Authors: Randolph M. Nesse
While the sickle-cell allele is the most frequently cited example of a gene that is selected for even though it causes disease, it is unusual for three reasons. First, it is not widely distributed, being originally found almost exclusively in people of tropical African descent. Second, the hemoglobin alteration is a simple sort of adaptation. Most adaptations, such as color vision or the capacity for fever, are complex, closely regulated systems whose assembly requires many genes. By contrast, the sickle-cell allele differs from that for normal hemoglobin only by a single T substituted for a single A. When this genetic code is translated into the protein hemoglobin, the amino acid valine ends up where glutamic acid should be. It is this molecular change that gives the blood cell its abnormal shape and other properties. Third, there is extraordinarily strong selection acting on one gene locus. It may well be that heterozygote advantage is common in human populations, but when selection against homozygotes is weak, the effect is hard to demonstrate.
In areas where malaria is rare, you would expect the sickle-cell allele to decrease in frequency. Indeed, African Americans, many of whom have lived in malaria-free regions for ten generations, show a lower sickle-cell frequency than Africans, lower than any admixture with Caucasian genes would explain. It appears that selection has been decreasing the frequency of the sickle-cell gene in regions where malaria is unimportant, as would be expected from evolutionary theory.
Several other inherited blood abnormalities also protect against malaria, the most dramatic being a deficiency of the enzyme glucose-6-phosphate-dehydrogenase (G6PD). Patients with this abnormality get very sick when exposed to oxidizing medications such as quinine, the original and still effective antimalarial drug. When a malarial parasite uses oxygen in a red blood cell, a lack of G6PD causes the cell to burst, thus interfering with the reproduction of the malarial organism. The ability of some malarial parasites to make their own G6PD illustrates the prevalence of the host-parasite arms race.
One in twenty-five northern Europeans has a copy of the recessive gene that causes cystic fibrosis, and 70 percent of cases are accounted for by a single mutant allele (∆F508). According to Francis Collins, director of the Human Genome Project, this “suggests that there may have been some heterozygote selection or a very strong founder effect for this particular mutation in the northern European population.” Exactly what benefits might maintain the frequency for the gene for cystic fibrosis remain unknown, but decreased death from diarrhea has been suggested.
Tay-Sachs disease kills all homozygote individuals before they reproduce but the gene is present in 3 to 11 percent of Ashkenazic Jews. Maintenance of this high a frequency would require an overall reproductive advantage of 6 percent for heterozygotes compared to homozygotes for the normal gene. Data on infection rates and population distributions suggest that the benefit to heterozygotes may have been protection against tuberculosis, historically a major selective force in Ashkenazic Jews. Fragile-X syndrome is still another common genetic disease, which causes mental retardation in about one out of every two thousand males born. For this syndrome there is direct evidence of increased reproductive success of heterozygous women.
University of California physiologist Jared Diamond recently emphasized another mechanism that can explain the unexpectedly high frequency of some genes that cause disease. He says that as many as eight out of ten conceptions end in early abortion or later miscarriage. The majority are never noticed because they occur before or just after implantation of the embryo. If a gene were to decrease the chances of miscarriage even slightly, it could be selected for even if it also increased the risk of developing a disease. Diamond gives the example of childhood-onset diabetes, which can be caused by a gene called DR3. If one parent is heterozygous and the other is homozygous
for the normal allele, 50 percent of the babies would be expected to have the DR3 gene, but the observed rate is 66 percent! It seems that the presence of the DR3 gene in a fetus greatly decreases the miscarriage rate and thus it perpetuates itself, despite causing diabetes.
Phenylketonuria (PKU) may be another example of disease caused by a gene maintained by frustrating the mother’s uterine selectivity. When homozygous it causes mental retardation because the body cannot handle normal levels of phenylalanine, an amino acid found in many foods. The retardation can be prevented if the child is given a diet free of this common component. PKU is a fine example of a disease that is completely genetic yet whose effects are completely preventable by environmental manipulation. It is so common (one person in a hundred has the gene) that most states require screening at birth. Why is it so common? Like the diabetes-risk gene, the PKU gene seems to reduce the likelihood of miscarriage and thus to perpetuate itself despite causing disease.
O
xford biologist Richard Dawkins has viewed the body as the gene’s way of making more genes. Genes cooperate to form cells, organs, and individuals only because that is the best way of making more copies of themselves. The body’s cells are factories, each with specialized functions, that must cooperate in order for the individual to survive and reproduce. There isn’t any way for genes to get into the next generation except by doing their part for the whole organism. Or is there? Given the stakes, one would expect that any gambit that would get a gene into the next generation would be used, even if it decreased the viability of the individual. Does this occur?
Certain genes do compete to get into a sperm or egg, even to the detriment of their carriers. There are several examples, the best known being the T-locus gene in mice. Two copies of the abnormal allele are lethal in males, but males with only one copy transmit it to more than 90 percent of their offspring, instead of the usual 50 percent. This is a fine example of an outlaw gene whose actions benefit itself but harm both the individual and the species. We know about
it because it produces a striking effect and because we can do carefully controlled experiments on mice. Might there not be minor human defects that owe their existence to a biased transmission of genes from parent to offspring that balances the decrement of fitness from the defect?
One possibility is polycystic ovaries. This disorder, which accounts for 21 percent of all visits to infertility clinics, is characterized by menstrual irregularity, obesity, and signs of masculinization. A recent study found that 80.5 percent of sisters of women with polycystic ovaries were also affected, a number far too high to be explained by an autosomal dominant or an X-linked gene. Researcher William Hague and his colleagues in Adelaide, Australia, have considered the possibilities that the condition results from transmission of DNA in the cytoplasm of the ovum or from genes that distort the process of meiosis in ways that increase their own chances of getting into an egg, a phenomenon called meiotic drive.
T
he above diseases result from the specific effects of one gene, but susceptibility to many diseases is determined by the complex effects of many genes. Hardly a week goes by without a newspaper report on the genetics of heart disease, breast cancer, or drug abuse. In most of these polygenic diseases we don’t know how many genes are responsible or what chromosomes they are on. We know only that the risk increases if close relatives have the disease. Such associations become especially convincing when people who were adopted as infants show closer resemblances to their biological families than to those in which they grew up, thus reducing the likelihood that the similarity is due to environmental factors.
Susceptibility to coronary artery disease is a good example. The risk of having a heart attack depends considerably on genes. A man whose father had a heart attack before the age of fifty-five has a risk of early death from heart attack five times that of other men. Twins
with identical genes have heart-attack rates more similar than those of nonidentical twins, even when all the twin pairs share the same environment. Does this mean that heart attacks are caused by a genetic defect? In some cases, yes. Several abnormalities of cholesterol metabolism have been discovered, one of which is an early candidate for treatment by genetic engineering in which a new gene is inserted into the cells of blood vessel walls. But we also know that heart disease results from eating a high-fat diet. Japanese immigrants to the United States who adopt the high-fat diets of this country have heart attacks more than twice as often as their relatives back home. The rate of premature death from heart disease is high enough that natural selection must be steadily weeding out any genes that contribute to the risk. People often want to know what proportion of heart disease results from genes and what proportion from the environment, but this is not the way the question should be asked. To find out why, let’s return to the mystery of myopia.
As the professor said, myopia is a genetic disease. If one identical twin has myopia, the other will almost certainly have it. We have also argued that such a harmful genetic defect would not be expected to persist. Yet about 25 percent of Americans have myopia, often so severe that they would have a hard time in a hunter-gatherer society. How well could they avoid predators, fight in a battle, or recognize a face at fifty paces? Recall poor Piggy, the castaway in
Lord of the Flies
, who without his glasses was trapped “behind the luminous wall of his myopia.” Given the disadvantage, it is perhaps no surprise that present-day hunter-gatherer populations have a low incidence of myopia. So why is it
so
common in modern populations?
When we look carefully at the transition from hunter-gatherer to industrial societies, we see that myopia does not result from a new gene. Native people in the Arctic were seldom nearsighted when they were first contacted by Europeans, but when their children began attending school, 25 percent of them became myopic. It would seem that learning to read and prolonged confinement to classrooms may permanently impair the vision of a substantial proportion of children. Why should this be?
Imagine, for a moment, the difficulty of accurately growing an eye. The cornea and the lens have to focus an image exactly on the retina, even as the eyeball grows steadily during childhood. How exact does the length of the eyeball have to be? The leeway is 1 percent of the
length of the eyeball, about the thickness of a fingernail. Is it possible to program the growth of the cornea, the lens, and the eyeball so that the image stays exactly in focus? Unlikely. Yet somehow, even as it grows, the eye keeps images in focus. How?
In a series of experiments, scientists at several laboratories are trying to work out the mechanisms that lead to nearsightedness. First, they noted that an eye with a clouded view grows longer than a normal eye, whether the clouding results from inherited disease, from injury, or from wearing foggy glasses. This is the case for chickens, rabbits, some monkeys, and some other animals, as well as humans. Next, they cut the nerve that carries information from the eye to the brain and found that in some species this stopped the excessive growth of the eye. They began to suspect that whenever a blurred image falls onto the retina, the brain sends back a message, in the form of a growth factor, that induces expansion of the eyeball. The clincher: when only one part of the visual field is blurry, only that part of the eye grows. This kind of asymmetrical growth results in astigmatism.
This mechanism is as necessary as it is elegant. In order to ensure coordinated development of the parts of the eye, the brain processes a signal from the retina, detects blurring, and sends back a signal to increase growth at the particular spot where it is needed. When growth is sufficient, the stimulus stops, and growth does too—except in some people. For 25 percent of us, there is something about reading or other close work that causes the eye to keep growing. Perhaps it is the blurred edges of letters or the plane of focus on a book held close with distant objects all around. It seems possible that printing children’s books with especially large, sharply defined letters on oversized pages could prevent some nearsightedness.
Myopia is a classic illustration of a disease whose cause is simultaneously strongly genetic and strongly environmental. To become myopic, a person must have both the myopia genotype and exposure to early reading or other close work. Many other diseases also result from complex gene-environment interactions. For instance, some people eat all the fat they want and never get heart disease, while others eat the same amount of fat and drop dead at age forty. Similarly, some people go through all kinds of losses and never become seriously depressed. For others, the loss of a pet can set off a severe episode of melancholia. Remember also the gene-environment inter-action
in PKU. For such diseases, it is a mistake to ask what proportion of the cause is genetic and what proportion is environmental. They are both completely genetic and completely environmental.
Can conditions such as myopia and clogged arteries be blamed on defective genes? In our current environment the genes that cause these conditions can certainly create a disadvantage, but in the ancestral human environment many of them might have caused no trouble at all or might even have conferred some real benefits. Perhaps hunter-gatherers with the myopia gene have better vision during childhood. A craving for fatty foods might have been thoroughly adaptive in an environment where such foods were scarce. For this reason we prefer to call such genes not defects, but
quirks
. They have no deleterious effects except in people who are exposed to novel environmental influences. Dyslexia may be another example, difficulty in reading not being a problem for hunter-gatherers.
Susceptibility to drug or alcohol addiction likewise depends on historically abnormal conditions. There are strong genetic influences on susceptibility to alcoholism, but they were a relatively modest problem before the reliable availability of beverages with at least several percent alcohol. Before the rise of agriculture and the vintners’ and brewers’ development of yeast strains tolerant of high alcohol concentrations, these genes probably were no problem at all. It may prove fruitless to search for a “gene for alcoholism.” There may be many such genes on different chromosomes that can make a person susceptible to alcoholism. Many of these genes probably have some positive effects—for instance, a tendency to continue pursuing sources of reward despite difficulties, or a tendency to experience strong reinforcement in response to stimulation of certain brain areas. While it may be tempting to postulate genetic defects in people who abuse drugs, we think it is more likely that the genetic factors that influence drug use will turn out to be a diversity of genetic quirks.