This skill is essential because many chemical reactions inside an organism require energy. Much of this energy comes from molecules that store energy in the chemical bonds linking their atoms. But to use it, an enzyme—protein or RNA—must first attach to an energy-rich molecule so that it can harvest this energy. Hence the question these researchers asked: Which among these trillions of short RNA molecules can perform this first step of energy harvesting?
In their experiments, the researchers used an energy-rich molecule called guanosine triphosphate, or GTP, that is recognized by molecules in every living organism, and they found not just a few but thousands of such GTP-binding RNA molecules.19 What’s important is that these RNA molecules did not form a single peak in the fitness landscape of energy-harvesting molecules. Instead, they clustered into fifteen different fitness peaks that differed in height—the higher the peak, the greater an RNA’s attraction to GTP. What is more, the peaks were scattered far and wide throughout the landscape. A population of molecules stuck at a low peak could be stuck there forever.
Not all creations of biological evolution rely on new molecules like the energy-hungry RNAs or the antibiotics-devouring beta-lactamases that I described. Some require nothing but changes in the place and time where an old molecule is made. That’s because the unfolding of new life in a developing embryo follows a program, a recipe a bit like that in a cookbook, but sophisticated beyond belief, requiring thousands of protein ingredients and exquisite timing when adding these ingredients to the simmering stew. By changing only the when and where of adding these ingredients, evolution can bring forth entire new life-forms—four-legged animals from fish, feathered birds from dinosaurs, and so on. Here is how.
Our bodies harbor trillions of cells, but only a few hundred different kinds of cells, like those that transmit electrical signals in your brain, contract muscles in your arms, or transport oxygen in your blood. Each cell type harbors different molecules—many of them proteins—that are unique to it, much like a fingerprint is unique to a person. In other words, each cell transcribes and translates into protein only some of the twenty thousand genes of our genome. Some genes are only turned on in the liver, others only in the brain, yet others only in muscle, and so on. Where, when, and how often genes are transcribed and translated is regulated by specialized proteins known as transcriptional regulators. These cooperate with the biochemical machinery that transcribes a gene into RNA. The details of this cooperation are complex, but the basic principle is simple: to make their influence felt, these regulators need to be close to where the biochemical machinery starts transcribing, which is at the gene’s beginning. They achieve this with a very simple mechanism. Each such regulator protein can recognize and latch onto short DNA words with specific sequences of letters, like CATGTGTA or AGCCGGCT, and if such a word occurs near a gene, the gene’s transcription gets turned up or down. What is more, many of the thousands of genes in our genome contain DNA words recognized by the same regulator. In this way, one regulator can regulate a multitude of genes.
When our bodies grow and develop from a fertilized egg, these regulators—hundreds of them—are like cooks that follow the immensely complex recipe needed to build a body. They ensure that thousands of genes are turned on to just the right level, helping to manufacture the right proteins in the right amount. Nothing illustrates the importance of these regulators better than the birth defects that happen when they fail to follow this recipe to the letter. A minor glitch will create a mild birth defect like a cleft lip or fused fingers, while major deviations will lead to serious defects like malformed hearts or even to lethal ones where a mangled body dies before it is born.
Because gene regulation helps sculpt all multicellular organisms, from primitive jellyfish to complex primates, from microscopic algae to gargantuan redwood trees, a new kind of body—or even just a new body part—requires new regulation. The tubular bodies of snakes are braced by a grotesquely elongated thorax containing hundreds of ribs, the long legs of horses are supported by a massively enlarged third toe that helps them outrun predators, and in some orchids a simple whorl of petals is transformed into an elaborate lure whose resemblance to a female insect is uncanny, attracting male pollinators on the prowl for females. All these products of life’s creativity require altered regulatory recipes that guide genes to making a bit more or less of their proteins a bit earlier or later, subtle alterations that manipulate the numerous ingredients needed to create new life.
Evolution can manipulate such recipes easily because any one regulator recognizes not just one but hundreds of different DNA words. It binds to some of them tightly, stays on the DNA for a long time, and turns a nearby gene on strongly, like a volume dial cranked up to the max. Others it binds to loosely, falling off soon after it has latched on, such that the gene is barely turned on, the volume barely audible. Single-letter changes in one of these words can modulate a regulator’s binding and fine-tune how often a gene is transcribed. Many such small changes can add up to a new body architecture, a new creative product of evolution.
Together, all these DNA words—each a special kind of genotype—form a landscape of gene regulation. But unlike some of the hopelessly vast landscapes we have encountered, we can map this landscape completely because the DNA words bound by most regulators are short, usually fewer than a dozen letters. For example, compared to the astronomical number of proteins with one hundred amino acids—more than 10100 texts—there are only sixteen million twelve-letter DNA words.
Such smaller numbers are a relief for those of us who study molecular landscapes because we have a technology that allows us to measure how strongly a regulator binds to every single DNA molecule in such a landscape. It is known as microarray technology or DNA chip technology. Like the chips in a computer, which perform many simple calculations simultaneously, a microarray allows scientists to perform many measurements in one go. Think of a microarray as a rectangular grid with as many locations—grid points—as there are DNA words to be studied. Each location harbors many copies of a DNA molecule with one specific letter sequence. When bathing the chip in a solution containing a regulator protein, the regulator will bind to some DNA words, and the strength of this binding can be measured on each grid point.20 In short, a single DNA chip experiment can map an entire fitness landscape. If an orchid’s flower is maximally seductive, if a fruit fly’s wing creates maximum lift, or if a horse’s leg provides optimal support only if some regulator turns a specific gene on to the max, then the peaks of this landscape are the regulator’s most tightly bound words. And because microarray technology makes it so easy to map an entire landscape, it has been used to map the landscapes of not just one or a few, but more than a thousand regulators, from organisms as different as plants, fungi, and mice.21
Such microarray data allowed Joshua Payne and José Aguilar Rodríguez, two researchers in my Zurich laboratory, to ask a question that will sound familiar by now: How many peaks do these landscapes have?22
The answer mirrors what we learned from other adaptive landscapes: the landscapes of gene regulation can be somewhat rugged but not impossibly so. Many of them even have only a single peak—easy to conquer by natural selection—but others have dozens of peaks with different heights. The peaks correspond to different DNA words that are able to turn on genes to varying extents—some more, some less—but from each peak, higher peaks cannot be reached by walking only uphill. The landscapes on which evolution can explore new body architectures are no different from those that brought forth its other creative products, including proteins that disarm new antibiotics and RNA enzymes that can splice new RNA strings.
Biology has come a long way since Sewall Wright’s day, when he could only speculate on the topography of evolution’s landscapes. His hazy guesswork has been supplanted by high-resolution maps containing the finest molecular details, like satellite imagery that can resolve single grains of sand. But even more important than these details is an idea that is central to any science o
f creation. Wright was unaware of its generality, but we will encounter it again and again in later chapters: the difficulty of a problem can be encapsulated in the topography of its landscape. Single-peaked smooth landscapes correspond to easy problems. Their peaks—harboring the single best solution—can be conquered through exclusively uphill steps. Multipeaked landscapes correspond to harder problems. The more peaks, the harder the problem. The hardest problems need the most creative solutions, and a big part of finding these solutions is getting off of dead-end peaks and finding higher ones.
Every example of a fitness landscape so far in this book has revolved around a problem that nature solved—from efficient swimming by bulky ammonites to disarming novel antibiotics unleashed on bacteria. For easy problems, a straight uphill march will do, but unlike von Helmholtz, who could retrace his steps from a dead-end peak, populations driven uphill by natural selection cannot. This is an important lesson for those who call natural selection all powerful: selection’s relentless uphill drive—like that of a hypercompetitive human who always strives for the faster, better, superior—is a fatal impediment to solving truly difficult problems. Selection and competition alone are impotent to solve such problems.
Which leaves us with a crucial question: How does nature get off those dead-end peaks?
Chapter 3
On the Importance of Going Through Hell
In 1922 British archeologist Howard Carter discovered the tomb of King Tutankhamun, where he found 130 walking canes, even though King Tut died at only nineteen years old. Somebody had clearly thought that King Tut could use these canes in the afterlife, but it would take almost a century to find out why. That’s when a research team led by the Egyptian archeologist Zahi Hawass discovered through a CT scan that King Tut had suffered from various deformities, including a left club foot and a missing toe on the right foot, as well as a cleft palate and signs of a hereditary bone disease. Tut could really have used those canes. Historical records and DNA testing show that incestuous marriages were rampant in the bloodline leading up to King Tut. Tut’s parents, for example, were brother and sister. Further DNA work also resolved another mystery: the identity of two stillborn fetuses interred in King Tut’s tomb. They turned out to be Tut’s children—failed attempts to continue the royal bloodline.1
Perhaps King Tut’s was the first royal bloodline that succumbed to the dangers of inbreeding, but it certainly was not the last. Three thousand years later, a similar fate befell the European Habsburg dynasty.
The Habsburgs were an ugly bunch. Walk into any major European museum of art, and you will recognize their portraits without reading a label. What gives them away is that oddly jutting lower lip, well known as the Habsburg lip.
Their misfortune, it turns out, was more than cosmetic, an ominous sign of much deeper maladies. The Habsburg lip resulted from mandibular prognathism, a protruding lower jaw so large that the lower and upper teeth no longer align. This deformity itself emerged from six centuries of incestuous intermarriage among a small circle of royal families. Their marriages helped forge political alliances, prevent wars, and gain new territories. Long before twentieth-century geneticists would explain why inbreeding has disastrous consequences, these and other aristocratic families experienced its effects firsthand. In the Spanish Habsburg line alone, nine of eleven marriages, from King Philip I in the fifteenth century to Charles II in the seventeenth, were consummated between cousins or between a man and his niece. Even an observer completely ignorant of genetics would have known that something was amiss: childhood mortalities in the royal family were as high as 50 percent, more than twice the level found among ordinary Spaniards.2
At the end of this multigenerational experiment stood King Charles II, who had much more wrong with him than just his lower lip. He was barely a notch above a drooling idiot. Unable to speak until age four, unable to walk until age eight, he was short, weak, and thin, with a tongue so large his speech was difficult to understand and a lower jaw that protruded so far that he could not chew. He was uninterested in the world around him and, worst of all, was plagued by a disease fatal for any royal bloodline: impotence. And so with him the Spanish Habsburg line went extinct.3
Deformities of body and mind are among the multiple consequences of inbreeding, but don’t be misled into thinking that inbreeding is all terrible. It can also be a force for good. Animal breeders, for example, rely on it to enhance coveted features of livestock or pets. That’s because inbreeding—unlike natural selection—is indifferent to whether a feature is good or bad. It merely brings out extreme features. A cattle breeder may breed a single Texas Longhorn bull that has exceptionally long horns with multiple females. All of their offspring are half-brothers or half-sisters, and some of them may have horns just as long as their dad. The breeder might mate those exceptional animals with each other and cull the rest. Continue this selective breeding for several generations, and the exceptional horns of the founding father can become ordinary in his distant descendants.
Selective inbreeding like this—among livestock, pets, or plants—is a time-honored procedure. But as the Habsburgs and King Tut show, it also has unintended consequences. These consequences can be understood from basic genetics. Let me explain.
The genome of a bull has two copies each of its more than twenty thousand genes, one from its mother, the other from its father. Because mutations constantly sprinkle any genome with DNA errors, the DNA sequences of the two copies will differ for many of these genes, and in some genes a mutation may have damaged one of the copies. That’s usually no problem, as long as the other, intact copy is around. But when two damaged copies come together in the same genome, the result is a genetic disease.4 This is usually rare. It can happen, for example, if the bull had already inherited a damaged copy from its parents and the other copy suffers a mutation during the bull’s lifetime. But it becomes frequent when the descendants of the same family breed repeatedly. Each of the bull’s offspring—all of them half-siblings—has a 50 percent chance of inheriting a damaged copy of the same gene from its father, and if these half-siblings themselves sire offspring with each other, the chances that their offspring get the disease are large—25 percent, to be precise, as geneticists can calculate. Breeders use various tricks to reduce the number of diseased animals, such as periodically outbreeding some individuals or culling the sickest animals, but they cannot completely avoid disease.
Because there are so many genes that could be damaged in any one family line, inbreeding brings out different defects in different lineages. Some of them may be mild, such as a missing tail switch in cattle. (It is considered an aesthetic flaw by cattle breeders, but to a cow it surely would be a handicap in keeping those pesky blood-sucking insects away.) Others are less innocuous, like shorter fur in cattle. It means poor thermal insulation in the winter, resulting in calves that gain weight more slowly—not exactly a desirable trait in beef cattle. Yet other defects are disastrous. They include premature death and low fertility, and neither of these is rare, just like in those royals.
Because inbreeding indiscriminately affects both good and bad traits, various races of cattle, dogs, cats, and other domestic animals are defined as much by their flaws—often hidden—as by their strengths.
German shepherds frequently suffer from a malformed hip joint, whose ball does not fit tightly into the joint’s socket. As a result, many of these beautiful animals walk in pain, labor to climb stairs, strain to get up, or become lame as they age.
Even worse off are American Burmese cats. They are a gorgeous breed of cat with wide-set and large, vaguely childlike and curious eyes. Sadly, some of them also suffer from a horrific and lethal deformity called the Burmese head fault, where some kittens develop two upper muzzles, and the top of the head is incompletely formed. Their misfortune can be traced to a single stud cat that sired many prizewinning offspring, but also introduced the deformity to the breed. The stud’s name was Good Fortune Fortunatas. It surely didn’t bring any good fortun
e to those pitiful kittens.
In some animals, the very gene that causes a coveted trait also brings about its opposite. A case in point is Ojos Azules, a stunning Mexican breed of cats with—the name says it all—deep blue eyes. The blue color, really a lack of pigment in the iris, results from a mutation in one gene, and as long as only one of the gene’s two copies is mutated, everything is fine: eyes are blue and kittens are healthy. But woe unto the cat whose copies are both mutated. It suffers from a deformed skull and will not even survive birth.5
Nature has developed mechanisms to avoid inbreeding wherever it can. In some animals, offspring scatter far and wide, as in prides of lions, where all males and some females leave the natal pride and disperse to find their reproductive luck elsewhere.6 In other animals—mice, quail, and voles—familiarity must breed contempt, or at least lackluster sexual attraction, because littermates or nest mates rarely mate.7 In humans, this phenomenon is called the Westermarck effect, named after a Finnish anthropologist who first proposed that growing up together suppresses sexual attraction among adults. Case in point: kibbutzim, the Israeli communities that raised children collectively, where very few marriages occurred among the thousands of men and women raised in the same kibbutz.8 Inbreeding is even bad for plants, where it is a prelude to stunted plants, seeds that do not germinate, and “albino” seedlings unable to harvest light.9 The flowers of some plants detect the molecular fingerprint of an incoming pollen grain, and, if this fingerprint is too similar to their own—the grain comes from a close relative or even from the same plant—the pollen grain fails to fertilize, and no seed develops. To be sure, some of these behaviors may also exist for other reasons. Roaming lions, for example, also avoid sexual competitors. But all of these behaviors help avoid inbreeding.10
Life Finds a Way Page 5
