Five Questions for Dawn Thompson
Dawn Thompson has spent much of her career studying yeast. The experimental biologist, who is assistant director of the Broad’s Cell Circuits Program, and group leader in core member Aviv Regev’s lab says that she fell into the field “by accident.” While interviewing for her first post-college research position, she happened to hit it off with a yeast geneticist. After working in that lab, she went on to graduate school. While she didn’t intend to continue working on yeast, she found during her lab rotations that the biological questions raised by yeast were the ones that most interested her. She’s been active in yeast research ever since.
That happy accident brought Thompson to the Broad, where she and her colleagues use comparative genomics to learn what biological characteristics have been conserved or have diverged in various species of yeast throughout the evolutionary process.
Thompson answered five questions about yeast research, its relevance to human biology, and her recent experience with scientific publishing.
Photo by Maria Nemchuk,
Q1. Why study yeast?
DT: The cool thing about yeast is – and it’s the reason the NIH and other funders give labs money for yeast research – is that yeast share almost a third of their genes with humans. At the same time, the yeast genome is more tractable: yeast have roughly 6,000 genes, whereas humans have 20,000, so yeast genomes are easier for computational biologists to study.
Q2. What do you and your colleagues in the Regev lab study in your work on yeast?
DT: Like many researchers at the Broad, we’re tackling the question: how do we look at an organism’s DNA sequence – its genotype – and understand how it produces the organism’s traits or behaviors, or phenotype. Humans share a lot of genes with other organisms – dogs, apes and even yeast. We have the same genetic toolbox, but clearly we’re pretty different, so it must be gene expression – when genes get turned on and when they’re turned off – that gets rewired over the course of evolution to produce new phenotypes.
Our lab has been studying gene expression patterns of several yeast species to see how they have been rewired. We’re using a new algorithm, Arboretum, which was developed by a postdoc in our group, Sushmita Roy. It’s called Arboretum because you can apply it to a tree structure, such as an evolutionary tree, and infer what happens at branching points in that tree. In our case, those points are ancestors. If we can infer what those ancestors looked like, we can understand how each yeast species evolved.
Q3. What did you find in your recently published research?
DT: We published two papers this spring: one came out in Genome Research. That paper describes Arboretum in detail and applies it to a dataset of stress responses in yeast. We did some stressful things to the yeast like heat shock them and then looked to see how the response of these species changed over the course of evolution. For example, one species we tested was Candida albicans, which lives on us. In most people, it’s harmless, but in immunocompromised people it can be dangerous. In our study we were able to see that parts of the environmental stress response were specific to candida, which might help explain how it turns pathogenic while other yeast don’t.
We also published a big comparative study of 15 different yeast species in eLife. One of the main things we did for that study was look at two types of yeast — respiratory yeast, which produce CO2 and water, and respiro-fermentative yeast, like the yeast used to make bread and alcohol, which produce ethanol. Using Arboretum, we were able to predict that the common ancestor of respiratory and respiro-fermentative yeast was not an ethanol producer, but a respiratory yeast, and we could see how changes in gene expression helped create this new alcohol-producing phenotype.
Q4. eLife is a relatively new publication. It launched in December, as a new approach to scientific publishing. What was your experience like writing for the journal?
DT: eLife was started by a group of funding agencies that saw that publishing papers was becoming really arduous; it would take months and months to get a paper published. What’s cool about eLife is that their review process is supposed to only take four weeks, and instead of returning separate comments from each reviewer, the reviewers talk to each other and produce a composite review that summarizes the main points the researchers need to address before publication. It’s also a completely online journal so length wasn’t an issue (which was nice because we were writing about 15 yeast species), and they encourage you to embed movies and cool graphics to make reading the paper a more interactive experience.
Q5. Do the findings from your eLife paper relate at all to human biology?
DT: Biologists may be able to learn more about cancer by looking at what causes yeast cells to proliferate rapidly. Normal cells don’t divide a lot, whereas cancer cells divide wildly. That cell growth in cancer is similar to yeast reproduction during respiro-fermentation, except instead of making ethanol, cancer cells make lactate. It’s called the Warburg effect. What seems to be happening in both cases is that cells funnel glucose metabolism into making lots of nucleotides, which form the DNA building blocks needed for new cells, and this promotes rapid cell division.
When we traced the evolution of respiro-fermentation in yeast, we found that some of the genes that had been rewired in all respiro-fermentative yeast species were in pathways that fed nucleotide synthesis. Some research on the Warburg effect made connections to genes in the same pathways. I bet that as they look more closely at the orthologs, or human versions, of the genes that we found in our yeast study, they’ll find that they’ve been rewired in cancer cells as well. If you could find a way to block those pathways you could potentially stop cancer cell growth.