In article <3207BF49.5D1B at santafe.edu>,
Martijn Huynen <mah at santafe.edu> wrote:
>Woese(1987, Microb. Rev. 51 pp 221-271) pointed out that
>the 16S ribosomal RNAs of hyperthermophilic Archaea have evolved
>>Does anybody know other references to this fact, and to possible
It appears that the SSU dendrograms may be artifactual because of compositional
bias. The apparent slow rate of evolution could be due to the fact that the
branches leading to the hyperthermophilic bacteria are incorrectly rooted
deep in the tree. When they are positioned correctly, higher up, the rate of
evolution appears more normal. See,
Woese, C.R., Achenbach, L., Rouviere, P. and Mandelco, L. (1991)
Archael Phylogeny: Reexamination of the Phylogenetic Position of
Archaeoglobus fulgidus in Light of Certain Composition-induced
Artifacts. System. Appl. Microbiol. 14, 364-371.
The problem of compositional bias is well known and has been discussed in
the literature. It's one of the reasons for preferring amino acid sequences
over nucleotide sequences in the construction of phylogenetic trees. (The
bias tends to be damped out when one uses amino acid sequences and the
alignments of coding regions are easier.) The problem is particularly
acute with thermophilic bacteria. As Olson and Woes (1993) put it;
"The phylogenetic placement of thermophilic species by rRNA
analysis is often problematic. The problem lies in the fact that
that rRNAs of thermophilic species tend to be considerably higher
in G+C content that (sic) do those from mesophiles, which tends
to bring about an artificial clustering of the thermophilic
species. Although a casual inspection of the base ratios might
lead one to describe the compositional shift as small (several
mole percent G+C), it must be realized that this may constitute
a large fraction of the sites that have changed."
Olsen, G. and Woese, C.R. (1993) Ribosomal RNA: a key to phylogeny.
The FASEB Journal 7, 113-123.
>I tend to think that their slow evolution has to do with extra
>thermodynamic constraints on their secondary structure, but
>I am probably not the first one to suggest that.
It's likely that the thermophiles have not evolved more slowly than any
other species so there really is no reason to construct hypotheses to
account for this. The idea that the last common ancestor (LCA) was a
thermophile is based largely on the SSU dendrograms and is not supported
by the analyses of protein-encoding genes. For an interesting discussion
on the nature of the LCA see,
Forterre, P. (1996) A Hot Topic: The Origin of Hyperthermophiles.
Cell 85, 789-792.
Forterre is one of the growing number of scientists who question Woese's
Three Domain Hypothesis, which is largely based on SSU trees.