CREB Provides Clues in Search for a Memory Trace

Author:  Tetyana Pekar
Institution:  University of Toronto

The search for a memory trace', the chemical and biophysical changes that occur in the brain when we form memories, has been one of the most researched areas in neuroscience. Despite this, many important questions remain unanswered, such as: what factors influence whether a neuron will be incorporated into a memory network, and indeed, is there a network of neurons responsible for a specific memory?

In the March 13th issue of Science, Dr. Sheena Josselyn and colleagues at the Hospital for Sick Children in Toronto report that they were able to induce and then selectively ablate a specific fear memory in mice. This finding, which builds upon the lab's previous research, shows that there is a subpopulation of neurons in the amygdala (a region of the brain involved in fear memory formation) that encodes a highly specific memory trace. Furthermore, this memory can be erased without damaging the structure or impairing the animal's ability to form new associative fear memories.

The researchers used a standard tone-shock fear conditioning paradigm to induce a fear memory. In this paradigm, mice learn to associate a tone with a foot shock and subsequently freeze upon exposure to the tone in the absence of the shock. In this study, prior to training, the researchers over-expressed cAMP Response Element Binding (CREB), a transcription factor in the amygdala. It has been previously shown that neurons over-expressing CREB are preferentially selected to store a memory since these neurons are more likely to be activated during subsequent memory tests.

In the Science podcast, Dr. Josselyn explained the rationale, "We thought, Well, maybe these neurons with high levels of CREBs are the ones that are actually storing the memory.' So... what we did is we went in and just killed those neurons with high levels of CREB, and we saw that the memory, in fact, was basically erased."

The same neurons that were over-expressing CREB were also expressing a receptor for diphtheria toxin. After the mice were trained in the fear-conditioning paradigm, they were injected with the toxin. This selectively targeted and killed only the cells that were expressing the receptor and over-expressing CREB (mice do not normally express a receptor for diphtheria toxin). Remarkably, subsequent memory tests revealed that mice could no longer remember the robust tone-shock association. In contrast, mice that were not injected with diphtheria toxin were able to remember the tone-shock pairing for at least one month following the training. Importantly, mice injected with diphtheria toxin were able to re-learn the tone-shock pairing, suggesting that damage to the amygdala did not impair its function.

In the future, the researchers want to find out whether "this is a generalized sort of feature [the role of CREB in memory formation], or if this is very specific to amygdala and fear memories." Although this procedure is not feasible in humans, understanding the formation of associate fear memories may provide avenues for treating post-traumatic stress disorder by eliminating or reducing the association of previously salient stimuli with pervasive and debilitating fear.

Reviewed by Brandy Sullivan, Matt Getz

Reviewed by Lois Alexander