Out of Eden: The Peopling of the World

by Oppenheimer, Stephen

  Mendel was careful to choose simple, common, easily distinguished characters and to study them individually. In reality, the expression of some physical characters is determined by more than one gene, and we all have around 30,000 pairs of functioning genes because we are rather complex organisms. Thus the visible random variation between siblings in the same family is not the result of any vagueness in the process of heredity, but arises because there are large numbers of gene pairs being randomly chosen during sexual procreation. With so many gene pairs in different combinations, there is huge potential for variety. By way of contrast, the extreme similarity of identical twins gives us a glimpse of how precise the conversion of heredity into physical form really is. The extraordinary achievement of Watson, Crick, and Rosalind Franklin was to translate Mendel’s discoveries into biochemistry – or molecular biology, as this aspect of biochemistry came to be known.

  Cardboard keys to life

  ‘We are the products of our genes.’ The secret keys to this Edwardian truism were traced and cut out on bits of cardboard by two adventurer-scientists, Jim Watson and Francis Crick, in 1953.34 Their ‘keys’ were scale diagrams of four chemicals (nucleotide bases), whose unique interlocking relationship, set in the double-stranded zip-locking deoxyribonucleic acid (DNA), holds the code for life on Earth. Those bits of cardboard unlocked the mechanism linking Mendel’s work to the theory of evolution by natural selection as set out by Darwin in the Origin of Species. Watson and Crick explained exactly how thousands of unique characteristics, varying from one individual to another, are passed on intact from generation to generation. In short, it was the greatest advance in biological understanding in the twentieth century.

  Within each of the cells of our bodies we all have incredibly long strings of DNA. It is the stuff of the genes. It stores, replicates, and passes on all our unique characteristics – our genetic inheritance. These DNA strings hold the template codes for proteins, the building blocks of our bodies. The codes are ‘written’ in combinations of just four different chemicals known as nucleotide bases (represented by the letters A, G, C, and T), which provide all the instructions for making our bodies. We inherit DNA from each of our parents, and because we receive a unique mixture from both, each of us has slightly different DNA strings from everyone else. Our own DNA is like a molecular fingerprint.

  During human reproduction, the parents’ DNA is copied and transmitted in equal proportions. It is important to know that although most of the DNA from each parent is segregated during reproduction, small bits of their respective contributions are shuffled and mixed at each generation. The mixing here is not that of mass random allocation of genes inferred by Mendel, but tiny crossovers, duplications, and swaps between maternal and paternal DNA contributions. This is known technically as recombination. Luckily, for the purposes of genetic researchers, there are two small portions of our DNA that do not recombine. Non-recombining DNA is easier to trace back since the information is uncorrupted during transmission from one generation to the next. The two portions are known as mitochondrial DNA (mtDNA) and the non-recombining part of the Y chromosome (NRY).

  Mitochondrial DNA: the Eve gene

  To say that we get exactly half of our DNA from our father and half from our mother is not quite true. One tiny piece of our DNA is inherited only down the female line. It is called mitochondrial DNA because it is held as a unique circular strand in small tubular packets known as mitochondria that function rather like batteries within the cell cytoplasm. Some molecular biologists say that, aeons ago, the mitochondrion was a free-living organism with its own DNA, and possessed the secret of generating lots of energy. It invaded single-celled nucleated organisms and has stayed on ever since, dividing, like yeast, by binary fission. Males, although they receive and use their mother’s mitochondrial DNA, cannot pass it on to their children. The sperm has its own mitochondria to power the long journey from the vagina to the ovum but, on entry into the ovum, the male mitochondria wither and die. It is as if the man had to leave his guns at the door.

  So each of us inherits our mtDNA from our own mother, who inherited her mtDNA intact from her mother, and so on back through the generations – hence mtDNA’s popular name, ‘the Eve gene’. Ultimately, every person alive today has inherited their mitochondrial DNA from one single great-great-great-. . .-grandmother, nearly 200,000 years ago. This mtDNA provides us with a rare point of stability among the shifting sands of DNA inheritance. However, if all the Eve chromosomes in the world today were an exact copy of that original Eve mtDNA, then clearly they would all be identical. This would be miraculous, but it would mean that mtDNA is incapable of telling us much about our prehistory. Just knowing that all women can be traced back to one common ancestral Eve is exciting, but does not get us very far in tracing the different geographic lives of her daughters. We need something with a bit of variety.

  This is where DNA point mutations come in. When mtDNA is inherited from our mother, occasionally there is a change or mutation in one or more of the ‘letters’ of the mtDNA code – about one mutation every thousand generations.35 The new letter, called a point mutation, will then be transmitted through all subsequent daughters. Although a new mutation is a rare event within a single family line, the overall probability of mutations is clearly increased by the number of mothers having daughters. So, within one generation, a million mothers could have more than a thousand daughters with a new mutation, each different from the rest. This is why, unless we share a recent maternal ancestor within the past 10,000 years or so, we each have a slightly different code from everyone else around us.

  Using mutations to build a tree

  Over a period of nearly 200,000 years, a number of tiny random mutations have thus steadily accumulated on different human mtDNA molecules being passed down to daughters of Eve all around the world. For each of us this represents between seven and fifteen mutational changes on our own personal Eve record. Mutations are thus a cumulative dossier of our own maternal prehistory. The main task of DNA is to copy itself to each new generation. We can use these mutations to reconstruct a genetic tree of mtDNA, because each new mtDNA mutation in a prospective mother’s ovum will be transferred in perpetuity to all her descendants down the female line. Each new female line is thus defined by the old mutations as well as the new ones. As a result, by knowing all the different combinations of mutations in living females around the world, we can logically reconstruct a family tree right back to our first mother.

  Although it is simple to draw on the back of an envelope a recent mtDNA tree with only a couple of mutations to play with, the problem becomes much more complex when dealing with the whole human race, with thousands of combinations of mutations. So computers are used for the reconstruction. By looking at the DNA code in a sample of people alive today, and piecing together the changes in the code that have arisen down the generations, biologists can trace the line of descent back in time to a distant shared ancestor. Because we inherit mtDNA only from our mother, this line of descent is a picture of the female genealogy of the human species.

  Not only can we retrace the tree, but by taking into account where the sampled people came from, we can see where certain mutations occurred – for example, whether in Europe, or Asia, or Africa. What’s more, because the changes happen at a statistically consistent (though random) rate, we can approximate the time when they happened. This has made it possible, during the late 1990s and in the new century, for us to do something that anthropologists of the past could only have dreamt of: we can now trace the migrations of modern humans around our planet. It turns out that the oldest changes in our mtDNA took place in Africa 150,000–190,000 years ago. Then new mutations start to appear in Asia, about 60,000–80,000 years ago (Figure 0.3). This tells us that modern humans evolved in Africa, and that some of us migrated out of Africa into Asia after 80,000 years ago.

  It is important to realize that because of the random nature of individual mutations, the dating is only approximate. T
here are various mathematical ways of dating population migrations, which were tried with varying degrees of success during the 1990s, but one method established in 1996, which dates each branch of the gene tree by averaging the number of new mutations in daughter types of that branch,36 has stood the test of time and is the main one I use in this book.

  Figure 0.3 Real maternal gene tree of 52 randomly selected individual people from around the world. Note the age of Mitochondrial Eve. Branch dating by author based on complete sequence data; the chimp–human coalescent date arises from analysis, i.e. not assumed from fossil evidence – see note 22 in Chapter 1.

  Y chromosome: the Adam gene

  Analogous to the maternally transmitted mtDNA residing outside our cell nuclei, there is a set of genes packaged within the nucleus that is only passed down the male line. This is the Y chromosome, the defining chromosome for maleness. With the exception of a small segment, the unpaired Y chromosome plays no part in the promiscuous exchange of DNA indulged in by other chromosomes. This means that, like mtDNA, the non-recombining part of the Y chromosome remains uncorrupted with each generation, and can be traced back in an unbroken line to our original male ancestor.

  Y chromosomes have been used for reconstructing trees for less time than mtDNA has, and there are more problems in estimating time depth. When these are solved, the NRY method may have a much greater power of time and geographical resolution than mtDNA, for both the recent and the distant past. This is simply because the NRY is much larger than mtDNA and consequently has potential for more variation.

  Yet Y chromosomes have already helped to chart a genetic trail parallel to the mtDNA trail. At the major geographical branch points they support the story told by mtDNA: they point to a shared ancestor in Africa for all modern humans, and a more recent ancestor in Asia for all non-Africans. In addition, because men’s behaviour differs in certain key ways from women’s, the story told by the Adam genes adds interesting detail. One difference is that men have more variation in the number of their offspring than women: a few men father considerably more children than the rest. Women, in contrast, tend to be more even and ‘equal’ in the number of children they have. The main effect of this is that most male lines become extinct more rapidly than female lines, leaving a few numerically dominant male genetic lines.

  Another difference is in movement. It has often been argued that because women have usually travelled to their husband’s village, their genes are inevitably more mobile. Paradoxically, while this may be true within one cultural region, it results in rapid mixing and dispersal of mtDNA only within that cultural region. For travel between regions, or long-distance intercontinental migrations, by sea for example, the burden of caring for children would have limited female mobility. Predatory raiding groups would also have been more commonly male-dominated, resulting in increased long-distance mobility in the Y chromosome.

  A final point on the methods of genetic tracking of migrations: it is important to distinguish this new approach to tracing the history of molecules on a DNA tree, known as phylogeography (literally ‘tree-geography’), from the mathematical study of the history of whole human populations, which has been used for decades and is known as classical population genetics. The two disciplines are based on the same Mendelian biological principles, but have quite different aims and assumptions, and the difference is the source of much misunderstanding and controversy. The simplest way of explaining it is that phylogeography studies the prehistory of individual DNA molecules, while population genetics studies the prehistory of populations. Put another way, each human population contains multiple versions of any particular part of our genome, each with its own history and different origin. Although these two approaches to human prehistory cannot represent exactly the same thing, their shared aim is to trace human migrations. Tracing the individual molecules we carry is just much easier than trying to follow whole groups.

  Naming gene lines

  In this book I refer interchangeably to maternal or paternal clans, gene lines, lineages, genetic groups/branches, and even haplo-groups. All these terms mean much the same thing: members of a large group of genetic types sharing a common ancestor (usually through their mtDNA or Y chromosome). The size of the group is to a certain extent arbitrary and depends on how far back the base of the branch is on the genetic tree. One thing that quickly becomes apparent from study of genetic trees is the lack of uniformity in the nomenclature of these branches. For the Y chromosome in particular, each new scientific paper proposes a new system of scientific nomenclature based on different ‘in-house’ genetic markers used by different research laboratories. This can make comparison between different studies tedious and repeatedly tests the limits of the reader’s memory. The underlying tree, however, is more or less the same from lab to lab. Recently, a consensus Y nomenclature was published, with letters from A to R describing the main genetic branches in the tree.37 I use this nomenclature as far as possible in my referencing and figures. The trouble is that even these eighteen letters and their location on the tree are difficult to hold in the mind – at least in mine. The fact that they are just letters makes it worse.

  Luckily, our memory is often aided by context and association. For this reason, and this reason alone, I have introduced names for the major branches and for other branches I refer to frequently. Some of the names are regional, such as Ho for one Chinese/Southeast Asian Y-branch (after the Southeast Asian explorer Admiral Cheng Ho and also for Ho Chi Minh, the revered Vietnamese nationalist hero). Others are biblical, like the out-of-Africa Y-line founder Adam and his three descendent lines, Cain, Abel, and Seth (see Appendices 1 and 2). There is no intention with these names to infer any deeper meaning – they are simply aides-mémoires.

  The mtDNA picture is slightly easier. Many of the different labs agreed at an early stage to try to use a single nomenclature. (Perhaps there was less testosterone involved in the process!) For instance, there are two agreed non-African daughter lines, ‘N’ and ‘M’ from the single out-of-Africa line L3. I have called them Nasreen, in keeping with a southern Arabian origin, and Manju, to be consistent with an Indian subcontinental origin.

  1

  OUT OF AFRICA

  ONE OF THE MOST ENDURING media images of popular genetics in the 1980s was a cover of Newsweek showing sophisticated and attractive nude portraits of a black Adam and Eve sharing the apple, with the snake looking on approvingly between (see Plate 5). This cover sold record numbers of the magazine. But in spite of the media hype, Newsweek was reporting a major advance. There were two stunning insights portrayed in the picture and described in the lead article. The first was some new genetic evidence published in 1987, using genes that could be passed down only through our mothers. This work, by American geneticist Rebecca Cann and colleagues, resolved an old argument about the birthplace of modern humans. The new evidence said that we, ‘the modern human family’, had originated as a single genetic line in Africa within the last 200,000 years, and not as multiple separate evolutionary events in different parts of the world. This single line, which leads back eventually to the ancestor we share with the Neanderthals, gave rise to the half-dozen major maternal clans (or branches) that are, even today, clearly of African origin.1

  The second reason for using a biblical allegory on the Newsweek cover was that this new genetic approach used only maternally transmitted mitochondrial DNA. Ten years later, a small group of geneticists would use this newly discovered method to identify a single twig from those dozen or more ancestral African maternal genetic branches as forming the sole founding line for the rest of the world.2 In other words, there was a single common ancestor or ‘Mitochondrial Eve’ for all African female lines and then, much later, came a subsidiary ‘Out-of-Africa Eve’ line whose genetic daughters peopled the rest of the world. It proved to be an extraordinary discovery.

  The label ‘Eve’ or ‘First Lady’, so celebrated by the media at the time, was not the exact truth as geneticists saw it. Using mitoch
ondrial DNA, they had identified a root female genetic line for only a tiny part of our genetic heritage. Mitochondrial Eve was not a sole individual ‘First Mother’ as often implied by the media. Tracing a maternal genetic marker back to a shared ancestral type does not mean that all our genes literally derive from one woman. In each of our cell nuclei we have tens of thousands of genes, each with their own history, that can be mapped. Any one of these genes could be used as a marker system to trace back to a common ancestor.3 Because much of our DNA mixes around at each generation, the gene trees would not necessarily all go back to the same ancestor. In fact, the genetic heritage of modern humans may be derived from a core of 2,000–10,000 Africans who lived around 190,000 years ago.

  Although the reality of parallel ‘gene trees’ rather than ‘people trees’ may seem to take the romance out of the Eve tale, it does not diminish the exceptional power of such genetic tracing to tell the grand story of human wanderings over the past 200,000 years. The ability to tell the same story using a number of different genes confirms and enriches the tale. Male gene lines, for instance, tend to show a rather more adventurous intercontinental spread than do female ones. In South Africa, for example, a self-proclaimed ‘lost tribe of Israel’ was recently identified by geneticists as having Jewish ancestors, but only through the male line.4

  Objections from multiregionalists and geneticists

  The bold and clear trail of spreading humankind drawn by the mitochondrial markers was bound to raise objections. The multiregionalists certainly dissented. These were mainly palaeoanthropologists who still believed that the different world ‘races’ each broke out of pre-modern ancestral forms – from various parts of the world such as ‘Java Man’, ‘Peking Man’, and the Neanderthals. A major problem here is that, in many ways, the regionally defined modern human peoples resemble one another more than they do their supposed ancestors. To cope with this observation, some multiregionalists now argue that different regional ‘human species’ subsequently mixed and exchanged genetic material with one another to form the complex modern races we have today (Figure 1.1).