Shining a Light on How We Remember

For an agent of Men in Black , a neuralyzer is standard issue equipment. The “alien technology”, looking a bit like a silver cigar could be used to “isolate the electronic impulses in your brains, more specifically the ones for memory”. The device allowed Will Smith or Tommy Lee Jones to simply and safely erase the memory of anyone they want, simply by making them look at a flash of light.

The technology is pretty impressive, very useful for MiB agents, but from a science standpoint perhaps a little far-fetched. However, a finding from a new study by researchers based in Japan and the USA bears a surprisingly striking resemblance to this technique. Essentially, by shining a light into the brain of a mouse, the researchers were able to erase specific memories that the animals had just formed. However, rather than aiming to construct a science fiction memory eraser, the group were instead trying to answer some very important, fundamental, questions about how the brain forms memories.

For many years, neuroscientists have believed that memories are generated and stored by altering the strength of the connections, or synapses, between neurons. Put simply, the output part of one neuron (or the axon) sends a signal to the input area of the next neuron (or the dendrite). Repeated firing between the same cells strengthens the connection these two cells have, meaning that in the future they can communicate more efficiently – forming a proposed cellular basis of memory.

One way in which these connections could be strengthened is by the enlarging of small swellings on the dendrite of the input neuron – called dendritic spines. The size of dendritic spines (which are usually just a couple of micrometers) correlates closely with the strength of the synapse, and it is well described that spines can form, disappear or change in size during learning. Never before though, have we been able to directly implicate these tiny protrusions on a neuron as a direct physical correlate of memory.

Theoretically, pinpointing the function of dendritic spines is simple. You watch them grow as an animal is forming a memory, artificially erase them, and check to see if the animal has forgotten what it just learned. In practice though, this is far from easy.

In order to investigate the role of dendritic spines in memory the researchers first designed a protein to be expressed in neurons of the mouse motor cortex – the area of the brain involved in learning new motor skills. This protein had two parts. The first ensured it was transported selectively to newly growing dendritic spines. The second was something called Rac1, activation of which can collapse dendritic spines. The clever part, Rac1 would only be active when a laser of the correct wavelength is shined on it though – the researchers called this AS-pRac1.

With the protein lying in wait in their motor cortex, mice were trained to run on either on a rotating rod (or rotarod) or a narrow balance beam. Both of these tasks are picked up very simply by mice, but they do require an amount of learning to become proficient. Once the mice had learned these skills, the researchers could watch new dendritic spines formed in their motor cortex.

The AS-pRac1 was transported to these growing regions, then simply by shining a light the researchers collapsed the newly formed dendritic spines. After the laser was turned on, and the spines collapsed, the mice were no longer able to run on the rotarod or the balance beam. So amazingly, simply by removing some micrometer (a thousandth of a millimeter) sized swellings from the dendrites of a few neurons, mice completely forgot a motor skill they had just recently mastered.

In a further advance in our understanding of memory, by training mice sequentially on either the rotarod or the balance beam it was possible to erase the dendritic spines associated with just one of the tasks. Here, when collapsing only the rotarod dendritic spines the animals forgot how to run on the rotarod, but were still able to run on the balance beam – clearly showing that different neurons and dendritic spines were required for different memories.

So, what does this mean? Are we just one step closer to a Men in Black-style neuralyzer? Thankfully, that’s still a long time away. This study does, however, make a much more important contribution to science. We now have a much clearer understanding of what changes in the brain when we form a memory. Crucially, only by understanding how memories are formed and stored in the brain will be able to one day understand what causes memory to be lost or degraded in disorders like dementia.


Hayashi-Takagi, A., Yagishita, S., Nakamura, M., Shirai, F., Wu, Y., Loshbaugh, A., Kuhlman, B., Hahn, K., & Kasai, H. (2015). Labelling and optical erasure of synaptic memory traces in the motor cortex Nature, 525 (7569), 333-338 DOI: 10.1038/nature15257

Lu J, Zuo Y. (2015) Neuroscience: Forgetfulness illuminated. Nature. 525, 324-325. doi: 10.1038/nature15211
Lu J, & Zuo Y (2015). Neuroscience: Forgetfulness illuminated. Nature, 525 (7569), 324-5 PMID: 26352474

Image via Patrice6000 / Shutterstock.

Andy Murray, PhD

Andrew Murray, PhD, is a research scientist at the Sainsbury Wellcome Centre for Neural Circuits and Behaviour, University College London. Andy received a BSc and PhD in neuroscience from the University of Aberdeen in the UK, and carried out postdoctoral work at Columbia University in New York. He studies how neural circuits generate behaviour, with a focus on the vestibular system. Twitter @andymurray000 Website:
See All Posts By The Author