Comment on this article
Evolution in a Petri Dish
Paul Turner grows deadly viruses—deadly to bacteria, that is—to find out how life evolves.
by Carl Zimmer '87
May/June 2006
Carl Zimmer '87 writes regularly about science for the New York Times. His books include Soul Made Flesh: The Discovery of the Brain and How It Changed the World and Evolution: The Triumph of an Idea.
Charles Darwin never thought he could witness evolutionary change. He relied instead on indirect clues. He looked at its effects after millions of years—in the fossil record and in the similarities and differences among living species. He got clues to the workings of evolution from the work of pigeon breeders, who consciously chose which birds could reproduce and thus created birds with extravagant plumage. But that was artificial selection—not natural selection that had been operating long before humans came on the scene. Darwin was pretty sure that natural selection worked too slowly for him or anyone else to witness.
Darwin got a great many things right, but on this score, he was most definitely wrong. Just ask Paul Turner. In his lab at the department of ecology and evolutionary biology, Turner and his colleagues watch evolution play out in a matter of days. They observe organisms acquire new traits, adapt to new habitats, and become new species in the making.
| |
In a day, a single virus can produce a billion offspring. |
Turner can hold one of these experiments in his hand. It’s a sealed petri dish. “So here we have a lawn of bacteria,” he says, gesturing to a cloudy smear in the dish. Then he points out a large clear spot in the middle of the lawn, where millions of the bacteria have died. “That’s a beautiful example of a plaque,” he says. There’s a tinge of admiration in his voice. The plaque was made by an organism Turner is particularly fond of: a virus known as phi-6. The virus invades bacteria and uses their cellular machinery to make hundreds of copies of itself. The bacteria rupture, and the new viruses escape. As the bacteria die, they leave behind a clear spot.
The plaques are evidence of the virus’s staggering powers of reproduction. In a day, a single virus can produce a billion offspring. From generation to generation their genes mutate, creating opportunities for the viruses to evolve. They evolve so quickly, in fact, that scientists can set up experiments to test ideas about how evolution works. And because viruses carry just a handful of genes, scientists can identify exactly which mutations provide an evolutionary edge. “They’re really one of the few organisms we can study in the lab from nuts to bolts,” says Turner. “We can see the molecular changes.”
Turner has become a leader in a relatively young field: experimental evolution. He is using bacteria and viruses—especially phi-6—to investigate some of the most profound questions about life on Earth. How do new species emerge? Why do so many species reproduce sexually, when they could just clone themselves? Why do organisms evolve into peaceful cooperators in some cases and ruthless competitors in others?
| |
Viruses acquire mutations that allow them to shift from an animal host to humans. |
In many of these experiments, Turner and his colleagues are searching for rules that may govern the evolution of all living things. But the research also has a practical side. While phi-6 infects bacteria, viruses similar to phi-6 like to infect humans—including HIV and influenza. New viruses such as SARS are also now emerging. Their emergence is a case of evolution in action: the viruses acquire mutations that allow them to shift from an animal host to humans. Turner’s research may help scientists better understand how that transition happens. “Within our lifetime we’re going to see more and more viruses shift onto humans,” he says. “What are the next likely pathogens to emerge? That’s something we'd like to predict.”
Turner, now 39, started out with an interest in larger fauna and flora. He grew up in upstate New York, where he loved to wander through the forests. “I was the kid who always liked to go to the zoo,” he says. As an undergraduate at the University of Rochester he thought he might like to be a biologist, and in 1989 he started graduate work at the University of California-Irvine. It was when he got to know Richard Lenski that he made the jump from macroscopic to microscopic.
Lenski had started his career studying beetles in North Carolina forests, but he had been frustrated by how long it took to run experiments to tease out the forces controlling their populations. Many of the same basic forces also govern the existence of microbes, which breed far faster. And microbes are an ideal lab organism—so small and fast-breeding that scientists can run many trials of the same experiment simultaneously to make sure their results are valid.
So Lenski began running experiments on harmless strains of the gut bacteria Escherichia coli. In one series of experiments, he founded 12 colonies from the genetically identical offspring of a single microbe. Each colony was allotted only a meager supply of glucose. Lenski expected that, with food so scarce, natural selection would favor individuals that grew faster than others. He froze samples of bacteria from many generations; he would thaw them out later to compare them with their descendants. The experiment is still running today, some 40,000 generations later. The bacteria in all 12 colonies have evolved to the point where they can reproduce nearly twice as fast as the microbe Lenski started out with.
| |
"Paul had some ideas about the ecological interactions that might take place in rotting meat.” |
Lenski (who now teaches at Michigan State University) interviewed Turner at Irvine when he was a prospective student. “Paul had some ideas about investigating carrion—rotting meat—and some of the interesting ecological interactions that might take place there,” he remembers. “Anyone who was attracted to that system, I figured, must be more interested in the questions themselves than in nature and all its beauty.” Lenski suggested that hauling rotting meat into a laboratory might make for difficult experiments, and described his own work. Turner was immediately interested. “I knew then that Paul was my kind of scientist,” Lenski says.
At first, admits Turner, “the faith you put into working with things you cannot see was a very foreign concept for me.” But he started running experiments on E. coli to track how genes were gained and lost over the generations. It was only when he had almost finished graduate school that he shifted his research down to the even smaller scale of viruses. A scientist named Lin Chao, then at the University of Maryland, gave a talk atIrvine about his research on phi-6. Chao was using the virus to answer a particularly deep and difficult question in biology: why does sex exist?
Although sex is the only way humans naturally reproduce, some other species do well enough without it. Whiptail lizards in the southwestern United States, for example, are all female. Their eggs require no sperm to begin developing into healthy baby lizards; in essence, they just clone themselves. The late great biologist John Maynard Smith once pointed out that sex should put organisms at an evolutionary disadvantage. It takes two individuals to reproduce sexually, but just one to clone. Over a few generations, that difference should allow a population of cloners to become far bigger than one of sexual reproducers. “Our problem is to explain why sex arose, and why it is today so widespread,” Maynard Smith wrote in 1999. “If it is not necessary, why do it?”
Viruses, Chao recognized, would allow scientists to explore this question like never before. Billions of them can reside on a dish, they reproduce quickly, and some, including phi-6, sometimes engage in their own sort of sexual reproduction.
| |
“If sex is not necessary, why do it?” wrote Maynard Smith in 1999. |
When a single phi-6 invades a host cell, it makes clones of itself. Its genetic material is inserted into the host, and the host begins producing copies of the virus’s genes and pieces of the virus’s protective proteinshell. These chunks of genes and shell float around inside the microbe before assembling themselves into new viruses. All the new viruses are clones of the original invader, differing only by whatever mutations emerged as their genes were produced.
If two or more phi-6 viruses invade the same cell at the same time, their host produces new copies of both sets of genes—which can then mix together. The new viruses carry combinations of genes from the original invaders. In other words, the new viruses have two (or more) parents. They are the product of viral sex.
Turner was fascinated by the way Chao was using viruses to study a big evolutionary question. And he was struck by the fact that phi-6 could serve as a model for similar viruses that have sex and infect humans—viruses such as influenza and HIV. “I’m living in southern California, and the AIDS crisis is starting to become a big deal,” he recalls. “It was becoming clear that HIV was doing major damage. It all dovetailed.”
Turner joined Chao’s lab as a postdoctoral student. They proceeded to design a series of experiments to explore the interplay of sex and evolution in phi-6. They created lines of promiscuous viruses and celibate ones. To make celibate viruses, they kept the ratio of viruses to hosts low, so that each microbe was invaded by just one virus. For the promiscuous line, they made sure the viruses outnumbered their hosts, so that each microbe was infected on average by five different viruses. They allowed the viruses to invade new hosts and replicate for many generations. They then measured how quickly the evolved lines of viruses could replicate compared with their ancestors.
Turner and Chao discovered that the promiscuous viruses sped up their replication. Turner suspects that one important factor behind their success was their ability to trade genes. Imagine that two viruses invade a single microbe. One of them carries an inferior gene that slows down its replication, and the other virus is slowed down by a different inferior gene. When they invade their host, they can produce a superior virus by combining their good genes and leaving the bad ones behind.
Viruses may not be alone in benefiting from stripping out bad genes with sex. Other researchers have been examining sexual reproduction in other organisms, and they’ve found similar patterns. Susanne Paland and Michael Lynch at Indiana University recently published a study on water fleas. Some species of water fleas reproduce sexually, while some do not. Paland and Lynch discovered that the asexual water fleas accumulated harmful mutations four times faster than the sexual ones. This sort of genomic hygiene may have played a role in the evolution of our own distant ancestors as well, as they shifted permanently to sexual reproduction.
But the mystery of sex is far from solved. Turner and Chao’s work is proof of that. Viruses that have lots of sex, their experiments revealed, evolve into cheaters. Natural selection favored viruses that could use the proteins made by other viruses in the same cell. By exploiting their neighbors, these cheaters could put more resources into reproducing quickly. “Why do something for yourself, if you can get someone else to do it for you?” says Turner.
| |
Cheating is a classic puzzle of science. |
Cheating is a classic puzzle of science. In 1968, the ecologist Garrett Hardin wrote an influential essay known as “The Tragedy of the Commons.” Hardin asked his readers to picture a pasture open to all the herders in the region. The rational choice for each herder would be to add more animals to his herd. But since all the herders are increasing their herds, they’re making a collective demand on the commons larger than it can support. The herders might try to stave off destruction of the commons by limiting their herds. But this solution can easily come undone, since individuals may still be tempted to cheat. “Ruin is the destination toward which all men rush, each pursuing his own best interest,” Hardin wrote.
The evolutionary parallel might be a species of birds that live on a remote island, eating seeds from a single species of plant. For their long-term survival, it would make sense for the birds not to gorge themselves on the seeds and drive the plant extinct. But natural selection cannot shape instincts to reach some long-term goal. It can only shape the behavior of individuals based on their reproductive success.
Turner and Chao demonstrated this in their virus work when they showed that too much sex may be a bad thing (at least from an evolutionary point of view). Viruses that had evolved with lots of sex, they found, became too good at cheating.When viruses that had adapted to a promiscuous life were forced to reproduce on their own, they reproduced far more slowly.
“It’s a beautiful study,” comments Lenski. “It’s like the tragedy of the commons on a microscopic scale.”
Unlike other life forms, viruses lack the means to reproduce themselves. Viruses typically get into their hosts by latching onto proteins on the surface of cells and managing to gain passage inside. Different species have different proteins on the surface of their cells. If a virus’s key doesn’t fit a species' lock, it cannot make that species a host.
| |
The 1918 Spanish flu pandemic killed 20 million people. |
Turner is intrigued by how that key sometimes changes. A virus may mutate in such a way that allows it to slip into cells of another species. It often takes many of these mutations for a virus to complete such a transition. But it’s clear that viruses do manage to make the transition fairly frequently. Influenza viruses reside in birds and other animals; when a strain evolves the ability to spread quickly from human to human, it can become a pandemic—like the 1918 Spanish flu pandemic, which killed 20 million people before it was over. Scientists are now watching a new strain of bird flu spread across the world, acquiring mutations that allow it to infect humans. It still can’t spread from human to human, but it may be just a few mutations away.
Meanwhile, we are also encountering entirely new diseases thanks to host-shifting viruses. HIV-1 began as a chimpanzee virus. Hunters likely contracted the virus through cuts, and while most of the viruses died off, a few survived. In the 1930s strains of the virus began establishing themselves in humans, and eventually became specialized on our own species. SARS Coronavirus appears to have emerged from civet cats sold in Chinese markets.
It’s just going to get worse, Turner predicts. “As the human population continues to grow, we’re a target. We’re also creating agricultural landscapes where there were wild landscapes. We’re driving native species out into the open, and those native species can be reservoirs for viruses.”
As serious as the threat of emerging viruses is, scientists still know relatively little about how viruses shift hosts. They cannot, for example, confidently predict which viruses in animals are most likely to colonize human hosts in the future. They still need to understand some of the basic rules of this particularly dangerous sort of evolution.
Turner believes that phi-6 can shed light on some of those rules. Like HIV, it has its own host of choice. The strain that Turner studies lives on plant bacteria called Pseudomonas syringae. Turner is carrying out experiments to see which conditions favor its shift to other bacteria. “Right now we’re at the early stages of looking at those questions,” he says.
| |
Scientists still know relatively little about how viruses shift hosts. |
Turner’s graduate student Siobain Duffy has been studying phi-6 to track the earliest stages of host-jumping. Previous research suggested that viruses face some serious challenges in jumping from one host to another. Instead of a clean leap, viruses apparently had to make a gradual shift. Early in a transition, the virus needed to live in both its new host and its old one. Yet being a jack-of-all-trades may not make much evolutionary sense for a virus. A mutation that made a virus able to invade a new species might interfere with its ability to invade its traditional host. Many studies on evolution hinted at such trade-offs.
Duffy began to search for signs of a trade-off. She prepared lawns of bacteria from 14 different types of Pseudomonas. She then added phi-6 to their dishes and allowed the viruses to infect their new hosts for one day. The vast majority of viruses failed miserably at the task. But Duffy identified 30 mutant viruses that succeeded in creating plaques.
Duffy then looked at the genes of the mutants. She focused on a gene that encodes a protein called P3. The virus uses its P3 protein to attach to its host. Since each type of host has different proteins on its surface, it seemed likely that P3 would likely undergo mutations in viruses that could attach to new hosts. She discovered that each of the host-shifters did indeed carry a mutant P3 gene. Remarkably, the mutant genes differed only by a single “letter” from the normal code. That’s all it takes for phi-6 to invade a new host: one random mutation in a single gene could do it. And while the host species Duffy studied were all in the genus Pseudomonas, many are separated by millions of years of evolution.
| |
All it takes for phi-6 to invade a new host is one random mutation in a single gene. |
All told, Duffy identified nine mutations that allowed host-shifting. In some cases only one virus carried a particular mutation; in others, nine shared the same one. To measure the cost of these mutations, Duffy then infected the original Pseudomonas host with viruses carrying each of the nine mutations. When she checked how quickly they reproduced, she found that seven out of the nine mutations caused the viruses to grow more slowly on their original host. The discovery, which she and Turner and their colleagues reported in the journal Genetics earlier this year, marks the first time that scientists have precisely measured the cost of being a jack-of-all-trades.
But the other two mutations go against conventional wisdom: phi-6 strains with these mutations can still grow quickly on their old host. In other words, sometimes a virus can be a jack-of-all-trades for free. Duffy also discovered another unexpected result: some mutations discovered in a virus infecting one new host could also let it infect another new host that it had not yet seen.
Does this mean that we’re vulnerable to any virus with a host-shifting mutation? Not quite, says Turner. On their own, these sorts of mutations are not enough to allow a virus to spread into a new species: “It has to mutate and sustain itself long enough to take off.” Turner and his students have been closely observing one host-shifting strain of phi-6, and they find that it grows ten times slower in the new host than the old one.
Slow growth can put a new strain of virus at risk. If the viruses aren’t producing enough offspring, they might not find new hosts to infect, and the new strain could become extinct. But Turner and his students have now shown that we shouldn’t take too much comfort in that fact.
Turner and postdoctoral researcher John Dennehy are exploring what it takes for a new strain of virus to survive this dangerous passage. Dennehy put a host-shifting strain to the test by forcing it to shift back and forth, between its old and new hosts, four times.
| |
Mutations that allow viruses to infect rodents may give them the ability to infect us. |
The virus proved remarkably resilient. Dennehy and Turner had expected that, in the trials with small founding populations, the virus strain would become extinct in its new host. Instead, it survived and managed to expand its numbers. Somehow, reproducing in the old host gives viruses an extra boost when they infect the new host. Based on what scientists know about viruses, that shouldn’t make a difference. But it does. “And that’s mysterious,” says Turner. “There’s no good reason for that. It’s like an ecological hangover.” It’s possible, he thinks, that the virus grabbed one or more key proteins from the previous host.
These experiments are just the first steps in Turner’s study of host-shifting. Dennehy is now trying to create experiments that mimic natural conditions more accurately. He hopes to create dishes in which different species of bacteria live side by side. He’s curious to see how the viruses “decide” which species to infect. Duffy meanwhile has been allowing host-shifting mutants to evolve. She wants to see whether they can shift completely to a new host—becoming, in other words, a new species.
The research going on in Turner’s lab hints that viruses are even more adept at shifting hosts than previously thought. They may not have to sacrifice their ability to breed in their old host to begin breeding in their new one. And some viruses may not even need to be “trained” on human hosts. Mutations that allow them to infect rodents or other mammals may give them the ability to invade our cells as well. “It may be easy for viruses to enter a new host type even if they haven’t seen that new host type before. If something is hanging out in a mouse and jumps into a human, maybe that shouldn’t be so surprising,” Turner says.
Turner pauses for a moment, noticing that he has caused a visitor some distress at the thought of viruses easily gliding into our species. Witnessing evolution is not always a happy thrill. “Scary,” he says. “Stay healthy. Wash your hands.”
Given the urgency of these risks, Turner finds the continuing debate over creationism versus evolution a dangerous distraction. “I could take somebody into the lab, and over the course of a week, I could prove to them that evolution actually happens in microbes,” says Turner. “And it has been profoundly important in the rise of antibiotic resistance and our inability to make effective anti-HIV drugs. We'd better be aware of it.”  |
