The scientists provide powerful new evidence for a 30-year-old theory, proposed in a classic paper from Mary-Claire King and Allan Wilson of Berkeley. That 1975 paper documented the 99-percent similarity of genes from humans and chimps and suggested that altered gene regulation, rather than changes in coding, might explain how so few genetic changes could produce the wide anatomic and behavioral differences between the two.
Using novel gene-array technology to measure the extent of gene expression in thousands of genes simultaneously, this study shows that as humans diverged from their ape ancestors in the last five million years, genes for transcription factors -- which control the expression of other genes -- were four times as likely to have changed their own expression patterns as the genes they regulate.
Because they influence the activity of many "downstream" genetic targets, small changes in the expression of these regulatory genes can have an enormous impact.
"When we looked at gene expression, we found fairly small changes in 65 million years of the macaque, orangutan, and chimpanzee evolution," said study author Yoav Gilad, Ph.D., assistant professor of human genetics at the University of Chicago, "followed by rapid change, along the five million years of the human lineage, that was concentrated on these specific groups of genes. This rapid evolution in transcription factors occurred only in humans."
"For 30 years scientists have suspected that gene regulation has played a central role in human evolution," said Kevin White, Ph.D., associate professor of genetics and ecology and evolution at Yale and senior author of the study. "In addition to lending support to the idea that changes in gene regulation are a key part of our evolutionary history, these new results help to define exactly which regulatory factors may be important, at least in certain tissues. This helps open the door to a functional dissection of the role of gene regulation during the evolution of modern humans."
To measure changes in gene expression from different species, White and Gilad developed the first multi-species gene array. This allowed them to compare the level of expression of more than 1,000 genes between humans, chimps, orangutans and rhesus macaques - representing about 70 million years of evolution. To make the samples comparable, the researchers studied tissue from the liver -- one of the most homogeneous sources -- from five adult males from each of the four species.
They focused their search on expression levels of two sets of genes, those that remained largely unchanged across all four species, suggesting that there was little room -- or need -- for improvement, and those that changed most dramatically, usually in the human lineage -- an indication of powerful incentives to adapt to a changing environment.
Of the 1,056 genes from all four species, 60 percent had fairly consistent expression levels across all four species. "The expression levels of these genes seem to have remained constant for about 70 million years," the authors wrote, "suggesting that their regulation is under evolutionary constraint."
Many of these genes are involved in basic cellular processes. The authors suggest that altering the regulation of these fundamental and ancient genes may be harmful. In fact, five of the 100 most stable genes have altered expression levels in liver cancer.
When they also looked for human genes with significantly higher or lower expression levels, they found 14 genes with increased expression and five with decreased expression. While only ten percent of the genes in the total array were transcription factors, 42 percent of those with increased expression in humans were. None of those with lower expression were transcription factors. This pattern, the authors note, is consistent with "directional selection."
Previous studies have found that many of these same genes have also evolved rapidly in humans, accumulating changes in their coding sequence as well as in expression rates. "Together," they add, "these findings raise the possibility that the function and regulation of transcription factors have been substantially modified in the human lineage."
This is a very efficient way to make big changes with very little effort, according to Gilad. By altering transcription factors, the entire regulatory network can change with very few mutations, increasing the impact and minimizing the risk.
"The big question," he said, "is why are humans so different? What sort of changes in the environment or lifestyle would drive such a rapid shift in the expression of genes -- in this case in the liver -- in humans and in no other primate?"
Part of the answer, he suspects, is rapid alterations in diet, probably related to the acquisition of fire and the emerging preference for cooked food. "No other animal relies on cooked food," he said. "Perhaps something in the cooking process altered the biochemical requirements for maximal access to nutrients as well as the need to process the natural toxins found in plant and animal foods."
This is just the first of a series of similar studies, said Gilad, that will look at changes in gene expression over evolutionary time. The next steps are to look at larger arrays of genes and to focus on other tissue types.
Additional authors include Alicia Oshlack, Gordon Smyth and Terence Speed from the Hall Institute in Parkville, Victoria, Australia. This study was supported grants from the Keck Foundation, the Beckman Foundation and the National Human Genome Research Institute to Professor White.