The human brain is an organ of staggering complexity, and the subject of intensive study by tens of thousands of researchers. Encased in the skull, it is largely inaccessible; most scientists investigate its workings only indirectly, using techniques such as functional neuroimaging and electroencephalography.
As director of the Cognitive Neurophysiology Laboratory and the Epilepsy Surgery Program at the University of California, Los Angeles, though, Itzhak Fried has unparalleled access to this mysterious organ. During a career that spans more than 40 years, Fried has pioneered the use of microelectrode arrays to record the activity of single cells directly from the brains of patients being evaluated prior to undergoing neurosurgery.
“We have been using this particular method since 1993,” says Fried. “We place electrodes in patients who have pharmacologically resistant epilepsy, to try to find out where the seizures are coming from. After implanting the electrodes, we wait for them to have spontaneous seizures, which usually takes between 7 and 10 days.”
The patients remain fully conscious throughout their waking periods. By performing certain tasks in the lab while they wait for a seizure, they provide Fried and his colleagues with the rare opportunity to probe the activity of individual neurons.
“Fried is a very theoretically-minded neurosurgeon,” says Robert Knight, director of the Helen Wills Neuroscience Institute at the University of California, Berkeley, “so besides his clinical work he’s also very interested in understanding how the brain works.”
“He uses unique micro-wire electrodes to get recordings from single cells, and there aren’t many groups that can do this effectively. He has capitalized on this to do some really interesting brain and behavior correlations on a whole variety of topics, and is clearly a leader in this field.”
Fried’s work–which is partly funded by the Dana Foundation–has yielded some remarkable results, and provided insights into the mechanisms underlying the voluntary control of movement and the mechanisms of memory formation, among other things.
In 2005, Fried and his colleagues reported that single cells in the hippocampus fire highly selectively in response to images of famous landmarks such as the White House and the Eiffel Tower, or to well-known celebrities such as Jennifer Aniston and Halle Berry.
In a follow-up study, they showed another group of patients a series of film clips, and confirmed that the hippocampus contains neurons that respond selectively to very specific stimuli, such as footage from "The Simpsons." Importantly, they found that the same cells also fired when the patients were asked to freely recollect the clips at a later time.
These discoveries led to the revival of the “Grandmother cell” concept. Originally proposed by Jerry Lettvin in the 1960s, this refers to a hypothetical neuron that responds only to a highly complex and meaningful stimulus, such as one’s grandmother.
They also led some researchers to argue that abstract concepts are encoded by single cells. This is, however, unlikely. Each cell is likely to be a node in a network of perhaps several million cells distributed throughout the hippocampus, which together encode a specific abstract concept.
“These aren’t really Grandmother cells,” says Fried. “They are cells which show a remarkable capability of abstraction. They are not bound by the physical constraints of the stimulus but respond to the idea of the stimulus.”
The hippocampus is deep inside the brain in the medial temporal lobe, and is known to be critical for memory formation [See: One Man’s Continuing Contribution to the Science of Memory]. Neuroscientists have made much progress in understanding the cellular basis of learning and memory during the past three decades, yet very little is known about how networks of neurons encode memories.
The psychologist Karl Lashley tried to locate the memory trace, or engram, in a series of classic experiments beginning in the 1930s. He trained rats to find their way through a maze, then damaged different parts of their cerebral cortex and had them run the maze again. The animals’ memories of their route through the maze remained intact, regardless of the location of the damage. Lashley concluded that memories are not localized to discrete parts of the cortex but are instead distributed throughout it.
Distributed networks of hippocampal cells such as those identified by Fried’s group, and the electrical activity that takes place within these networks, might therefore constitute the elusive memory trace. Individual cells are likely to contribute to thousands or millions of such networks, each of which encodes a different abstract concept.
“In the same way that the cells respond to celebrities,” says Fried, “we see that the patients develop these representations for the postdocs and grad students who work with them for a short period of time. The same cell fires when the memory comes spontaneously to mind.”
More recently, Fried replicated a classic study at the center of an ongoing debate about what neuroscience might tell us about free will. In 1983, Benjamin Libet and colleagues reported that brain activity associated with voluntary movement can be detected fractions of a second before a person became conscious of the intention to act.
These early findings proved to be highly controversial—they seem to show that the brain prepares for actions before our intention to act enters conscious awareness. Last year, Fried replicated Libet’s work at the cellular level, by showing that the activity of individual neurons in the medial prefrontal cortex predicts the impending decision to perform a movement approximately three-quarters of a second before awareness of the intention to move.
Fried’s latest study, published in the New England Journal of Medicine in February, shows that the experimental technique called deep brain stimulation can enhance spatial memory. Fried and his colleagues implanted stimulating electrodes into the entorhinal cortex, which lies next to the hippocampus, or the perforant path, a bundle of nerve fibres that connects the two structures.
They then asked the patients to find their way around a virtual environment and memorize the route to particular locations. The researchers applied electrical stimulation while patients navigated to some locations but not others. They found that the patients subsequently had a better memory for those routes that were initially navigated during the electrical stimulation.
“Some of the patients had low baseline memory but others had high baselines,” says Fried. “We improved memory capacity in all of them by stimulating during encoding of the information. This seemed to make the encoding better, maybe by enhancing formation of the engram. Exactly how is a mystery, but one possibility is that we influenced the brain’s oscillatory rhythms.”
The results suggest that memory function could be enhanced not only in those with conditions such as Alzheimer’s disease, but also in healthy people. The findings need to be confirmed, however, and it remains to be seen if deep brain stimulation could also enhance other types of memory.
Despite these many advances, the techniques being used to probe the human brain are still in their infancy. Miniaturization and scaling-up of the microelectrode arrays will allow for simultaneous recordings from multiple cells across larger expanses of the cortex, enabling researchers to gain a better understanding of the rhythmic activity within neuronal networks.
“We have to consider the oscillations and rhythms which probably reflect the synchronous work of populations of cells,” says Fried. “We also need to combine single cell recordings with stimulation or modulation of these systems. The technology to stimulate and record from the same electrode exists, but still needs to be developed.”