Time to abandon the progenote hypothesis!!

arlin at ac.dal.ca arlin at ac.dal.ca
Wed Nov 3 09:40:17 EST 1993

Sorry, folks, but I'm not convinced.  Its scarcely tenable that the appearance
of 'progenote' in the literature is due to a pervasive inability to spell the
word 'progenitor'.  Instead, 'progenote' is typically accompanied by a reference
to the original work of Woese & Fox, or to some more recent reference by Woese
[by the way, my first post left the impression that the 'progenote' concept was
attributable to Woese alone, when in fact the theoretical concept (if not the
term) was the brainchild of both Woese and Fox (1977, J.Mol.Evol 10: 1-6)].

Several people have defended Woese's progenote hypothesis on the grounds that
they haven't seen enough evidence to refute it conclusively, but this hesitancy
is (I hope) based on a misunderstanding, since:

  "the progenote could neither have nor have evolved 'modern' proteins.
   Its proteins would have been small or of non-unique sequence or both."
   [Woese 1987, Microbiological Reviews 51: p. 263]  

In fact, most of us *have* seen *more* than enough evidence to refute the idea
that the ancestor of eukaryotes, eubacteria and archaebacteria had such
properties.  If this ancestor was progenotic (rather than genotic), then euks,
eubs and archaes must have each evolved many of their *apparently* homologous
genes by convergence, since they have long (1-2 kb, typically) DNA genes,
whereas progenotes have short (100 bp? 500 bp?) RNA genes.  Does anyone really
believe that the genes for EF-Tu proteins (for instance) of eubacteria,
eukaryotes, and archaebacteria are not *really* homologous along their entire
length, but instead evolved by convergence from a short ancestral RNA gene?

If so, then it is pointless to use such molecules to draw phylogenetic
conclusions about eukaryotes, eubacteria and archaebacteria, since the
phylogenetic inferences are based on the assumption of homology (i.e., they use
aligned sites, assumed to be homologous).  If the progenote hypothesis is true,
such trees tell us little.  Likewise, it would be pointless to make
trees from rRNAs (large or small subunit), shared ribosomal proteins (a dozen of
them), RNA polymerase subunits (2 of them), GAPDH and other dehydrogenases
(several), DNA polymerase, ATPase subunits, etc.  It doesn't really salvage the
underlying logic of the progenote hypothesis to say that only one of the groups
made the progenote-genote transition, then generously passed a vast array of
long DNA genes to the other still-struggling progenotes, since the whole point
of the progenote hypothesis was that archaes, euks and eubs are so dissimilar
(not!) that they must have evolved their molecular biology independently.


More information about the Mol-evol mailing list