Neurons, when not responding to stimuli or involved in generating a behavior, emit electrical impulses spontaneously to an average of 3-2 frames per second. The way this occurs is random and irregular and constitutes what is known as neuronal variability. For years, this variability has been the subject of both experimental and theoretical studies, trying to determine how it is generated and its implications on information processing in the brain. IDIBAPS researchers have focused on studying whether the variability of each neuron is independent or not, and what is the cause of a possible interdependence, through the study of the activity measured simultaneously in groups of about 100 neurons in the auditory cortex of rats under anesthetic conditions.
The results show that neurons have a correlated activity because the network they conform tends to be silenced for very short periods of about 100-200 milliseconds, which occur randomly. "It's like the circuit of thousands of neurons had two operation modes: one in which neurons fire 2-3 pulses per second independently and another in which all neurons are silenced", explains Dr. de la Rocha. "Our hypothesis is that when the information must be processed reliably, the number and duration of silences decreases and neurons emit pulses independently and, when the demands for processing are not high, the network of neurons allows itself these little silences as a way off”, he adds.
These results provide insight into the mechanisms responsible for neuronal variability and its effect on brain function. Although the experiments were conducted in conditions of anesthesia, silence periods also occur when the animal is awake. This opens the door to the study of whether silence periods in awake subjects may have an impact on the behavior leading to a worse representation and processing of sensory stimuli.
Article reference:
Stochastic transitions into silence cause noise correlations in cortical circuits
Gabriela Mochol, Ainhoa Hermoso-Mendizabal, Shuzo Sakata, Kenneth D. Harris, and Jaime de la Rocha
PNAS 2015 ; published ahead of print March 4, 2015, doi:10.1073/pnas.1410509112