The 16,000 Generation Yeast
On the fourth floor of Northwest Labs lives ye olde yeast colony, propagating for over 16 thousand generations. Neither a medieval fungus problem nor an ancient sourdough factory, the yeast has been carefully cultivated by scientists for over ten years in one of the longest-running experiments of its kind.
Saccharomyces cerevisiae, commonly known as baker’s or brewer’s yeast, grows quickly and easily in the lab, making it a common research tool. It’s so popular among scientists that yeast boasts a whopping five Nobel Prizes in the 21st century alone. But it has never been continuously grown for so long.
Michael M. Desai, professor of Organismic and Evolutionary Biology and of Physics, hopes to leverage the common fungus to answer an age-old question: How did we come to be? Is it inevitable that life on Earth took this form?
Most biologists approach the search for life’s raison d’être by studying fossils and comparing them to modern organisms, a process that is primarily observational: Let’s look at what happened in evolutionary history and reason about why.
For Desai, this approach is insufficient. “We don’t have many replicates. We can only see what happened in the one example that we have in nature,” he says.
Desai imagined that growing an organism in the lab for a long, long time could be a stronger approach to studying evolution. He could multiply yeast in a myriad of controlled microcosms — in effect running evolution on repeat — and compare results to determine which adaptations are inevitable and which occur by chance.
“Here in the lab, we can repeat the experiment. We can see what happens if you do it again,” he explains. “We can manipulate things in ways that are impossible in nature.”
The approach is based on a similar experiment, started in 1988 at the University of California, Irvine, called the Long-Term Evolution Experiment. Professor Richard E. Lenski and collaborators have been growing a non-pathogenic strain of E. coli, the bacteria best known for causing stomach bugs, in 12 different flask environments—and have now surpassed 75 thousand generations.
“It’s been so successful that people have drawn really broad conclusions about how evolution works from that,” Desai says.
These conclusions should be extrapolated cautiously, he adds.
“It’s one strain of one particular species of a bacterium in one particular laboratory environment, and they only have a dozen replicant lines,” he says. “There have been a lot of surprising things that have come up, but we don’t really know how general they are.”
This mystery has followed Desai since he was a graduate student focused on theoretical physics.
“I started to get fascinated by evolutionary problems,” he recalls. “It seems like it should be completely unpredictable, but somehow, collectively, you get statistically predictable behavior at certain levels.”
Hoping to build on the LTEE’s findings, Desai began ideating his own experiment with yeast. Unlike the E. coli bacteria used by the LTEE, yeast is a eukaryotic organism, meaning it organizes its DNA in a way that more closely approximates animal and human cells.
On a fateful day in April 2015, identical yeast cells were carefully pipetted into each spot on a 96-well-plate, a plastic array of miniature test tube environments that each correspond to a distinct evolutionary history-to-be. The clock started ticking.
Every week, members of the lab group take turns babysitting the yeast: bringing the plates from temperature controlled rooms to a robot that carefully dilutes the concoction and adds more nutrients. It’s like feeding a sourdough starter in a choreographed 30-minute routine.
“This is really a team effort,” says Shreyas Pai, a graduate student in Systems, Synthetic, and Quantitative Biology.
Each day, the yeast multiplies ten generations, and at the end of the week, the progress is saved and stored at minus 80 degrees Celsius, creating a record every 70 generations.
“When you look at populations in the wild, you look for fossils to fill in missing links in evolution,” Pai says. “We have the complete record right here in the freezer.”
The lab group dubbed the project the Very Long-Term Evolution Experiment in a jesting nod to the inspirational work of Lenski and the LTEE — and hedged a bet.
“We do 10 generations a day. They do around six generations a day,” Pai says. “In principle, sometime in the future, like 20 or 30 years from now — you could do the calculation — we should be able to overtake them.”
Wilting balloons hang listlessly in the lab office, left over from the lab group’s celebration of the 15,000 milestone just a couple months ago.
It’s hard to fathom what our own ancestors were like so many generations ago. Scientists estimate that humans started farming only a couple hundred generations ago. About 1,500 generations ago, Homo sapiens roamed the planet alongside our Neanderthal cousins, and the common ancestor of us and chimpanzees likely lived 250,000 human generations ago.
“We can simulate hundreds of millions of years in a matter of months or years,” Pai says. “We can look at long evolutionary processes in the lab just in the space of a PhD or an academic career.”
The experiment currently has no end-date.
“We want to keep this experiment going in perpetuity, see if we get to 80,000, 100,000, or 150,000 generations,” Pai says. “We’re trying to understand long-term evolutionary dynamics. Are things repeatable? If you replay the tape of life many times, would it look the same?”
By sequencing DNA from different generations of yeast, Desai and his collaborators are starting to find answers.
Though each sample represents a different evolutionary lineage, the yeast’s outward characteristics have changed in remarkably similar ways. But simultaneously, their DNA codes have diverged. “Somehow, they’re doing similar phenotypic things in very different genetic ways,” Desai says.
Both the LTEE and the VLTEE have also challenged the notion that evolution is inherently slow.
“People have this intuition that change is slow, that it happens over millions of years, and that beneficial mutations are really rare events. One of the things we’ve learned from experimental evolution is that evolution can happen rapidly,” Desai says. “There’s lots and lots of mutations happening in large populations all the time.”
These compounding mutations are not always beneficial to the organism.
“Because there’s so much going on, they can’t optimize everything,” Desai says. “There’s all kinds of trade offs, like hitchhiking of bad things along with the good mutations.”
Taken together, these findings reshape our understanding of evolution. The simplicity of the experiment’s design coupled with the power of modern genetic tools and the dedication of a lab group over more than a decade strong make the VLTEE a testament to what’s possible with experimental evolution.
“This struck me as spectacular because you could manipulate the course of evolution itself and then see what happened,” Pai says. “You’re in some sense playing God.”
— Magazine writer Dina R. Zeldin can be reached at dina.zeldin@thecrimson.com.