Serotonin transporter and BDNF genetic variants interact to predict cognitive reactivity in healthy adults

Abstract

Background

Cognitive theory and empirical evidence both suggest that cognitive reactivity (the tendency to think more negatively when in a sad mood) is an important marker of depression vulnerability. Research has not yet determined whether genetic factors contribute to the expression of cognitive reactivity.

Methods

The present study examined associations between the 5-HTTLPR polymorphism of the SLC6A4 gene, the Val66Met polymorphism of the brain-derived neurotrophic factor (BDNF) gene, and cognitive reactivity in a never depressed, unmedicated, young adult sample (N = 151).

Results

The interaction between 5-HTTLPR and Val66Met polymorphisms significantly predicted change in dysfunctional thinking from before to after a standardized sad mood provocation. Cognitive reactivity increased among S/L_G 5-HTTLPR homozygotes if they were also homozygous for the Val Val66Met allele. In contrast, presence of a Met Val66Met allele was associated with attenuated cognitive reactivity among S/L_G 5-HTTLPR homozygotes.

Limitations

The sample size of the current study is relatively small for modern genetic association studies. However, results are consistent with previous research demonstrating biological epistasis between SLC6A4 and BDNF for predicting connectivity among neural structures involved in emotion regulation.

Conclusions

The BDNF Met allele may protect S/L_G 5-HTTLPR homozygotes from increased dysfunctional thinking following a sad mood provocation. Study results are the first to demonstrate an epistatic genetic relationship predicting cognitive reactivity and suggest the need for more complex and integrative models of depression vulnerability.

Introduction

Cognitive theories of depression vulnerability (Beck, 1967) suggest that underlying negative beliefs and attitudes characterize individuals vulnerable to depression. Furthermore, these negative beliefs and attitudes may not be apparent when vulnerable individuals are in a euthymic state, but will be best observed in the context of sad mood states (Segal and Ingram, 1994, Teasdale, 1988). The increase in dysfunctional thinking from euthymic to dysphoric mood is termed cognitive reactivity.

Cognitive reactivity, typically measured as self-reported dysfunctional attitudes, has been associated with depression vulnerability. For example, Miranda and Persons (Miranda and Persons, 1988) found that women who had previously experienced an episode of depression were more likely to endorse dysfunctional attitudes when in a sad mood state than never depressed women in a similarly sad mood. Several other studies, many using standardized methods to provoke a sad mood, have found similar changes in dysfunctional thinking among depression vulnerable populations (Lewinsohn et al., 1999, Miranda et al., 1990, Van der Does, 2002a).

More recently, longitudinal studies have documented that cognitive reactivity prospectively predicts depressive relapse in previously depressed patients (Segal et al., 2006, Segal et al., 1999). In these two important studies, remitted depressed patients were administered the dysfunctional attitude scale (Weissman and Beck, 1978) before and after a negative mood provocation consisting of sad music combined with autobiographical recall of a sad event. In both studies, Segal and colleagues found that increased dysfunctional attitudes following the sad mood provocation predicted depressive relapse during the follow-up period, even when controlling for number of previous depressive episodes (Segal et al., 2006). Further, preventative treatments designed to reduce cognitive reactivity have been shown to decrease depressive relapse, particularly among individuals with three or more episodes of depression (Coelho et al., 2007, Teasdale et al., 2000). In summary, there is considerable evidence that cognitive reactivity is an important marker of vulnerability to depression (Lau et al., 2004, Scher et al., 2005).

Despite growing evidence supporting cognitive reactivity as a risk factor for depression, few studies have examined mechanisms that increase cognitive reactivity. Recently there has been an increase in interest and research examining the link between genetic risk for depression and cognitive vulnerability factors such as cognitive reactivity (Beevers and Wells, 2009). Variants of the serotonin transporter-linked polymorphic region (5-HTTLPR) of the SLC6A4 gene and the Val66Met polymorphism of the brain-derived neurotrophic factor (BDNF) gene are two well-studied genes implicated in the pathophysiology of depression (Wray et al., 2009, Schumacher et al., 2005). In addition, variation in these genes has been associated with neural function in brain structures involved in the processing of emotional information (Schofield et al., 2009, Paaver et al., 2007, Pezawas et al., 2005). As such, these are intriguing candidate genes for association with cognitive vulnerability to depression.

The 5-HTTLPR is characterized by two common variants: a long (L) allele and a short (S) allele (Heils et al., 1996). The S allele is associated with decreased serotonin (5-HT) transporter expression and serotonin uptake compared to the L allele (Lesch et al., 1996). A recently discovered single nucleotide polymorphism (SNP) rs25531 results in two L variants (L_A or L_G) with the L_G allele acting as a functional equivalent to the S allele (Hu et al., 2005). This results in a triallelic classification of the 5-HTTLPR (L_A, L_G, and S). Caspi and colleagues (Caspi, 2003) found that the S allele of the 5-HTTLPR was associated with increased risk for depression in the context of life stress. Although a number of studies replicated or partially replicated this effect across a variety of paradigms (Uher and McGuffin, 2008), recent meta-analyses have called this gene by environment effect into question (Munafo et al., 2009, Risch et al., 2009). Given the inconsistent findings, more research is needed to better understand the relationship between the 5-HTTLPR and depression (Rutter et al., 2009).

One helpful strategy to better understand the relationship between candidate genes and psychiatric outcome is to identify intermediate phenotypes (Meyer-Lindenberg and Weinberger, 2006). A potential intermediate phenotype for depression is cognitive vulnerability to depression and there is an increasing amount of evidence that 5-HTTLPR variation is associated with cognitive factors associated with depression. For example, individuals with S/L_G alleles of the 5-HTTLPR show increased attention to anxious word stimuli (Beevers et al., 2007a, Beevers et al., 2007b) and increased difficulty disengaging attention from happy, sad, and fearful facial stimuli (Beevers et al., 2009b). In addition, S allele carriers may lack an attentional bias for positive information that is found in individuals homozygous for the L allele (Fox et al., 2009). Children with S/L_G 5-HTTLPR alleles also show greater depressogenic attributional styles than children with L_A 5-HTTLPR alleles (Sheikh et al., 2008).

Several small studies have also found an interaction between genetic status, mood, and cognitive factors related to depression. For example, after a sad mood provocation, 7-year-old children homozygous for the S allele were found to endorse and recall more negative words in a self-referent encoding task compared to L allele carriers (Hayden et al., 2008). Similarly, young adults homozygous for the S 5-HTTLPR allele endorsed more negative thinking after viewing a sad film clip than L allele carriers but there were no differences between allele groups after viewing a neutral film clip (Beevers et al., 2009a). In short, the presence of S/L_G 5-HTTLPR alleles appears to be associated with increased negative thinking in response to a sad mood. However, no studies to date have examined the relationship between the 5-HTTLPR and change in dysfunctional attitudes elicited by a sad mood provocation.

The effects of BDNF Val66Met variation on depression may be complex and the overall contribution of Val66Met variation on the development of major depression may be small (Chen et al., 2008, Martinowich et al., 2007). However, Val66Met Val/Val genotype status has been associated with increased levels of cognitive and affective symptoms of depression in healthy participants (Duncan et al., 2009). In addition, Val66Met variation has been associated with rumination, a risk factor for depression (Spasojevic and Alloy, 2001). Among adolescent girls at risk for depression, Val/Val genotype status was associated with increased depressive symptoms and increased rumination (Hilt et al., 2007). In contrast, among mothers with a history of adult-onset depression, Val/Met genotype status was associated with increased depressive symptoms and increased rumination (Hilt et al., 2007). In healthy young adults, Val/Met genotype is associated with increased rumination compared to individuals with Val/Val genotype (Beevers et al., 2009c). Clearly, more research is needed to determine the impact of Val66Met allele status on factors related to depression.

In addition to the individual effects of 5-HTTLPR and Val66Met variability on risk factors associated with depression, there is evidence of interaction between serotonin and BDNF systems at a neural and genetic level. For example, BDNF regulates the development and function of serotonergic neurons while increases in extracellular serotonin increase BDNF expression (Martinowich and Lu, 2008). Furthermore, there is evidence from both mutant mouse studies and human studies of a genetic epistasis between 5-HTTLPR and BDNF (Martinowich and Lu, 2008). In a recent study, Pezawas and colleagues demonstrated epistatic effects on neural structure. In humans, the 5-HTTLPR S allele is associated with decreased amygdala and anterior cingulate volume compared to L allele homozygotes (Pezawas et al., 2005). However, the presence of the Val66Met Met allele protects against the structural volume reduction associated with the 5-HTTLPR S allele (Pezawas et al., 2008). The expected reduction in amygdala and anterior cingulate volume is observed in individuals with Val66Met Val/Val genotype with the 5-HTTLPR S allele, but is absent in individuals with Val/Met genotype and the 5-HTTLPR S allele.

Given previous findings of increased negative thinking and negative self-referent processing after a sad mood provocation among 5-HTTLPR S allele carriers (Hayden et al., 2008, Beevers et al., 2009a) we predicted that 5-HTTLPR S/L_G allele carriers, compared to L_A allele homozygotes, would show greater change in dysfunctional attitudes (i.e., cognitive reactivity) following a sad mood provocation. In line with Pezawas et al. (2008), we also hypothesized that the presence of the Val66Met Met allele would protect against increased cognitive reactivity among S/L_G 5-HTTLPR allele carriers.