I just logged onto this listing for the first time in months, and read
with great interest the postings
from L . Moran and P Gogarten on the challenge to the three domains
caused by the similarities
between archaeal and Gram-positive genes. I should state at the outset
that I am a believer in the
Woesian worldview, but I think that there needs to be some reconciliation
between it and some
"inconvenient" facts.
We have been looking at nitrogenase genes from methanogens. We (graduate
student Y-T Chien
and myself) recently sequenced the nifD gene from Methanosarcina barkeri
(Mb) which encodes
the nitrogenase subunit containing the FeMo cofactor and its analogs. We
have also sequenced
part of nifK and the late L. Sibold and collaborators sequenced nifH
(1991. Res. Microbiol. 142:
5-12) which encodes the iron protein. Analyzing the amino acid
sequences encoded by these
genes (after removing gaps) using PROTDIST and any of the distance
programs in the PHYLIP
package (as well as PROPARS) leads to the clustering of nif from Mb with
nif genes of C.
pasteurianum (Cp) to the exclusion of other nitrogenases. Specifically,
the "normal" eubacterial
nifs all cluster tightly together, reflecting the high degree of sequence
conservation for which
nitrogenases are famous. Nitrogenase from Frankia, a high-%G+C
Gram-positive, does not
cluster with Cp but is instead in the normal cluster. This is similar to
glutamine synthetase 1.
The Cp nitrogenase has always been considered divergent from other
eubacterial nitrogenases. Its
components form inactive complexes with complementary components from
other nitrogenases.
The nifD product has a 50 AA insert which was not found anywhere else
until we found it in
Methanosarcina, and the nifH product lacks about 15 amino acids at the
C-terminus which are
present in other eubacterial Mo nitrogenases. I should also mention that
both Mb and Cp also have
alternative nitrogenase ORFs which cluster separately in a cluster
containing both eubacterial and
methanogen genes.
So here is yet another example of clustering of archae sequences with
clostridial ones. Moreover,
Frank Robb and collaborators (DiRuggiero, J., et al. 1993. Cloning and
sequencing of glutamate
dehydrogenases from hyperthermophilic archaea. Abst. 93rd Gen. Mtg. Amer
Soc. Microbiol. H-
103 (p 208)) found that GDH from archaes clustered most closely with GDH
from Clostridium
difficile.
Thus, if there was a genetic exchange event, it was a biggie, including
nitrogenase, GDH, GS,
HSP70, and probably other genes. Maybe it was closer to cell fusion
forming a chimera. But,
truth be told, I do not favor genetic exchange. For one thing the %G+C
of the clostridial and
methanogen genes are about 12% different, and the methanogen gene has two
small ORFs
separating the nifH and nifD genes, an arrangement typical of methanogen
nif genes. These ORFs
are not present in Cp. While these can be explained by genetic drift and
rearrangement, Occam's
razor gets duller.
It is also possible that the ancestor of the archaea and eubacteria (I'm
against calling them just
bacteria) had copies of both the clostridial type and the alternative
nitrogenase, and that the
eubacterial Mo-nitrogenase cluster represents a more recent radiation
from the clostridial type, to
which it is more closely related than to alternative nitrogenases.
In any event, I think there is more to this relationship between archaes
and low-G+C Gram-
positives than meets the eye. Could it be that the clostridial sequences
resemble the ancestral
eubacterial sequences more than do those of the proteobacteria most often
sequenced? At least a
partial answer to this would come from getting more sequences of the
earlier-branching oddball
bacteria such as Thermus, Thermotoga, and Aquifex. Already, there is a
V-like ATPase in
Thermus. It could be that the proteobacterial, cyanobacterial, and
actinomycete sequences represent
a Cambbrian-like explosion of microbial evolution.
Well, enough ranting and raving. I would be interested in anyone's
comments on this.
