The Tree of Life: Cold Start?

Click on image to enlarge. Phylogenetic tree (Stetter 1996) modified from Woese et al. (1990), demonstrating the thermophilic root of the last universal common ancestor (in red) but also the putative mesophilic progenote (Forterre and Philippe 1999). The"primitive" nature of FeIII reducers (green) are also evident (Liu et al. 1997; Vargas et al. 1998).

In the late 1970s, Carl Woese, at the University of Illinois at Urbana-Champaign, revolutionized our view of life on Earth. Comparing among many different organisms the sequence of a gene that encodes a ribosomal RNA (rRNA), Woese drew the first comprehensive tree of life.

One surprise in the tree was the appearance of bacteria that thrive in high temperatures (hyperthermophiles), near or at the root of the tree. The tree, and supporting research from other fields, led to the speculation that life originated in very hot environments, perhaps in the hydrothermal vent systems found deep underwater around the globe. Woese’s original tree, and the resulting speculation that life arose in a hot environment have become widely accepted among researchers, and have taken on the status of textbook explanations of the origin of life.

"I’ve always felt uncomfortable with that," says Arizona State University geologist Jack Farmer. "You look at the RNA tree and you see that all of the deep branches are organisms that grow at temperatures above 80 C (176 F). A lot of people have taken that to mean life must have originated at high temperatures. But, for several reasons, I’ve felt that it’s more likely that life originated at low temperatures. If we could find a low-temperature organism that’s even more primitive, that would prove the point."

Early in 2002, Celine Brochier and Herve Philippe may have given Farmer just that low-temperature organism, and a surprising one at that. They report their results in the May 16, 2002 issue of the journal Nature.

Brochier and Philippe, then both of Universite Pierre et Marie Curie, Paris, revisited the rRNA tree of life, correcting for an error that can creep into the computer programs used to convert the raw gene sequences into a tree diagram. Dubbed "long-branch attraction," the error can shift the apparent positions of certain organisms in the tree of life.

Fixing the Artifact

"The effort to address long-branch attraction is enormous," says Philippe, now at the Universite de Montreal. "Presently, most phylogenetic papers discuss, at least a little, the problem. Much theoretical work is also done on this question."

Brochier and Philippe approached reconstruction of the tree of life by paring down the parts of the rRNA gene to include in the analysis. They chose parts of the gene that changed the least over time. This decision sidesteps some of the errors that can affect less conservative approaches.

Long-branch attraction arises because the assumptions in computer models used to construct evolutionary trees underestimate the rate of DNA change. If the part of a gene under study evolves more quickly than the average, the problem is even worse. Choosing slowly evolving locations helps match the data to the assumptions. The approach reduces systematic errors.

The downside of using less data is that the approach increases random error. "This is why we plan to use more genes in the future," Philippe says.

The tree they came up with differs in some striking ways from the one that has become standard. First, the hyperthermophilic bacteria do not appear near the trunk of the tree. Instead the reanalysis shows them up among the leaves, part of a large, bushy crown group that includes most of the bacteria.

In a way, this result does not come as a surprise to Farmer. "There’s a lot of adaptations required to live at high temperatures, designed basically to keep the molecules together. They tend to fall apart at high temperatures. So I’ve always felt more comfortable that it was a secondary set of adaptations some time after life had originated that allowed these organisms to radiate into high temperature environments," Farmer says.

Aquifex Aeolicus, an example of a hyperthermophilic organism, is known to survive in temperatures as high as 96 C.
Credit: K.O. Stetter & Reinhard Rachel University of Regensburg

Philippe agrees. "One could have guessed from the discovery of hyperthermophilic organisms that they are specialized, and thus very evolved." In Philippe’s analysis, the statistics for finding hyperthermophiles near the trunk of the tree are weak.

Other research supports the doubt cast on hyperthermophiles as common ancestors. Research from a French and American group, Nicolas Galtier, Nicolas Tourasse and Manolo Gouy, suggests that the subunit, or nucleotide, makeup of rRNA in the earliest ancestral organisms is incompatible with life at high temperatures.

Furthermore, a key enzyme that helps protect bacterial DNA in high-temperature environments was acquired by horizontal gene transfer from Archaea. When the genome of the hyperthermophilic bacterium Thermotoga maritima was completed, 24 percent of its genes showed more similarity to archaeal genes than to bacterial genes. Similar studies on other hyperthermophilic bacteria produced similar results.

But the real surprise is the organism that did place close to the trunk of the tree of life in Brochier and Philippe’s reanalysis: an unusual bacterial group called planctomycetes, which thrives only in moderate temperatures.

Planctomycetes are intriguing because they combine features found in all three domains of life: Archaea, Bacteria and Eukarya. The textbook distinction among the three domains of life goes like this: Eukaryotic cells all have their DNA packed into a nucleus with a double membrane; bacterial and archaeal cells have no nuclear membranes. And many biochemical differences separate Archaea and Bacteria, including a key component of their cell walls.

Planctomycetes reproduce by budding, like the yeast shown above.
Image Credit: SPL

Planctomycetes play differently. Instead of dividing in two like most bacteria, they reproduce by budding, a little like yeast. They lack that key cell-wall chemical found in all other bacteria, but not found in archaeans and eukaryotes. The punch line, though, is the presence in planctomycetes of a nuclear membrane. In some species, it’s a single membrane, in others a double membrane.

Brochier and Philippe point out that no common evolutionary origin, or homology, can be proved between the nuclear membrane in planctomycetes and the nuclear membrane in eukaryotes. But, he says, "Homology has not been ruled out." At the very least, the existence of the nuclear membrane in planctomycetes should change the textbook definition of "eukaryote" to include complexity beyond the mere presence of a nuclear membrane.

What’s Next

Some bacteria are so easy to grow in a lab, or culture, that microbiology classes routinely culture species from the environment, as well as commonly studied laboratory strains. But our knowledge of bacteria is, in general, very narrow. Only about 5 percent of all known bacteria have been cultured. In many phyla, no species have been cultured. Scientists know about so-called "uncultured" bacteria only from the study of RNA contained in environmental samples. Philippe’s current work excluded the uncultured groups.

"Several people have asked us where these phyla would emerge in our reanalysis. We plan to do this soon. This is potentially an important problem, since at least 50 percent of the bacterial phyla contain only uncultured organisms," Philippe says. The team also plans to study more genes when the complete genomes of some planctomycetes species become available. Another possible approach, Philippe says, is to study the genome of a non-hyperthermophilic member of a group that is normally hyperthermophilic, to see whether the group is primarily or secondarily adapted to low temperature.

And, of course, the reanalysis Brochier and Philippe present in the May Nature paper must be confirmed. If it is, the odd, cool-temperature planctomycetes may have more to teach us about Earth’s primitive organisms than do hyperthermophilic bacteria. "If our finding is verified," Brochier and Philippe conclude in the paper, "the origin of Bacteria should be seriously reconsidered."

Further Reading

"A Non-Hyperthermophilic Ancestor for Bacteria", Brochier and Philippe, Nature, 16 May, 2002
"A Nonhyperthermophilic Common Ancestor to Extant Life Forms", Galtier, Tourasse and Gouy, Science, 8 January, 1999
"Ancient Phylogenetic Relationships", Gribaldo and Philippe, Theoretical Population Biology, 61, 391-408 (2002)
"Cell compartmentalization in planctomycetes: novel types of structural organization for the bacterial cell", Fuerst, et. al., Arch. Microbiol (2001) 175: 413-429
"Early-branching or fast-evolving eukaryotes? An answer based on slowly evolving positions", Philippe, et. al., Proc. R. Soc. Lond. B (2000) 267, 1213-1221
"Reverse gyrase from hyperthermophiles, probable transfer of a thermoadaptation trait from Archaea to Bacteria", Forterre, et. al., TIG (Trends in Genetics) April 2000, volume 16 No. 4
"Where is the root of the universal tree of life?", Forterre and Philippe, BioEssays 21:871-8790, 1999