One of the most widely studied polymorphisms in psychiatric behavioral genetics is the serotonin transporter-linked polymorphic region (5-HTTLPR), of which the ‘short’ (S) allele is 43 base pairs shorter than the ‘long’ (L) allele (Wendland et al.
). Given its location in the promoter region upstream from the serotonin transporter genetic locus, this polymorphism affects the efficiency of DNA transcription into messenger RNA; the S allele is associated with decreased transcription efficiency, which leads to less production of the serotonin transporter protein and subsequently to less reuptake of serotonin (Lesch et al.
). The S allele has been associated with a number of psychiatric conditions, including anxiety and neuroticism (Lesch et al.
), psychosis (Shcherbatykh et al.
) and perhaps most notably, depression (Lotrich and Pollock, 2004
Although a large volume of research has linked the 5-HTTLPR to depression, scientists are just beginning to discover possible mechanisms by which the 5-HTTLPR confers risk for depression. Several recent studies have revealed particular environmental inputs that interact with the S allele to lead to depression. The first of these studies showed that stressful life events (e.g. illness, financial problems) were more depressogenic for carriers of the S allele than they were for individuals homozygous for the L allele (Caspi et al.
). Similar interactions have been reported among other adult samples (Kendler et al.
) and among children (Kaufman et al.
Recent studies in this area have begun to identify the more proximal effects of this genetic variability on human behavior, physiology and neural systems. Gotlib and colleagues, for example, have identified effects of the 5-HTTLPR on cortisol levels, both upon awakening (Chen et al.
) and in response to a stressor (Gotlib et al.
). Recent neuroimaging studies have begun to delineate the neurobiological pathways whereby the 5-HTTLPR increases risk for depression. A handful of studies have shown that certain brain areas—especially the amygdala—are selectively affected by 5-HTTLPR variability. The first such study (Hariri et al.
) found that carriers of the S allele showed greater right amygdala reactivity in response to fearful or angry faces [stimuli known to reliably produce amygdala activity; Morris et al.
)]. Independent research labs have replicated the greater amygdala reactivity among carriers of the S allele (e.g. Canli et al.
; Furmark et al.
). Other groups have reported greater resting
amygdala activity associated with the S allele (Canli et al.
; Rao et al.
). A meta-analysis of these reports concluded that the average effect size of 5-HTTLPR genotype on amygdala activation is d
= 0.54 (Munafò et al.
), which by convention falls in the ‘medium’ range (Cohen, 1992
). The effect of 5-HTTLPR genotype on amygdala activity is intriguing given the evidence that amygdala activity is elevated during depression (Drevets et al.
) and experimentally induced negative affective states (Posse et al.
), and that the amygdala had been hypothesized to be part of a limbic network that is dysregulated during depression (Seminowicz et al.
A recent review (Hariri and Holmes, 2006
) summarized the importance of these findings and presented an integrated model of the effects of 5-HTTLPR variability on neural emotion regulation networks. The authors presented evidence that the inhibitory feedback circuits in prefrontal cortex are less effective in carriers of the S allele, resulting in dysregulated limbic emotion centers [see also Heinz et al.
(2005) and Pezawas et al.
)]. The result is a cascade of behavioral and neuroendocrine effects that may lead to clinical syndromes including mood and anxiety disorders.
One question that is not answered by the existing work in this area is whether these findings for greater amygdala reactivity among carriers of the S allele extend to depression-related stimuli. As Hariri and Holmes (2006
; see their ) demonstrated, most of the studies that have addressed emotion-related effects of the 5-HTTLPR on amygdala activity use stimuli that may or may not be associated with depressed mood. Examples of stimuli used in these studies included emotional faces (Hariri et al.
), negative words (Canli et al.
), and a public speaking paradigm (Furmark et al.
). These stimuli do not bear a substantial resemblance to the events that have been reported to interact with 5-HTTLPR genotype to predict depression onset, such as financial hardship (Kendler et al.
). The current study set out to explore the response of the brain under conditions that likely are more similar to the events commonly associated with depression risk.
Neural activations: sadness—baseline contrast
One of the most common factors that precedes a depression onset is the experience of loss (Kendler et al.
), including loss of relationships (Monroe et al.
), employment and financial resources (Kendler et al.
) and the death of a loved one (Kendler et al.
); therefore, manipulations that induce the thoughts and feelings associated with loss may be well-suited to reveal depression diatheses. Moreover, because low mood is a hallmark of depression, evoking sad mood states is likely to get closer to the mechanisms by which the 5-HTTLPR affects depression risk compared to, for example, briefly viewing fearful faces. The current study tested whether the 5-HTTLPR influences neural activity during a sad mood induced by the imagined loss of a loved one.
The ability to regulate
a sad mood also has been suggested to play a key role in depression risk (Gilboa and Gotlib, 1997
); without the ability to recover from low moods, episodes of normal sadness potentially could lead to a bout of depression (see Segal et al.
). Therefore the current study also examined whether 5-HTTLPR genotype modulates neural activity during sad mood regulation. The conscious regulation of negative emotion is significantly associated with amygdala activity (Schaefer et al.
), and decreases in amygdala activity predict decreases in negative affective responses (Phelps et al.
Given the time course of the events of interest—moods lasting several minutes rather than more ephemeral emotional responses measured on the order of seconds—the current study used perfusion functional magnetic resonance imaging (fMRI), which is better suited to capture longer-term brain changes than is blood oxygen level-dependent (BOLD) fMRI (Wang et al.
). Our primary region of interest (ROI) was the amygdala, based on existing studies reviewed above. The subgenual anterior cingulate cortex (subACC) was included as a secondary ROI, for several reasons. First, activity in subACC varies as a function of brief, non-pathological mood states (Damasio et al.
) as well as with frank depression (Seminowicz et al.
). Furthermore, there is evidence suggesting that electrical stimulation of this region may relieve treatment-refractory depression (Mayberg et al.
), and that 5-HTTLPR genotype predicts the degree to which subACC and amygdala activity are ‘functionally coupled’ (Pezawas et al.
). Hariri and Holmes (2006
) incorporate subACC into a neural emotion regulation circuit that may be dysregulated among carriers of the S allele. Importantly, the cingulate cortex has the densest concentration of serotonin transporter sites in the human cortex (Gurevich and Joyce, 1996
). Although fewer studies have tested for main effects of 5-HTTLPR genotype on subACC activity, existing data suggest that carriers of the S allele will show higher subACC activation during sad mood and during sad mood regulation, including at least one study that found higher ACC activation for S allele carriers during masked viewing of emotional faces (Dannlowski et al.
). For these reasons we tested for effects of 5-HTTLPR genotype on subACC activation as well as amygdala activation during sad mood and recovery from sad mood.
On the basis of the studies reviewed here, we made two primary hypotheses: 1(a) S vs L participants will show greater amygdala reactivity during a sad mood; and 1(b) S vs L participants will show greater amygdala activity during recovery from a sad mood. We made the following secondary hypotheses for subACC activity: 2(a) S vs L participants will show greater subACC reactivity during a sad mood; and 2(b) S vs L participants will show greater subACC activity during recovery from a sad mood.