Monday, January 19, 2009

Scientists Discover Homer Simpson Neuron?

How does the brain encode and recall memories? Is there a specific neuron that fires when we recognize and remember Homer Simpson?

A recent publication by Gelbard-Sagiv et al. in the journal Science addresses the role of neuronal firing in the formation and recall of memories in humans (1). Subjects of the study were epilepsy patients who underwent surgical implantation of electrodes into their brains in order to localize the focus of their seizures. This created a unique opportunity for scientists to gain insights into memory formation and recall by directly measuring the neuronal activity in the human hippocampus, a region of the brain thought to play a role in processing memories. The authors studied the activity of individual neurons in the hippocampus and surrounding brain areas by recording electrical activity while subjects viewed clips from television episodes such as the Simpsons. They also recorded from these neurons while the subjects freely recalled the same episodes. They discovered that some neurons fired selectively while the subjects viewed specific episodes. Interestingly, the same neurons were active later during verbal reports of memories of the same episodes. Researchers hypothesize that this reactivation is an internally generated neuronal mechanism of memory recollection.

This study is among the first to measure neuronal activity in memory recall in humans. Previous studies performed on animals with implanted electrodes have linked neuronal activity with the memories of behavioral tasks. In a classic study by Wilson and McNaughton, electrical activities of hippocampal neurons were recoded in rats while they searched for a food reward (2). The authors also recorded from the same neurons while the rats slept, both before and after they executed the task above. They discovered that pairs of neurons that fired together during the task also exhibited correlated activity during sleep following training, but not before. This led them to hypothesize that specific neuronal activity was activated during the behavioral task and then later reactivated, as if the task was being replayed during sleep. However, this represents only indirect evidence, as one can only infer that the animals were recalling task related memories during sleep recordings. In contrast, humans in experiments conducted by Gelbard-Sagiv and coworkers were able to describe what they were thinking at the time of the neuronal recording.

The work by Gelbard-Sagiv et al. provides us with one of the best glimpses into memory formation in humans to date. However, these new experiments are not devoid of limitations. Is it possible that the increased neuronal activity observed during presentation and recall of certain clips is not specific to responsive cells? From a purely casual inspection of their data, this does not appear to be the case. However, a more rigorous approach would have been to compare (normalize) the cells’ activity to the local neuronal firing rate or field potentials during the recording. Also, unlike the work by Wilson and McNaughton, the authors only measured firing rates and did not examine the role of timing of electrical spikes, either within a single cell or between cells. These correlations are thought to be key to understanding neuronal coding and learning (3, 4, 5). Lastly, could the cells’ firing reflect an emotional state rather than a response to a particular memory? For example, the neuron that responded to the Simpsons clip also showed a significant response when a clip of the show Seinfeld was presented. Perhaps that neuron or area of the brain becomes active during feelings of laughter or joy. In fact, we know that the mesial temporal lobe, where these recordings were taken, is part of the limbic system, which has been implicated in processing emotional content (6). What do recordings of responsive cells show when subjects are shown content with similar emotional significance? These important control experiments were not done.

The above discussion points to an important related question. What information content do these memory-selective neurons code for? Did Gelbard-Sagiv and colleagues discover a neuron that codes for Homer Simpson, or perhaps a sitcom neuron? The authors point out that the study was not designed to answer this question, but to demonstrate that a subset of individual cells become reactivated during recall of specific memories. Does the concept of a Homer Simpson-specific cell make sense? The study measured only a small number of the 10^11 total neurons in the brain. What is the probability that any single neuron should respond specifically to one of the handful of clips chosen by the authors, such as the Simpsons? Exceedingly small. That is, unless vast numbers neurons not recorded respond to each clip, and each neuron responds to many clips that were not presented. Therefore, it is unlikely that there exists a Homer Simpson cell. But could there be a Homer Simpson brain cloud?

You can learn more about how large neuronal networks code for information in the brain in Brain Clouds: The Science of Collective Phenomena and the Brain.


1. Gelbard-Sagiv, H., R. Mukamel, M. Harel, R. Malach, and I. Fried. 2008. Internally Generated Reactivation of Single Neurons in Human Hippocampus During Free Recall. Science 322:96 - 101.
2. Wilson, M. A., and B. L. McNaughton. 1994. Reactivation of hippocampal ensemble memories during sleep. Science 265:604-604.
3. Rieke, F., D. Warland, R. d. R. v. Steveninck, and W. Bialek. 1998. Spikes: Exploring the Neural Code. MIT Press, Cambridge, Massachusetts.
4. Bi, G.-q., and M.-m. Poo. 1998. Synaptic Modifications in Cultured Hippocampal Neurons: Dependence on Spike Timing, Synaptic Strength, and Postsynaptic Cell Type. The Journal of Neuroscience 18:10464-10472.
5. Bi, G.-q., and M.-m. Poo. 1999. Distributed synaptic modification in neural networks induced by patterned stimulation. Nature 401:792-796.
6. Meletti, S., L. Tassi, R. Mai, N. Fini, C. A. Tassinari, and G. L. Russo. 2006. Emotions Induced by Intracerebral Electrical Stimulation of the Temporal Lobe. Epilepsia 47:47-51.