Model Demonstrates the Importance of Randomness to Brain Adaptation
In a paper published last month in the Journal of Neuroscience, Prof. Erik De Schutter and his former postdoc Gabriela Antunes report that by constructing a computational model of many molecules in a part of the neuron that regulates synapse strength, they have gained new and surprising insight into what determines that strength.
When you get a new pair of glasses, take up a musical instrument, or play a new sport, it feels awkward at first, but your muscles and brain soon adapt to integrate the new tools and movements seamlessly into your life. In the brain, that adaptation comes about in the cerebellum as meeting points between neurons, called synapses, grow stronger or weaker. As with many processes in the brain, there is much about this strengthening and weakening process that is still unknown. In a paper published last month in the Journal of Neuroscience, Prof. Erik De Schutter and his former postdoc Gabriela Antunes report that by constructing a computational model of many molecules in a part of the neuron that regulates synapse strength, they have gained new and surprising insight into what determines that strength.
Antunes and De Schutter modeled a spine, a structure in neurons that receives signals from another neuron in the synapse and relays them to the rest of the cell. Spines can increase or decrease the number of receptors on their surface, which affects their response to future signals—and thus strengthens or weakens the connection across the synapse. Because spines are so small, containing only about 50 molecules of each type of chemical, the researchers were able to model the actions of each individual molecule for about 20 different chemicals. This was a change from previous models, which had used approximations of how groups of molecules should behave. Another contrast with previous models was that rather than using deterministic equations with always the same answer, this one allowed for uncertainty and randomness in the molecules’ reactions. This messiness more accurately reflects the reality of biological systems, De Schutter explains.
The two researchers used the model to predict spines’ response to a quick pulse of calcium ions entering the synapse. It was already known that these brief pulses cause a relatively long-lasting loss of receptors in the spine, weakening the synaptic connection—the question was how. “We have to explain how a one-second signal drives a process that lasts 20 minutes,” De Schutter says. He and Antunes confirmed that the key is a positive feedback loop that amplifies the signal and keeps it going. But the process is unpredictable, as it is set off to different extents—or sometimes not at all—by identical signals. “We do not fully understand why this would be useful to the system,” says De Schutter. “That’s one of the things we want to study in the future.”