A proposed model suggests habituation requires a whole neural network
One day when I came in to the office, my air conditioning unit was making a weird rattling sound. At first, I was slightly annoyed, but then I chose to ignore it and get to work. In another 30 minutes, I was completely oblivious to the noise. It wasn’t until my cubicle neighbor Meghan Rosen came in and asked about the racket that I realized the rattle was still there. My brain had habituated to the sound.
Habituation, the ability to stop noticing or responding to an irrelevant signal, is one of the simplest forms of learning. But it turns out that at the level of a brain cell, it’s a far more complex process than scientists previously thought. In the June 18 Neuron, Mani Ramaswami of Trinity College Dublin proposes a new framework to describe how habituation might occur in our brains. The paper not only offers a new mechanism to help us understand one of our most basic behaviors, it also demonstrates how taking the time to integrate new findings into a novel framework can help push a field forward.
Our ability to ignore the irrelevant and familiar has been a long-known feature of human learning. It’s so simple, even a sea slug can do it. Because the ability to habituate is so simple, scientists hypothesized that the mechanism behind it must also be simple. The previous framework for habituation has been synaptic depression, a decrease in chemical release. When one brain cell sends a signal to another, it releases chemical messengers into a synapse, the small gap between neurons. Receptors on the other side pick up this excitatory signal and send the message onward. But in habituation, neurons would release fewer chemicals, making the signal less likely to hit the other side. Fewer chemicals, fewer signals, and you’ve habituated. Simple.
But, as David Glanzman, a neurobiologist at the University of California, Los Angeles points out, there are problems with this idea. “If an animal shuts down the input from a pathway, what if there’s a case where that sensory information needs to get through?” he explains. “You can’t instantly override that signal.” For example, if my hearing pathway was habituated to the sound of the air conditioner, that would be fine for tuning out the rattle. But let’s say the fire alarm goes off. If my hearing pathway is habituated and signals aren’t getting through, I might not hear the alarm in time.
Ramaswami proposes that a different mechanism is responsible for habituation, based on years of work from his lab and others. Instead of each synapse releasing more or fewer chemical messengers, he imagines a network system made of many more cells, where inhibitory brain cells surround many neurons in a signaling pathway. When my rumbling air conditioner needs to be ignored, the original signal from my ear doesn’t decrease. Instead, the inhibitory neurons in my auditory tract increase their chemical release. This inhibitory signal competes with the excitatory signals coming in from the air conditioner, and dampens the signal.
This mechanism, which Ramaswami calls a “negative image” model, is much more flexible than the idea that a single neuron decreases its signals. Not only can inhibitory neurons fight to tune out the noise, but Ramaswami proposes that these brain cells also predict the particular patterns associated with, say, my air conditioner’s clatter, releasing inhibitory chemicals that block out the sound long before it reaches higher centers where I might notice. At the same time, because these inhibitory neurons are attuned to a particular noise pattern in my brain, they allow other signals through. If the fire alarm goes off, I will still know immediately.
Ramaswami says to think of it like the pixels on your screen. “Each pixel on a TV screen can be part of many images,” he explains. “The synapse depression model will make each individual pixel fade. But the negative image model acts at the level of the whole picture. The picture itself is faded, but each pixel can robustly display something else.”
A new framework to describe habituation could be important for people with autism spectrum disorders, who often have problems with habituation. That inability to filter out background colors, noises or people can make them feel overstimulated, especially in complex environments with many irrelevant sights and sounds (a carnival, for example). Understanding how habituation might work could help scientists to understand how it might go wrong and lead to those overstimulated feelings.
But first scientists need to determine just how well this new idea fits with what is happening in our brains. Ramaswami says that he has already seen this concept at work in the olfactory system in fruit flies, but it needs to be expanded to other systems in the brain. He also needs to determine whether habituation would be eliminated if inhibitory neurons were no longer able to change their responses to block out irrelevant signals.
Despite the long road ahead, these overhauls of previously accepted ideas are a part of the way science progresses — it’s how scientists put their small, separate findings into a larger context. “It’s the way we do science,” Glanzman says. “One of the things that drives a field forward is coming up with a model that integrates different lines of evidence. It presents a big picture model of the way things are. Then we can decide how well the big picture fits reality.” He says that with Ramaswami’s new model, “people will start to see that this basic form of learning is not simple. It engages a whole network of neurons.”