In a lab at MIT, a rat enters a T-shaped maze, hears a tone, and runs down the left arm towards a piece of chocolate. It’s a habit. The rat has done the same thing over so many days that once it hears the tone, it’ll run in the same direction even if there’s no chocolate to be found. Humans are driven by similar habits. Every morning, I hear my alarm go off, put some clothes on, and shamble into the kitchen to brew some coffee.
Habits, by their very nature, seem permanent, stable, automatic. But they are not, and the MIT rat tells us why. Earlier, Kyle Smith had added a light-sensitive protein to one small part of its brain – the infralimbic cortex (ILC). This addition allows Smith to silence the neurons in this one area with a flash of yellow light, delivered to the rat’s brain via an optic fibre. The light flashes for just three seconds, and the habit disappears. The rat hears the tone, but no longer heads down the chocolate arm.
The experiment shows that even though habits seem automatic, they still depend on ongoing supervision from the ILC and possibly other parts of the brain. They’re ingrained and durable, but subject to second-by-second control. And they can be disrupted in surprisingly quick and simple ways.
“We were all stunned by how immediate and on-line these effects really are,” says Smith. “Changing the activity of this small cortex area could profoundly change how habitual behaviour was, in a matter of seconds.”
By cutting out bits of a rodent’s brain, or inactivating them with chemicals, other scientists had already identified parts of the brain, including the ILC, that are important for habits. But these are somewhat clumsy methods. Smith’s team wanted some more refined, something that could inactivate the ILC on demand for short bursts of time.
They turned to optogenetics. This revolutionary technique takes light-sensitive proteins from around the tree of life, and uses viruses to introduce them into an animal’s neurons. By choosing the right protein, and targeting the right part of the brain, scientists can now excite or silence a chosen group of neurons with astounding precision, using little more than flashes of coloured light.
Working with supervisor Ann Graybiel and optogenetics founder Karl Deisseroth, Smith filled his rats’ ILCs with halorhodopsin – a protein that comes from salt-loving microbes, and silences neurons when hit by yellow light.
Smith then trained rats to run down one arm of his T-shaped maze towards some chocolate, or down the other towards a sugary drink. One tone set them off, and a second one told them which arm to run down. After days of practice, the habits became ingrained. Even when Smith started “devaluing” one of the rewards by lacing it with a nauseating chemical, the rat would still run towards it the next time it was tested, even if it didn’t drink. That’s a habit – a behavioural reflex. The tone sets off a chain of actions, regardless of what the rat gets out of them.
Then, Smith inactivated the rat’s’ IL, while they were running through the maze. The effect was dramatic: almost immediately, they behaved as if they had never acquired their habit in the first place. When cued to run towards the devalued reward, they ran in the opposite direction instead. Bernard Balleine from Sydney University is impressed with how quickly the pulse of yellow light changes the rat’s behaviour. “The rapidity… is astonishing,” he says.
What’s happening here is quite subtle. It’s not that the yellow light wipes a rat’s memory. The animal still knows what the tone means. It still knows where the rewards are. Instead, disrupting the IL restores the importance of outcomes in the rat’s decision-making. Before, its habits meant that it ran wherever the tone told it to run. Now, it weighs up what happens when it reaches its destination. And knowing that one arm ends in a nauseating mouthful, it runs down the other.
So, is the ILC is a switch that flicks habitual behaviour on and off? No, says Smith, it’s more complicated than that.
After more trials, all the rats lost their old habit of running towards the nauseating reward, and developed a new habit: they always ran down the other arm. Then, Smith inactivated their IL again. As he expected, they lost their new habit. More surprisingly, they regained their old one!
They reverted back to their original behaviour, and ran towards the devalued reward when instructed, even drinking from it despite the nausea. “This was totally unexpected, but very robust and consistent. It was a big “What the…?” moment for me,” says Smith.
Science thrives on “What the…?” moments. Through further tests, the team showed that silencing the ILC doesn’t change the rat’s motivation to drink, or its memory of the foul-tasting reward. Instead, they think that the ILC acts as a behavioural overseer that toggles between different habits. The old ones are still encoded by neural networks, but the overseer ignores these in favour of networks that encode new habits. “This result suggests something new about habits,” says Smith. “They can be basically stacked upon one another, and the IL promotes the newest one, the one most appropriate to the current environmental conditions.”
To him, the idea makes intuitive sense. “We are all familiar with the idea that old habits might never really be lost, and that new habits can replace old ones, but I think this unexpected result gives us some insight into how this happens in the brain.”
Could optogenetics help us to break our own habits? Could it help people with obsessive-compulsive disorder, or those suffering from drug addictions? “I’m asked this at nearly every presentation – everyone has bad habits they want to get rid of!” says Smith. “But I think far more basic research needs to be done at this early discovery stage. There is quite a lot left to figure out.”
Balleine agrees. He notes that the ILC isn’t directly connected with other parts of the brain that are involved in the creation of habits, like the putamen or striatum. “Exactly why infralimbic inactivation works is difficult to say,” he says.
Smith agrees that no one knows what is actually going on in the ILC, or how its neurons affects other parts fo the brain. We do know that it connects to areas involved in learning, rewards, emotion, and forming plans of action. “The ILC is in a prime position to supervise these basic processes and perhaps influence how much access they have over behaviour, whether animals are flexible and learning something new, or whether they are just running on autopilot as they had been,” says Smith.
The team are now going to track the activity of neurons in the IL for months at a time. The plan is to see how they change and how they affect the rest of the brain, as habits are formed, lost and resurrected. They’re also going to look at the striatum. “We really need a better understanding of what the neurons are actually doing,” says Smith. “The IL contains many subclasses of neurons and many different output pathways, so we are also keen to figure out which of these is really crucial.”
Smith adds: “As one colleague said, the result was “humbling”; the brain is quite surprising sometimes, and we’re probably only scratching the surface on how it controls something complex like habits.”
Reference: Smith, Virkud, Deisseroth & Graybiel. 2012. Reversible online control of habitual behavior by optogenetic perturbation of medial prefrontal cortex. PNAS http://dx.doi.org/10.1073/pnas.1216264109
Image by Roland