I have been thinking about the various comments concerning the timing of the
divergence of the members of Bacillus (Gram positive bacteria) from the members
of the Proteobacteria (Purple bacteria). As has been pointed out, it is very
unlikely that the 16s rRNA chronometer is completely smooth running. It is
likely that the split occurred after 3.5 billion years BP, but before 1.5
billion years BP. What does this tell us? Not much. Some comments in this
discussion refer to the appearance of stromatolites at 3.5 billion yr BP. It
is suggested that such formations, containing filamentous bacteria forms,
contain Cyanobacteria. I doubt this very much. First, I challenge anyone to
distinguish between fossilized Cyanobacteria and fossilized Beggiatoa and
fossilized Chloroflexus, etc. Those of you with a general microbiology
background will recognize that Beggiatoa is not a phototrophic organism, but a
chemolithotrophic organotroph (mixotroph); Chloroflexus is a green sulfur
bacterium, which only has photosystem I. My point is that it is very difficult
to distinguish from the fossil record just when a bacterial group evolved.
In fact, I would say that it is impossible. As such, we are left with the
imperfect chronometer of rRNA.
It should also be pointed out that "modern" cyanobacteria, that is those with
both photosystem I and II could not have been around at 3.5 billion yr BP, but
must have evolved by 1.5 billion yr BP, since O2 began to exist in the
atmosphere by that time.
So back to the original question, when did Bacillus (or Gram + bacteria)
diverge from the Proteobacteria? Well, by the most recent trees that I have
seen (see J. Bacteriol. 1994. #1, pg. 1-4), it would appear that the Gram +
split before the Cyanobacteria did. Of course, maybe even this interpretation
is based on whose algorithm was used to generated the tree. Who knows?
Finally, we always have the question of how much genetic exchange still goes on
in the microbial community. How promiscuous are the various plasmids and the
assorment of insertion elements/transposons that reside on them and else where?
I bring this up because these processes (transformation, conjugation,
transduction) could really screw up an incorrectly picked chronometer, if they
occur as frequently as some fear. Well, this point may launch another
David R. Nelson
Department of Biochemistry, Microbiology, and Molecular Genetics
University of Rhode Island
Kingston, RI 02881