"We were attempting to model a disease that in humans begins in mid-life, and do it in a mouse that only lives for two years in the laboratory, " says Gillian P. Bates of the GKT School of Medicine at King's College, London. "So we had to find some way of accelerating the onset." Photo: Ben Woodman

Science is known to be a collaborative endeavor, but few collaborations have ever been so spectacularly successful as the search for the Huntington’s disease (HD) gene, which took ten years, involved fifty researchers from six laboratories, and finally paid off in 1993. Progress since then on understanding the havoc wreaked by the progressive and often fatal neurodegenerative disorder has been almost exponential. Since the identification of the HD gene, researchers have created mouse models of the disease, unraveled the degenerative mechanisms behind the disorder, and are now in the process of screening possible pharmaceutical compounds.

   While Huntington’s research has remained extraordinarily collaborative, one researcher who has helped move the field dramatically forward is Gillian P. Bates of the GKT School of Medicine at King's College in London. Bates and her collaborators developed the first mouse models of Huntington’s disease. In just two years, Bates's1997 Cell paper on the work–"Formation of neuronal intranuclear inclusions underlies the neurological dysfunction in mice transgenic for the HD mutation"–has already amassed over 160 citations (see paper #1 in the table on page 4). Earlier this year, this paper and four other highly cited reports coauthored by Bates over the last two years placed her on this publication's annual list of the "hottest" authors in science (see Science Watch, 10[2]:1-2, March/April 1999).

   Bates, 43, did her undergraduate work in genetics at the University of Sheffield, from which she graduated in 1979. She then went on to obtain her master’s degree in biomolecular organization from the University of London, and in 1984 joined Bob Williamson’s laboratory at St. Mary’s Hospital Medical School, also in London. She did her thesis on the cystic fibrosis gene and obtained her doctorate in 1987. (Her collaboration on a candidate CF gene resulted in what is actually her most-cited paper to date–X. Estivill, et al., Nature, 326:840-5, 1987, cited more than 330 times–unexpected news to Bates, who notes , "That's my most-cited paper? It turned out that it wasn't the gene after all! I'm surprised it was so heavily cited....") For the next six years, she worked with Hans Lehrach at the Imperial Cancer Research Fund in London. In 1994, she moved to the GKT School of Medicine at King’s College, where she is now professor of neurogenetics.

Bates spoke with Science Watch correspondent Gary Taubes from her office in London.

Let’s begin more or less at the beginning: What is the nature of the genetic mutation that causes Huntington’s disease?

   Bates: The HD gene maps to chromosome 4. The gene is a long one, and a DNA motif close to the very beginning reads "CAG, CAG, CAG, CAG…" We know from cloning the gene that people who don’t get HD can have between 6 and 35 CAG repeats in their HD gene. When you get to 36, the trouble starts. There are people who carry 36 repeats who have gone on to develop HD. That’s the only difference between these genes. And it’s dominant, so you only need one copy that has the mutation–if just one of your chromosome 4s has it, you develop Huntington’s disease. Today, we know eight neurodegenerative diseases caused by this type of mutation. In each case the CAG repeat is in the part of the gene that codes for the protein. Since CAG encodes the amino acid glutamine, the effect is that there is a longer stretch of glutamine in the protein when it causes the disease than when it doesn’t.

What’s the normal function of this protein?

   Bates: It doesn’t look like anything described before. We really don’t know much about its normal function yet. The only thing that all these proteins have in common is this poly-glutamine stretch.

When did you start the mouse work?

Bates: When I came here in early 1994 and set up my own lab. When I left Hans Lehrach’s lab we agreed that I would take the mouse-modeling side of the project with me and he would continue to work on the protein.

Were there any particular challenges in making the HD transgenic mouse?

   Bates: Well, we were attempting to model a disease that in humans begins in mid-life, and do it in a mouse that only lives for two years in the laboratory. So we had to find some way of accelerating the onset. We were fortunate in this sense with HD and the other CAG diseases that there was a correlation between the number of CAG repeats and the onset of the disorder. There are childhood and adolescent cases of HD in which the gene has 70 or more CAG repeats. So we wanted to put the longest repeats into the mouse to try and accelerate the phenotype. The difficulty is when you’re trying to piece together the DNA you want to introduce in a mouse, you do it by cloning it in E. coli, and these are extremely unstable in E. coli. They change size all the time. It’s difficult to get the number of repeats you want. So we went through some torturous cloning procedures. We ended up basically running two experiments in parallel–one was the complete gene, trying to get about 150 CAGs into it, and the other was a short fragment, just the first exon of the HD gene with some of the control elements and the CAG repeats.

Which worked better?

   Bates: It turned out that the short fragment gave us a wonderful model. With it we were able to establish a number of different transgenic lines containing different repeat lengths. One line started with about 150 repeats, although it changes size slightly from one mouse generation to the next. These develop a progressive neurological and cognitive disorder that first becomes apparent around two months of age, which is equivalent in humans to young adults.

Tell us what the model taught you about the disease process.

Bates: Well, for starters, we analyzed these mice at 4 weeks, 8 weeks, and 12 weeks. If the disease was being caused by neuronal cell death, we should really be able to see it by 12 weeks when the mouse is really sick. But even at 12 weeks we couldn’t find any real evidence of cell death. This was work done with my collaborator Stephen Davies, a neuropathologist at University College London.
   The first time Steve saw a major difference between the transgenic mouse brain and the normal litter mate controls was when he started to use antibodies to the very beginning of the HD protein; these antibodies became available from several laboratories in late 1996. Up until then, nobody had noticed any difference between autopsy brains from HD patients and autopsy brains from normal people, in terms of where the HD protein was localized. When Steve used these antibodies on our transgenic mouse brains, we saw an intense focus of staining in the nucleus. Under electron microscopy one could see this granular, fibular structure.

Did you know what it was?

   Bates: We had no idea. It looked like a deposit. But researchers had always said that you don’t see deposits in HD like you do in Parkinson’s or Alzheimer’s. When we saw this in mouse brains we didn’t know how to relate it to the disease because nobody had described anything like this in HD before.
   Meanwhile Hans Lehrach had moved to Berlin, where he is director of the Max Planck Institute for Molecular Genetics. Working with Erich Wanker, Hans started to study just this small protein in vitro. When they tried to purify this, they found, remarkably, that these proteins stayed soluble if they had only 20 or 30 CAG repeats, but that if they contained 51, 83, or 122 they spontaneously formed fibers. These fibers are essentially amyloid, which means they have a particular type of structure–the same as you find in plaques in Alzheimer's disease brains. Many different proteins can form this kind of structure. Obviously it’s hard for cells to deal with. So, that all fit in fairly nicely. It looked like HD and the other CAG-repeat diseases are amyloid diseases after all, although much simpler than Alzheimer’s because it’s just one mutation causing it and every patient has the same mutation.
continued