Erasing a memory with light


Memories make you who you are! Your memories give you a personal narrative, an ability to recall the past, and ultimately, lay the foundation for your thoughts, joys, fears, opinions, and skills. The ability to form, store, and recall memories is one of the most remarkable phenomena, and puzzling mysteries, of the brain. How does a piece of biological tissue, the brain, store so much information? This enormous database in your head, and the ability to access and edit it, are the result of hundreds of billions of small cells called neurons that communicate with each other by passing electricity back and forth at connections called synapses; the exact mechanism of memory, however, has yet to be revealed.

Many scientists and philosophers over the last 100 years have attempted to answer this question, from many different points of view. One leading hypothesis posits that memories are stored in patterns of activity across many neurons “speaking” with each other at synapses; the formation of a new memory thus causes the stabilization of new patterns of activity and new connections between neurons. While this is an attractive theory, it has been hard to prove rigorously in an experimental setting: there are many ways to visualize synapses, but no techniques to selectively manipulate them in a behaving animal…until now!

In 2015, Dr. Haruo Kasai led an international group of molecular biologists, optical engineers, and neuroscientists to offer the first mechanistic understanding of how the cells in our brain create, store, and recall a memory. Their discovery confirms over 100 years of theoretical speculation that the storage of a new memory creates new physical connections between neurons.

            If there are 100 billion neurons, and each of those neurons connect to many other neurons, then the number of connections between those 100 billion neurons is astoundingly high: more than the galaxies in the universe! This is why it has been hard for researchers to find the synapses that form with a memory: how would you know which ones to look for? Kasai and his team engineered a powerful segment of DNA that was able to locate synapses that were recently activated and destroy them (when they are bathed in blue light). To do this, they used fluorescent proteins and hijacked intracellular signaling processes to zap away synapses at will. They introduced this DNA into the brains of mice, and performed surgeries on them to replace their skull with glass, so they could permanently see into their brains with sophisticated laser-scanning microscopes. With the ability to see and destroy newly formed connections between neurons, Kasai and his team decided to ask the question of whether or not memories were the product of new connections in the brain. How would you use this technology to test whether or not the hypothesis was true?

            Kasai and his team took these mice (with the engineered DNA segment in neurons, and skulls replaced with glass) and taught them a new task. The mice were placed on a rolling pin which was constantly spinning. At first, the mice couldn’t do the task and fell off the rolling pin, onto the ground beneath it. Eventually, however, the mice learned how to walk on the rolling rod, and could stay on it for longer periods of time. This change in behavior must have been accompanied by the formation of a memory: the memory of how to walk on the rotating rod. Looking into the brains of these mice, Kasai and his team saw that lots of new synapses were labelled from the DNA they introduced and that many others had grown in size (Figure 2 in the paper). This confirmed the hypothesis that memories form new connections in the brain, but was the memory actually stored in these new connections somehow? Kasai and his team showed, amazingly, that when they bathed the brain in blue light and destroyed all of the new synapses, the mouse could no longer perform the task of walking on the rotating cylinder (Figure 3 in the paper). They showed that these newly-formed synapses were essential for the animal to access and integrate the information in that memory. While this may seem obvious to a neuroscience aficionado, no one had actually proved this assumption before Kasai and his team performed this experiment.

            Kasai and his colleagues tested the hypothesis that new memories in the brain are made by creating new connections between neurons. They developed amazing new technology that allowed them to label, then destroy, new synapses that were created during the formation of a memory. In their most important experiment, Kasai et al. showed conclusively that destroying connections which were formed during the formation of a memory also destroyed the memory itself. This impressive work sounds like science fiction, but it really happened, and is an important step in understanding how the brain works. The findings in this study generate many other hypotheses and new avenues for researchers to pursue: for example, if the destruction of connections associated with a specific memory destroys memory itself, maybe disease like Alzheimer’s can be reversed by figuring out a way to save connections that would otherwise be destroyed.


Hayashi-Takagi, A., Yagishita, S., Nakamura, M., Shirai, F., Wu, Y. I., Loshbaugh, A. L., ... & Kasai, H. (2015). Labelling and optical erasure of synaptic memory traces in the motor cortex. Nature525(7569), 333.