Research NewsEvolution

Genome Data Shake Tree of Life

See allHide authors and affiliations

Science  01 May 1998:
Vol. 280, Issue 5364, pp. 672-674
DOI: 10.1126/science.280.5364.672

New genome sequences are mystifying evolutionary biologists by revealing unexpected connections between microbes thought to have diverged hundreds of millions of years ago

Over the past 3 years, the deciphering of the complete genetic codes of more than a dozen microbes has opened the way to a whole new understanding of how bacteria live and cause disease. “It's like being in a candy store,” says Richard Stevens, at the University of California, Berkeley. But on one front—the study of evolution—the information pouring out in the genome sequences has so far proved more confusing than enlightening. Indeed, it threatens to overturn what researchers thought they already knew about how microbes evolved and gave rise to higher organisms.

For more than 2 decades, systematic biologists, led by evolutionist Carl Woese at the University of Illinois, Urbana-Champaign, have been using the sequences of RNA from the ribosomes—the cell's protein-making factories—to classify bacteria. One of the stunning successes of rRNA analysis was Woese's identification of one group of microbes, now called the archaea, as a third kingdom of life, adding to the two kingdoms already known: the true bacteria, which like the archaea don't have cell nuclei, and the eukarya, including higher plants and animals, which do (Science, 2 May 1997, p. 699).

Since then, he and others have used rRNA comparisons to construct a “tree of life,” showing the evolutionary relationships of a wide variety of organisms, both big and small. According to this rRNA-based tree, billions of years ago a universal common ancestor gave rise to the two microbial branches, the archaea and bacteria (collectively called prokarya). Later, the archaea gave rise to the eukarya. But the newly sequenced microbial genomes and comparisons with eukaryotic genomes such as yeast have been throwing this neat picture into disarray, raising doubts about the classification of all of life.

For one, because genes don't evolve at the same rate or in the same way, the evolutionary history inferred from one gene—say for rRNA—may be different from what another gene appears to show. “Before, people tended to equate rRNA trees with the [life history] tree of the organism,” says John Reeve, a microbiologist at Ohio State University in Columbus. “From the whole genomes, you very quickly come across [genes] that don't agree with the rRNA tree.”

Even more perplexing, the newly unveiled genomes often contain a mix of DNAs, some seeming to come from the archaea and others from bacteria. “Features of both bacteria and archaea are turning up in eukaryotes, and to a surprising degree,” says Russell Doolittle, a molecular evolutionist at the University of California, San Diego.

Many evolutionary biologists are coming to believe that these mosaics arose because genes hopped from branch to branch as early organisms either stole genes from their food or swapped DNA with their neighbors, even distantly related ones. The genetic oddballs may simply mean that the branches of the tree of life intertwine, but that the basic shape is sound. But if the gene swapping was extensive enough, the true branching pattern may be quite difficult to discern. Worse, the tree's “base” may turn out to be indecipherable: a network of branches that merge and split and merge again before sprouting the modern kingdoms. It may be, Woese concedes, that “you can't make sense of these phylogenies because of all the [gene] swapping back and forth.”

Confounding genes

The just-completed sequence of the bacterium Aquifex aeolicus, which lives at near-boiling temperatures, embodies the problems that molecular evolutionists are now confronting. To assess Aquifex's kinship to its fellow microbes, molecular geneticists Ron Swanson and Robert Feldman of Diversa Corp. in San Diego and their colleagues, who described the sequence in the 26 March issue of Nature, compared several of its genes with their counterparts in a range of species from the archaea and eukaryotes as well as other bacteria. The conclusion, Feldman reported in early February at the Conference on Microbial Genomes in Hilton Head, North Carolina, is that “you get different phylogenetic placements based on what genes are used.”

The gene for a protein called FtsY, which helps control cell division, placed Aquifex close to the common soil microbe Bacillus subtilus, even though the two supposedly come from different branches of the bacterial tree. Even worse, a gene encoding an enzyme needed for the synthesis of the amino acid tryptophan linked Aquifex with the archaea. That wasn't the only anomaly the Diversa team found regarding the archaea, however. Their analysis of the gene encoding the enzyme CTP synthetase, which helps make the building blocks of DNA, spread the archaea out among all the other organisms evaluated, suggesting that they may not be as coherent and distinct a group as the rRNA tree implies. “It points to caution in terms of interpreting the 16s [rRNA],” Feldman concludes.

The gene for FtsY and perhaps the gene for CTP synthetase might have been expected to tell a different story from the rRNA genes, because they have probably evolved at different rates. Woese picked the rRNAs to study because they are part of one of the cell's most basic activities—protein synthesis—and thus are unlikely to change radically. He hoped they would serve as a slow, steady clock. But genes not involved in such core activities, including those for FtsY and CTP synthetase, may evolve fast or slowly, depending on the different conditions microbes live in. “Each gene has its own history,” says Feldman.

If differences in the way genes evolve account for some of the disparities between the patterns Feldman and his colleagues traced, the real tree of life might be worked out by overlaying different gene trees to come up with a consensus. “The sum of all these trees makes up the organism,” suggests biochemist Dieter Söll of Yale University in New Haven, Connecticut.

But the tree Feldman derived from the gene for the tryptophan synthesis enzyme implies a more insidious problem: the possibility of widespread gene swapping among organisms, which could make arriving at a consensus tree quite difficult. Few researchers think Aquifex is kissing cousin to archaea, in spite of the similarity of their genes for this enzyme. The enzyme might be a relic from the ancestor common to both kingdoms, which has evolved unexpectedly slowly since then. But more likely, at some point, Aquifex took on the archaeal gene, substituting it for its own version, a process called lateral transfer.

Gene swapping

Not too long ago, attributing an unusual result to lateral transfer would have raised quite a few eyebrows. For years molecular evolutionists tended to use this idea to excuse irregularities in their attempts to construct phylogenetic trees, when actually their methods were at fault. But the microbial genomes have made the idea respectable.

In one case last year, for instance, W. Ford Doolittle, an evolutionary molecular biologist with the Canadian Institute for Advanced Research in Halifax, Nova Scotia, scanned the genome of an archaeon called Archaeoglobus fulgidus, newly sequenced by The Institute for Genomic Research (TIGR) in Rockville, Maryland, with a computer program that searches new genomes for resemblances to genes in existing databases. The scan turned up an enzyme called a reductase that was much more like the reductases seen in bacteria than like comparable enzymes in other archaea and eukaryotes—supposedly the closer relatives of A. fulgidus.

Similarly, an unusual gene that Söll, Yale biochemist Michael Ibba, and their colleagues first spotted in the genome of the archaeon Methanococcus jannaschii has since turned up in the Lyme disease pathogen, a spirochete called Borrelia burgdorferi. Because spirochetes are thought to be descendants of bacteria that had a different version of the gene, “we believe [it] arose from a lateral transfer,” Söll says, in which the spirochete took up an archaeal gene and lost its original one.

Conversely, other researchers are finding archaealike genes in microbes classified as bacteria. Take Treponema palladium, the spirochete that causes syphilis. After Steven Norris of the University of Texas School of Medicine in Houston, working in collaboration with TIGR scientists, completed the Treponema genome last year, they noticed that its DNA contains genes for two particular ATPases—enzymes that break down adenosine triphosphate, often to release energy—known before to exist only in the archaea. And Treponema also has other genes that look suspiciously archaeal in origin, he reported at the microbial genomes meeting.

Revising history

In an upcoming issue of Trends in Genetics, Ford Doolittle proposes a new mechanism for this kind of gene swapping. He suggests that early eukaryotes may have gotten a significant part of their genomes from genes picked up by their predecessors from their food. As he puts it, “You are what you eat.”

Assuming that the current tree of life is correct, he asks, how else can one explain Russell Doolittle's conclusions that 17 of 34 families of eukaryotic proteins that date back to early cell evolution look as if they come from bacteria, while only eight show a greater similarity to archaea, the supposed ancestor of eukarya. Terry Gasterland's team at Argonne National Laboratory, outside Chicago, Illinois, has made a similar finding: Twice as many yeast nuclear genes match up with bacterial genes as with archaeal genes.

Although some modern bacteria are quite adept at taking up new genes—many pathogens develop antibiotic resistance this way—the successful incorporation of genes from food bacteria into eukaryotic genomes would be accidental and infrequent. But “we've got hundreds of millions of years for it to happen,” Ford Doolittle points out. Also, these genetic morsels are consumed with each meal, so that an incoming gene can have many opportunities to get into the genome and replace its native counterpart. In contrast, once a native gene happens to get removed from the host's genome, “it's lost forever,” he adds. Over evolutionary time, these processes would favor the loss of native genes and their replacement with borrowed ones.

Woese thinks gene swapping was rampant even among life's earliest organisms. In his view, the organisms that lived before archaea, eukarya, and bacteria went their separate ways lived communally. “It was more like a consortium,” Woese says of this very early world. The ability to make use of a neighbor's genes would have proved an important advantage, he asserts.

Members of this consortium may even have had different genetic codes. But the organisms that outlasted the rest would have been those that could make use of their neighbor's genes to adapt to changing conditions, says Woese. Over time, this advantage “ensured that the [DNA] code was universal,” he says, because those not able to read DNA-based genes could not survive as well as those organisms using DNA.

This prehistoric commune might have worked well for early life, but it adds to the challenge for biologists trying to make sense of it all. With each descendent from this community “having taken up different things from the ancestor, you won't be able to draw clear trees,” Woese points out. He still has faith, however, that organisms roughly followed the patterns of evolution seen in changes in rRNA and that the three kingdoms will remain intact.

However, the existence of so many genes that seem out of place has led some researchers to question whether eukarya descended from archaea. These researchers are also wondering whether archaea really are distinct from true bacteria, noting that although archaea were once considered limited to extreme environments, they are also turning up in the milder surroundings favored by true bacteria (Science, 24 April, p. 542). “I think it's open whether the three domains will hold up,” says Feldman.

Despite the current ferment, however, Woese and others are confident that eventually a consistent picture of microbial evolution will emerge—even if what it might look like is uncertain, and even if its base is the mix of communal organisms Woese envisions. Within a year, some two dozen more genomes will be complete. At the same time, new software programs are refining researchers' ability to trace the heritages of different genes and discover more links between the three kingdoms. All this, says Texas's Norris, “will lead to a much better understanding of evolution as a whole.”

View Abstract

Navigate This Article