Thursday, March 24, 2016 | 2 a.m.
When people make risky decisions, like doubling down in blackjack or investing in volatile stocks, what happens in the brain?
Scientists have long tried to understand what makes some people risk-averse and others risk-taking. Answers could have implications for how to treat, curb or prevent destructively risky behavior, like pathological gambling or drug addiction.
Now, a study by Dr. Karl Deisseroth, a prominent Stanford neuroscientist and psychiatrist, and his colleagues gives some clues. The study, published Wednesday in the journal Nature, reports that a specific type of neuron or nerve cell, in a certain brain region helps galvanize whether or not a risky choice is made.
The study was conducted in rats, but experts said it built on research suggesting the findings could be similar in humans. If so, they said, it could inform approaches to addiction, which involves some of the same neurons and brain areas, as well as treatments for Parkinson’s disease because one class of Parkinson’s medications turns some patients into problem gamblers.
In a series of experiments led by Kelly Zalocusky, a doctoral student, researchers found that a risk-averse rat made decisions based on whether its previous choice involved a loss (in this case, of food). Rats whose previous decision netted them less food were prompted to behave conservatively next time by signals from certain receptors in a brain region called the nucleus accumbens, the scientists discovered. These receptors, which are proteins attached to neurons, are part of the dopamine system, a neurochemical important to emotion, movement and thinking.
In risk-taking rats, however, those receptors sent a much fainter signal, so the rats kept making high-stakes choices even if they lost out. But by employing optogenetics, a technique that uses light to manipulate neurons, the scientists stimulated brain cells with those receptors, heightening the “loss” signal and turning risky rats into safer rats.
“We know from other work that this is all relevant to human addiction and gambling,” said Trevor Robbins, chairman of the psychology department at the University of Cambridge, who was not involved in the new research. “This study has zeroed in on the area precisely where this occurs. They’ve tried to show that not having this signal biases you toward risky judgments in the future, and they’ve done a lovely job on that.”
Step by step, the researchers built evidence that neurons with a dopamine receptor called D2 in the nucleus accumbens, a region integral to brain reward circuitry, play a critical role in risky-or-not decision-making. Strikingly, they found they could alter the message those neurons send.
Rats were given a choice of two food levers. One released a consistent amount of sucrose each time; the other often delivered a tiny amount, but in 25 percent of presses, it unleashed a delicious sucrose flood. Over time, both levers gave the same quantity, so rats did not go hungry and their choices came down to whether or not they were gamblers.
Risky rats gambled on the iffier lever more than half the time. Risk-averse rats were strongly influenced by their last choice; if they picked the risky lever and received a trickle, they picked the consistent lever next time.
“Some are very sensitive to losing, and if they take a risky option and lose, they’re very likely to not go back to it again,” said Paul Phillips, a professor of psychiatry and pharmacology at the University of Washington and a co-author of a commentary about the study. “That’s very common in human behavior. An analogy is a slot machine in Vegas.”
To identify the brain location involved in these decisions, the researchers gave rats a drug used to treat Parkinson’s disease, pramipexole, marketed as Mirapex, which acts on D2 receptors and seems to dampen some patients’ ability to restrain risk-seeking behavior. Risk-averse rats receiving pramipexole turned into risk-taking rats, but the drug had much greater effects when piped directly into the nucleus accumbens than when it was administered to another brain area researchers had thought might be involved.
The scientists used a technique Deisseroth helped invent called fiber photometry, which uses light particles to track activity of neurons tagged with certain proteins. They found that neurons in the nucleus accumbens with D2 receptors transmitted a signal when rats were making their decisions. That signal was much larger if the choice the rat had just made had been a loser, yielding just a dribble of sucrose. The signal only spiked in non-risky rats, however; it was negligible in rats that always gambled for the sucrose windfall.
So, what to do with those risky rats? Using optogenetics, which Deisseroth also helped develop, the team stimulated nucleus accumbens neurons with D2 receptors at the very moment of the fateful food-lever decision. That caused the receptors to send strong loss signals to the rats, apparently making them weigh recent losses more heavily, and prompting them to play it safe with their next lever choice.
“It turns out you can explain a large part of whether rats were risky or not by this particular signal at this particular time,” Deisseroth said. “We saw it happen, and then we were able to provide that signal, and then see that we could drive the behavior causally.”
Human brains are more complex, of course, and “are not only affected by immediate recent losses,” Deisseroth said, but “your appetite for risk in many circumstances might be at least possibly reducible to what a particular set of cells in a particular brain area is doing.”
Robbins said that might yield insights for drug addiction, since it “clearly involves the dopamine system and these areas of the brain,” and in addicts, as in risky rats, the same receptors produce weaker signals.
For Parkinson’s patients, if versions of drugs like pramipexole could be developed to skip the nucleus accumbens and focus on brain areas responsible for movement, “it would be a much more effective therapy,” Phillips said. “It’s because it gets to the nucleus accumbens that it has this gambling effect.”
He added, “Now, not only do we know the part of the brain, but we know the particular cells in the brain, and we know that if you manipulate them you can change the behavior.”
Deisseroth said optogenetic manipulation is too invasive to be done in humans, but findings from optogenetic studies in animals are now being used to identify brain areas to target with noninvasive brain stimulation for problems like cocaine addiction.
Finding the roots of risk in the brain also “helps us understand what might be making people different in terms of their risk appetites,” he said. “It may help us see them differently, maybe in a more tolerant way, to realize that there’s a real biological basis for their behavior.”