Dna: The Secret of Life

by Watson, James

  Bryan Sykes at Oxford has done much to reveal the complexity of the genetic map of modern Europe. Conventional wisdom had held that modern Europeans were largely derived from the Middle Eastern populations that invented agriculture in the Fertile Crescent, between the Mediterranean and the Persian Gulf. Sykes, however, has found that most European ancestry can be traced not to the Fertile Crescent but to older indigenous lines predating the incursions of Middle Easterners and to migrant groups from Central Eurasia. Such groups include the Celts and the Huns, who swept into Europe from the East around 500 B.C. and A.D. 400 respectively. And taking his analysis of mtDNA a step further, Sykes has argued as well that virtually all Europeans are descended from one of seven "daughters of Eve," his term for the surprisingly few major ancestral nodes in the European mtDNA family tree. A company he founded, called Oxford Ancestors, will, for a fee, sequence part of your mtDNA to determine from which of the seven "daughters" you are descended.

  Another key to understanding the human past may rest with an observation fruitfully exploited by Cavalli-Sforza and others: patterns of genetic evolution often correlate with those of linguistic evolution. There are, of course, the obvious parallels between genes and words. Both are transmitted from one generation to the next; both undergo change, which in the case of language can be particularly fast, as any parent of a teenager knows. Likewise, American English is similar to but distinct from British English even though the two have been evolving separately for only a few hundred years. On the basis of the similarities and differences, then, the family tree of languages can be reconstructed in much the same way the genetic family tree can. But even more important, in many cases, as Darwin himself first predicted,* we can identify instructive correspondences between the two trees, such that what we learn about the one can deepen our understanding of the other. Both the Celts and the Basques offer dramatic cases in point: each people is genetically isolated from the rest of Europe, and each one's languages are correspondingly distinct from those of the rest of the continent. As for the New World, a controversial linguistic theory proposes that there are but three major language groups native to the Americas, and two of these correlate with the two early immigration events discerned in the Amerindian Y chromosome data. The third, by far the smallest, involves the isolated Inuit.

  *In The Origin of Species, Darwin notes: "If we possessed a perfect pedigree of mankind, a genealogical arrangement of the races of man would afford the best classification of the various languages now spoken throughout the world,"

  The availability of sex-specific genetic data – mtDNA for women, Y chromosomes for men – invites comparisons between male and female history. Mark Seielstad, a graduate student of Cavalli-Sforza's, chose to compare patterns of migration between the sexes. The logic is simple. Imagine a mutation that arises on a Y chromosome in Cape Town, South Africa. The speed with which it reaches, say, Cairo, is an index of rates of male migration. Similarly, the speed with which a Cape Town mutation in mtDNA reaches Cairo can be said to measure the rates of female migration.

  For good or ill, history has been much more the chronicle of men, rather than women, on the move. Typically they were in search of plunder or empire: think of Alexander the Great's march from Macedonia into the northern reaches of India; of the Vikings and their sea-borne rampages from Scandinavia to Iceland and America beyond; or of Genghis Khan and his horsemen pouring across the steppe of Central Asia. But even without warfare as an excuse for travel, we still think of men as the more mobile members of human society. Men traditionally do the hunting, an activity that can often take them a long way from the hearth, whereas women in traditional hunter-gatherer societies stay close to home, gleaning food locally and raising the children. Therefore, Seielstad had reason to expect that men would be our species's genetic prime movers. The data proved him startlingly wrong. Women, on average, are eight times more mobile than men.

  In fact, counterintuitive though it may be, the pattern can be simply explained. Almost universally, across all traditional societies, we humans engage in something anthropologists call "patrilocality": when individuals from two different villages get married, the woman moves to the man's village, and not vice versa. Imagine that a woman from village A has married a man from village B, and she moves to B. They have a daughter and a son. The daughter marries a man from village C and moves to C; the son marries a woman from village D and she moves to join him in B. Thus the male line stays put in B whereas the female line has moved, in two generations, from A to C via B. This process is carried out generation after generation, and as a result female migration proves extensive, but male migration does not. Men do indeed occasionally rush off to conquer distant lands, but these events are unimportant in the grand scheme of human migratory patterns: it's actually that step-by-step village-to-village migration of women that has shaped human history, at least on the genetic level.

  Detailed regional studies of Y chromosome and mtDNA variation may also reveal something of the patterns of sexual relations and mating customs promoted in the course of colonization. In Iceland, for instance, which was uninhabited before the arrival of the Vikings, we find a marked asymmetry when we compare mtDNA and Y chromosomes. Most Ys are predictably Norse, but a large proportion of the mtDNA types are derived from Ireland. Apparently, the Norsemen colonizing Iceland took Irish women with them. Unfortunately, how the Irish women felt about this cannot be extracted from the mtDNA data.

  A recent study of Y chromosome and mtDNA variation in Colombia shows a similar effect. In most segments of society, Colombian Y chromosomes are Spanish Y chromosomes, a direct biological legacy of the European conquest of the Spanish Main. In fact, approximately 94 percent of the Y chromosomes studied have a European origin. Interestingly, however, the mitochondrial pattern is quite different: modern Colombians have a range of Amerindian mtDNA types. The implication is clear: the invading Spaniards, who were men, took local women for their wives. The virtual absence of Amerindian Y chromosome types reveals the tragic story of colonial genocide: indigenous men were eliminated, while local women were sexually "assimilated" by the conquistadors.

  Sometimes, however, enduring asymmetries are more a matter of cultural continuity than violent clash of cultures. The Parsees, a minority group in India, believe themselves to be descended from the Zoroastrians, an Indo-European Aryan people who fled religious persecution in Iran in the seventh century. Genetic analysis of modern Parsees indeed reveals that they have retained "Iranian" Y chromosomes, but their mtDNA tends to be of the "Indian" type. In this case the asymmetry is maintained by tradition. To be accepted as a true Zoroastrian Parsee, one has to have a Zoroastrian Parsee father.

  Thus membership in the Parsee community is paternally transmitted together with a Y chromosome. Here genetics confirms the hold of tradition.

  Tradition has informed patterns of genetic variation among Jews as well. A recent study has shown that members of the priestly caste, the kohanim (and their descendants, usually identifiable today by the surname Cohen), have a Y chromosome distinctive enough to set them apart from all other groups. Even among the most obscure populations, those flung farthest by the Jewish Diaspora, such as South Africa's Lemba, the Cohen Y has been preserved – almost like a sacred religious text. Its source is thought to be Aaron, according to Scripture the founder of the kohanim caste and the brother of Moses. It is certainly not impossible that the kohanim Y chromosome sequence was indeed his and that it has been passed down intact, father to son, in every generation since. Such have been the rigors of tradition over the course of Jewish history.

  Hammer and others have been able to use Y chromosomes to track the entire Diaspora with interesting results. The Ashkenazim, for example, who have lived in Europe for the past twelve hundred years (and now the United States and elsewhere), have nevertheless maintained the genetic indications of their Middle Eastern origins. In fact, molecular studies have made plain that the Jews, genetically at least, are virtually indistinguishable from a
ll other Middle Eastern groups, including the Palestinians. So, too, is it written. Abraham, the great patriarch, is said to have had two sons by different women: Isaac, from whom the Jews are descended, and Ishmael, forefather of the Arabs. That such a deadly enmity should have arisen between the descendants of one man is an irony that grows only more bitter when genes seem to verify tradition's narrative.

  A simple stroll down a Manhattan street would suggest that ours is the most genetically variable species on the planet. In fact, though, the human genome is markedly less variable than those of most species for which we have genetic information. Only about 1 in every 1,000 human base pairs varies among individuals. Genetically, then, we are 99.9 percent alike, a minute degree of difference by the standards of other species. Fruit flies – even if they all look the same to us – have levels of variation some 10 times higher. Even Adélie penguins, those icons of sameness in their vast Antarctic colonies of indistinguishable individuals, are more than twice as variable as we are. Nor is this lack of variability found in our nearer relatives: chimpanzees are about 3 times as variable as we are, gorillas 2 times, and orangutans 3.5.

  With the mtDNA and Y chromosome family results at hand, it is readily apparent why we humans are so alike. It's because our common ancestor was so recent; 150,000 years is a blink of an eye by evolutionary standards – insufficient time for substantial variation to arise through mutation.

  Another counterintuitive finding about human variation, what little there may be, is that it does not correlate, for the most part, with race. Prior to Cann and Wilson's demonstration of humankind's surprisingly recent flight out of Africa, it was assumed that different groups had been isolated from one another on different continents for ages and ages, up to two million years. This would have permitted the accumulation of substantial genetic difference, in accordance with the Pauling-Zuckerkandl model, whereby the extent of genetic divergence between isolated populations is a function of the time over which they have been isolated. In light of Cann and Wilson's conclusion that we all share a much more recent common ancestor, it is clear that there has simply not been time enough for geographically separate populations to diverge significantly. Thus, though genetic differences, like skin color, are manifest across groups, race-specific genetic differences tend to be very limited. Most of our scant variation is actually spread rather uniformly across populations: one is as likely to find a particular genetic variant in an African population as in a European one. One is left to surmise that much of the genetic variation in our species arose in Africa before the out-of-Africa event, and so was already present in the groups that went forth to colonize the rest of the world.

  As a final blow to any pride we may take in our own genetic variety: the Human Genome Project's conclusion that only about 2 percent of our DNA encodes genes would suggest that at least 98 percent of our variation falls in regions of the genome where it has no effect. And because natural selection very efficiently eliminates mutations that affect functionally important parts of the genome (such as genes), variation accumulates preferentially in noncoding (junk) regions. The difference between us is small; the difference it makes is even smaller.

  Because of the short evolutionary timescales involved, most of the consistent differences we do see among groups are probably products of natural selection: skin color, for one.

  Under their dense matted hair, the skin of our closest relatives, the chimpanzee, is largely unpigmented. (Chimpanzees, you might say, are white.) And presumably the common ancestor of chimpanzees and humans from which the human lineage spun off five million years ago was similar. And so we infer that the heavy skin pigmentation characteristic of Africans (and of the earliest modern humans, in Africa born) arose in the course of subsequent human evolution. With the loss of body hair, pigment became necessary to protect skin cells from the sun's damaging ultraviolet (UV) radiation. We now know at a molecular level how UV rays can cause skin cancer: they make the thymine bases of the double helix stick to one another, creating a kink, so to speak, in the DNA molecule. When that DNA replicates itself, this kink often promotes the insertion of a wrong base, producing a mutation. If, by chance, that mutation is in a gene that regulates patterns of cell growth, cancer may result. Melanin, the pigment produced by skin cells, reduces UV damage. As anyone with as hopelessly fair a complexion as mine knows too well, sunburn, though typically not lethal, can be a much more immediate health threat than skin cancer. Thus it is easy to imagine natural selection favoring the acquisition of dark skin in order to prevent not only cancer, but also the infections that can easily result from a severe sunburn.

  Why did people living in higher latitudes lose melanin? The best explanation involves vitamin D3 synthesis, a process carried out in the skin and requiring UV light. D3 is essential for calcium uptake, which in turn is a critical ingredient of strong bones. (A deficiency of D3 can result in rickets and osteoporosis.) It is possible that, as our ancestors moved out of Africa into highly seasonal environments, with less year-round UV radiation, natural selection favored pale-skinned variants because they, with less sun-blocking pigment in their skin, synthesized D3 more efficiently with the limited UV available. The same logic may apply to the movements of our ancestors within Africa. The San, for instance, in South Africa, where UV intensities are similar to those of the Mediterranean, have a strikingly pale skin. But what about the Inuit peoples, who live in or close to the hardly sunny Arctic but are surprisingly dark? Their opportunities for producing the vitamin would appear to be further limited by the need to be fully clothed all the time in their climate. In fact, the selective pressure favoring lightness seems not to have asserted itself among them, and the reason appears to be that they have solved the D3 problem in their own way: a diet with plenty offish, a rich source of the essential nutrient.

  Given what a powerful determinant, mostly for ill, skin color has been in human history and individual experience, it is surprising indeed how little we know about its underlying genetics. This deficit, however, may have less to do with the limitations of our science and more with the intrusion of politics into science; in an academic world tyrannized by political correctness, even to study the molecular basis of such a characteristic has been something of a taboo. What little we understand about it depends on old studies of mixed-race children, which established that several genes contribute to pigmentation. But our knowledge of other species and the similarity of basic biochemical processes among all mammals suggest a more complicated picture. We know, for instance, that many genes affect coat color in mice, and it is likely that these have direct human equivalents. So far, though, we have managed to identify only two genes involved in human pigmentation: the one that, when mutated, causes albinism, and the other, the "melanocortin receptor," associated with red hair and a pale (often freckled) complexion. The melanocortin receptor gene is variable among Europeans and Asians, but invariant among Africans, suggesting that there has been strong natural selection in Africa against mutations in the gene, i.e., against red-haired, fair-skinned individuals. Albinos, who lack pigment altogether, occasionally appear today in African populations (probably through de novo mutation) but their acute sensitivity to sunlight puts them at a severe disadvantage.

  Another morphological trait likely determined by natural selection is body shape. In hot climates, where dissipating body heat is a priority, two basic types have evolved. The "Nilotic form," represented by the East African Masai, is tall and slender, maximizing the surface-area-to-volume ratio and thus facilitating heat loss. The Pygmy form, on the other hand, though still lightly built, is very short. In this case, a physically strenuous hunter-gather lifestyle has selected for small size to minimize the energy expended in movement – why lug a big body around to look for food? In high latitudes, by contrast, selection has favored body forms that promote heat retention: those with the lower ratios of surface area to volume. Neanderthals from Northern Europe were therefore heavily built, and so too on average are today's inhabitants
of the same boreal climes. Some of the variation in athletic performance we see among groups is presumably attributable to these body-form differences. It should come as no surprise that in the high jump, for instance, a tall Nilotic body is better adapted than a short robust one.

  If there is a trait whose distribution among human populations is hard to fathom, it is lactose intolerance. Mammalian milk, including the human variety, is rich in a sugar called lactose, and newborn mammals typically produce a special enzyme, lactase, to break it down in the intestine. Upon weaning, however, most mammals, including humans – at least, most Africans, Native Americans, and Asians – stop making lactase and so as adults cannot digest lactose. "Lactose intolerance" means that drinking a glass of milk can have unpleasant consequences, including diarrhea, gas, and abdominal bloating. Most Caucasians and the members of a few other groups, on the other hand, continue to produce lactase throughout their lives, and can therefore handle a lifelong dairy diet. The explanation has been advanced that lactose tolerance evolved in those groups historically most dependent upon dairy products, but the pattern of the trait is by no means fully convincing; there are, for example, groups of Central Asian animal herders – cheese for everyone – who are lactose intolerant. And despite belonging to an ethnic group that is typically lactose tolerant, I am intolerant. If natural selection had favored tolerance in a particular group, why would it leave its job undone? The most compelling evidence yet in support of the standard explanation is the presence of lactose tolerance in African groups traditionally associated with livestock. We may never fully understand the adaptive dimension of this trait, but molecular biologists working on a Finnish population have recently identified the mutation responsible for it. And so while we are by no means fighting a killer here, it is now possible, with a simple genetic test, to determine whether a newborn will grow up to face a choice between ice-cream deprivation and chronic gastric cramps.