Molecular Memories – Manipulating the Brain

Memories are pretty important to most us. As collections of our experiences they anchor us to a personal sense of reality. Memories are pretty important to science, too; some of the earliest studies in psychology focused on their capabilities and qualities and they remain a major research topic today.

At a functional level, we know that memories are more like reconstructions than photographic records. Recall is an active process, and an imperfect one, so what we remember is rarely an accurate copy of the original event. We forget, we repress, and we confuse elements of some memories with elements of others (characteristics that make eyewitness testimony so unreliable).

What physical and biochemical processes in the brain can account for these functional properties? What actually happens when we form memories, and what happens when we forget?  Two new studies, both from MIT, tackled these questions directly, and an effect of this work is to point out some weighty questions about how firmly our memories are tied to “reality” and what we could (or should) do to change them.


One team was interested in exploring the molecular processes of forgetting.  They focused on one gene, Tet1, that’s involved in changing neurons to extinguish memory, and employed two sets of mice: a normal group and another group which had their Tet1 genes inhibited.

They first placed the mice in a cage and gave mild shocks to their feet until they learned to fear the cage setting. Later, they put the mice in the same cage again but didn’t shock them. In other words, the cage was now “safe.”

The mice with the normal Tet1 gene grew comfortable during their second stay in the cage, once they learned that there was no longer any shock. They had formed a new memory that “overwrote” the original one. The mice with the inhibited Tet1 gene, however, remained fearful during their second exposure. Without the gene action to extinguish it, the original shock memory was still intact in these mice.

Losing old memories, therefore, may be necessary to developing new ones and, in this experiment, that scientists were able to control that process by controlling a particular gene, Tet1. The memories in this study were both real. Another MIT team took a different direction: to create a new memory for something that never actually happened.

The researchers in the second study used optogenetics methods for their work. This is a way to activate individual brain cells by shining a blue light on them. Optogenetics involves transplanting a gene found in algae (channelrhodopsin, or ChR2) into selected neurons. When a blue light shines on these neurons, the gene causes those neurons to fire. The science team performed this transplant on cells in the hippocampus of mice, a brain structure that encodes memories.

They first established a nominal memory in mice by placing them in a box where nothing happened. Then, they placed the mice in a different box and shocked them. During the shocks, however, they activated the memory of the first box by shining light on those neurons. Their purpose was to create a memory of shock in the same neuron network that formed the memory of the original (no-shock) box.

When the mice were put back in the original box, they were fearful even though they had never been shocked in that setting. In other words, a new (false) memory of shock had been created.

These experiments show that memories are fundamentally just activities of cells. Manipulate those activities at the neuron level and you can manipulate memory itself, losing a real memory or creating a new one. This is a long way from manipulating the kind of complex memories in humans, of course, but even that possibility raises important technical and ethical questions.

Scientists suggest, for example, that knowledge like this could help generate new treatments for people with anxiety or psychiatric disorders by altering information in the brain. So, do we “edit” memories for the benefit of a patient? Do we implant better ones? Should we? Memories are likely woven into a narrative pattern, to give us a consistent sense the world and our place within it. Altering some memories may very well alter that larger pattern.

There are sure to be strong debates about these issues as knowledge and capabilities evolve. Regardless of how successful this work turns out to be, though, it’s sobering to think that our memories are increasingly understood in terms of their underlying biology, and that they may someday be the subject of physical intervention.

Leave a Reply

Your email address will not be published. Required fields are marked *