Julie Williams: Using genetics to understand Alzheimer's

Chris Smith asked Julie Williams how she has leveraged genetics to help us understand more about Alzheimer's disease...

Chris - The problem is that, the situation you arrived at where you wanted to grapple with the genetics of this very important and very common condition, but there weren't the tools there when you started that we really very much take for granted today to do genetic work. When we want to find genes today, it's very easy in comparison to the situation in which you would have found yourself 30-40 years ago.

Julie - That is true. I've never been in an area where we were learning new methods, technologies, perspectives, almost on an annual basis. So yes, when we started, we would spend perhaps nine months trying to understand a bit of a gene to try and describe its variation and compare it between people who had Alzheimer's and people who didn't. Now we can do whole genomes in a very short period of time and really interrogate those differences between those with a disease and those without. And what we also needed was to have very powerful samples - that was another thing that allowed us to find these things out because there's a lot of variation in there, and you could be finding things by chance a lot of the time. You need to build a lot of power and that's what took a lot of time in collecting enough people with Alzheimer's, comparing them to people without.

Chris - How did you go about that then? So you recruit families who have a premature onset of Alzheimer's? Or just lots of people who get it? What were you selecting and then how were you trying to find out what genes might be involved? Because in a genome of 20,000 genes, which we now know there are in a human, that's a huge number of moving parts.

Julie - My interest was in understanding the common forms of Alzheimer's disease. That was the initial focus: to collect hundreds of people so that we could just sample their DNA through a blood test or a saliva test and describe their disorder. That took a lot of time, but we spent about 20 years collecting, throughout the UK, Northern Ireland, these samples. More lately, collecting rarer samples of people with very early onset disease, which is more severe and probably more genetic in its basis. But there's going to be hundreds of genes, thousands possibly, that could influence the disease.

Chris - And how does one connect a gene that someone carries to a disease? How do you then square that circle so you can understand what its contribution is?

Julie - So we would take an individual's DNA, and the DNA has markers throughout it that can tell you whether there is a variant that is actually that variant or something very close to that variant that is associated with the disease. So we would compare DNA from individuals with Alzheimer's and without Alzheimer's, and we would look at simple changes that can occur in one or two forms, these snips or single nucleotide polymorphisms and if, for example, people with Alzheimer's had more of the first sort of that than those without, we could say with some confidence that there was something in that region that was associated with Alzheimer's disease. It may not be that particular snip, but it may be something close to that.

Chris - So it's like a signpost that says, in this region of the genome, there is something linked to this disease. And if you have that particular marker, then probably in that bit of the genome you have something that's going to increase or decrease your chance of getting that particular condition. So look at enough people enough times with enough outcomes and you will begin to hone in on what those areas are, even though you don't have to know exactly what the gene is yet?

Julie - That's true. So what we were seeing were patterns actually that were quite surprising when we first did our initial very large, what we called genome-wide association studies - these are these big studies of all the variants that tag most of the genome - we were seeing patterns that implicated our immune system. We knew from Alzheimer's first drawings that glial cells, immune cells, were there around the damage in the brains, but we'd always thought that this was just a normal reaction to something going wrong in the brain. But what the genetics was telling us is, no, these changes in the immune system are actually part of the pathway to disease.

Chris - If you look at people who have down syndrome, who have an extra copy of chromosome 21, they also get an Alzheimer's like change in their brain, don't they? So is the same thing going on or is that a red herring?

Julie - That's a very good question. I think there are a number of components to this complex disease and the component that is probably implicated there is more to do with the amyloid pathway, if you like. And that's maybe what is common, but it's not exactly the same as common Alzheimer's disease which is contributed to by a number of different components going wrong at the same time. It's the accumulation of risk in those various areas that pushes you into a disease process.