More than glue: Glia cells found to regulate synapses
December 30, 2011
Glia cells: the brain's supervisors (credit: Gray's anatomy/Wikimedia Commons)
Glia cells are central to the brain’s plasticity, Tel Aviv University researchers have found, controlling how the brain adapts, learns, and stores information — and their design can be implemented in neuromorphic computer chips.
Glia cells (Greek for “glue,” also known as glial) hold the brain’s neurons together and protect the cells that determine our thoughts and behaviors. But glia cells have now been found to do much more: a mechanism within the glia cells also regulate the synapses, sorting information for learning purposes, according to Ph.D. student Maurizio De Pittà of TAU’s Schools of Physics and Astronomy and Electrical Engineering.
“Glia cells are like the brain’s supervisors. They control the transfer of information between neurons, affecting how the brain processes information and learns.”
De Pittà’s research, led by his TAU supervisor Prof. Eshel Ben-Jacob, along with Vladislav Volman of The Salk Institute and the University of California at San Diego and Hugues Berry of the Université de Lyon in France, has developed the first computer model that incorporates the influence of glia cells on synaptic information transfer.
The model can also be implemented in technologies based on brain networks such as microchips and computer software, Prof. Ben-Jacob says, and can aid in research on brain disorders such as Alzheimer’s disease and epilepsy.
Regulating the brain’s “social network”
The brain is constituted of two main types of cells: neurons and glia. Neurons fire off signals that dictate how we think and behave, using synapses to pass along the message from one neuron to another. Scientists theorize that memory and learning are dictated by synaptic activity because they are “plastic,” with the ability to adapt to different stimuli.
But Ben-Jacob and colleagues suspected that glia cells were even more central to how the brain works, particularly the astrocytes (a form of glia cells) in the hippocampus. Glia cells are abundant in the brain’s hippocampus and the cortex, the two parts of the brain that have the most control over the brain’s ability to process information, learn and memorize. In fact, for every neuron cell, there are two to five glia cells. Taking into account previous experimental data, the researchers were able to build a model that could resolve the puzzle.
Astrocyte (glial cell) regulation of synapses in the hippocampus (credit: Tel Aviv University/PLoS Computational Biology)
The brain is like a social network, says Prof. Ben-Jacob. Messages may originate with the neurons, which use the synapses as their delivery system, but the glia serve as an overall moderator, regulating which messages are sent on and when. These cells can either prompt the transfer of information, or slow activity if the synapses are becoming overactive. This makes the glia cells the guardians of our learning and memory processes, he notes, orchestrating the transmission of information for optimal brain function.
New brain-inspired technologies and therapies
The team’s findings could have important implications for a number of brain disorders. Almost all neurodegenerative diseases are glia-related pathologies, Prof. Ben-Jacob notes. In epileptic seizures, for example, the neurons’ activity at one brain location propagates and overtakes the normal activity at other locations. This can happen when the glia cells fail to properly regulate synaptic transmission. Alternatively, when brain activity is low, glia cells boost transmissions of information, keeping the connections between neurons “alive.”
The model provides a “new view” of how the brain functions. While the study was in press, two experimental works appeared that supported the model’s predictions. “A growing number of scientists are starting to recognize the fact that you need the glia to perform tasks that neurons alone can’t accomplish in an efficient way,” says De Pittà.
The model will provide a new tool to begin revising the theories of computational neuroscience and lead to more realistic brain-inspired algorithms and microchips, which are designed to mimic neuronal networks, the researchers say.
Ref.: Maurizio De Pittà et al., A Tale of Two Stories: Astrocyte Regulation of Synaptic Depression and Facilitation, PLoS Computational Biology, 2011 [doi: 10.1371/journal.pcbi.1002293]
Comments (1)
by Phil Osborn
What I find distressing is how such a major fact could sit there, undiscovered for decades. Was this due to a lack of technology? Not likely. If over half of the cells in the brain are involved then surely there should have been multiple avenues that would have pointed to this.
That’s the meta-problem. Why is data ignored? I recall, a decade or more ago, attempting to have a discussion of how the brain/mind system must actually work with someone who thought that he knew more than me – which was largely true, I’m sure – and that therefore I had nothing to say and his proper attitude was contempt for my ignorance.
Turns out, my model is proving to be the reality. I postulated to start that there were at least two major dimensions to brain function. One of them was the synaptic transference of precise data, neuron to neuron. The other was the setting up of the brain on a macro level, such as the fight or flight orientation, involving multiple neurotransmitters targetted at particular brain areas. That second dimension – dynamic organization of resources triggered by “digital” events – was something that would be nearly impossible to build into or emulate on the kind of computers we have today. His reponse was that my picture was nonsense, because their were only four neurotransmitters and everything was handled by synaptic communcations using those transmitters.
Similarly, there was a time when evolutionary selection of groups was thought to be extremely rare and mostly insignificant. As a consequence, evidence to the contrary was ignored while entire avenues of valuable research – especially research aimed at aging – were not pursued. This was not something so esoteric that an intelligent layman couldn’t validly consider the issues involved, but getting listened to is an entirely different matter.
On that note, I recall an article on how the Pentagon’s Space Plane was designed. It went something like this: The estimates from the professional engineers for just the design process would have bankrupted the nation. So, try something different. They invited a bunch of science fiction writers to join in with the engineers over a weekend at a hotel. The result was a 90% reduction in the design estimates – still WAY too high. So, go with what works. They added a fourth grade class to the mix, and with their ignorance that forced everyone to learn how to communicate with a fourth grader, the design was budgetted. The need to be correct can be a major impediment to knowing anything.