So for the last 2 or 3 years a good chunk of the work in our lab has been to look at these molecules to try to understand what they're doing in the brain in terms of how they help stick cells together when memories are being made, how they go wrong when the bits of the external sections of the molecules get broken off, and, very excitingly, what you can do to reverse that, or prevent that from happening. One of the things that I was doing at Cold Spring was talking about a new molecule that we've discovered — a little peptide , five amino acids long, which seems to be able to rescue the memory loss that you get with the disorder of the Alzheimer proteins. What started as a sheer intellectual excitement also looks like it's going to have rather significant human payoff, and that's good news.

JB: How did you make this discovery?

ROSE: It came about in a rather classical scientific way. What we did first was to show that in order to make long-term memory, that is memory that persists for more than half an hour or so, you need to make a new class of proteins. Then, using standard biochemical techniques, we were able to identify what the class of proteins were. They turn out to be a group called cell adhesion molecules. That is, they're molecules whose job is to stick together the two sides of the synaptic junction, the business end of the relationship between one cell and another. And that was interesting in itself; you can discover how they work, you can show how you have to unstick them and restick them in new configurations.

I was looking at this, and then I suddenly realized that one of the key proteins which is a major risk factor for Alzheimer's Disease, is itself a cell adhesion molecule. The question was could that be involved in memory as well? And it turns out that the normal functioning of this molecule is necessary for long-term memory to be made; if we stop the molecule from functioning — you put an antibody into the brain which binds to the molecule, or a specific bit of RNA which stops it being synthesized — then the memories can't be made.

Then if you look at the structure of this molecule, the amyloid precursor protein, it turns out that there is a very small section of it which is just a few amino acids long which seems to have some very special properties. It's those properties which you can mimic by making an artificial peptide, and it's that it turns out will rescue the memory which is lost otherwise. Clearly that's a long way from having a drug which will cure or protect from Alzheimer's Disease. But nonetheless, being able to rescue memory in this sense seems to me to be a step which is potentially in the right direction.

JB: What steps are necessary to make this available for human use?

ROSE: There's a lot more work to be done in animals first of all, the standard sort of drug development stuff would have to be done, and then you would have to show that you can take it orally without it breaking down — at the moment we have to get it in by injection. Or else you have to find a way of protecting it so it can get into the brain. Then there are various other bits and pieces of peptide controls that we need to do, and so on. You're talking a few years downstream, but you're moving in that direction.

JB: What has been the reaction among scientists?

ROSE: People are pretty excited about it. The formal scientific paper is just in press.

JB: When will it come out?