(none) (Horizontal gene transfer)

Andrew J. Roger roger at evol5.mbl.edu
Fri Jun 20 15:13:38 EST 1997


Simon Topp wrote:

         Although I am not claiming horizontal gene transfer does not
occur, I
> believe it is and has always been a very rare event. 

I actually don't think that it has been investigated carefully enough to
really know how rare it is. In some special cases such as organelles in
eukaryotes derived from endosymbiotic bacteria (mitochondria and
chloroplasts)
, lateral transfer of genes from organelle to the nucleus is VERY
COMMON.
How common it is in a different context is less clear.  I would argue
that
there is preliminary evidence that it is quite common. For instance,
Gupta and Golding have published a few papers where they claim that
eukaryotes are the result of
a fusion of an archeabacterium and a "gram negative" eubacterium. 
Although
I strongly disagree with their conclusion, I believe that they have
drawn
attention to the fact that there is deep incongruence between gene
phylogenies
when considering the relationships between the eukaryotes/eubacteria and 
Archaebacteria. I believe that these incongruencies are the result of
lateral transfers and/or extensive gene paralogy.  The genes tree for
which this
is a necessary explanation are: glutamine synthetase, glutamate
dehydrogenase,
FGARAT, hsp70...V-type ATPase amongst quite a few others.

>The main seminal
> papers on the subject are by R.F.Doolittle et al, where he suggests a
> number of likely and a few possible occurences of this in evolution.
> Last year I repeated one of his firmest test cases
> (glyceraldehyde-3-phosphate dehydrogenase in E.coli) and found his
> conclusions to be no longer valid. A phylogeny covering the current
> content of the protein databases (much expanded since the original work)
> showed clearly that the E.coli gene in question was in it's perfect
> taxonomic position and that no horizontal gene transfer had taken place.

I totally disagree with you.  E.coli has three genes that are closely
related to GAPDH named Gap A-C.  E. coli GAPA  falls with an enteric
gamma-proteobacterial clade of GAP genes which are together located
within the eukaryotic GAPDH clade.  This is not at all a straightforward
placement for these enzymes- I believe that two explanations are still 
open. 1) the enteric GAPA clade is derived from a lateral gene transfer
from the eukaryotic lineage OR 2) there is a eubacterial family of GAPA
proteins which has been lost from many taxa but retained in enterics,
cyanobacteria, and bacteroides- the eukaryotic enzyme is in this case
suggested to be derived from a lateral transfer from the proteobacterial
lineage (the mitochondrial endosymbiont perhaps).  On either explanation
I doubt one could suggest that the enteric group is in a "perfect
taxonomic position". There are deep problems with GAPDH trees (look at
the position of the Trichomonas vaginalis GAPDH for instance)-- some of
this
may be due to lateral transfer and some of it probably is due to
paralogy.


Cheers
Andrew J. Roger



More information about the Mol-evol mailing list