In Praise of Yeast

We do a pretty good job at appreciating the visible intricacies of nature: the antennae and legs and claws of a lobster, the geometrical order of the spots on a butterfly's wings. But a lot of nature's intricacies are hidden away inside single-celled creatures, such as the baker's yeast that makes bread rise and beer ferment. At an audition for a David Attenborough documentary, a yeast cell guzzling away on sugar is bound to do a lousy job. ("Thanks, don't call us; we'll call you. Send in the King Cobra!") But the intricacy of its metabolism is no less impressive. What's more, scientists know how to manipulate yeast in ways they can't with animals, and that power lets them set up experiments that yield clues to how that intricacy evolved.

The latest study of yeast's intricacy comes from the University of Wisconsin lab of Sean Carroll. Carroll has become the public's go-to guy for evo-devo, or the evolution of development, thanks to his book Endless Forms Most Beautiful. Carroll and his colleagues have carried out path-breaking experiments that reveal how relatively small changes in DNA can lead to dramatic changes in how animals grow into adults. A key point of Carroll's work, as well as that of many other evo-devo researchers, is that evolution is not just about the mutations that alter the way proteins work. The genes that encode those proteins are controlled by intricate switches, which determine where and when they make proteins. Change those switches, and you can change how an animal develops. For example, there's a circuit of genes that specifies the coordinates of a insect's overall body plan. Carroll and his colleagues have demonstrated that this same mapping system was borrowed to determine where spots go on butterfly wings.

Recently Carroll has been moving away from lovely butterflies and other insects, to the less lovely yeast. But many of the same principles are at work in yeast too. Instead of spots and wings, yeast use genetic circuits to control how they eat. One of the best studied of all genetic circuits in the world is the one yeast uses to feed on a sugar called galactose. Yeast can break down galactose by making a group of special proteins. But it usually doesn't make them. That's because galactose is not a particularly good sugar to eat. Give yeast a choice between high-energy glucose and galactose, and it will stick with the glucose. If there's no glucose, however, it will switch over to galactose. It uses a series of proteins to break galactose down and extract energy from it. The first protein in this pathway adds a cluster of phosphorus and oxygen atoms to the sugar. It is made by a gene called GAL1.

As with most genetic networks, we're going to be dealing with a bunch of genes with painfully unmemorable names. From here on out, I'll give them nicknames. Here are the Dramatis Personae in this Biochemical Melodrama, in order of appearance:

GAL1: The Digester
GAL4: The Activator
GAL80: The Repressor
GAL3: The Sensor

So why doesn't yeast make The Digester all the time? Because the only way to make it is with the help of The Activator. The Activator clamps onto the DNA near The Digester gene and attracts proteins to the gene that can make a copy of the gene that can then be used to build The Digester protein. But normally The Activator can't do its job. That's because The Repressor grips the Activator and gets in the way.

Yeast needs to receive certain signals in order to start making the Digester. The most important of those signals is galactose itself. The sugar causes Gal3 (the Sensor) to grab the Repressor. Now the Repressor can no longer clamp onto the Activator.
Once liberated, the Activator can help switch on the production of the Digester.

Elegant, you may be thinking. Astonishing, even. Yeast can respond with exquisite precision to its changing environment, retooling itself to make the best use of the food it can find. Remove a part of this genetic circuit, and the yeast is crippled. It may still be able to eke out a living on galactose, but it will be quickly outcompeted by yeast with intact circuits. So how did this elegant circuit evolve?

An important clue comes from a closely related species of fungus known as Kluyveromyces lactis. There's good reason to think that galactose-feeding system is a lot like that of the distant ancestors of today's yeast.

As its name suggests, K. lactis feeds on lactose, the sugar in milk. When lactose is split into pieces, one of the pieces is galactose. Like baker's yeast, K. lactis can also switch to feeding on galactose. It uses a nearly identical set of genes, too. But there's one remarkable difference. Instead of two genes for The Sensor and The Digester, it only has one. The same gene acts as a Sensor--responding to galactose by pinning down The Repressor--and the Digester--taking the first step in breaking down galactose.

It may come as a surprise that a gene can do such different jobs. But the fact is that many genes do. That's because the proteins they encode are big, messy molecules that can interact with other molecules in a number of ways. Sometimes a protein is equally good at two jobs; other times, it's good at one and mediocre at another. Sometimes its potential to do a second job actually goes unappreciated. In a recent experiment, for example, scientists studying E. coli found that if they disabled an essential gene, they could sometimes keep the microbe from dying by adding extra copies of another gene that did a completely different job.

So how did the ancestor of yeast go from The Sensor-Digester to having The Sensor and The Digester? The answer begins with the duplication of genes. About 100 million years ago, the entire genome of the ancestors of baker's yeast became duplicated by accident. Some Most extra genes later disappeared thanks to mutations that snipped them out. Others continued doing their old job. And others changed their line of work. The Digester and The Sensor in yeast are actually 74% identical, one sign that they descend from a common ancestor. But how did the initially identical genes evolve into two such specialized descendants?

Carroll and Chris Hittinger (his former grad student, now at Washington University) ran a series of experiments to investigate how this transition took place. They replaced parts of the Sensor with parts of the Digester and vice-versa. They also swapped parts with the Sensor-Digester gene from K. lactis. Then they observed how well the engineered yeast could reproduce on a diet of galactose compared to normal yeast. From these tests, they reconstructed the molecular details of this transition.

A key change, they conclude, took place not in the protein-coding parts of genes. Instead, it took place near the genes, where other proteins can bind to the DNA and switch the genes on or off. (These regions are called Promoters.) At first, both copies of the Sensor-Digester gene could only be switched on and off in the same way, since they had the same binding sites. But then the binding sites began to evolve. One copy lost the sites where the Activator once attached. Now it could no longer be switched on to act like a Digester. At some point, it also lost the ability to transform galactose at all. All that was left for it to do was to respond to galactose by grabbing the Repressor.

Meanwhile, the other copy of the gene evolved as well. The binding sites changed to make the gene switch on and off much more efficiently. The Repressor could stop the production of the Digester almost completely. And when the Repressor was removed, the Activators could drive the production up a thousand-fold. It changed from a mediocre Sensor-Digester to a fabulous Digester. (See the illustration at the end of the post for a visual summary.)

What's particularly striking about this transition is just how unimpressive it shows the ancestors of yeast to have been. The promoter for yeast's Digester actually makes K. lactis to a better job of growing on galactose. So why didn't K. lactis ever evolve a better promoter? Because the same gene also has to act as the Sensor. The switches for one function conflict with the other. Once the one gene became two, each could specialize in one function without harming the other.

The lesson from yeast, Hittinger and Carroll suggest, may apply to animals and other organisms. Give identical genes different marching orders, and they can change from mediocre jacks-of-all-trades to exquisite specialists.

[Update: I've fixed some parts thanks to Hittinger's responses to the post.]