Tracking evolution can be tough. Consider the giraffe, an 18-foot herbivore, the tallest animal on earth. The giraffe wasn't always so elongated--it evolved that way over millennia. Relatively long-necked giraffes grazing the African savannas could reach foliage growing on high branches, giving them an edge over their lowlier brethren and a better shot at living and reproducing. Occasionally they produced offspring with even longer necks, who passed on the genetic inclination for still greater height to their children. At least that's the theory. The truth is we don't know for sure, because the changes happened over great stretches of time in animals long gone. Biologists and paleontologists are reduced to inferring the past from the flotsam of the fossil record, a record full of gaps. We can't see precisely how evolution works.
But Gerald Joyce can. In his lab at the Scripps Research Institute in La Jolla, California, Joyce is witnessing nothing less than evolution in action. He's even making a permanent record of the steps along the way, creating a documentary of the evolutionary process. In fact, for a 37-year-old biochemist, Joyce is doing a pretty convincing job of impersonating Mother Nature.
"We say to our players, 'We'll select you if--and only if--you meet the challenges of life in our lab,' " he explains. " 'If you do really well, we'll give you a reward. We'll let you make lots of babies.' "
Joyce's approach mimics the time-honored strategies of Darwinian evolution: selection, amplification, and mutation. Selection is the process by which those organisms best suited to any particular environment (long- necked giraffes, for example) survive to have children--that is, to amplify their numbers. Genetic mutations in their offspring, meanwhile, ensure variety, so that new generations will be subtly different from those preceding them.
There are differences between Joyce and nature, however. One is that of scale. Whereas nature directs a huge and varied cast of characters- -giraffes to ants, whales to worms, flu viruses to humans--the players in Joyce's laboratory drama are all tiny molecules of the genetic material RNA (single strands of ribonucleic acid as opposed to the deoxyribonucleic acid that forms DNA's double helix). Another major difference is that of number: "We can't do nearly as many generations as nature has done," says Joyce. "Nature has had a 4-billion-year head start. So we compensate by having a very, very large population."
The result resembles science fiction: trillions upon trillions of RNA molecules busily evolving into creatures unknown in nature, all under Joyce's demanding tutelage. The ramifications are fascinating. Joyce's test-tube evolution not only lets us look at the genetic maneuverings within the evolutionary process, it suggests a visionary approach to developing medical drugs. Already nearly a dozen biotech firms are trying their hand at "evolving" new medications.
According to Joyce, the seeds of these ideas first came to him way back in 1977 when, as a biology major at the University of Chicago, he read the big, baroque novels of Thomas Pynchon, V. and Gravity's Rainbow. Joyce was struck by a recurrent theme in Pynchon's writing: how random natural destruction is counteracted by human striving toward order. "I found that inspiring and wondered what the mechanism of the counterforce might be," he recalls. "How can order arise out of randomness? One way is through evolution. That's what got me into all of this. Ever since Pynchon, the whole game for me has been to make evolution happen."
In 1982, while Joyce was pursuing dual medical and doctoral degrees at the University of California at San Diego, a discovery a thousand miles away provided him players for his game. Thomas Cech, a biochemist at the University of Colorado, had discovered something surprising about RNA. Generally RNA plays a complementary role to DNA: its prime function in a cell is to carry genetic instructions from the nucleus, the cell's DNA repository, to the ribosomes, the factories that make the proteins that will translate those instructions into action. But in some cases, Cech found, RNA can also act as an enzyme, a type of protein. It can deftly cleave away unnecessary parts of itself before delivering its message. Cech called his newfound RNA a ribozyme.
For Joyce, the discovery was a revelation. This versatile form of RNA was the perfect player for the grand experiment he had in mind. Like DNA, it carried the genetic information needed for its own replication. But unlike DNA--and more like a protein--it also performed a specific task, a task that might be altered through evolution.
It wasn't until 1990 that Joyce could give it a try. By then he had moved to his own lab at Scripps, a white building perched on a perennially sun-drenched hillside overlooking the Pacific. Joyce found Scripps's casual style to his liking, as he did its innovative research. "What Scripps is about," he says, "is the never-never land between biology and chemistry--collaborations between the two just aren't very usual." It was a land he was about to visit. Using Cech's original ribozyme as a template, he produced 10 trillion versions of it, each with slight differences, and geared up to do what he called directed evolution.
First, Joyce had to decide what his selection criterion would be. For giraffes, developing a longer neck was apparently a passport to survival. But now Joyce, not natural selection, was calling the shots. The task he chose had to be attainable given the existing abilities of RNA, but he also wanted it to be useful. "So we took these ribozymes, which know how to cut RNA," says Joyce, "and asked them instead to cut DNA. That's something they don't do in nature. But if they could evolve the new trait, it might be medically useful--for instance, in attacking the genes of harmful viruses."
So Joyce dumped strands of DNA into ribozyme-filled test tubes and sat back to see what would happen. Not much. True to form, most of the molecules ignored the DNA. But a few nonconformists showed interest--they eventually snipped the DNA strands apart. "One in a million RNAs was able to do it," Joyce says. "The cleaving was slow and miserable, so we set them a time limit of one hour, but the fact that they did it even a little bit was encouraging."
The RNA molecules that successfully cut DNA were recognizable by the DNA residue that stuck to them, so Joyce was able to "select" those evolutionary winners, thus satisfying the first of the three evolutionary strategies. Now he was faced with the second: amplifying his few DNA- snipping ribozymes into many. In nature, of course, that's what sex is for: put a male and a female giraffe together and they make more giraffes. But put two RNA molecules together and they just sit there. RNA isn't alive like an organism; it can't reproduce itself without help. Joyce had to lend a hand, with fancy tricks borrowed from the tool kit of molecular biology.
In his role as midwife, Joyce supplied his successful ribozymes with four specialized molecules--two primers that prepared the RNA for copying, and two enzymes that expedited the replication process. In a matter of hours, he was awash in the progeny of his DNA-cleaving ribozymes. "Those one in a million that cut DNA each got to make a million babies," he explains.
But amplification alone, as Joyce well knew, wasn't enough. Producing ribozymes identical to their parents would have given him nothing more than an army of relatively inept DNA cutters. To evolve, these ribozymes had to become better at their job. So, taking another leaf from nature's book, Joyce introduced mutations into his ribozymes' descendants. Only by mutating could the molecules change and have a chance of becoming better DNA snippers.
To achieve this, Joyce purposely used sloppy reproductive enzymes. The enzymes copied the ribozymes, all right, and in huge numbers, but they made mistakes along the way. The results of these mistakes were molecules that were subtly different from the original ribozymes. Some were no more proficient at slicing DNA than their mostly inept forebears; some were even less so. But a few of them, purely by chance, were better at the job. They had begun evolving, just as Joyce had hoped.
Now Joyce repeated the cycle of selection, amplification, and mutation all over again. He took the most successful offspring and made them parents of the next generation, introducing more mutations along the way. And so it went, for generation upon generation. "In each case, the babies didn't turn out the same as their parents," he says. "And in each generation, the babies got better at their job." It was indeed evolution in action.
Two years later, Joyce had completed 27 of these cycles. His ribozymes were now much better at their task; whereas their successful ancestors took an hour to cut DNA, they could do it in less than five minutes. In fact, they'd evolved to cleave DNA almost as well as they did RNA, which is the function they'd evolved naturally over 4 billion years. "For me, that's an important threshold: when they do better at what we want them to do than at what nature wants them to do," says Joyce, who's now continuing the experiment well past the 27-generation mark.
Needless to say, so many generations, amounting to hundreds of trillions of RNAs, are a lot to keep track of. And Joyce, in contrast to nature, doesn't dispose of his intermediary evolutionary stages. "The fun of it is that we put each of these generations in the freezer," he says. "That means we can pluck out individuals from generation 20 and see in detail why they're better than RNAs in, say, generation 10." The detail Joyce refers to is genetic detail. He can track down the evolution of his ribozymes to variations in their building blocks, four recurring chemical bases, or nucleotides, designated by the letters A, C, U, and G. Each ribozyme consists of a string of 393 such nucleotides. When the molecule mutates during replication, certain nucleotides are randomly replaced by others--an A may replace a G, for example. These changes can modify the ribozyme's function.
One of Joyce's insights into evolution is that these mutations actually compete with one another--survival of the fittest within the molecule itself. A striking example involves three nucleotides toward the end of the RNA string. As the generations evolved, a mutation from G to U at position 313 and from A to G at position 314 helped bring about more efficient DNA slicing. But it's only together, Joyce found, that they did any good. (On a graph, the rapid accumulation of these mutations came to resemble two spires that dwarfed the lower frequency of the mutations around them--hence their nickname, the "twin towers.") "In this experiment the twin towers only worked together," says Joyce. "Tower 313 alone is pretty lousy. Tower 314 alone is mediocre. But towers 313 and 314 together are always better than either alone."
There was also a mutation right next to the towers, a G-to-A shift at position 312, that improved slicing. But mutation 312 and the towers didn't get along. "They're mutually exclusive," says Joyce. "If 312 occurs, no twin towers. If the twin towers occur, no 312."
Joyce's description of this genetic competition over 27 generations has the flavor of an announcer calling a tight horse race: "Early on, 312 is leading, but here come the twin towers. The twin towers outcompete 312 and beat it. That's because the value of the twin towers together is greater than that of 312 alone. By the eighteenth generation the situation reverses, and the twin towers are down and 312 is up. But by the twenty-seventh generation the twin towers are once again up, and 312 is down--only tower 313 now displays a different mutation from the one before." He laughs. "That's evolution. Today's loser may turn out to be tomorrow's winner."
And today's loser may disappear for good. Another insight from Joyce's various test-tube experiments--one that strongly mirrors conditions in the larger world--is that mutations that are useful in one context may become extinct when conditions change. Such was the fate of a mutation at position 260, which flourished between the second and eighth generations, waned as other, more beneficial mutations began to accumulate elsewhere, and was history by generation eleven. It's a jungle in there.
This ebb and flow goes on against a background of cooperating and competing mutations all along the RNA string, all of which similarly come and go, influencing the success or demise of other mutations. And this dynamic and interdependent process is going on in just a tiny molecule. One can only imagine the complexity of evolution within whole living organisms.
"Scientists are beginning to see that evolutionary traits are highly dependent on each other," says Joyce. "You can't simply say, 'Here's a gene, it does this. Here's another gene, it does that.' It's more how they interplay with each other--synergistic effects, mutually exclusive effects." That may well have implications for the budding field of gene therapy. "It's not known what adding or removing one gene in a whole system of genes will do," Joyce cautions. "It all depends on how its effect plays out in a network of interactions."
Still, that hasn't stopped Joyce from exploring the use of ribozymes for potential gene therapy. He has, for example, inserted genes for his RNA into cells and seen them produce an RNA army capable of spotting and cutting the DNA of invading viruses. So far he's protected only simple, single-celled bacteria in this way--a far cry from protecting multicellular humans. "It's nothing to write home about yet," he says, "but at least it demonstrates that a ribozyme can potentially prevent infection by DNA viruses."
Meanwhile a number of biotech companies are forging ahead with directed evolution to create new drugs. (One of them, Darwin Molecular Corporation, in Bothell, Washington, uses Joyce as an adviser.) The beauty of this Darwinian approach is that scientists need not foresee every twist and turn of drug design along the way--the molecules themselves do the work. For example, at NeXagen, a biotech firm in Boulder, Colorado, Larry Gold and his colleagues are evolving RNA to recognize a wide variety of substances. Recently they've produced an RNA molecule that targets a growth factor, a type of hormone that's implicated in certain cancers. By grabbing hold of the growth factor and neutralizing it, the molecule may help prevent disease. Gold considers the approach the wave of the future: "You can evolve RNA to do virtually anything. No one knows the limits. Everyone is working feverishly to find out."
Even DNA can be persuaded to evolve within limits. No one has yet reported that DNA can act as an enzyme, as RNA does, but it is capable of interacting with proteins--it binds with "protein switches" that turn a gene on or off. At Gilead Sciences near San Francisco, biologists have produced a DNA molecule that sticks to the blood-clotting protein called thrombin. The result is a promising new anticoagulant--it has already worked in animals--for use during heart surgery. Other biotech companies are trying a variant of directed evolution with proteins or other, smaller organic molecules. These substances can't evolve on their own, of course, so the researchers' methods are a lot more labor-intensive. At Affymax Research Institute in Palo Alto, California, collections of molecules with various structures are assigned a task, such as grabbing a protein in the human body that researchers want to block. The most successful molecules are identified and selected for reproduction. Researchers then chemically copy these molecules, altering their structure here and there in the hope of introducing mutations that will make them perform even better--and start the whole process again. By this kind of trial and error, biotech companies are hoping to devise drugs that will treat a wide range of disorders, from septic shock to inflammation, arthritis, and cancer.
Evolution, though, as Joyce has found out, is full of perverse and delightful surprises. His original ribozymes, recall, required a pair of primers and two enzymes to reproduce. In a recent experiment, Joyce was trying to get his ribozymes to do without one of these reproductive enzymes when a few of his molecules struck out on their own. "All of a sudden something quite a bit smaller than the parent ribozymes popped up," Joyce says. "Through subsequent generations it got a little bigger, then a little smaller, kind of shifting around in size. But each time it got better and better at replicating." He dubbed this free-spirited, pint-size RNA the "minimonster," or, less charitably, "the beast."
The thing Joyce immediately wanted to know was how his wayward new molecule was replicating. Could the minimonster somehow be doing without enzymes and primers altogether? If so, Joyce and his colleagues really had something on their hands. A substance that reproduces itself by itself for its own purposes--that's a pretty good description of being alive. Was their minimonster a brand-new example of what nature had first accomplished all those billions of years ago? Were they seeing the spontaneous flowering of life?
Well, not exactly. "We found out that it did require the two enzymes," says Joyce. "So it wasn't self-sustaining." But although the molecule still needed the enzymes, it eventually dispensed with both the primers that allow the enzymes to go to work. The minimonster had figured out a way to accomplish that part of the reproductive process itself. It wasn't alive, but it sure was heading in that direction.
"This is what it will look like when it happens," says Joyce, imagining the moment of genesis. "Someone will be doing an experiment with molecules that are capable of evolving, doing the kind of useful chemistry we're after. And the molecules will be changing very rapidly, exploring all sorts of possibilities. But instead of developing some clever little mechanism to fake us out, which is what the minimonster did, they'll develop some clever little mechanism to obviate the need for our enzymes, to do without our help altogether."
When that happens, the process of directed evolution will be complete. Whoever pulls it off will have mimicked nature's greatest achievement, life itself. As far as Joyce is concerned, it's only a matter of time.