“But we always go there!” And so begins another Friday night. When it comes to choosing where to go to dinner, my husband likes to stick with the tried and true. I like trying out new places. A new study suggests that the roots of this conflict could spring partly from our genes. It suggests that a DNA variation affecting the neurotransmitter dopamine influence a person’s willingness to explore new options instead of sticking with the status quo. The finding comes from a study by Michael Frank and colleagues from Brown University and the University of Arizona. The researchers focused on how people learn from positive and negative feedback. Subjects were confronted with a clock face that counted down five seconds. Before time was up they had to push a button to receive points. In some trials, the experiment was set up so that the faster they pushed the button, the more points they got. In other trials, waiting longer got more points. To the researchers’ surprise, people showed wide swings in response speed within each type of trial as they adjusted their timing in an attempt to maximize their scores. Computer models showed that a likely reason for these swings is that people change their strategy (pressing the button faster or slower) in proportion to how uncertain they are that a new strategy (speeding up or slowing down from what they’ve been doing) will yield better results. It makes sense: If you think a new restaurant might be only marginally better than the one you usually go to (and could be worse), you’re probably not that likely to vary from the usual routine. Why risk it? But if you really have no idea how good a place might be — who knows, it could blow your mind – you’d probably be more inclined to give it a whirl. Further analysis of the data, which will appear in the August issue of Nature Neuroscience, showed that the extent to which a person tried out new strategies correlated with variations in the COMT gene. People who carried the “Met” version of the gene were more exploratory in the face of uncertainty about what strategy to try next than people with two copies of the “Val” version (“Met” and “Val” refer to particular amino acids encoded by different versions of the gene). People with two copies of the Met version were the most adventurous, but even those with only one copy were statistically different in their exploration of new strategies from the people with two copies of the Val version. (The different versions of the COMT gene are determined by rs4680, which is available to 23andMe customers in the Browse Raw Data feature. A=Met, G =Val) The protein encoded by the COMT gene is involved in dopamine signaling in the prefrontal cortex, an area of the brain involved in planning and decision-making. The Met version of the gene leads to increased dopamine activity in this region and has been linked to more efficient information processing. So does this explain my date-night drama? Well, there’s undoubtedly more to it than genes alone, but I do have one copy of the more exploratory Met version of the COMT gene. And my husband? Two copies of the stuck-in-a-rut Val version. Dopamine and Learning In a region of the brain called the basal ganglia, dopamine helps us internalize positive and negative feedback in order to develop those “gut” feelings of what strategy will work and what won’t. The effects of dopamine in the basal ganglia have been shown in experiments that use drugs to raise or lower levels of the neurotransmitter in the brain. Higher dopamine levels help people learn to repeat rewarding behaviors, while lower dopamine leads to better learning from bad experiences. In a game where “A” usually yields more points than “B,” people with boosted dopamine levels learn to choose A. People with decreased dopamine levels learn to avoid B. In non-medicated test subjects, genetic variations that influence dopamine signaling in the basal ganglia also impact so-called “Go” (choose A) and “NoGo” learning (avoid B). People with two copies of the A version of a variant in the DARPP-32 gene, which increases dopamine signaling, tend to be better at Go learning than their G-version-carrying friends. Those with two copies of A at rs6277 in the DRD2 gene, which decreases dopamine signaling, tend to be better NoGo learners than people with one or two copies of the G version of this SNP. The clock-and-button experiments Frank et al. conducted further tested the association of these two variants with Go and NoGo learning. Trials that rewarded faster responses measured Go learning. Trials that rewarded holding off on the action of button pushing measured NoGo learning. As expected, people with two As at the DARPP-32 variant tended to be better at Go learning than people with one or two Gs, and people with two As at rs6277 in the DRD2 gene were better at NoGo learning than people with AG or GG at this SNP. (23andMe customers can see their data for rs6277 in the DRD2 gene using the Browse Raw Data feature. Data for the DARPP-32 variant is not available at this time.) Parkinson’s Disease Connection Understanding the role of dopamine in learning from experience may have important implications for treating people with Parkinson’s disease, which is characterized by a loss of dopamine producing neurons in the brain. Studies have shown that people with Parkinson’s have trouble with Go learning. It’s thought that the lack of dopamine in their brains prevents the dopamine spikes needed to learn from positive feedback. This fits with evidence that drugs that increase dopamine help people with Parkinson’s improve their performance on tasks that require Go learning. But there is a downside: because they flood the brain with dopamine, the normal dips in signaling that are needed to learn from negative feedback are blocked by these drugs. This might explain why some people with Parkinson’s disease who take dopamine-increasing medications develop gambling problems — they’re overly attuned to winning, but incapable of learning from their losses.