'Chumanzee' evolution: the urge to diverge and merge
© BioMed Central Ltd 2006
Published: 24 November 2006
Skip to main content
© BioMed Central Ltd 2006
Published: 24 November 2006
A recent analysis of the human and chimpanzee genomes compared with portions of other primate genomes suggests that the divergence of the human and chimpanzee lineages beginning around 6 million years ago was not a simple clean split.
The popular and scientific press gave extensive coverage to the recent analysis by Patterson et al.  of the human and chimpanzee genomes, in which they conclude that after initially splitting, our lineage continued to hybridize with chimpanzees for more than a million years. While the Washington Post noted that "Human ancestors may have interbred with chimpanzees" , Slate.com asked more bluntly: "Did humans mate with chimps? And are we their offspring?" .
Given the extraordinary similarity of the chimpanzee and human genomes, scientists and the public alike have often asked such questions. An extensive review of the literature has yet to turn up a credible report of such crosses. In the 1920s, a Soviet scientist, Il'ya Ivanovich Ivanov, with the assistance of the Institut Pasteur at one of their field stations in French Guinea, unsuccessfully artificially inseminated three chimpanzees with human sperm . He then tried to continue his experiments at the primate center at Sukhum in the then Soviet Republic of Georgia, where he intended to artificially inseminate human volunteers with ape sperm. He was arrested by the Soviet secret police on charges unrelated to this project and was never able to carry it out .
Through their own sequencing efforts and data mining, Patterson et al.  have put together an alignment of human, chimpanzee, gorilla, orangutan, and macaque sequences that covers almost 20 Mb, which is 800 times larger than any previous analysis. But it is not just the size of the dataset that is important, it is the phylogenetic distribution. Most recent analyses of the human and chimpanzee genomes compare them with the mouse genome, which seems to be evolving at a different rate and under different constraints. By adding the very closely related gorilla, moderately close orangutan, and somewhat more distant macaque, the timing and processes of primate evolution can be more effectively studied. It is difficult, to nearly impossible, to infer whether an evolutionary event occurred on the human or chimpanzee lineage unless relatively closely related primate sequences are available for comparison.
They found a considerable amount of variation in the amount of divergence among different regions of the genomes of humans and chimpanzees. Applying molecular dating techniques to each of these regions, they inferred that human and chimpanzee speciation occurred less than 6.3 million years ago. Depending on the calibration points used to estimate this date, it could be as recent as 5.4 million years ago. This could be important if the current most favored interpretation of the fossil record holds up. In this interpretation, the fossil species Sahelanthropus tchadensis, dated to 6.5 to 7.4 million years ago, is considered to be a hominin . That is, it falls on the human lineage after the divergence of chimpanzees and humans. It has dental features similar to other fossil hominins and is inferred to be bipedal like all other hominins, and unlike chimpanzees. Another fossil species, Orrorin tugensis, is also inferred to be a bipedal hominin dating to around 5.8 million years ago. Thus, if either or both of these species are indeed true hominins, they would contradict a 6.3 million year or younger date for the split between humans and chimpanzees. However, the hominin status of these fossils is not absolutely certain and several researchers dispute their bipedality.
If chimpanzees and humans were hybridizing for over a million years after their 'split', this might imply that the early human lineage still maintained the 2n = 48 karyotype found among all the great apes (modern humans have 2n = 46). Such a speculation might also explain the apparent lack of hybridization found between modern humans and the very closely related extinct Neanderthals . If the population leading to the modern human lineage subsequently underwent a chromosomal fusion event, giving us our 2n = 46 karyotype, while the Neanderthal lineage retained 2n = 48, perhaps modern humans could not successfully interbreed with Neanderthals.
Back on firmer ground, a potentially messy split between humans and chimpanzees should not be surprising given other examples from the order Primates. Interspecies crosses and hybrids are very common among the Old World monkeys. For instance, the species Macaca arctoides may have formed by the hybridization of two other species, Macaca fascicularis and the species that gave rise to M. thibetana and M. assamensis . The different species of baboons, which initially split nearly two million years ago, regularly hybridize in the wild wherever their adjacent ranges meet , and almost all possible combinations of crosses are known. Fertile intergeneric hybrids are also known. In one case, the offspring of a Theropithecus gelada and a Papio hamadryas baboon subsequently produced offspring in a zoo setting and such hybrids are also known to occur naturally . Even more distant crosses between Papio hamadryas and Macaca mulatta have been purposely produced in captivity, but the resulting offspring, while healthy, were infertile . Thus, the potential hybridization of two newly split lineages, even if they belong to two different genera, should not be so shocking. What is more interesting is why lineages do not merge, rather than continuing on their own separate evolutionary trajectories.
The conclusions drawn from the analysis of Patterson et al.  now await testing with the completion of additional primate genomes. Sequencing of the genomes of a gorilla, orangutan, gibbon, baboon, marmoset and bushbaby is planned or in the works. However, improving on our theories of human evolutionary history also requires the continued discovery of new fossils and better ways of interpreting them. Inferences extrapolating backwards in time not only require fossils to calibrate the molecular clocks used, but can also be tested by the only hard evidence we have for ancient events, the bones and teeth of the ancestors we are hypothesizing.