Can We Really See the Formation of Memories?

formation-of-memory-linkboxIn two words, not yet…


A recent article published in the journal Science by Robert Singer’s lab at the Albert Einstein College of Medicine in New York has been picked up by the media (particularly by blogging websites such as IFLS and ios9) and promoted as being the first time we can see the formation of memories. Without trying to reduce the importance of the author’s findings, which I find fascinating, I don’t believe that what we are seeing is memory formation but rather the media blowing things out of proportion and here I’ll tell you why.


I will start with a bit of background information on neuronal communication and memory formation which will help those unfamiliar with neuroscience to follow. It is well known that neurons, arguably the most important cells in the brain, signal using both electrical and chemical means. These electrical signals travel through neurons, down the axons towards synapses. At the level of the synapse this electrical signal is converted into a chemical signal whereby chemicals called neurotransmitters are released, crossing the synaptic cleft and activating receptors on the target neuron. If receptor activation on the target neuron is strong enough, this can generate an electrical signal which can pass down this neuron and continue the signalling process.



This process can be likened to a relay race; we need to have at least two people (or two neurons) to participate. The first runner (the electrical signal in neuron A) races down the axon towards the synapse. At the synapse, runner A passes the baton (neurotransmitter) to runner B. Only once the baton (neurotransmitter) has been passed (across the synapse) can runner B (the electrical signal in neuron B) continue the race. Let us imagine that the runner B is quite lazy and that he doesn’t always do his part. So we’ll say that the majority of the time runner B receives the baton he doesn’t continue the race. This is the case for neurotransmission, most of the time the second neuron doesn’t respond to neurotransmitter release from the first.

Over many years of research in neuroscience we have now come to accept that the synapses are the places where we store our memories. This idea began with the Canadian neuropsychologist Donald Hebb in 1949 when he proposed that the connection between two neurons, at the synapse, can become strengthened if both neurons are active at the same time. This theory of memory formation became known as Hebbian learning.


We can think of this theory of Hebbian learning like two runners being ready for a race. If both are focused and want to win, the passing of the baton should be a lot smoother and more efficient.



How does the brain manage to create new synapses and reinforce others, in order to form memories? This feat is achieved by producing specific proteins and delivering them at the right time to the right place, which in our case is the synapse. One of the first structural proteins to be delivered to the synapse when we form new memories is a protein called b-actin. This protein accumulates to form an internal scaffold in the synapse. B-actin makes the synapse bigger, stronger and helps to hold receptors at the synapse as well as holding other signalling proteins on the inside of the cell which are vital to the memory formation process.


We can compare this process of memory formation to training of our second runner. If these runners carry out a lot of training, runner B will become a lot fitter and stronger. Runner B will then be better able to continue the race when he receives the baton.

Now back to the study in question…


The experiments by Singer and his colleagues, rather than showing memory formation, as cited by the media, have in fact demonstrated that following the stimulation of neurons, there is an increase in the delivery of the b-actin messenger RNA (mRNA), i.e. the protein ’blue-print’, to the synapse. This b-actin mRNA will be crucial for the learning process because it is the mRNA which gives us the instructions on how to make the protein. What this study has elegantly demonstrated is that following neuronal activity, there is delivery of b-actin mRNA to the synapse which will increase the chances of b-actin protein being produced which may (or may not) start a cascade of events leading to a memory being stored in this synapse.

Coming back to our runners… this b-actin mRNA is like a training schedule. Just because runner B receives his training schedule doesn’t mean that he will follow it!

[youtube id=”memories” mode=”normal” align=”left” autoplay=”no”]

Here is a video from this study. What we are looking at is a neuron, the little balls you can see moving around the inside of the neuron are b-actin mRNA. Normally we can’t see them moving around, one of the highlights of this study is that they managed to tag native mRNA with a green fluorescent protein (GFP).





Although I think that the attention received by this study is just (this really is good research), there is simply no evidence of memory formation in this work. First of all, this study was carried out in brain slices and not in living animals (a common practice in modern neuroscience research), so how can we possibly see memories being formed? The protocol used by authors to stimulate neurons does not resemble the protocols commonly used to induce memory processes. The authors have indeed shown that the mRNA is delivered to the synapse following neuronal activity but we do not know the outcome… This leaves me with several questions;


  1. Was this b-actin mRNA actually translated into functional b-actin protein? This is an important question to address. Just because the mRNA is delivered to the synapse does not tell us whether it was made into protein and used by the neuron.
  2. Is this mRNA delivery specific to the synapse activated? It is believed that memories are stored in specific synapses so if we could show mRNA going to the synapse which was activated, we would learn a lot about memory formation and how this may malfunction in disease.
  3. And finally the big question… Does this process occur in the intact brain? This is a big question in all studies of this kind (including my own research). The authors have used a wide array of approaches which leads me to believe that this phenomenon will hold true in the brain. However that remains to be seen (Reading the press release by the Albert Einstein College of Medicine, it is clear that the authors have this experiment in mind).


I really don’t want to berate this article because I do like it, so I would like to finish on a high note. For your work to be published in the journal Science it must be of very high quality and have a big impact in the field. Furthermore, this lab published two articles in the same edition of Science which is impressive; having even one paper published in Science in your entire career is a big deal. This paper is the accumulation of many years of hard work and was carried out to very high standards. The conclusions drawn by the authors in the text are conservative, as opposed to the sensationalist conclusions in the media.


Studies taking advantage of this novel technique are likely to explode in the near future and reveal a wealth of information regarding biological processes. As I mentioned earlier, further studies could help us determine whether this process malfunctions in disease, for example it has already been shown that our scaffold protein b-actin is destabilised in Alzheimer’s disease, which leads to the collapse of synapses (Those of you looking for more detail can take a look here for a nice review). Maybe the delivery of b-actin mRNA has a role to play, or this process could be a candidate for a potential therapy? Like any good scientific study, we are left coming away with even more questions than before we read the paper.


So coming back to my initial question, Can we really see the formation of memories? Not quite but if the authors said they are thinking about it in their press release, it’s likely that this work has already started.