Nearly a century after the discovery of strange star-shaped cells in the brain, scientists say they have finally begun to unravel their function.

Researchers from the Massachusetts Institute of Technology report in Science that it appears astrocytes—named for their stellar form—provide nerve cells (neurons) with the energy they need to function and communicate with one another, by signaling blood to deliver the cell fuels glucose and oxygen to them.

When astrocytes were first discovered, it was believed that they were bit players in the brain. But the new research indicates they may actually be major operators that, when out of whack, may help trigger mental disorders such as autism and schizophrenia.

Study coauthor Mriganka Sur, a neuroscientist and head of MIT's Department of Brain and Cognitive Science, says his team saw astrocytes in action while examining brain activity in ferrets.

Using technology called two-photon microscopy, Sur and his colleagues observed that astrocytes in the visual cortex (part of the brain responsible for vision) activated and blood flow increased to nerve cells just seconds after the neurons had fired or sent out signals.

Sur believes the astrocytes—which are as plentiful in the brain as neurons—may control the strength and length of nerve cell communications. Consequently, he says, if astrocytes fail, so, too, may nerve cell connections, potentially leading to still largely unexplained neurological disorders.

"A great many genes that have been linked to autism and schizophrenia are likely to be active in astrocytes," Sur says. "We believe astrocytes will have a huge role in understanding certain brain diseases."

He says astrocytes may also shed light on brain activity captured on scans such as fMRIs (functional magnetic resonance imaging), which measure blood flow allowing scientists to diagnose strokes (areas that are deprived of blood or oxygen) and regions activated during tasks and activities from solving problems to daydreaming.

The reason, according to a Science editorial by neuroscientists Fred Wolf and Frank Kirchhoff of Germany's Max Planck Institute: the findings indicate fMRI "reflects the responses of both cell populations [neurons and astrocytes] in the brain."

Sur says that when the team blocked astrocytes, blood flow did not increase to firing neurons. That means, he says, that "[fMRI is] really measuring astrocyte activation. Thus, anything that influences astrocytes is likely to influence fMRI readings."