Rewriting Life’s Cookbook

The variations that Darwin saw among his pigeons and his barnacles–and which he was at a loss to explain–come into existence as DNA’s sequence changes. Cells can duplicate their DNA almost flawlessly, but every now and then a mistake creeps in. Proofreading proteins can find most of these mistakes and fix them, but a few slip through. Some of these rare changes–known as mutations–may alter a single letter in the recipe of DNA, but others can be far more drastic. Many pieces of DNA can spontaneously cut themselves out of one position and splice themselves back in elsewhere, altering the gene where they make their new home. Sometimes when DNA is being copied as a cell divides, an entire gene can get duplicated, or even an entire set of genes.

As early as the 1920s, scientists began to realize that mutations had huge ramifications for evolution. These researchers–foremost among them the British mathematician Ronald Fisher and the American biologist Sewall Wright–synthesized natural selection and genetics, putting Darwin’s theory on a far more solid foundation.

When DNA mutates, a cell may simply malfunction and die, or it may multiply madly into a tumor. In either case, the mutation disappears when the organism that carries it dies. But if the mutation happens to alter the DNA in an egg or a sperm, it gets a chance at immortality. It may get carried into the genes of an offspring, and that offspring’s offspring. The effects that a mutation has–favorable, unfavorable, or neutral–will influence how common it becomes as generations pass. Many mutations have harmful effects, often killing their owner before it is born or interfering with its ability to reproduce. If a mutation cuts down reproductive success, it will gradually disappear.

But sometimes instead of harm, a mutation does some good. It may change the structure of proteins, making them more efficient at digesting food or breaking down poisons. If a mutation’s effects allow an organism to have more offspring, on average, than the organisms that lack it, it will gradually become more common in a population. (Biologists would say that this mutant has a higher fitness than the others.) As the mutant’s offspring thrive, the mutation they carry becomes more common, and it may do so well that it drives the older version of the gene to extinction. Natural selection, Fisher and Wright showed, was largely a matter of the changing fortunes of different forms of genes.

Fisher made a particularly important breakthrough when he demonstrated that natural selection progresses by the accumulation of many small mutations rather than by a few giant ones. Fisher used some esoteric math to prove this point, but a simple hypothetical example can make it clear. Consider a dragonfly’s wings. If a dragonfly has particularly short wings, it may not be able to generate enough lift to stay off the ground, but if it has wings that are too long, they may be too heavy to flap. Somewhere between short wings and long ones is the length that brings the greatest fitness. If you chart out this relationship between length and fitness on a graph, you draw a hill, with its peak at the optimal length of the wings. If you were to actually measure the wings of dragonflies and plot them on this graph, they might cluster as points near the hill’s peak.

Now imagine that a mutation cropped up that changed the length of the dragonfly’s wings. If the insect’s fitness gets lowered as a result, insects with a better wing design may outcompete its offspring. But if the mutation pushes the dragonfly closer to the peak of fitness, natural selection will favor it. In other words, natural selection tends to push life up the hillside of fitness.

On such a landscape a giant mutational leap might seem like the best strategy for evolving quickly. Instead of natural selection’s gradual creep, it could catapult a dragonfly’s fitness to the top of its hill. But mutations are catapults with no sense of aim. They occur randomly, hurling the dragonflies in random directions across the evolutionary landscape. Rather than landing squarely on the hilltop, they might end up somewhere far away, their wings much too long or much too short. Mutations with small effects, on the other hand, can nudge the dragonflies uphill much more reliably. Even a slight advantage, translating into just a few extra offspring, may allow a mutation to spread through a population after a few dozen generations.

Of course this uphill journey is only a metaphor, and a simplistic one at that. For one thing, the terrain of evolution is not fixed. As the environment changes–as temperatures rise and fall, as competing species invade or retreat, as other genes evolve–hills can become valleys and valleys hills. The landscape is more like the surface of a slowly surging ocean.

Nor does evolution always produce the best possible combination of genes. Genes can, for example, sometimes spread without any help from natural selection whatsoever. Heredity is like a ball on a roulette wheel. If the ball is thrown on the wheel often enough, it will land half the time on red and half the time on black. But if you play only a few spins, your ball might end up landing on red every time. The same thing happens with genes. Say two smooth hybrid peas produce four new plants. Each new plant has a 25 percent chance of inheriting two smooth genes, a 25 percent chance of inheriting two wrinkled ones, and a 50 percent chance of being hybrids. But that doesn’t mean they will turn out to be one smooth pea plant, one purebred wrinkled one, and two hybrids. They might all end up smooth, or even all wrinkled. Every pea plant is a genetic roll of the dice.

These statistical flukes don’t happen in big populations, but small ones can defy Mendel’s probabilities. If a few dozen frogs living on a mountaintop breed only among themselves, a mutant gene may emerge among them and spread without any help of natural selection thanks only to an odd spin of the evolutionary roulette wheel. And once the mutant gene has spread through the entire population, the gene it replaced is gone for good.