Some lifestyles have been invented repeatedly on different islands—plants, for example, turned woody more than once—whereas others are one-of-a-kind.32 Among these unique lifestyles is the feeding habit of the vampire finch, which pecks the tail of blue-footed boobies to feast on their blood. Or take the remarkable skill of the woodpecker finch, yet another peculiar Galápagos inhabitant. Evolution has taught it to use tools like cactus spines and twigs to help it scare insects from their hideouts inside trees. Another example is the split-jaw snake from Round Island near Mauritius. Its upper jaw is not a rigid bone like ours, but rather it is hinged by a flexible joint, which allows the snake to devour large lizard prey. And there is the marine iguana from the Galápagos, the only lizard alive today that can live and forage in the sea. Among its multiple innovations are glands to extrude excessive sea salt from its body. And the lifestyle of some Hawaiian moth larvae from the genus Eupithecia couldn’t be further from that of harmless leaf-munching caterpillars.33 They are the stuff of horror movies. Disguised as leaves or twigs, they assassinate insects landing near them, snatching their unsuspecting victims lightning-fast with specialized, pincer-like legs.34
All these are a smattering of nature’s myriad creative solutions to the same problem: how to survive and make a living when you get stranded on an island.
Through bursts of innovation, islands teach us that creativity can blossom when competition’s grip is not too powerful. However, they showcase only what’s possible in a short interval of evolutionary time, when genetic drift meets new and empty island environments. That’s because these innovation bursts on islands usually unfold over a mere few million years, a tiny fraction of the four billion years that have passed since life’s origin.
In the unimaginably large time span since that origin, something even more profound than the creation of unusual island faunas has taken place. The power of drift has increased slowly but steadily as evolution has created ever more complex and larger organisms. With this increasing power, genetic drift has transformed the very genetic substrate that enables nature’s creativity: it has fashioned genomes whose architecture is primed for innovation. Here is how.
A hundred square meters can host a population of ten trillion microbes, whereas even a hundred square kilometers—a million times as much space—is barely enough for a handful of large mammals, some forty lions, fifteen tigers, or two polar bears.35 The larger an organism is, the more space it needs. And this also means that larger organisms generally live in smaller populations. As a rule of thumb, bacteria live in groups of some one-hundred million individuals, small invertebrates like insects or some worms cohabit with ten million others, and the populations of vertebrates and trees typically lie below ten thousand members.36 Population sizes vary, of course, within these groups—elephants and mice are both vertebrates, but our planet surely hosts more mice than elephants. However, the general trend—larger organisms, smaller numbers—is clear, and so are its consequences: in the smaller populations of larger organisms, the influence of genetic drift is greater, and that of selection is weaker than in larger populations of smaller organisms. (If our billion-strong population of humans seems an exception to the rule that large organisms live in small populations, it is worth remembering that the human population was below a few million for most of our evolutionary history. What is more, during this time the human population was subdivided into many small, isolated tribes. Our population has exploded so recently that the evolution of our bodies and our genomes has not yet caught up.37)
In the life of large organisms, drift is ten thousand times stronger—and selection that much weaker—than in tiny bacteria with their huge populations. Many bad alleles that would be wiped out quickly in a bacterial population are invisible to natural selection and can persist in large animals or plants. This increasing power of drift in larger and more complex organisms has several surprising consequences. Most important, it has allowed the size of our genomes, the number of its DNA letters, to creep upward, steadily, over many millions of years.
A typical vertebrate genome like ours has three billion letters, almost a thousand times as many as that of a bacterium like E.coli. And while we need more genes to build and maintain our complex bodies, we don’t have that many more genes—less than seven times more than E.coli’s four and a half thousand. In other words, our greater number of genes cannot explain why our genomes are so much larger. Most of the difference in genome size comes, in fact, from the DNA outside of genes. Such DNA is also called non-coding because it does not encode any proteins.
To be sure, the genome of a bacterium like E.coli also harbors a bit of non-coding DNA. Much of it consists of short DNA words necessary for gene regulation. These words are recognized by the protein regulators of transcription, which I first mentioned in Chapter 2. A regulatory protein can latch onto such a word and turn on or shut off a gene’s transcription—the essential prelude to protein synthesis. In other words, most of E.coli’s non-coding DNA has a specific purpose. It helps regulate genes.38 The amount of this DNA is quite modest, covering only about 12 percent of E.coli’s genome.
Our genome couldn’t be more different. It is a vast sea of non-coding DNA in which our genes are but small islands that occupy a mere 3 percent of the genome. Two human genes can be separated by thousands or millions of non-coding letters, whereas two average E.coli genes are only separated by some 120 non-coding letters.39 What is more, only a tiny fraction of human non-coding DNA regulates genes. We do not know yet what—if anything—most of the rest is doing, but as we shall see, it is a giant playground for evolution’s creativity.40 Genetic drift is crucial to understanding where it is coming from.
As a genome is passed on from generation to generation, its size can increase in several ways. One of them is DNA duplication, a kind of DNA mutation that happens no less frequently than the single-letter changes—point mutations—we encountered earlier. It occurs when cells aim to repair damaged DNA and commit a particular kind of error, a bit like an editor who proofreads an electronic manuscript and copy–pastes a paragraph of text by error. These errors are not rare, because DNA constantly gets damaged, and cells thus incessantly edit their DNA.41
The DNA text copied in a duplication may comprise a few letters, thousands of letters, or large parts of a chromosome with millions of letters. It may also comprise one or more genes. In this case, a gene duplication has occurred.
A duplicated gene is exposed to the same constant drizzle of DNA-changing mutations that falls onto the rest of the genome. If it is lucky, one of these mutations will teach it a new trick—perhaps it can help digest a new kind of food molecule or defend the cell against some toxin. But most of these mutations do what mutations usually do: muck things up. They impair or destroy a gene’s ability to make a useful protein. The mutation then has turned the duplicate into a stretch of inert DNA known as a pseudogene, and when a pseudogene is born, a genome’s content of non-coding DNA has grown.
To understand the fate of duplicated DNA in evolution, consider that duplication is not free. It costs energy. That’s the energy a cell needs to manufacture the building blocks of the duplicated DNA. And if that DNA includes one or more genes, it also requires additional energy to decode the DNA’s information and manufacture the encoded protein. In a 2007 study, I used data from complex experimental measurements of this energy to calculate that a gene duplication typically consumes some 0.01 percent of a microbial cell’s energy budget.42 That energy is no longer available for other purposes, such as reproduction.
A hundredth of a percent does not sound like much, and indeed it escapes the notice of our scientific instruments. But natural selection is a more discriminating judge. Recall that microbes live in huge populations, where tiny differences in fitness matter. A microbe hosting such a duplication would be slowly outcompeted by its more efficient brethren. That might take thousands or millions of generations, but eventually and inexorably, its descendants would disappear.
&nbs
p; Most animals and plants with the same duplication would not. They live in smaller populations where drift is stronger and the same kind of difference is invisible to selection.43 As a result, duplicated DNA can accumulate, one by one, millennium after millennium, during the unimaginably long time evolution needs to unfold. The end result? Vast stretches of non-coding DNA. They include some fifteen thousand pseudogenes that litter our genomes, and probably untold thousands more whose features have been washed out beyond recognition in the steady rain of DNA mutations.44
Genes become duplicated passively, through no effort of their own, but another kind of DNA—mobile DNA—aggressively promotes its own duplication. A stretch of such DNA is usually a few thousand DNA letters long, and it encodes one or more proteins with a peculiar talent: the ability to copy and paste their coding DNA to some, usually arbitrary location elsewhere in the genome. From there, this DNA can get copied again and again, ad infinitum.
Mobile DNA is a quintessential example of what Richard Dawkins described in The Selfish Gene.45 Not serving any higher purpose, it multiplies within its host’s genome without regard for the host’s well-being. If nothing were to hold mobile DNA in check, a genome could become overrun by its copies.46
Fortunately, something does keep mobile DNA in check, and that something is natural selection.47
If you were to paste an arbitrary paragraph of a novel’s text into a random new location, chances are that the novel would get worse (unless it was awful to begin with). By the same token, bad stuff also happens when mobile DNA inserts itself haphazardly into a new genomic location. When it gets pasted into another gene, it can disrupt the gene’s information string and damage the instructions to make a useful protein. If that gene is required, for example, by a developing embryo, the result may be the embryo’s death. Also, when mobile DNA gets pasted near a gene, it can inadvertently turn that gene on. That’s because mobile DNA contains regulatory sequences necessary to turn on its own genes for relocation, and this regulatory DNA can activate any gene that happens to be nearby. When this happens to a gene involved in an embryo’s development, and the gene gets turned on at the wrong time and place, development can be altered dramatically or subtly—two nerve cells do not connect properly, a blood vessel does not form in the right place, or a bone is not quite as strong as it needs to be. Indeed, subtle changes are more frequent than dramatic ones, and so the damage is often slight, reducing the host’s fitness by less than 1 percent.48 And while natural selection will quickly eliminate any insertions that wreak havoc, selection alone does not decide the fate of insertions with subtle effects. Genetic drift may also have a word to say.
In organisms with huge populations, like E.coli, the influence of drift is weak, and selection has no trouble wiping out most damaging mobile DNA insertions. That is why most microbial genomes contain little mobile DNA—usually 1 percent or less of their genomes.49 In large organisms, however, mobile DNA can steadily accrue because populations are too small, genetic drift is too strong, and selection is too weak to weed out mobile DNA with subtle effects. The end result: more than 50 percent of our genome—and that of other large animals and plants—has mobile origins.50 Our genomes contain millions of copies of mobile DNA.51
And while mobile DNA has been steadily accumulating in our genome, it was exposed to the same rain of mutations as all the rest of our DNA. Because these mutations can destroy its ability to copy and move over time, the vast majority of our mobile DNA—more than 99 percent—is crippled and inert.52 It can no longer relocate, and its genes have been washed out to become pseudogenes. Even when crippled, though, it still contributes to the ocean of non-coding DNA in our genomes.
The upshot is that larger organisms generally have more complex genomes, courtesy of genetic drift. As organisms get larger, they live in smaller populations—selection becomes weaker and genetic drift stronger—and their genomes acquire more genes, more duplicate genes, more pseudogenes, more active mobile DNA, more defective mobile DNA, and overall more non-coding DNA, so much of it that their genomes’ size has steadily increased more than a thousand-fold.
Indiana University biologist Michael Lynch was among the first to show that genetic drift is important to explaining our ever-increasing genome complexity. Lynch made his case by comparing the genomes of hundreds of different organisms, and his data reveals not only that drift helps increase genome size, but also that drift increases the complexity of individual genes.53
When the DNA of some genes is transcribed into RNA, parts of these genes, known as introns, are eliminated, while other parts, known as exons, are joined or spliced together. Only the joined exons are translated into proteins. In other words, genes can come in pieces that are assembled only when the genes’ information is decoded.54 During life’s ascent, the number of these pieces per gene rose steadily, just like genome size did. Whereas a typical gene in a microbe comes in one or two pieces, the genes of mice and men have more than seven.55 What is more, the size of the discarded intronic DNA steadily increased, such that in our genomes, more than 98 percent of a gene’s transcribed DNA is discarded, and less than 2 percent is translated into protein.56
The result of all this genomic complexity is a giant playground for evolution’s creativity. The more pieces a gene has, the more kinds of proteins can be created by mixing and matching these pieces in new ways. Our genes come in so many pieces that our body can make fifty thousand more proteins than a fruit fly can, even though we have fewer than twice as many genes.57
In addition, because typical vertebrate genes are separated by thousands or millions of non-coding DNA letters, random mutations stand a far greater chance of creating new DNA words that can be bound by regulator proteins and bring about new gene regulation in genomes like ours than they do in E.coli, with its minuscule amounts of non-coding DNA.58 That’s why changes in gene regulation have been crucial to the evolution of large and complex organisms—perhaps more so than changes in genes themselves. Most genetic differences between humans and chimpanzees, for example, are changes in non-coding DNA that can alter gene regulation.59 Such changes modulate life’s recipe in ways that appear subtle but that can have effects as dramatic as creating a new species capable of symbolic language, art, and literature.
Unlike the genome of a microbe, which is like the spartan, barely furnished cell of a monk, the genome of a multicellular organism resembles the workshop of an inventor, filled to the rafters with spare parts, tools, abandoned projects, disassembled machines, and half-finished designs—in short, the kind of junk that is the seed of the next breakthrough invention. And central to filling this workshop with useful parts is genetic drift. As long as our inventor can keep natural selection—an overzealous janitor—in check, he can keep on tinkering.60
In sum, genetic drift affects life’s evolution in two fundamental ways. In the short run—the few million years needed to form new species—it helps evolution attain new and higher peaks in nature’s genetic landscapes. In doing so, it accelerates the creation of new species with unique lifestyles. And in the long run, drift alters the architecture of genomes and increases their potential for future innovation.
As different as these manifestations of genetic drift are, they share a common principle: good things can happen when evolution is free to explore the landscapes of nature’s creativity, temporarily liberated from selection’s shortsighted and relentless uphill drive.
As it turns out, genetic drift is not the only help evolution has in traversing these landscapes.
Chapter 4
Teleportation in Genetic Landscapes
As difficult as it is to imagine traversing a mountain range in four or more dimensions, I can think of something that is even more difficult: traversing a mountain range in only two. The problem in two dimensions, however, is not our lack of imagination. Darwinian evolution would be much harder in two dimensions because a flat mountain range is more difficult to traverse than one in three or more dimensions—far more difficult.
Imagine that the upper part of Figure 4.1 shows a fitness landscape in a flat world. A population that lives on the left plateau and needs to get to the right plateau would face an already familiar problem: natural selection would prevent the population from traversing the valley separating the plateaus because it forbids even a single downhill step. But what if that figure were really a two-dimensional slice through a three-dimensional landscape, the profile of a crater like that shown in the lower part of Figure 4.1, shaped, perhaps, by a meteor strike? The two plateaus in the flat plane would then become a single three-dimensional plateau, really a circular ridgeline, undulating in altitude, perhaps, but easily circumnavigated by a population propelled only by the modest amount of genetic drift present even in large populations. No need for the strong drift—and tiny population—required to go through the crater’s bottom.
Although our three-dimensional brains cannot visualize it, the same idea works in higher dimensions. If the Rocky Mountains, or the Alps, or the Andes were just a three-dimensional profile of a mountain range in four dimensions, some of their peaks—separated by deep three-dimensional valleys—could be connected by a ridge in the fourth dimension. And some of those peaks that were still inaccessible in four dimensions could be accessed via an easy stroll in five dimensions, leaving fewer inaccessible peaks, which could become accessible in a sixth dimension, and so on.
Life Finds a Way Page 7
