Connections Between Neurons Act As Information Filters In The Brain

Armen Hareyan's picture

Brain and Information

For the first time, researchers at the Salk Institute for Biological Studies have demonstrated that cell-cell contacts in the brain play an active role in processing information: called synapses, these interfaces act as precise filters that sense and amplify meaningful information, Salk researchers report in the current issue of PLoS Biology, available online.

Neurons are often considered to be the primary computational units of the brain. But it was unclear whether the connections between neurons actively participated in the computational process, or merely acted to convey information.

"Our study shows that synapses not only ensure the flow of information but actively modify their properties to help with computations," says Howard Hughes Medical Investigator Charles Stevens, a professor in the Molecular Neurobiology Laboratory and senior author of the study.

Lead author Vitaly Klyachko, a post-doctoral researcher working with Stevens explains, "Brain cells produce a lot of background chatter. Synapses filter this random noise and enhance relevant information. They work as very fine-tuned filters that do exactly what you would want them to do."


Brain cells signal by sending electrical impulses along axons, long, hair-like extensions that reach out to neighboring nerve cells. They make contact via synapses, from the Greek word meaning "to clasp together." When an electrical signal reaches the end of an axon, the voltage change triggers release of neurotransmitters, the brain's chemical messengers. These neurotransmitter molecules then travel across the space between neurons at a synapse and trigger an electrical signal in the adjacent cell.

Scientists had postulated that synapses play a major role in information processing in the brain. But not all signals are transmitted. Just like cell phone calls are dropped in areas of spotty coverage, synapses drop up to 90 percent of all incoming signals. "The unreliability of neuronal connections that presumably convey and process information throughout the brain was difficult to reconcile with the fact that brain as a whole is very reliable," says Klyachko.

Trying to get to the bottom of this enigma, the Salk researchers relied on naturally occurring activity patterns that were recorded in living animals from a part of the brain known as the hippocampus - a structure critical for memory formation and learning. They used these recorded patterns to stimulate isolated groups of neurons and measured which signals synapses transmitted to neighboring cells and which ones they dropped.

In the past, similar studies were commonly performed at room temperature. Since scientists had found that results were often too complex to interpret, the Salk researchers recorded data in warmer conditions, slightly below body temperature. "Intuitively, I recorded at physiological temperatures instead of room temperature and that turned out to be the key," remembers Klyachko. "I found that synaptic transmission is highly temperature-dependent."

From there it was only a small step to the discovery that the two major types of synapses, excitatory and inhibitory ones, that were previously thought to always work against each other, act in concert to identify patterns carrying relevant information in an incoming signal. Stevens explained: "Synapses recognize bursts of neuronal activity and turn up their strength, acting like a switch." As a result, meaningful patterns are amplified, while stray noise disappears into some sort of "synaptic abyss."

Until now, experimental evidence for a filtering function of synapses has been elusive. "Our work is the confirmation that everybody has been waiting for," explains Klyachko. "It is the precisely-tuned filtering properties of the two major types of synapses and their collaboration that makes the information processing reliable."