of depression severity on myelin integrity in the UF and SLF.