Bigfoot, Yeti, and the Last Neanderthal, page 12
But they are not. The Neanderthal fragments were slightly more similar to their equivalent fragments in European than in African genomes. The null hypothesis predictions of 50/50 balance were shifted faintly in favour of a closer genetic match to Europeans than to Africans. In one comparison, for example, Neanderthal fragments had closer matches to a modern French genome 52.5% of the time, and to a Yoruba from Nigeria in 47.5% of comparisons. Though the difference is small, so many millions of comparisons were made for each analysis that the final result is highly statistically significant.
Less clear, though, are the alternative interpretations. For example, if Neanderthals were descended from an ancient African population in, say, East Africa that was also the source of later Homo sapiens dispersals, that shared ancestry may have contributed to the closer genetic affinities to Europeans apparent in the sequence comparisons. Or could it be that the genome of Homo sapiens changed to meet the much colder conditions of Ice-Age Europe in the same way as the Neanderthal genome became adapted hundreds of thousands of years earlier? An intriguing paper published in 2011 noted that the DNA segment with one of the strongest similarities between Neanderthal and modern Europeans is located in the part of the genome that controls the immune response, vital for fighting infections.2 Having the same immune response genes as Neanderthals, perhaps through interbreeding or maybe through selection of pre-existing genetic variants, could well have been critical for the survival of Homo sapiens in Europe in the face of resident Neanderthal pathogens.
Svante Pääbo, who led the Neanderthal Genome Project, has written a candid first-hand account which covers all these possibilities, and also reveals the excruciating complexity of the venture.3 It wasn't only the bioinformatics that was difficult. Getting reliable data from the tiny percentage of Neanderthal DNA remaining in the Vindija bone fragments in the face of overwhelming amounts of bacterial and modern human contamination was a gruelling business indeed, taking its toll on budgets, collaborations and even on Pääbo's health.
One immediate puzzle that needed explaining was this. How is it that not a single Neanderthal mitochondrial DNA has been ever been found in a modern European? Several hundreds of thousands, maybe even a million, Europeans have had their mitochondrial DNA tested whether in research projects or as customers of genetic genealogy companies. Yet there has never been a whisper of Neanderthal mitochondrial DNA in any of them. Although mitochondrial DNA is inherited exclusively down the maternal line, all things being equal, we should expect the same proportion of mitochondrial DNA with a Neanderthal origin in modern Europeans and Asians as there is Neanderthal nuclear DNA. With the average amount of Neanderthal nuclear DNA in Europeans estimated to be 2.5%, then if a million Europeans have had their mitochondrial DNA analysed, which is a reasonable estimate after two decades of widespread testing, 25,000 of them would have had shown a Neanderthal result. There is not a single one.
This anomaly isn't quite as hard to explain as it first seems, though it remains something of a puzzle to me. The solution offered by statisticians is centred on the fact that mitochondrial DNA is far more likely than its nuclear counterpart to be lost as it travels through the generations. For example, a woman will pass her nuclear DNA to both her sons and daughters. However, thanks to its strict matrilineal inheritance pattern, only her daughters will pass on her mitochondrial DNA to the next generations. Although her nuclear DNA will live on in her sons' children, her mitochondrial DNA dies with them. The upshot of all this is that, on average, four times less mitochondrial DNA is passed on to the next generation compared to nuclear DNA. This same equation applies at each generation, so very soon there are fewer and fewer different mitochondrial DNAs in circulation.
Mitochondrial DNA can never be lost entirely because it is vital for aerobic metabolism. Nevertheless, after a few hundred generations it is theoretically possible for Neanderthal nuclear DNA to have got through to the present day and for all the Neanderthal mitochondrial DNA to have been lost. This eradication was never certain to happen, and the other scenario might have been that Neanderthal mitochondrial DNA did get through and lots of people had it. In The Seven Daughters of Eve I explain how I found that over 95% of native Europeans are matrilinear descendants of only seven ancestral clan mothers. If Neanderthal mitochondrial DNA had survived as well as its nuclear equivalent, at least one of these seven women might have been a Neanderthal. But that did not happen and, as the authors of the 2010 paper argued, it is just a matter of chance that no one these days carries the mitochondrial DNA of a Neanderthal.
The other solution was that the interbreeding which led to the Neanderthal DNA getting into the human genome in the first place was all between Neanderthal men and Homo sapiens women. In that frankly unlikely scenario, there would be no Neanderthal mitochondrial DNA in any of the offspring. This was diminished as an explanation when further work showed that there were no Neanderthal Y-chromosomes, which would have come from males, in modern Europeans either.
A few weeks before the Neanderthal genome paper was published in 2010, another astonishing genetic revelation found its way into the journal Nature.4 Here the same team that had sequenced the Neanderthal genome announced that they had identified a new human species. Mitochondrial DNA was extracted from a fragment of a little finger bone found alongside other human remains in Denisova Cave in the Altai Mountains of southern Siberia during an excavation in 2008. The sequence showed that this bone fragment belonged not to a Neanderthal, nor to a Homo sapiens but to an as yet unknown human species. Following the tradition of naming a species after the place where it was found, the authors were tempted to name this as Homo altaiensis. They wisely recanted as all that remained on which to base a description of the new species was the fragment of finger bone, two molar teeth found nearby and the DNA sequence. There are strict rules about naming new species, as Heuvelmans discovered to his cost when he tried to register the Minnesota Iceman as Homo pongoides (literally man-ape). Heuvelmans' attempt to file his new species name without a type specimen so enraged traditional taxonomists that they lobbied, successfully, to have it struck off the official register of species.
Pääbo wisely avoided any such controversy, so there is no Homo altaiensis, at least not yet. The new species is for the moment known simply as Denisovan.
Another great surprise was that the Denisovan bone was in such good condition. It was no bigger than two grains of rice, but contained more intrinsic DNA than all the Vindija Neanderthal fragments put together. Exactly why this should be is still a mystery. The sample is too small to be carbon-dated so we do not know how old it is. One possibility is that it is very much younger than the less well-preserved Neanderthal fossils. Pääbo even suggests, though not very seriously, that it might be from a modern alma.5 Now that would be something. It would also be vindication of a sort for Russian hominologists, from Porchnev onward, who always believed in the survival of Neanderthals. I am sure they would settle for Denisovans instead.
When the Denisovan sequence was compared to the same region in modern humans and in the six complete Neanderthal mitochondrial DNA sequences known by that time, there were twice as many differences between the Denisovan and Homo sapiens as there were between ourselves and Neanderthals. By this reckoning, the Denisovans were our considerably more distant relatives than the Neanderthals. An estimate can be made, based on the differences between the DNA sequences, of how long has passed since two species last shared a common ancestor. There are many provisos in such estimates and nobody relies too much on the precision of the figures but, roughly speaking, the last common ancestor we shared with the Denisovans lived about a million years ago while the same calculations split Homo sapiens from Neanderthals about half a million years back.
Thanks to the exceptional preservation of the Denisovan bone fragment it did not take long to get a good genome sequence, and a much better one in terms of quality than the Neanderthal. However, there were more surprises in store. A comparison of the Neanderthal and Denisovan nuclear DNA showed that they were much more closely related than the mitochondrial DNA had suggested. One possible explanation for the discrepancy is that the mitochondrial DNA in the Denisovans was actually from yet another, earlier human species with whom their ancestors had interbred. Its survival through interbreeding in Denisovans was just as much a matter of chance as the apparent extinction of Neanderthal mitochondrial DNA in modern humans, where we have the reverse outcome. The matrilineal lineage of the other ancestral species survived in Denisovans whereas most or all of the nuclear DNA had come from the other hybridising species.
There were yet more surprises to come when the signs of interbreeding between ourselves and Neanderthals, at least in Europe, was also detected between Denisovans and ourselves. Denisovan nuclear DNA was found not in natives of Europe but of Papua New Guinea, and subsequently in native Australians and Pacific Islanders, and at a slightly higher level, up to 4.8%. With some Neanderthal thrown as well, the total percentage of non-sapiens DNA in modern Papuan genomes rises to the substantial total of 7.4%. To explain the link between Denisovans living forty thousand years ago in Siberia and present-day occupants of Melanesian islands like Papua New Guinea, the only logical interpretation is that the ancestors of both human species had interacted elsewhere, probably as the ancestors of today's Papuans were en route to the islands of Melanesia. It is a remarkable story, completely unexpected from classical palaeontology and the sure sign, if one were needed, of the real contribution ancient DNA is now making to the understanding of our own evolution.
It is quite likely that there are more collateral hominids yet to be discovered. Interbreeding between different human species, once thought unlikely or impossible, is now all the rage, with DNA signals of mixing between the ancestors of modern Africans and some other archaic human species.6 Denisovan-like mitochondrial DNA was recently found in a 400,000 year-old ‘human’ bone excavated from a deep cave-shaft in northern Spain, making it look as though the ancestors of Europeans might have interbred first with Denisovans, then with Neanderthals!7 What exciting times we live in. Who knows what will turn up next?
11
Keeping it in the Family
Hybrids have always fascinated cryptozoologists and, as we shall see, they are still implicated in the creation of les bêtes ignorées. Hybrid appeal is nothing new. Between the thirteenth and fifteenth century fabulous creatures from the union of one or more different species inhabited all medieval bestiaries. Among the favourites were the griffon, with the body of a lion and the wings of an eagle, the leucrota, having the haunches of a stag, the breast and shins of a lion and the head of a horse, and the yale, sporting the tail of an elephant and the face of a boar.
However, contrary to what you might now think after reading the previous chapter or browsing through early manuscripts, successful hybridisation through interbreeding is actually very rare in most mammals, especially in the wild. Whereas it is, in theory anyway, comparatively uncomplicated for two closely related species to breed in captivity, the offspring are generally not as fit and healthy as their parents. In the wild, without the care and attention of the zookeepers, they would be eliminated in the face of competition from the two parent species who have had, after all, millions of years of adaptive evolution to come to terms with their environment. However in captivity, protected from this fatal competition, hybrids can thrive. Famous examples are the offspring of tigers and lions, the liger (lion father, tiger mother) and the tigon with the opposite parentage. They are healthy, indeed typically the liger is larger than either of its parents. The trouble begins when they come to breed, as the males of both hybrids have very low sperm counts, though the females are normally fertile. This follows what has become known as Haldane's rule, named after the evolutionary biologist J.B.S. Haldane who formulated it in 1922. Haldane's rule states that in a hybrid the heterogametic sex is disadvantaged by low fitness or sterility. It governs all sorts of hybrids, both plant and animal.
The mechanisms behind Haldane's rule are complex, and need not concern us here, but the consequences for ligers and tigons, not to mention theoretical hybrids between different human species, is that males (the heterogametic sex, as males have X and Y chromosomes while females have two identical X chromosomes) are usually infertile while female hybrids are not. In a Homo neanderthalensis x sapiens hybrid, whichever way round the parentage is arranged, the girls will have a better chance of being fertile than the boys.
Haldane's rule is not the only problem for hybrids. In all species, nuclear DNA is carried on chromosomes. While different species vary in their numbers of chromosomes, typically between ten and fifty, there is no tolerance of variation in chromosome count. One chromosome too many or one too few always leads to a serious medical condition, like Down's syndrome in humans, where sufferers carry an extra chromosome number 21. If the two parent species of a hybrid have different numbers of chromosomes, breeding is ruled out altogether as both sexes will be infertile. That female ligers and tigons can produce offspring at all is because their parents have the same numbers of chromosomes. It is not just that the lion and tiger parents are genetically fairly close, being two species in the same genus, but the equality in their chromosome count that allows the hybrids to breed. If the parental chromosome numbers of a hybrid are different, then it will be infertile. The most famous example is the mule, a hybrid between a horse and a donkey, each of which have different numbers of chromosomes. Although perfectly fit and healthy themselves, mules cannot produce viable germ cells – that is, eggs or sperm. The reason here is that the hybrid mule has an odd number of chromosomes. In this situation, any germ cells that are formed will have either one too many or one too few chromosomes. What generally happens is that the germ cells give up trying to sort this out and fail to form at all.
When it comes to humans and the sapiens x neanderthalensis hybrids that the DNA tells us have introduced the Neanderthal component into modern European genomes, both parents must have the same chromosome count. Although we and our Neanderthal cousins are far more closely related than tigers and lions, chromosomal compatibility does not necessarily follow from this evolutionary proximity. Our nearest primate relatives, chimpanzees, gorillas and orang-utans, have one extra pair of chromosomes compared to humans because, at some point in our evolution after we split from the great apes, two ancestral chromosomes fused to become our chromosome number 2. This chromosome imbalance is the reason why a chimp x human hybrid, the so far only theoretical humanzee, would certainly be infertile. Like the mule, a humanzee would not form sperm or eggs. Until very recently we did not know where on the tree of human evolution this chromosome fusion occurred. If it was during the last half million years, that is after humans and Neanderthal last shared a common ancestor, the two human species would have different chromosome counts and any hybrids would be infertile, like the mule. If the chromosome fusion occurred before the split between the two human lines, then both H. sapiens and H. neanderthalensis would have the same number of chromosomes and fertile hybrids would not be ruled out by numerical incompatibility. But what is the answer? The best way to find out is to look at the chromosomes under a microscope, but to do that requires having live cells which is, of course, impossible with Neanderthals – or at least it is until one is found alive.
Of more immediate importance to our own interest in the question of hybridisation is that the high-quality Denisovan genome sequence contained information about the chromosome number. Chromosomes are essentially very long linear strands of DNA made up of only four chemical units abbreviated A, T, C and G. DNA is a code which conveys instructions on how to build and run an organism from one generation to the next. As in any code, like a word, it is not so much the letters themselves but the order in which they occur that matters. Although the DNA alphabet has only four letters, the possible combinations are almost infinite. The sequence is all. At both ends of human and primate chromosomes there is a stretch of DNA with the sequence GGGGTT. When the ends of the two primate chromosomes fused to form human chromosome 2, these sequences of GGGGTT met head to head at the join to create the sequence GGGGTTTTGGGG. This joining segment has remained in the genome of Homo sapiens ever since. A search of the Denisovan genome found these head-to-head fragments, presumably from the fused chromosome 2, but the same search in the chimpanzee genome, where the chromosomes are still separate, found none. While perhaps not quite as conclusive as counting the chromosomes of living cells under a microscope, it is pretty good evidence that the ancestral chromosome fusion had already occurred by the time the Denisovans appeared on the scene. As that was probably half a million years before the Neanderthals, it looks as though the barrier of numerical incompatibility between the different hominids had never been erected. We were all free to breed with each other and live to see our daughters at least, remembering Haldane's rule, produce healthy grandchildren.
Even if hybrids between ourselves and our great ape cousins would not be fertile, would they ever be conceived, let alone born live? Bernard Heuvelmans was especially fascinated by the prospect, as I discovered in his archive in Lausanne when I looked through his bulging box files of press cuttings and scientific papers on the topic. One file, coloured pink and intriguingly entitled ‘Hybrides: Vrais, présumes et fabuleux’ (Hybrids: True, presumed and fabulous), contained a wide range of material from, at one extreme, academic papers, such as Richard Van Gelder's essay on the classification of genera and species written for the American Museum of Natural History in New York1 to, at the other, a tabloid French magazine's coverage of ‘Queen Kong: The Liberated Lady Gorilla’ illustrated by a picture of the firmly-bosomed pongid, hair swept back under a golden hairband, perched on top of a skyscraper under attack by fighter planes while clasping a hapless man in her giant hands.
