DNA Yields Details of Human–Chimp Split

Humans May Be a Million Years Younger than Believed, with X Chromosome the Youngest

Genomes are historical records—they contain a document of change that if read properly can yield clues about the evolutionary history of an organism. DNA sequence data has helped to shake up our knowledge of the trees of descent of many organisms. Recently, the sequencing of human and chimpanzee genomes has made it possible to guess how humans and their closest ancestor diverged.

Photo by Stephanie Mitchell

David Reich said that analyzing the genomes of modern-day humans and chimpanzees “allows us to go back to the time of human and chimpanzee final gene flow—speciation—and ask, What did the population look like then?”

In a study published online in Nature on May 17, scientists at HMS and the Broad Institute of Harvard and MIT conducted the most complete evaluation of human and chimpanzee genetic sequences, along with partial sequences of three other primates. They found that the split between humans and chimpanzees occurred about 6.3 million years ago at the latest, a finding that conflicts with humanlike fossils typically dated to seven million years old. Furthermore, the genetic divergence between the two species varies widely in different parts of the genome. The X chromosomes of the two primates are surprisingly alike compared to other genomic areas. The researchers, led by David Reich, HMS assistant professor of genetics, have put forward a hypothesis to explain their data: that humans and chimps may have separated initially but then interbred before separating for good.

Fossil vs. DNA Records
Scientists are able to infer the time of evolutionary events based on the idea that DNA acts as a “molecular clock,” in which random mutations happen at a more or less steady rate over time, at least in closely related species. The accuracy of molecular clocks is controversial, but Reich’s team took pains to make theirs as accurate as possible by limiting their analysis to portions of the genome that are more likely to change through random mutations. “The regions that we’re looking at are not inside genes, they’re random junk segments of the genome,” Reich explained. They calculated the relative lengths of the human and chimpanzee branches based on their relative divergence from other primates.

Using this method, the team estimated that humans and chimpanzees last exchanged DNA around 6.3 million years ago. This time is in line with previous genetic analyses using smaller portions of the genome. The genetics conflict with fossil evidence, however, particularly the Toumai skull, which is thought to represent the earliest known hominid, called Sahelanthropus tchadensis, dated to seven million years ago.

Beyond the average divergence time of the two species, the team calculated the range of divergence times within the genome. “If you take a tiny piece of genetic material, such as a single DNA base, it’s got a unique line of descent,” said Nick Patterson, senior computational biologist at the Broad and the paper’s first author. That means that different portions of the genome will have diverged at different times. The team found that the range of divergence times in the genomes of humans and chimpanzees is gapingly large, representing a period of four million years.

Reich said that this large variation suggests that the common ancestors of chimps and humans must have already been highly diverse when the populations finally split. It contradicts the classic, allopatric speciation, which was championed by Ernst Mayr. In this model, a population is divided, usually by geography, which ultimately causes their divergence. The two initial groups look very much like each other before splitting because both are random samples.

A Messy Speciation
The other striking pattern the team found is that “the divergence between humans and chimpanzees on the X chromosome is much smaller than on the autosomes,” said Patterson. “That’s true all along the X chromosome.” In fact, the X chromosome seems to be a million years younger than the non-sex chromosomes.

What is the explanation for this wide variation within the genome, particularly on the X? “These data say that at the time humans and chimpanzees separated, the population structure was nothing that we’d expect from an allopatric speciation,” Reich said. “That’s very solid. What actually explains the data is a different story.”

He and Patterson argue that if humans and chimpanzees separated initially but then hybridized before the final split, it would explain the wide divergence times and the relatively youthful X. Male offspring of hybrid populations are typically sterile, but females could still be fertile if they mated back to the males from the ancestral population with a matching X chromosome. If a hybrid population established itself, natural selection would work against the X-linked genes that tend to cause sterility in hybrids. “Hybridization would also explain the wide range of times to the common ancestor,” Reich said, “because in some places the genome would share a common ancestor with chimpanzees around the time of hybridization, and some places would be much older.”

John Hawks, assistant professor of anthropology at the University of Wisconsin, Madison, said that these data provide the best evidence yet that the genetic variation at the time of the split was very large. But he believes that this wide diversity could simply be caused by a large number of individuals in the ancestral population, not a period of interbreeding. And he thinks that hybridization is not needed to explain the X difference, since “we know that the X chromosome has more natural selection acting on it than any other chromosome,” because the single X chromosome in males can expose dangerous recessive traits that would be weeded out.

The challenge for geneticists has been to interpret a one-dimensional readout—the genome—with the added dimension of time, a task that is sure to produce conflicting explanations. Whether or not the hybrid hypothesis proves correct, “the model makes very testable predictions,” Reich said. The study has opened up new challenges for both geneticists and anthropologists to hunt for more evidence that can answer what really happened when humans and chimpanzees went their separate ways.

—Courtney Humphries

top