Our current results illustrate that alterations in DA signaling associated with functional polymorphisms in multiple DA-related genes contribute a significant proportion of the interindividual variability in reward- related VS reactivity, which itself covaries with self-reported impulsivity. Specifically, polymorphisms directly associated with relatively increased DA availability (DAT1 9-repeat) and release (DRD2 –141C Del) or decreased postsynaptic inhibition (DRD2 –141C Del and DRD4 7-repeat) in the striatum are associated with relatively greater VS reactivity. In contrast, genetic variation primarily impacting prefrontal DA availability (COMT Val158Met) was not associated with variability in VS reactivity. These findings implicate genetically driven variability in striatal DA neurotransmission and VS reactivity as a key pathway in the emergence of interindividual differences in impulsivity, which represents an important intermediate behavioral phenotype associated with risk for addiction and related disorders.
Although our BOLD fMRI measure of VS reactivity does not represent a direct index of DA neurotransmission, the observed pattern of relatively increased activity in individuals with polymorphisms associated with increased striatal DA is consistent with a recent in vivo
human study reporting a direct relationship between striatal DA synthesis, assessed with FDOPA positron emission tomography (PET), and brain activity, assessed with BOLD fMRI.42
Acute increase of DA release via oral amphetamine has also been linked with relatively increased extent of BOLD fMRI-assessed VS activity.43
More generally, acute pharmacologic increase of DA in both healthy volunteers44
and patients with Parkinson’s disease45
results in relatively increased BOLD fMRI-assessed activity in closely related limbic brain regions, namely the amygdala. In contrast, a recent [11
C]raclopride PET study reported relatively blunted striatal DA release in response to amphetamine challenge in subjects reporting above-median levels of trait impulsivity. 46
While superficially inconsistent with the direction of our genetically driven effects on VS reactivity, these data may, in fact, be biologically consistent if the relatively blunted drug-induced striatal DA release reflects higher baseline levels of DA neurotransmission, as would be associated with our genetic variants, in subjects with higher levels of impulsivity. Multimodal neuroimaging studies combining PET measures of DA function and BOLD fMRI measures of VS reactivity may offer a unique opportunity to more directly evaluate underlying molecular mechanisms regulating this circuitry. Despite the limitations inherent in attempts to directly compare neurobiological effects associated with constitutive genetic variation, which presumably impacts not only the functional response of neural circuitry but also its development, with those from acute pharmacologic challenge, which impact only the existing functional architecture, our current findings are largely consistent with those linking increased DA neurotransmission and VS reactivity.
The DAT is responsible for the active clearance of synaptic DA and, thus, plays a critical role in regulating the duration of postsynaptic DA signaling, especially in the striatum.47
Accumulating evidence indicates that the DAT1 polymorphism impacts the expression and availability of DAT.17
Although a genotype effect has not been consistently observed across all studies,48–51
several suggest that in comparison to the 9-repeat allele, the 10-repeat is associated with relatively increased levels of DAT both in vivo18,19
and in vitro
We found that the 9-repeat allele, presumably through lesser DAT expression and increased synaptic DA, is associated with relatively greater VS reactivity in comparison to the 10-repeat allele. This is consistent with a recent imaging genetics study demonstrating relatively increased memory-related midbrain activity in 9-repeat allele carriers.52
Our findings are also biologically consistent with observations that activity of midbrain DA neurons is associated with unconditioned receipt of reward as well as the presentation of conditioned stimuli predicting reward,53
and that subsequent DA release in the VS potentiates reward-related behaviors. 10,11
Concentrations of synaptic DA in the striatum are also affected by acute negative feedback mechanisms consisting of both the inhibition of mesencephalic DA neuron firing rates and blockade of neurotransmitter release through activation of somatodendritic and terminal DRD2 autoreceptors, respectively.54,55
The DRD2 –141C Ins/Del polymorphism may influence striatal DA release by affecting the availability of both types of DRD2 autoreceptors. The –141C Del polymorphism exhibits relatively reduced in vitro
Our current finding of relatively increased VS reactivity in –141C Del allele carriers is consistent with this in vitro
effect and may reflect reduced DRD2 autoreceptor-mediated negative feedback and subsequently increased striatal DA release. Our data are also consistent with a recent study reporting relatively increased reward- but not anticipation- related VS reactivity in carriers of the DRD2 TaqIA A1 allele,56
which like the –141C Del is associated with lower DRD2 density.57
DA modulation of striatal circuitry is further mediated by postsynaptic DRD2 and DRD4, both of which exert inhibitory effects via second-messenger signaling cascades.58
Although not definitive, available data suggest that the DRD2 is expressed on local postsynaptic striatal neurons and the DRD4 is expressed on both postsynaptic striatal neurons and presynaptic corticostriatal glutamatergic afferents.59–61
This localization pattern suggests that these DA receptor subtypes can exert either direct (D2 and D4) or indirect (D4) inhibitory effects on striatal neurons. While the –141C Del allele impacts the availability of the DRD2 through reduced gene expression, the 7-repeat allele results in a DRD4 which exhibits reduced cAMP-reduction potency and subsequent postsynaptic inhibition in comparison with the 4-repeat.23,24
Thus, decreased inhibition of striatal neurons through either direct or indirect effects of the DRD2 –141C Del and DRD4 7-repeat alleles, respectively, may underlie their association with relatively increased reward-related VS reactivity. The direction of the DRD2 –141C Del effect on VS reactivity is notably consistent with a recent report demonstrating that in rats, trait behavioral impulsivity predicting cocaine self-administration is associated with reduced striatal DRD2/3 density.62
Finally, to determine the relative specificity of genetically driven variability in striatal DA signaling on VS reactivity we examined the common, functional Val158Met SNP in COMT, an enzyme responsible for the degradation of prefrontal DA.63
The Met158 allele results in relatively diminished enzyme stability and subsequently greater prefrontal DA availability through decreased amine degradation. Although the Val158Met may impact striatal DA through indirect prefrontal feedback onto midbrain DA neurons,64,65
it is unlikely to manifest direct effects on activity-dependent striatal DA release and striatal-dependent behaviors.66
Consistent with our hypothesis, variability in DA associated with the COMT Val158Met did not have a significant impact on reward-related VS reactivity. This finding suggests that genetically driven variation directly impacting striatal DA function is uniquely associated with interindividual variability in VS reactivity related to behavioral impulsivity.
While all three polymorphisms impacted VS clusters associated with the main effects of reward (that is, reward-related voxels) and showed significant overlap with each other (), their spatial extent and distribution were not identical, suggesting that each may impact different aspects of functional VS circuitry. Of the three DA-related polymorphisms significantly impacting VS reactivity, the DRD2 –141C accounted for the largest effect in terms of extent of VS activation, exhibiting 54% overlap with main effects of reward (245 voxels). DAT1 and DRD4 exhibited 35 (161 voxels) and 23% (86 voxels) overlap with reward-related VS activation, respectively. A similar pattern of genetic effects existed in regard to overlap with VS activation that correlated with self-reported impulsivity. The largest proportion of inter-individual variability in reward-related VS reactivity was explained by DAT1 genotype (~12%), followed closely by DRD4 and DRD2 genotypes (~9% each). Consistent with this pattern, the DAT1 genotype alone predicted individual differences in BIS scores, with 9-repeat allele carriers exhibiting both increased VS reactivity and higher BIS scores in comparison with 10-repeat homozygotes. Although this convergence is intriguing we caution against overinterpretation of the current results because, while large for a neuroimaging study, our sample size is far too small to draw meaningful inferences with regard to genetically driven variability in much more distal behavioral phenotypes such as impulsivity. The absence of similar significant behavioral effects for the other genotypes under investigation, although exhibiting robust effects on brain function, speaks to this important limitation.
Despite the similarity of these independent genetic effects, simultaneous modeling of all three genotypes revealed that only the DRD2 –141C significantly impacted reward-related VS reactivity when controlling for the effects of other factors. Although these polymorphisms presumably impact interrelated components of the DA signaling cascade, there was no evidence for statistically significant genetic epistasis on reward-related VS reactivity. However, our sample size, while quite substantial for typical neuroimaging studies, may be underpowered to detect significant epistatic effects, and larger samples are needed to properly evaluate their likelihood. The unique effects of the DRD2 –141C on reward-related VS reactivity may reflect the potential of this polymorphism to impact both striatal DA release, through somatodendritic and terminal DRD2 autoreceptors, and inhibition of striatal neurons, through postsynaptic DRD2 heteroreceptors.
In selecting our candidate polymorphisms, we did not focus on available data from association studies with broad behavioral (for example, impulsivity) or clinical (for example, alcoholism) phenotypes. While such association studies are important in establishing the broader relevance of genetically driven variability in brain function, they are often inconsistent and preclude inferences regarding the effects of polymorphisms at the level of gene or protein function. Regardless, all three genotype effects on reward-related VS reactivity are consistent with some prior association studies implicating the alleles that resulted here in increased reactivity with higher levels of impulsivity67
and risk for addiction.68,69
The existence of convergent findings for positive genetic association with neural, behavioral and clinical phenotypes provides compelling support for the importance of these specific DA-related polymorphisms in the pathways mediating interindividual differences in these processes.
It is important to note, however, that the likely contribution of specific genetic and neural factors may vary as a function of the intermediate steps leading to addiction, such as initiation, persistence and abuse.70
As has been recently demonstrated in genetic studies of liability for mood disorders,71,72
the impact of genetic polymorphisms on complex behavioral and clinical phenotypes is likely to be unmasked in the context of neural responses to specific precipitating environmental factors such as access to drugs of abuse and presence of social stressors.73,74
In fact, drug-induced striatal DA release associated with impulsivity appears to be differentially modulated by the experience of acute and chronic stressors.46,75
Future imaging genetics studies should attempt to dissect the contributions of genetically driven variability in brain function to each component of the addictions pathway and account for potential moderating environmental factors.
Future studies should also examine the unique and shared effects of polymorphisms to anticipation- and reward-related VS reactivity, which may contribute differently to variability in impulsivity and risk for addiction. The pattern of VS activity seen in our prior work7
as well as the present study is generally consistent with that reported using the original version of this task, where each trial was associated with monetary feedback.36
The VS activations we observed are also consistent with prior studies using various other reward incentive paradigms and stimuli (cf. Breiter and Rosen;39
Knutson and Cooper40
). However, unlike many previous studies, each trial was not associated with a monetary outcome (that is, gaining or losing money per trial) within our modified blocked design. Rather, each trial occasioned a more general positive or negative feedback reflecting the ‘correctness’ of the subject’s guess. This aspect of our protocol limits our ability to identify VS activity associated specifically with trial-by-trial monetary outcomes,36,76,77
which might be related more closely to differences in reward-related behaviors. It is also possible that our VS activation may more generally reflect feedback learning associated with the probabilistic nature of our blocked design.37,38
Importantly, this potential limitation is largely avoided in our focused analyses on the differential VS activation associated specifically with positive versus negative feedback, where the probabilistic nature of each block is identical. Finally, our blocked design precludes the analysis of VS activity uniquely associated with anticipation and outcome components of each trial. Employing event-related fMRI paradigms that better allow for the dissection of anticipation and outcome may help further parse the relationship between individual variability in reward-related behaviors and VS function.
The candidate polymorphisms in our study were selected based on evidence from studies directly examining the impact of these variants on specific aspects of biological function (for example, in vitro
reporter gene assays, mRNA, and protein expression studies and in vivo
neuroimaging measures), which are relatively robust and provide basis for specific directional hypotheses regarding genotype effects on DA function and subsequent VS reactivity. Nevertheless, it is possible that the functional effects on VS reactivity associated with the DA polymorphisms in the current study reflect the influence of other variants such as those in linkage disequilibrium (LD) with our candidates. For example, while we have highlighted evidence for a direct functional effect of the DRD2 –141C Del allele on gene expression and DRD2 density, significant neurobiological effects have not always been demonstrated.78,79
Such inconsistency is commonplace in genetic association studies and may, at least in part, result from the effects of occult functional variation for which candidate polymorphisms serve only as indirect markers. The absence of a perfect relationship between such markers and occult functional variants (that is, 100% LD) across study populations can undermine replication of effects. One promising approach to overcome this limitation is through the establishment of informative haplotypes, comprised of ‘tagging’ SNPs distributed throughout the gene, which represent the majority of all possible genetic variation in any given gene. The adoption of haplotypes in future studies of DA-related gene effects on reward-related VS reactivity and correlated behaviors should improve our understanding of the detailed mechanisms leading to interindividual variability in these processes.
In conclusion, our imaging genetics results reveal that DA-related functional polymorphisms resulting in relatively increased striatal DA neurotransmission bias toward increased reward-related VS reactivity. Given the demonstrated relationship between increased VS reactivity and greater self-reported impulsivity, as well as preference for immediate over delayed rewards,7
such genetically mediated effects may contribute to the emergence of behavioral tendencies to favor immediate gratification with little regard for alternative responses or consequences, act without thought, fail to deliberate about decisions and neglect planning. As such, we anticipate that our findings will contribute to the development of a principled framework from which future studies can systematically examine how the dynamic interplay of genes, brain and behavior is sculpted by environmental factors to mediate interindividual variability in these aspects of personality and temperament as well as their contribution to risk for addictions.