Is Carl Woese losing a Kingdom?
lamoran at gpu.utcc.utoronto.ca
Wed Oct 2 15:49:37 EST 1996
Radhey S. Gupta wrote,
>The two main tenets of the current "three domain
>phylogenetic view" are:
>(i) Archaebacteria (or Archaea) constitutes a monophyletic domain
>which is completely distinct from the rest of the bacteria.
>(ii) Eukaryotic nuclear genome has directly descended from an
>However, both of these tenets, which were initially proposed
>based on either 16S rRNA data or elongation factor (EF-1,EF-2)
>sequences are not supported by other recent well characterized
This point cannot be overemphasized. There is very solid data that does
*not* support the two main postulates of the Three Domain Hypothesis.
We should not ignore this data.
>In terms of the relationship within the prokaryotes, if one
>examines all available protein sequences two common patterns are
>observed. (a) For a number of proteins that in general are not
>highly conserved in all prokaryotes (e.g. EF-1 /Tu, EF-2/G, RNA
>polymerase II and III subunits, GroEL/Tcp-1, etc.), the
>phylogenies based on such sequences generally indicate a distinct
>grouping of archaebacteria and eubacteria.
Gupta and others have pointed out that the main support of the Three Domain
Hypothesis comes from the analysis of genes that are not highly conserved.
Another way of putting this is to say that the dendrograms are questionable.
In some cases (eg. EF-G and RNAP subunits) there are several versions of
dendrograms that have been published. It is a mistake (IMHO) to assume that
the interpretation of this data unambiguously supports Woese's scheme.
>(b) On the other hand
>for several other protein sequences that are highly conserved
>viz. Hsp70, glutamine synthetase, glutamate dehydrogenase,
>FGARAT, FtsZ, etc., a specific relationship of archaebacteria to
>the gram-positive group of bacteria is observed. Such a
>relationship is supported by different phylogenetic methods as
>well as by the signature sequences in these proteins.
As Gupta points out, there is some very good data that simply refutes the
main tenets of the Three Domain Hypothesis. Analysis of these sequences
clearly shows that the archaebacteria cluster within the prokaryotes and
are most closely related to the gram positive bacteria. These papers have
been in the literature for many years but they are consistently ignored by
many workers who assume that the archaebacteria are related to eukaryotes.
The question is why has the Three Domain Hypothesis become so widely
accepted in spite of this data? Can anyone think of another scientific
hypothesis that has become *more* popular when contradictory data is
>a number of these gene sequences where sequence information from
>adequate number of archaebacterial species is available (viz.
>Hsp70, glutamate dehydrogenase), the various archaebacterial
>species do not form a monophyletic group as analysed by different
>phylogenetic methods. Instead they show a polyphyletic branching
>within the gram-positive bacteria. These results do not support
>the first tenet of the three domain hypothesis. It is of interest
>in this regard that before Woese's proposal that "the
>archaebacteria constitutes the third form of life",
>archaebacteria were traditionally classified with the gram-
>positive group of bacteria and many of them show a gram-positive
>staining characteristic. A close relationship between
>archaebacteria and gram-positive bacteria has also been suggested
>by Cavalier-Smith based on other characteristics.
I have nothing to add here. Statements such as this need to be repeated
again and again because the lesson doesn't seem to have sunk in. Why
do so many people believe that the archaebacteria constitute a third
domain of life when there is good evidence against it?
>It should also be pointed out that in the phylogenetic trees
>based on 16S rRNA on which the current prokaryotic phylogeny is
>almost entirely based, both gram-positive and gram-negative
>bacteria show polyphyletic branching within each other. However,
>most of the critical nodes pointing to such relationships
>generally have very low bootstrap scores ( 30-50%) and hence are
>statistically not supported. In contrast, phylogenies and
>signature sequences based on a number of protein sequences (viz.
>Hsp70, GroEL)indicate a clear distinction between gram-positive
>and gram-negative bacteria and such a relationship is
>statistically strongly supported.
In other words the SSU data is not nearly as reliable as most people
believe. Thanks to Gupta and Golding and others for having the courage
to publish such heretical views. Now if only people would listen ....
>On the other hand for a number of highly conserved protein
>sequences (e.g. Hsp70, glutamine synthetase, glutamate
>dehydrogenase, FGARAT, etc.) a very different sort of
>relationship is observed. In these cases the eukaryotic homologs
>show a closer relationship to the gram-negative bacteria, and in
>some cases (viz. Hsp70) are clearly derived from gram-negative
This is where I part company with Gupta. I am not a fan of the Three
Domain Hypothesis but I also don't agree with Gupta's interpretation
of some of the other dendrograms. For example, with respect to HSP70
sequences the data shows that the dnaK (HSP70) genes from gram negative
bacteria are more similar to those from other bacteria than eukaryotes.
Thus the tree looks like this according to my analysis,
|-------------- gram positive
| |-------------- archaebacteria
| |------------------------- gram negative
The rooting of this tree of life depends on the extent of overall similarity
and also on outgroups such as distantly related members of the HSP70 family.
If this tree is correct then the "problem" group is not the eukaryotes but
the archaebacteria. In order to reconcile the conflicting data it is possible
to imagine that the archaebacteria are mosaics of genes acquired by lateral
transfer from primitive eukaryotes and gram positve bacteria.
>in the case of Hsp70
>sequences, which constitute the most conserved protein known to
>date, and where all of the euaryotic homologs have been
>extensively characterized and clearly distinguished from each
>other based on numerous sequence signatures, it has been possible
>to establish beyond any doubt that the eukaryotic nuclear genes
>for this protein are derived from gram-negative bacteria and that
>this relationship is not due to any sort of horizontal transfer
>or genetic exchange between the sequences.
I disagree. The eukaryotic HSP70 genes are no more closely related to the
gram negative bacteria than to gram positve bacteria or archaebacteria.
The gram negative bacterial genes are more similar to those from other
bacteria than they are to eukaryotic genes. The dendrogram suggests strongly
that the last common ancestor of all life gave rise to all prokaryotes on
the one hand and all eukaryotes on the other.
My disagreement does not affect Gupta's point raised above concerning the
validity of the Three Domain Hypothesis. The Three Domain Hypothesis is
in serious trouble.
>To explain these mutually discordant phylogenies, we have
>proposed (an extension of Zillig's model) that the ancestral
>eukaryotic cell arose by a unique endosymbiotic fusion event
>involving an archaebacteria (eocyte, based on Lake's data on EF-1
>and EF-2 sequences) and a gram-negative bacteria. Following the
>fusion event, at an early stage, an assortment of genes from the
>two fusion partners took place and the resulting cell acquired
>and characteristics from each of the two parents. (see our recent
>TIBS article May, 1996, for further details of the model).
If we are forced to invoke fusion events to rescue what's left of the
original Woese model then I believe that it is more consistent with the
data to postulate that the archaebacteria arose from a fusion of a
primitive gram postive organism and a primitive eukaryote.
>The question may be asked as to why should the chimeric
>model be preferred over direct descendance of eukaryotic cells
>from archaebacteria? Well, there are several reasons for doing
>(a) It is the simplest and most parsimonius model to explain all
> of the genes/proteins phylogenies.
My chimeric model seems simpler and more parsimonius than yours! (-:
>(b) It readily explains why certain characteristics of
> eukaryotic cells (e.g. components of transcription and
> translation machinery) are similar to archaebacteria, while
> others are clearly derived from eubacteria (e.g. ester-
> linked straight chain membrane lipids, fatty acid
My model readily explains why certain characteristics of archaebacterial
cells are more similar to those in eukaryotes. It's because the
archaebacteria have acquired eukaryotic genes. The difference in models
depends on whether one tries to rationalize the eukaryotic genome or the
>(c) It provides a plausible explanation for the origin of the
> eukaryotic cell nucleus and endomembrane systems (Discussed
> in our recent TIBS article ) and is supported by the
> observed duplication of chaperones genes (e.g. Hsp70,
> Hsp90)which accompanied this event.
My suggestion accounts for the origin of the archaebacteria and explains
why they seem to be anomolous. The evolution of eukaryotes is not relevant.
>(d) It provides a plausible explanation for the enormous
> structural difference between the eukaryotic and prokaryotic
> cell types, and the absence of any species that are
> intermediates in this transition. These observations cannot
> be readily explained by simple evolutionary mechanisms. By
> contrast, a sudden and major evolutionary transition could
> readily be explained by fusion and subsequent gene
> assortment of two very different species.
In my model eukaryotes and prokaryotes are the fundamental divisions of
life and because this split took place before any other divisions it is
not surprising that there are enormous structural differences.
>(e) The inferred time of divergence of eukaryotic species from
> archaebacteria and the eubacteria (about 2 By ago) based on
> genetic distances between different proteins sequences
> (Doolittle et al , Science, 271, 470-477, 1996) and the
> fossil records, can also be satisfactorily explained by the
> chimeric model.
>(f) The phylogenies and signature sequences in various
> eukaryotic genes also provide compelling evidence that all
> extant eukaryotic species are monophyletic. This in turn
> suggests that the postulated fusion event that gave rise to
> the ancestral eukaryotic cell was unique and a successful
> fusion between the prokaryotic parents took place only once
> in the 3.8 By history of life on this planet. Had this
> chance fusion event not taken place, our landscape perhaps
> would have looked similar to that of Mars and its sole
> inhabitants would have been prokaryotes.
You are trying to explain the origin of eukaryotes by postulating a fusion
between a primitive archaebacteria and a primitive gram negative bacterium.
I am suggesting that all eukaryotes arose from a primitive ancestor that
diverged from its prokaryotic cousin a very long time ago. Subsequently
some eukaryotic genes were transferred to archaebacteria. It is the
archaebacteria that are chimeric.
>the basic division of living organisms into two primary
>groups - the Prokaryotes (ancestral) and the Eukaryotes (derived) is
>strongly supported by all available molecular and other data.
I agree that the basic division is prokaryote/eukaryote but I disagree
that eukaryotes are "derived".
>The above observations thus challenge the two main tenets of the
>current "three domain paradigm" which has resulted from Carl
>Woese work. However, all of the above observations which
>disagree with this view have not been dealt with by the
>supporters of this paradigm. This pattern is in accordance with
>what Woese has recently written
> "When facts or concepts arise that challenge (a
> paradigm), these tends to be ignored. If that is not
> possible, they are scoffed at and otherwise communally
> rejected- informally. Only as a last resort will the
> paradigm formally contest the novelty that threatens it
> (i.e. treat it scientifically)".
I have used this quotation myself in earlier discussions on bionet.molbio.
evolution. Let's hope that we move soon to the third stage of paradigm
More information about the Mol-evol