Australasian Science: Australia's authority on science since 1938

Memory Surgery

By Tim Hannan

Recent discoveries about memory modification open the way to erasing traumatic memories.

Many people have memories they would rather not possess. For some, it may be of a particularly embarrassing moment in adolescence; for others, it involves a vivid and terrifying re-experiencing of a traumatic event.

While psychological therapies can assist people to be less troubled by traumatic memories, the possibility of directly modifying their neural mechanisms has undeniable appeal. In the past year or so, a series of studies by MIT researchers has suggested that the notion of memory manipulation may one day be no longer restricted to the realm of sci-fi novels and movies like Inception.

In the first study, published in Nature, Steve Ramirez and Xu Liu sought to locate the specific group of cells in the brains of laboratory mice that could be reliably associated with a specific memory. They employed the novel technique of injecting a gene for a light-sensitive protein into the dentate gyrus, a region in the hippocampal formation of the temporal lobes long known to be part of the memory system. The cells that were active during the creation of the memory produced the protein, and thus became light-sensitive themselves. When viewed under a microscope, the active cells could be easily discriminated from others, suggesting that the researchers had discovered a means of identifying a specific neural memory trace, or engram.

To demonstrate that the light-sensitive cluster of cells truly reflected a memory engram, Ramirez and Liu developed a second and even more elegant experiment, the results of which were published in Science. A mouse was placed in a box and allowed to explore and develop a memory of this environment. A day later it was placed in another box of different shape, colour and aroma. The mouse then received a foot shock at the same time that a laser illuminated the light-sensitive cells of the dentate gyrus.

On another day, the mouse was placed again in the first box, in which no shock had ever been received, and froze in fear: it behaved as if it remembered a shock being associated with this box. After running studies to control for potentially confounding variables, the researchers concluded that the experimental procedure had succeeded in modifying the mouse’s initial memory engram, thereby creating the false memory of receiving a shock in the first box.

The technical wizardry of the methodology used in these studies was groundbreaking, but more astonishing is the result: the MIT researchers had shown not only that it is possible to identify the brain cells involved in the creation and storage of a single memory, but that the engram encoded in these cells may be modified to create a new, false “memory” of an event that never actually occurred.

In a recent extension of this work, the MIT team explored whether the emotions associated with a memory could be localised to a particular part of the brain. As it has previously been well established that the amygdala plays a role in emotional processing, they explored whether engrams in this region could be manipulated in the same way as those in the dentate gyrus had been modified in the earlier experiments.

Using two groups of mice, they found that while those whose dentate gyrus engram was targeted with illumination exhibited the change in emotional response or valence, those in the amygdala group did not. This result is interpreted to mean that, although the amygdala may provide the emotional input to the memory system, it is the neurons of the dentate gyrus that possess the plasticity to change the valence of a memory engram.

The demonstration that memories are directly modifiable, along with the identification of the functional neural circuits involved in changing the emotional quality of memories, has exciting implications for clinical psychology, with the possibility of surgical interventions for sufferers of traumatic stress disorders, depression and memory disorders such as dementia.

Yet should the practical difficulties of surgery be surmounted and the procedure found to be safe and effective, the possibility of directly manipulating engrams raises profound questions about the place of memories in people’s understanding of themselves, and of the implications of memory manipulation.

A/Prof Tim Hannan is Head of the School of Psychology at Charles Sturt University, and the Past President of the Australian Psychological Society.