It seems to me that Gupta (and others) are throwing out the child with
the bath-water. The finding that all archaebacteria studied so far contain many
genes that are similar to their eukaryotic counterparts and that these
archaebacterial genes are very dissimilar to their homologues found in other
prokaryotes certainly sets the archaebacteria as a group clearly and distinctly
apart from all of the eubacteria. One should keep an open mind as to whether
the currently known archaebacteria are mono- or paraphyletic, the above
mentioned genes strongly argue that a major component of the archaebacteria is
not polyphyletic with respect to the eubacteria. Given that deeper branching
archaebacteria are being discovered, it seems only a question of time that sooner
or later a group of archaebacteria will be discovered that is closer to the
eukaryotes than other archaebacteria (and many will argue that the eocytes
already fit this description). Yes, there are major distinction between pro- and
eukaryotes; however, this should not lead one to overlook the major differences
between archae- and eubacteria.
That some genes (e.g. HSP70s and glutamine synthetases) do not reflect a
"fundamental distinction" between archae- and eubacteria cannot and should not
be ignored; however, one can hardly take these genes and ignore all the
characters that define the archaebacteria as a distinct group and claim the
archaebacteria should be considered a part of the gram positives. [Following this
discussion it might be news to some that not all genes that encode enzymes
involved in biosynthetic pathways and bioenergetics group the archaebacterial
homologues among their eubacterial counterparts. Examples that group the
archaebacterial homologues as a distinct group include ATPsynthases,
cytochrome C oxidases, and argininosuccinate synthase.]
The genes that group the archaebacteria among the eubacteria can be put into
two categories (often it is not clear which one of these, because the rooting of the
respective phylogenies is controversial):
Category 1: genes that were contributed to the eukaryotic cell either via the
mitochondrial endosymbiont (R. Hensel and F. Doolittle think that many
glycolytic enzymes belong into this category) or via an earlier eubacterial
contributor to the eukaryotic cell (in my opinion the latter alternative is still mainly
based on wishful thinking). For this category the alpha purple bacterial genes
appear as the sister group to the eukaryotic homologues. The deeper branches,
which also contain the archaebacteria, often are not well resolved.
Category 2: well resolved phylogenies that group the archaebacterial genes
among the eubacteria. The longest internal branch connects all the prokaryotes
to all of the eukaryotes (e.g. HSP70). Midpoint rooting would place the root in
this longest internal branch. I think that the best explanation for these
phylogenies is horizontal transfer of genes from a eubacterium to the
archaebacteria. (Gupta et al.'s explanation needs to assume major variations in
substitution rate to fit the data, without solving the puzzle of the close
association between archaebacteria and gram positives).
Assuming horizontal gene transfer, or the fusion of formerly
independent lines of descent, certainly complicates the interpretation of
molecular phylogenies. However, given that some examples exist that
demonstrate cross-domin horizontal transfer, and that horizontal transfers and
symbioses between extant organisms occur frequently, it seems strange to
assume that these events did not play an important role also in the early
evolution. If one looks at the data from a less "eukaryocentric" perspective one
recognizes that these processes also played an important role in the evolution of
the other cellular lineages. Just because it complicates things, or because the
remnants of these events have not left distinct cell organelles (like in the case of
mitochondria and chloroplasts), does not mean it has not happened. Given the
limited resolution of molecular phylogenies it is at present difficult to pinpoint
these events on an organismal tree of life, or even to decide how many
independent events occurred or of which magnitude (i.e. how many genes were
simultaneously transferred) these events were. At one extreme one can envision
the formation of a chimera at the root of the archaebacterial domain, at the other
extreme is a scenario with many single gene transfers involving eubacteria
(mostly gram positives, but also a variety of gram negatives) and different
archaebacteria. The finding of extensive horizontal gene transfer between eu- and
archaebacteria should not distract from the major horizontal transfer (fusion)
events associated with the emergence of the eukaryotes. However, the
accumulated molecular data indicate that these events were not restricted to the
origins of eukaryotes.
Peter Gogarten
PS.: A more detailed discussion off the roles of horizontal transfer and
paralogous genes is/will be published by J. Peter Gogarten, Elena Hilario, and
Lorraine Olendzenski (1996) "Gene Duplications and Horizontal Gene Transfer
during Early Evolution." In: "Evolution of Microbial Life", Society for General
Microbiology Volume 54
