>I'm no expert on Molecular Biology, but my mentor in that field tells me that
>horizontal gene transfer has made Hox genes, and perhaps other genes,
>unreliable
>indicators of phylogenetic affinities. She cites a classic example in which
>the
>fruit fly, Drosophila, was found to have the "Eye" gene which controlled
>development
>of eyes in flies. The same gene was found to control eye development in the
>mouse.
>But an even greater surprise came to find that C. elegans, the nematode
>workhorse
>animal of Biologists, also had the Eye gene, even though that animal has no
>eyes.
>
> She said this strange development had to be explained by interspecies
>gene
>transfer, which has been observed in the laboratory. This phenomenon seems
>to have
>dimmed the bright hope that molecular affinities would enable certitude for
>determining relatedness in construction of phylogenetic trees.
>
Actually, nematodes such as C. elegans are more closely related to
arthropods such as Drosophilia than either is to mice, so the common use of
the gene in mice and Drosophilia is more surprising. Lateral gene transfer
does not seem likely in this case; rather one of two other possibilities.
On the one hand, the ancestral bilaterian may have used the Hox Eye gene
for eyes. It thus survives in most if not all bilaterians, whether they
have functional eyes or not. Hox genes are generally rather important in
development, so major mutations there are likely to cause serious problems.
Thus, the Eye gene can remain recognizable in nematodes that no longer use
it. The other possibility is that the ancestral bilaterian had the
ancestor of the Eye gene used for something that can easily evolve into
eyes-some sort of anterior sensory nerve control. Its use for eyes in both
vertebrates and arthropods would then be a case of parallel evolution.
More detailed study of the relevant genes in other groups of bilaterians
and in more basal animals, especially cnidarians and ctenophores, would do
a lot to answer the question.
Lateral gene transfer is certainly a problem for some molecular
phylogenetic work. It seems especially easy in bacteria, perhaps because
they do not have a nucleus and so cannot as easily distinguish stray DNA
from their own (although there are other tricks like methylation). I think
the current consensus for eukaryotes is that there was substantial fusion
of archaean and eubacterial genomes in our ancestry. There are other
possible confounding factors, such as organelle ancestry differing from
organismal ancestry. However, the problems I have encountered in my own
work on molecular phylogeny (in bivalves) seem to largely reflect lack of
data or random convergence (long branch attraction, which results because
DNA has only five options-A,G,T,C, deletion. Thus, two random strands will
have some similarity).
David C.
This archive was generated by hypermail 2b29 : Tue Jun 13 2000 - 09:34:48 EDT