New thinking about your grey matter


The bulk of a human's brain cells were thought to only be there to support the neurons, the "thinking" cells. But new research suggests the support cells may work more quickly and be more important than previously thought. Cathal O'Connell and Elizabeth Finkel report.


110814 astrocytes h.jpg?ixlib=rails 2.1
Star-shaped astrocytes are not just the support crew for neurons, they also communicate with them to perform cognitive functions.
Nancy Kedersha/SPL/Getty Images

There may be more cells in your brain than stars in the galaxy but only 10% of them do the thinking – the neurons that form circuits. The other 90%, disparagingly referred to as “glia” or glue cells, have long been credited with a mere supporting role. But startling new research reported in Proceedings of the National Academy of Science in July suggests that we may have underestimated glia – particularly the star-shaped variety known as astrocytes. Tamper with them in mice and a specific kind of memory is impaired – the ability to recognise new objects as being new.

“This could be the crack in the door that is going to allow us to really break the brain’s code and understand how we can see and remember things,” says computational neuroscientist Terence Sejnowski at the Salk Institute in California, one of the authors of the paper.

A cross-section of Einstein's brain. It was no larger than average but it contained many more astrocytes than a typical human brain. – National Museum of Health and medicine

Scientists were first given a clue that there was more to astrocytes than meets the eye in the early 1980s when Marion Diamond, a neuroscientist at the University of California, Berkeley, put Einstein’s brain under the microscope. To her surprise she found it largely unremarkable. It was no bigger than average nor did it have more neurons or more connections between them. But she did find one striking feature – Einstein’s brain had many more astrocytes. Tantalisingly, this over-abundance was located in a part of the brain important to mathematics and spatial reasoning, two of Einstein’s famous strengths. Suddenly a spotlight was shone on the humble astrocyte. Was it actually central to human intelligence?

The transplanted mice became better at navigating mazes and locating objects. In other words, if you put human astrocytes into a mouse, the mouse gets smarter.

Back then the findings stumped scientists. In contrast to neurons astrocytes cannot transmit electrical signals. “Everyone just concentrated on neurons because they're the guys that talk to each other,” says Ross O’Shea, a neuroscientist at La Trobe University in Melbourne. Neurons were having the important conversations and astrocytes were serving them refreshments and cleaning up the mess – literally. Neurons transmit messages via chemicals called transmitters and after each transmission astrocytes remove the transmitters to reset the system for the next message. Astrocytes also control the blood supply to neurons. By constricting or relaxing blood vessels they tailor supply to match neuronal activity.

But the more scientists have learned about astrocytes the more complex their function appears to be. Research in the 1990s and 2000s suggested that astrocytes were actually equipped to converse with neurons. They carry receptors for transmitter chemicals released by neurons as well as releasing similar transmitters themselves. And, as hinted by Einstein’s brain, scientists found a correlation between astrocytes and intelligence. The astrocytes of primates are colossal compared to those of other animals. Human astrocytes, for example, are 30 times bigger than rodent astrocytes by volume. And Einstein’s astrocytes were particularly large even for a human, with a more complex shape, according to a 2006 study by Argentinian neuroscientist Jeoge Colombo. Then, in 2013, Maiken Nedergaard and colleagues at the University of Rochester reported a dramatic link between astrocytes and intelligence. When they transplanted human astrocytes into mouse brains the mice became better at navigating mazes and locating objects. In other words they became smarter.

All this suggested that the astrocytes were doing something altogether more interesting than simple maintenance, but what exactly?

In the latest study Sejnowski and his collaborators engineered transgenic mice so that their astrocytes failed to release chemical transmitters. This gagged the astrocytes but did not interfere with their maintenance duties.

Astrocytes from the brain of a mouse. New research suggests they are important for recognising new objects. – SILVIA RICCARDI/SPL/Getty Images

At first glance the engineered mice appeared to behave normally. They could learn new mazes and be trained to fear objects. The differences showed up when new objects were introduced into their environments – they failed to recognise the objects as new. Something appeared to be awry with their “recognition memory” – the ability to compare a new environment against a stored memory of it – an important faculty that allows us to recognise people, places, facts and things that happened in the past, says Sejnowski.

It was a surprising finding on many levels. For one, the process behind recognition is lightning fast, taking milliseconds. But astrocytes typically respond on the timescale of seconds or longer.

The researchers got a clue to what was going on when they used an EEG to measure the chatter of the mice’s brains. When mice or humans are asleep this is comprised predominantly of low frequency waves. But when alert the frequency increases to the gamma range. Disturbances in the gamma waves are seen in schizophrenia, Alzheimer's disease, autism and epilepsy. The waves were also less powerful in the mice with gagged astrocytes, they found. And when the astrocytes were ungagged, allowing them to release transmitters again, the gamma waves returned to normal and the mice could once again identify new objects.

What’s clear is that the astrocytes’ input was changing the pitch of the conversation, says John Bekkers, an electrophysiologist at John Curtin School of Medical Research in Canberra. Although it is not clear exactly how.

Bekkers is impressed by the elegant techniques used by the Salk researchers but is surprised that turning off the astrocyte transmitters created a specific defect rather than a global one. The mice for instance could still learn mazes and fear. “It’s a very broad manipulation but produces a specific defect and that seems strange.” On the other hand he points out that recognition memory may recruit many more parts of the brain than learning a maze or fear.

While the details remain to be worked out, Sejnowski describes the finding as “a smoking gun”. “With this new discovery scientists can begin to better understand the role of gamma waves in recognition memory,” he says.

Latest Stories
MoreMore Articles