Over the weekend, Charles Darwin turned 202. I celebrated in Stony Brook on Long Island–which just so happens to be a very appropriate place to mark the event. Stony Brook University was the intellectual home of one of Darwin’s most important followers, the scientist George Williams. Williams may not be a household name, but for evolutionary biologists he looms large. Some fifty years ago, he framed some of the most important questions they are still seeking to answer today.
I was invited to Stony Brook University to give a Darwin Day lecture, and since Williams died last September at 84, I decided to make it a kind of scientific eulogy for him. It was an honor to have the chance to do so, but there was also a bittersweet irony to the experience. For the last few years of his life, Williams suffered from dementia (which may have been Alzheimer’s disease or Lewy body dementia, according his wife, Doris). The fact that millions of older people get dementia was exactly the sort of phenomenon that fascinated Williams throughout his career. Why, he wondered, do we get sick, and why do our sicknesses take their particular forms? Part of the answer, he realized, lies in our evolutionary past. Even as natural selection fine-tuned our ancestors for millions of years, it left us burdened with a susceptibility to many diseases. Evolution may be able to give rise to eyes, brains, and wings. But it’s not in the business of making us healthy.
At the end of The Origin of Species, Darwin made it clear that he did not see his theory of evolution as the last word on the subject. Instead, he saw it opening up a wide door, through which future generations of biologists could pass.”We can dimly foresee that there will be a considerable revolution in natural history,” he wrote. That revolution would extend to developmental biology, to paleontology, even to psychology. “Light will be thrown on the origin of man and his history,” he added.
But there’s one field that Darwin didn’t mention: medicine.
I am puzzled about why he left it out. Perhaps Darwin didn’t consider medicine as much of a science. His own experience with Victorian doctors certainly couldn’t have left him with a good impression. Darwin suffered from many ills during his life, which are summed up nicely on their very own Wikipedia page:
“For over forty years Darwin suffered intermittently from various combinations of symptoms such as malaise, vertigo, dizziness, muscle spasms and tremors, vomiting, cramps and colics, bloating and nocturnal intestinal gas, headaches, alterations of vision, severe tiredness, nervous exhaustion, dyspnea, skin problems such as blisters all over the scalp and eczema, crying, anxiety, sensation of impending death and loss of consciousness, fainting, tachycardia, insomnia, tinnitus, and depression.”
Darwin’s doctors had lots of theories for why he suffered all these symptoms. (They still sustain a cottage industry of speculations today, from Oedipal complexes to assassin bugs.) They prescribed him a long train of cures, including bismuth, laudanum, water therapies, and electric belts. They sometimes gave Darwin brief respites of relief, but he always relapsed sooner or later into suffering.
Evolution and medicine began creeping towards each other after Darwin’s death in 1882. It gradually became clear, for example, that infectious diseases were caused by rapidly-evolving viruses and other pathogens. But even if we could eradicate every last infectious microbe on Earth, we would still get sick, thanks to inherent weaknesses of our constitution. We would keep getting heart attacks. We would keep getting cancer. These diseases were more mysterious from an evolutionary perspective. In the late 1950s, George Williams came up with one of the most important concepts for explaining them.
At the time Williams would have seemed like the last person to forge such a bond. Williams was not a doctor. He was not a cell biologist. He was, rather, a newly-minted Ph.D. who had written his dissertation on the ecology of blennies, a kind of fish.
For Williams, blennies were just one small part of the diversity produced through evolution. And he wanted to find rules that could explain all of that diversity. He spent much of his time lost in these reflections. “George was gentle, quiet, and intensely thoughtful,” wrote James Bull, Eric Charnov, and Teresa Carlson wrote in the January 2011 issue of American Naturalist.
It was not uncommon to sit down with him and toss out an initial idea for discussion. Then, when his ensuing silence grew uncomfortably long, one would toss out another idea, ultimately another, and then another. When he finally spoke, it became clear that he was still thinking deeply about the first or second idea.
In the late 1950s, while he was a post-doctoral researcher at the University of Chicago, Williams was thinking a lot about natural selection. He saw some serious flaws in how leading scientists were thinking about it.
Many scientists believed, for example, that natural selection often produced adaptations that benefited entire groups.Why did animals get old and die, for example? Why didn’t animals just keep humming along until they were killed by a predator or a pathogen? Death had to be good for something, and one popular idea was that death benefited entire groups of animals. By dying, older individuals stepped out of the way for younger ones.
One day, Williams heard this idea for the umpteenth time,during a lecture by a renowned ecologist named A.E. Emerson. “My reaction was that if Emerson’s presentation was acceptable biology, I would prefer another calling,” he later wrote.
Williams believed that scientists like Emerson were making an error of reasoning. They were seeing natural selection at work in cases where it was both unnecessary and impossible. For a thought example, Williams liked to contemplate a fox paying a visit to a chicken coop on a snowy day. After the fox kills a hen, it trots back along the same path it took to the coop, simply because that’s easier than trotting through undisturbed snow. The next day the fox heads down the same path again for another meal. As it travels back and forth, it establishes a nicely cleared trail.
Now, does that mean that its legs are for clearing snow? Did natural selection drive the evolution of fox legs because a certain shape made them good snow plows? Of course not. They are for walking, not for snow clearing. The snow clearing is just a side effect.
Behaviors that seem to be adaptations for groups could also be side effects of natural selection acting on individuals, Williams argued. If you look at a school of fish escaping a predator, for example, it seems like it is behaving like a single body, all working as a unit. But Williams maintained that the behavior of a school could just be the result of individual fish trying to get deep into a group of fish because they’ve be less likely to be eaten.
Selection for group adaptations wasn’t just logically unnecessary in a lot of cases. It was also less likely to occur than selection on individuals. Imagine herds of wildebeest scattered across the savanna. In order for beneficial death to evolve, aging herds would have to do better than ones that didn’t. It would take a long time for winners to emerge in this process. Meanwhile, natural selection would be acting on individuals at a much faster rate in the opposite direction.
Imagine that some of the wildebeest in a herd have genes that program them to live for, say, 20 years. And others live for 25 years. Both kinds of wildebeest do what wildebeest do: eat, mate, stampede, and have calves. The 25-year wildebeest has an extra five years’ worth of calves. So in the next generation, there will be more calves with genes for living longer. Over successive generations, the wildebeest herd will live longer and longer (all other things being equal). A shorter lifespan might indeed make a herd more successful, but natural selection can’t reach such a happy solution, Williams decided.
So how did death evolve? “Walking home from the lecture with my wife, Doris,” Williams wrote later, “I regaled her with my unhappiness about the lecture and proposed the obvious idea that selection among individuals in any population would be biased in favor of the young, as long as the likelihood of living to age X was greater than to age X+1.”
Here’s what Williams meant: Imagine a fish–we’ll call it the Williams eel–that never ages. But despite its eternal youth, it is at a constant risk of getting killed by predators. In a population of Williams eels, the percentage of surviving eels will decline with age, simply because sooner or later a lot of eels get killed. When the eels reproduce, therefore, most of the parents will be young. As a result, mutations will have different effects if they strike at different times in a fish’s life.
Imagine there’s a beneficial mutation that makes a fish more resistant to diseases. If it makes fish resistant only in their old age, it won’t spread much, because so few old fishes will be around to pass it on to their offspring. It will spread much faster if it boosts the health of young fish, because they make up most of the parents. More fish will have it in the next generation.
The flip side is true for harmful mutations. If a mutation causes a devastating disorder when the Williams eels are still young, it will be very likely to wipe out a lot of their potential offspring. But if mutations only make eels sick when they’re older, they will have very little effect because most of the eels will already be dead anyway. In such a population, natural selection can act strongly on genes for survival and reproduction in youth. Mutations that cause harm in old age can pile up over the course of many generations.
Williams also observed that one mutation can have more than one effect on an organism. So imagine that a single mutation is beneficial in youth and harmful in old age. If the benefit is big enough, a mutation will spread through a population despite the harm it causes. Aging, according to Williams’s theory, was the price of youth.
Williams presented this explanation for aging (known as antagonistic pleiotropy) in papers and his influential 1966 book Adaptation and Natural Selection. It was an enticing idea, but was it true? The answer is, in a number of cases, yes.
One of the most striking examples from nature came to light in the 1990s. At the time, Steven Austad, then at Harvard, was studying the ecology of opossums. He would put bands on female opossums and then trap them every few months to count the babies in their pouches. He was amazed at how fast the animals would fall apart as they got older. In just a few months, a healthy opossum might develop cataracts, arthritis, and a host of other ailments.
Austad realized that opossums were part of a natural experiment in aging. Several thousand years ago, a population of opossums in Georgia were stranded by rising sea levels on what is now Sapelo Island. They had the good fortune to end up on an island with no predators. As a result, the percentage of older opossums still alive was higher. Williams had predicted that in such a case, natural selection would lead to slower aging. That’s because mutations that caused harm in old age would be more likely to interfere with reproductive success.
Austad compared the opossum of Sapelo to a population living on the mainland in which 80 percent of the animals get killed by predators. He discovered that the island opossum live 20 percent longer than their mainland cousins. They also enjoy better health for a longer time. Their tendons, for example, remain springier far later in life than the tendons of mainland opossum.
Some studies have failed to find evidence for antagonistic pleiotropy; it’s possible that the actual tradeoffs faced by natural selection are more complex than the ones Williams laid out. Nevertheless, evolutionary biologists still praise Williams for giving them a new way of thinking about health.
George Williams would go on to do pioneering work on a lot of other fundamental questions. Why do most animals have sex? Why are some species altruistic? He used the same elegant approach he had developed in Adaptation and Natural Selection, focusing on individuals and their genes, rather than the good of the species.
But Williams also wanted to expand the sphere of evolutionary biology. He believed that medicine, if done right, would be a kind of applied evolutionary biology. Doctors were trying to foster the health of their patients, but the biology they were contending with was the product of evolution. As he wrote in his book Why We Get Sick (co-authored by Randolph Nesse), an understanding of the ways evolution has shaped human biology would help doctors do a better job.
Antagonistic pleiotropy is particularly useful for medical researchers. We inherited our genes from a common ancestor with opossum and other mammals, and natural selection continued to negotiate a trade-offs between youth and old age in our primate forebears. Accidents, murders, and diseases all struck down our ancestors, leaving fewer older people to reproduce. Just as with opossum or chimpanzees, selection is biased towards youth in our species. George Williams was convinced that antagonistic pleiotropy had a lot to say about why we age. During his fifties, he chronicled his own decline at a track in Stony Brook, where he would time himself running a mile. With clinical detachment, he graphed his steady slowing.
Medical researchers have been tracking human aging at a molecular level, and they’re finding signs of antagonistic pleiotropy as well. The rate of most cancers is low in the young, for example, and rises steeply in old age. Yet pre-cancerous cells develop in our bodies every day, as dividing cells accidentally mutate in dangerous ways. If this was the whole story, cancer might well be far more common in young people. Low rates of cancer in the young are thanks to many lines of defense. One of the most important of these is made up of “gatekeeper” proteins that can response to DNA damage and abnormal growth. They can neutralize potentially dangerous cells by forcing them to commit suicide or by permanently pushing them into an enfeebled “senescent” state, where they can grow slowly at best.
It’s a pretty good strategy, as demonstrated by what happens when scientists shut down gatekeeper genes in mice: they get cancer a lot sooner than they would otherwise. But it’s only a stop-gap measure. Senescent and suicidal cells damage the surrounding tissue and make it harder for the body’s stem cells to regenerate new tissue. In other words, they cause our bodies to age. This aging can even lead to new cancer later in life. Gatekeepers show all the signs of antagonistic pleiotropy: they keep us healthy during our childbearing years, but we pay the cost later in life.
Not that long ago, a lot fewer people had to pay those costs, because so many people died young. Over the past 400 years, the life expectancy in Europe has more than doubled. As we live longer, we have to cope with more evolutionary side-effects. Williams himself appears to have suffered from one of the dreaded of these: the decline of the brain.
Alzheimer’s disease is associated with tangled clumps of protein in the brain called plaques. These plaques don’t pop up at random in the brain. They tend for form in a network of regions that plays a special function in our inner life. Known as the default network, it’s actually active when we’re not thinking about anything in particular. Neuroscientists suspect that the default network is vital for our very human sense of self, which we can project into the past and into the future. It’s a vital part of what it means to be human, but it is, in effect, always on. These hard-working neurons may make more proteins, and be at greater risk of making defective ones.
To understand how Alzheimer’s disease works, researchers have been investigating genes involved in it. One of these genes, called ApoE, comes in several different forms, and the form called ApoE4 appears to make people more prone to Alzheimer’s. In the latest issue of Functional Ecology, Lynn Martin of USF Health Byrd Alzheimer’s Institute, University of South Florida, and her colleagues step back to look at ApoE4’s full impact on people who carry it. It turns out to have several different effects. Early in life, it actually brings some benefits, including resistance to various infections, better memory recall, and enhanced learning. In old age, it makes people more prone to Alzheimer’s disease as well as heart disease. It is, in other words, just the sort of gene George Williams had in mind fifty years ago.
It would be a mistake to claim that evolutionary medicine can produce instant cures as soon as doctors open up The Origin of Species. The germ theory of disease did not lead immediately to antibiotics, either. But it certainly helped in the long run, by showing doctors what their proper target should be. Evolution provides doctors with new possibilities as well as unexpected warnings. A lot of research has gone into the possibility of reversing aging by manipulating our biochemistry. One of the most exciting results from this line of research has been the discovery that shutting down certain genes can make animals live longer. The worm C. elegans can live twice as long, for example, if scientists switch off a gene called daf-2.
Sounds great, right? Let’s start taking pills that shut down these genes getting in the way of eternity! Williams would have warned us to not leap so fast. Daf-2 might well shorten our lifespan when it’s working normally. But that might be because it has been shaped by antagonistic pleiotropy. Such a gene may harm us in old age, but it may also have benefits in our youth.
Studies on C. elegans have shown that this is, in fact, the case. Worms that can’t use daf-2 produce fewer offspring than ones that can. Austad has recently surveyed the medical literature and found a dozen genes in worms, flies, and mice with a similar double-edged effect: shutting them down lengthens life but also causes harm in youth. The side-effects range from sterility to delayed maturation to cognitive impairment.
These results don’t mean that the search for anti-aging medicine is a lost cause. Williams may not have escaped antagonistic pleiotropy, but perhaps his descendants will someday. But for that to happen, we will have to bear in mind Williams’s advice and get to know what we’re up against: namely, hundreds of millions of years of history.