Neurons apparently must work to the rhythm of a special pacesetter frequency to store memories properly, according to a new study in human subjects. The findings, obtained by implanting electrodes into the brains of people with epilepsy, highlight the role of a long-studied but little-understood feature of the memory-storage process.
“These experiments suggest that that the brain does not continuously record the surrounding world but is quite selective, and generates intermittent windows of opportunity,” says Gyorgi Buszaki, a neuroscientist at Rutgers who wasn’t involved in the study but works in this field.
The new study was reported in the March 25 issue of Nature by a team of researchers from the California Institute of Technology and two Los Angeles area research hospitals.
The study was designed to clarify what is believed to be a key aspect of the memory process: the so-called theta brain-wave rhythms, which are observed in many brain areas but appear to be especially important in memory-related areas. The “theta” designation refers to the frequency range of about three to eight cycles per second – typically measured by an electroencephalograph (EEG).
For more than three decades, researchers have been finding in various studies that theta rhythms in memory-related brain areas are somehow linked to an animal’s or a human’s ability to learn. A study published in Science in 1978 found “a significant predictive relationship between EEG frequency characteristics and the subsequent rate of learning” in rabbits.
Later studies have hinted that a neuron’s ability to store its share of a memory—for example by altering the strengths of its synaptic connections to other neurons—might depend on the precise timing of its electrical impulses, or spikes, relative to the surrounding theta rhythm.
In the new study, senior investigator Erin Schuman and her colleagues, who included neurosurgeons, placed electrodes into the brains of nine people who were planning to undergo surgery to treat their epilepsy. The electrodes’ main purpose was to help surgeons plot out the path of the surgery (in healthy subjects, such an invasive and dangerous implantation of electrodes would be unethical.)
Using the electrodes, the researchers recorded the electrical spikings of hundreds of individual neurons in the hippocampus and amygdala, two brain regions known to be crucial for memory. At the same time, they monitored the local electrical field fluctuations created by mass neuronal activity in this region. While they recorded, the researchers showed each subject a series of 100 never-before-seen images; 15 minutes later they showed another 100 images—50 of which had been shown in the first set—and asked the subjects to indicate whether they remembered each image or not.
In the electrode recordings taken at the time subjects first viewed the test images, there was one feature that tended to predict whether the person would later remember an image. The “spike-field coherence,” a measure of the degree to which the individual neurons spiked in sync with the local background theta rhythm, was about 50 percent higher for images that were later remembered, compared with images that were later forgotten.
Schuman, who recently joined the Max Planck Institute for Brain Research in Germany, plans to follow up with further experiments, but for now sees this finding as a strong confirmation that this theta-frequency neuronal synchrony seems to favor the encoding of memory. “Many biochemical pathways within cells are sensitive to inputs that arrive together,” she says.
She suggests that this synchrony of memory neurons is to some extent spontaneous, but also points out that other brain regions can push neurons towards this theta lockstep—including brain regions that mediate attention, arousal, and the perception of novelty or importance in a stimulus. Thus, a stimulus categorized by the brain as “important” might become more memorable, in part, by a stronger enforcement of theta synchrony as the memory is formed.
Buzsaki for his part suspects that the theta rhythm serves as an essential timing signal, akin to the wave of a conductor’s baton. “It likely allows a precise temporal sequence of neuronal spiking within each theta cycle, and this precise timing may be critical for writing the [stimulus] information into synapses,” he says.