News Release

Brain Cell 'Chorus' Appears as Attention Increases

The report, based on data acquired from monkeys, suggests that a baseball player tracking a fly ball through a cloud-cluttered sky, a driver reaching into a pocket to feel for keys, and a high-school student seeking a cafeteria dish that smells edible could all have something in common: Some of the nerve cells in the cortex, the sophisticated outer layer of the brain, may be sending messages in unison to allow them to pay attention to a single stream of sensory input.

"Every second, we get millions or hundreds of millions of bits of information coming in from our senses," says Ernst Niebur, assistant professor of neuroscience at the Krieger Mind-Brain Institute at The Johns Hopkins University. "And we have to decide, every second, which part of it is important and which part is not important."

"The nerve cells which represent the important information need a way to stand out from the crowd of other information," says Peter Steinmetz, lead author on the paper and a former postdoctoral scholar at the institute. "Firing synchronously – like singers in a chorus -- is one way to stand out from the crowd."

Scientists produced the new finding by re-analyzing data gathered over several years. Institute scientists Ken Johnson and Steve Hsiao had been monitoring brain cell activity in monkeys who were performing simple tasks that required them to focus their attention on visual or tactile stimuli. Tasks included identifying which of three white squares of light on a video monitor was beginning to dim, and comparing the shape of raised letters or figures pressed against a finger.

Applying a technique perfected by Hopkins neuroscientist Vernon Mountcastle, researchers used seven electrodes to simultaneously monitor individual brain cell activity in the monkeys as they worked. They originally analyzed the data they gathered for changes in the firing rate of brain cells as the animals switched attention between tasks.

When Niebur arrived at Hopkins a few years ago, researchers started talking about taking another look at the data.

Niebur and other theoretical neuroscientists were speculating that the brain might encode information both in the firing of individual brain cells and in the timing of those firings.

"It's been shown in animals that the firing rate of neurons can go up by a factor of 2 or 3 when they start to pay attention to a stimulus," Niebur says. "But it seems to run the risk of confusing signals if you try to code for two different things -- the stimulus itself and the degree that one should pay attention to it -- with one type of signal, the rate at which neurons are firing."

Niebur says the two different signals have to be connected. What your senses perceive will influence how much attention you pay to them, but, he said, "it seems like a good idea if you can have two different but related signals that you can use to represent these two things." An increase in the number of nerve cells firing in unison could represent just such an independent, but related, second signal.

Hsiao and Johnson had data from three earlier experiments appropriate for testing the theory. Steinmetz, now a post-doctoral scholar at Caltech, combined currently available computer power with a cutting-edge statistical technique to determine if nerve cells were firing synchronously and if the strength of that synchrony changed when the monkeys needed to pay attention.

"Detecting synchronous firing reliably has been difficult in the past because of the large amounts of data that need to be analyzed, but one outcome of the computer revolution has been the ability to perform this type of testing in reasonable timeframes," Steinmetz says. The results of the analysis, according to Niebur, suggested that when the monkeys were paying close attention to the stimuli, "the amount of synchronous firing appeared to increase in a sizable fraction of the neurons involved in these tasks."

Such a mechanism could have intriguing connections to basic nerve cell structure and function, Hsiao notes. Nerve cells frequently receive incoming signals from not just one but several different branch-like structures known as dendrites. Unless the signal is very strong, receiving a signal on any one dendrite doesn't necessarily guarantee that the nerve cell will pass on the message.

"If all the neurons upstream are firing synchronously, though, that strongly increases the possibility that the nerve cell will pass the message on downstream," says Hsiao, an associate professor of neuroscience.

"We were lucky that these three groups could come together for this team effort," Hsiao comments. "The Mind-Brain Institute is one of a very few places in the U.S. where you could see such a unique and close collaboration between experimental, theoretical, and computational neuroscientists."

All 3 research groups plan to follow up on the finding in the future.

"I'd like to go back to an earlier stage in this process, and look for some type of oscillatory signal that we're thinking could proceed these synchronized nerve cell firings," Johnson, a professor of neuroscience, says.

Niebur and Hsiao expressed interest in finding out what happens to synchrony rates if the test subjects fail to successfully complete the task they're concentrating on. Steinmetz's current research includes an investigation how strongly the neurons need to synchronize their firing to be "heard."

Funding for this study came from the National Institutes of Health, the Alfred P. Sloan Foundation and the National Science Foundation. Other authors were Arup Roy and Paul Fitzgerald, graduate students in neuroscience at Krieger Mind-Brain.