April 14, 1998
Tree of Life Turns Out to Have Complex Roots
By NICHOLAS WADE
rom Yellowstone Park to the ocean's abysses, researchers are in hot pursuit of the universal ancestor. Not the sort that is painted in oils and hung proudly in the hallway, but a single-celled creature with few distinctive features save a fondness for living in near-boiling water.
Though the universal ancestor probably lived more than 3.5 billion years ago and was too small to be seen, it was far from contemptible. From this Abraham of microbes sprang the three great kingdoms of evolution, the bacteria, the archaea and the eukarya. All the world's visible forms of life are slender shoots at the tip of the eukaryan branch.
Biologists have long aspired to paint a genetic portrait of the ancestor by running the tree of evolution backward, going from its leaves -- the living creatures of today -- down to the point where all its branches coalesce in a single trunk. Defining the organism that existed at this point, and when and where it lived, might help toward one of biology's major goals, understanding the origin of terrestrial life.
The longstanding road map for finding the universal ancestor, however, turns out in the light of new data to have given misleading directions, and the road map's chief author, Dr. Carl Woese of the University of Illinois, is proposing a new theory about the earliest life forms.
Working back to the ancestor, an exercise based on the sequence of DNA letters in genes, resembles the way that linguists reconstruct the words of vanished mother-tongues from their living descendant languages.
Genes that perform the same role in human cells and in bacterial cells, say, may have a recognizably similar spelling of their DNA letters, reflecting the genes' descent from a common ancestor. In one such gene the human-bacterium similarity is as high as 45 percent.
Hope of reconstructing the ancestor from its inferred genes received new impetus three years ago when the first full DNA, or genome, of a bacterium was decoded. Since then, the genomes of a dozen microbes have been sequenced, including at least one from each of the three main branches of the evolutionary tree.
The three kinds of genome offered a broad basis for triangulating back to the ancestral genome. But the emerging picture is far more complicated than had been expected, and the ancestor's features remain ill-defined though not wholly elusive. "Five years ago we were very confident and arrogant in our ignorance," said Dr. Eugene Koonin of the National Center for Biotechnology Information. "Now we are starting to see the true complexity of life."
Despite the quagmire in which their present efforts have landed them, biologists have not in any way despaired of confirming the conventional thesis, that life evolved on earth from natural chemical processes. But a ferment of rethinking and regrouping is under way.
Until now, searchers in the universal-ancestor treasure hunt have followed a hallowed chart known as the ribosomal RNA phylogenetic tree. This is a family tree drawn up by Woese and based on a gene used by all living cells to specify ribosomal RNA, or ribonucleic acid, a component of the machinery that translates genetic information into working parts.
It was this tree that led Woese to recognize the tripartite division of living things and to realize that one of the three kingdoms belonged to the archaea, previously assumed to be a weird sort of bacteria.
Many of the deepest branches in Woese's tree, those that join nearest to the three-way junction of the kingdoms, turned out to belong to organisms that live at high temperatures, as in the fuming springs in Yellowstone Park or the volcanic vents that gash the ocean floor. That clue fit well with new ideas holding that life originated at volcanolike temperatures.
With the new ability to decode the full DNA of a microbe, it is these high-temperature microbes that biologists have chosen for some of their first targets. Aquifex aeolicus, a denizen of Yellowstone Park that lives at 5 degrees below the boiling point of water, is the deepest branching of all known bacteria.
In the light of evidence suggesting that the oldest region of the ribosomal RNA tree lies on the branch leading into the bacterial kingdom, Aquifex provided grounds for the claim that it was the nearest living cousin of the universal ancestor.
But the sequence of the Aquifex genome, reported last month in the journal Nature, has yielded only disappointments. For one thing, the microbe appears to have only one gene, called a reverse gyrase, that is not found in organisms that live at ordinary temperatures.
That suggests it may be quite easy for microbes to switch between high and normal temperatures, said Dr. Ronald Swanson, a member of the Aquifex team who works at Diversa Corp. of San Diego.
A second blow is that with the full genome sequence in hand, for Aquifex and a dozen other microbes, biologists can draw up family trees based on other genes besides the ribosomal RNA gene that provided the original map. And the trees based on other genes show different maps that do not agree with the ribosomal RNA map. "Each picture is different, so there is tremendous confusion," Woese said.
A basic source of the confusion is that in the course of evolution whole suites of genes have apparently been transferred sideways among the major branches. Among animals, genes are passed vertically from parent to child but single-celled creatures tend to engulf each other and occasionally amalgamate into a corporate genetic entity.
It has long been argued that mitochondria, the tiny organelles that handle the energy metabolism of eukaryotic cells, were once free-living bacteria that were enslaved by an early eukaryote. Mitochondria still possess their own bacteriumlike DNA but many of their genes have emigrated into the eukaryotic cell's own DNA in the nucleus.
Horizontal transfer of genes between kingdoms would severely tangle up the lines in family trees. "What impresses me is that the pattern of genes we see among organisms is not reduced to total chaos," said another member of the Aquifex team, Gary Olsen of the University of Illinois.
Presumably because of sideways gene traffic in the distant past, both archaea and eukarya seem to rely on bacterial-type genes to manage much of their general chemical metabolism. (The eukarya, thought to be descended from the archaea, rely on archaean-type genes to manage their DNA and to translate its genetic information into protein products.)
"It's possible that bacterial genes have swept all over the world and replaced everything else that existed, so some of the features of the last common ancestor may have been erased from the face of the planet," Koonin said.
But no one is abandoning the search for the ancestor. "My biggest fear is that evolution would be indecipherable because of all the random changes that took place," said Craig Venter of the Institute for Genomic Research in Rockville, Md. "The good news is that that is clearly not the case. I think it will be completely decipherable but because of horizontal transfer the tree may look more like a neural network," he said, referring to the criss-cross pattern of a neural computing circuit.
Venter, who pioneered the sequencing of microbial genomes, estimated that 50 to 100 more genomes needed to be sequenced to help triangulate back to the last common ancestor.
Evolutionary biologists are working on several approaches for seeing beyond the confusion caused by lateral transfer. Computational biologists like Koonin believe that it is already possible to identify 100 or so genes that the common ancestor must have possessed -- mostly ones that manage DNA and its translation into proteins -- and that others can be added with varying degrees of certainty.
Most biologists still favor the standard view that the universal ancestor, already a quite sophisticated organism that had come a long way since the origin of life, first branched into the bacteria and the archaea.
Later the eukarya branched off from the archaea, but accepted many genes from the bacteria. Koonin describes the eukaryotic cell as a "palimpsest of fusions and gene exchanges," referring to a manuscript that has been written over with new text.
But some important eukaryotic genes have no obvious predecessors in either the archaean or the bacterial lines. The family of genes that make the stiff framework of eukaryotic cells, known as the cytoskeleton, seems to appear out of nowhere.
"The absence of sequences closely related to the slowly changing proteins of the eukaryotic cytoskeleton remains unsettling," Dr. Russell Doolittle of the University of California, San Diego, wrote in the March 26 issue of Nature.
Another evolutionary biologist, Dr. Ford Doolittle of Dalhousie University in Halifax, Nova Scotia, has an explanation, though one that he concedes does not yet enjoy the company of evidence. He argues there might have been many lost branches of the tree of life before the universal ancestor. One of these branches, a fourth kingdom of life, might have contributed the cytoskeleton genes to the eukarya before falling into extinction.
A new and far-reaching theory about the universal ancestor has been developed by Woese. Though he declined to discuss it, because his article is due to be published in the Proceedings of the National Academy of Sciences, colleagues said the theory envisages that all three kingdoms emerged independently from a common pool of genes.
The pool was formed by a community of cells that frequently exchanged genes among themselves by lateral transfer. The price of membership in the community was to use the same genetic code, according to Woese's theory, which is how the code came to be almost universal.
The community of proto-genomes quickly shared innovations among themselves, in Woese's new view, and the system evolved by producing more complicated proteins, the working parts of the cell. The genetic code was at first translated rather inaccurately, so the proteins it produced were short and limited in capability. But the code became more accurate, and the proteins more complex, driven by the advantage that more capable proteins conferred.
At a certain stage of complexity, design decisions may have limited cells' ability to exchange genes, and the ancestral pool would have split into the three kingdoms seen today, the new theory suggests.
It is possible, of course, that evolution's early traces have become too faint to decipher. And at the back of researchers' minds is another worry, one that makes them throw up their hands since it cannot be addressed scientifically: that life may have arrived on earth from elsewhere.
Life seems to have popped up on earth with surprising rapidity. The planet is generally thought to have become habitable only some 3.85 billion years ago, after the oceans stopped boiling off from titanic asteroid impacts. Yet by 3.5 billion years ago, according to the earliest fossil records, living cells were flourishing, and there are indirect signs of life even earlier, in rocks that are 3.8 billion years old.
"There's the gee-whiz point of view, how can life possibly have evolved in 300 million years, which I think is still a problem," said Doolittle of Halifax. But life arriving from outer space is a hypothesis, he said, that "leaves you stunned -- there is nothing more you can say after that."
This narrowing window of time may be less embarrassing than it seems. Biologists are warming to the view that the emergence of life from chemical precursors is a quite probable event which does not require billions of years to get under way.
"You put a selective hammer on it and it happens fast," said Norman Pace, an evolutionary biologist at the University of California, Berkeley, referring to the force of natural selection "It's shockingly fast, maybe just tens of millions of years."
Still, many more years of evolution presumably passed before the universal ancestor, a quite sophisticated genetic system, attained its final form.
If the ancestor was a pool of organisms as Woese suggests, and not a definable species, it may be even harder to capture its likeness. But knowledge about this distant era at the dawn of life is moving so fast that few biologists are troubled by setbacks like the Aquifex dead end or the discordant family trees. "I'm unwilling to say we'll never know about anything, because we have come so far in the last two decades," Pace said.
Some family-tree problems, after all, have exact solutions. For example, Doolittle of Halifax wrote recently in commenting on an article by Doolittle of San Diego that they had discovered the reason for their common name: They shared a common ancestor eight generations back.
Other Places of Interest on The Web
TIGR, the Institute for genomic research , Web links. The Age of Science and Techology --Is It Really What We Want?, column by Dr. Carl Woese W. Ford Doolittle home page, Director, CIAR Program in Evolutionary Biology, Professor of Biochemistry, Dalhousie University.