Healthcare and Science thomsonreuters.com

Our research is all about how proteins are made in cells and how the process is controlled. This process is called translation. We look at the protein molecules that co-ordinate and help to store memories.

There are repressor proteins in the brain that inhibit memory. We knew of a protein that can control one of these repressors so we deleted this molecule from the brain, therefore eliminating the repressor and getting better memory.

The basic question that people want to know is what is memory and how is memory stored? Everyone knows that it has to do with the communications between the synapses, but we want to understand the molecular basis of memory..

If you memorise something, there is a signal that comes in and you hit let's say 500 of the synapses, all the rest are not affected. So when you do that, how do you control the proteins that are synthesised? You control the proteins that are made in the synapses at the level of translation.

If you train a mouse to do a task and at the same time you administer an inhibitor of translation, he won't remember what he has been taught. So, there is a requirement  for the synthesis of new proteins in order to memorise.

We were working on a protein that  controls  the rate of translation called GCN2. In humans it activates  genes, which are important in stress responses. Mauro Costa-Mattioli PhD, a post-doctoral fellow in my laboratory had read papers from Eric Kandel, a Nobel Laureate who noted that as well as translation you need to also have new transcription for memory and learning. Kandel identified a protein, ATF4, which inhibits transcription and  is therefore a suppressor of memory. So it could be the case that if you have a bad memory it's because you produce more of this protein.

Dr. Costa-Mattioli did a simple deduction. He said that  we know that GCN2 increases the amount of  the ATF4 repressor. We have a knockout mouse, where the GCN2 gene is deleted. It would therefore be interesting  to see what happens to the mouse memory, because it no longer has the repressor. Strikingly, we found that, depending on the protocol given, this mouse is smarter than a normal mouse. How do you measure the smartness of the mouse? You take a swimming pool with a platform submerged under water, but the water is opaque white, so the mouse cannot see this platform. There are clues all around the pool; a balloon here, an octagon there and a star there. So he can orientate himself. The mouse doesn't like to swim but you put it in the water and it swims around. It finds the platform, stops and sits on it. And of course, he looks around, and sees the ball, and understands where the platform is in relation to the ball by spatial memory. So the mouse knows where the platform is in relation to different clues. Now what you measure is the time it takes to get to this. Once he knows where it is, he'll find it straight away.

You have to do training each day as the mouse has to learn where the platform is located. After day six he's getting pretty quick so you do it three times a day for half an hour. If he doesn't find it then you guide him to it. The first part of the experiment was a disappointment because this mouse, which we expected to be smarter, was in fact worse.

So, I consulted the Nobel Prize winner, Eric Kandel, and asked what do we do about this and he said  to train the mouse once a day instead of three times a day,  which might be too much. When you do this the normal mouse it doesn't learn as fast but now the knockout mouse learned faster. It was smarter.

Collaboration is very important in my work. We didn't know electrophysiology so we went to see Jean-Claude Lacaille at the Universite de Montreal. We didn't know about behavior so we spoke to Karim Nader.

With simple ideas, good implementation and rigorous research you can achieve very important advances in a field as complex as memory and the study of the brain. We use Web of Knowledge among other search engines to get the very best information for our studies.