PASADENA, Calif-In the brain, as in sports, sex, and life, timing--and teamwork--are everything. Such is the message of a series of studies by researchers at the California Institute of Technology that offer insight into the processes by which memories are stored in the brain and that may someday guide the development of new therapies to prevent epileptic seizures.
Using computer models of neuronal circuits and experiments on live rats, Athanassios Siapas, assistant professor of computation and neural systems at Caltech, and his postdoctoral researcher Evgueniy Lubenov are revealing the curious mechanism by which the brain spontaneously tips itself toward a state balanced between order and chaos. The driving factor in the brain's self-regulation, they say, is the timing of neural pulses.
The researchers looked at how the timing of pulses fired by neurons in a simulated network (and, later, within the brains of freely roaming rats) interact with the plasticity in the synaptic connections between those neurons to influence the system as a whole. The studies revealed that when neurons fire in synchronized bursts, their harmony is fleeting; over time, the very act of synchrony tends to decouple the neurons, so that they become less organized, and their subsequent firing patterns more random.
Conversely, when neurons initially fire in a more random pattern, the randomness leads to strengthening of connections that drive the system toward a more synchronized firing pattern.
"Networks self-organize to an intermediate state, in between the two extremes," Siapas says.
Beyond its relevance to a basic understanding of neuronal circuits, the study may also prove significant for other, more clinically important, research. For example, overly synchronized neuronal firing is a characteristic of seizures in patients with epilepsy. Researchers have recently begun studying the effectiveness of deep brain stimulation in epilepsy patients. In the procedure, a device called a brain pacemaker is inserted into the brain, where it delivers electrical pulses to targeted regions.
"If this stimulation translates into the generation of synchronous events, it could decouple a possible locus for synchronous activity, while guiding the selection of targets for deep brain stimulation," thus reducing seizures, says Siapas. "One can fight synchrony with synchrony," he says, although he stresses that this is merely a conjecture and not based in experimental evidence.
The research also points to a mechanism by which short-term memories could be transferred from the hippocampus, a brain region involved in memory formation, to the neocortex, the area where long-term memories are held.
Neuroscientists have long known that during slow-wave sleep, the hippocampus exhibits a surge in synchronized neural firing directed to the neocortex. "This simultaneous activity is very effective at driving cortical neurons and strengthening the interactions between them," Lubenov says, and thus consolidating that information in the neocortex. In essence, a permanent memory is formed.
"We believe those same synchronous bursts also have a consequence for the memory trace in the hippocampus itself," Lubenov says, which is related to the self-organization that he and Siapas found in the system.
Their idea is that because synchronized neural firing in the hippocampus during this information transfer acts to desynchronize the system, it would reduce the strength "and lead to the gradual weakening of the hippocampal memory trace," Siapas says. Indeed, in experiments on rats implanted with electrodes, the researchers found a reduction in the simultaneous firing of neurons over the course of slow-wave sleep, indicating a move from synchrony to asynchrony. Eventually, through this process, a memory trace--unless reinforced through experience--would be erased from the hippocampus, freeing up neural resources there that could then be used to store new memories.
The paper, "Decoupling through synchrony in neuronal circuits with propagation delays," appears in the April issue of the journal Neuron.
