lamoran at medisun.med.utoronto.ca (Laurence Moran) wrote:
>In article <49cj03$5jm at news.nstn.ca>, Dave Edgell <dedgell at ac.dal.ca> wrote:
>>dasnyder at uci.edu (David Snyder) wrote:
>>The 16S rRNA has been sequenced from Thermotoga which places this
>>organism near the base of the eubacteria. A few other genes have
>>been sequenced but none as phylogenetically useful as rRNA. Check
>>out any publication by Woese or cronies for a tree which shows
>>where Thermotoga branches.
>>>>There are several genes that appear to be more phylogenetically useful than
>16S RNA. One of these is the glutamine synthetase gene and dendrograms
>constructed using these sequences tend to show that Thermotoga maritima
>clusters with the gram postive bacteria and the archaebacteria - well
>within the prokaryotic group. Brown et al. (1994) state,
>> "The close proximity of the GSI genes of Thermotoga to those of
> archae and Gram-positive bacteria is an additional and very
> striking conflict with 16S rRNA (ref.) and EF-G (ref) -based
> phylogenies, which tend to place this species as one of the
> deepest branches of the bacterial tree."
>
True, the glutamine synthetase dataset does show an affinity for Thermotoga
and the gram positive bacteria. However, there are very few GS sequences
for eubacterial compared to 16 sRNA, and no thermophilic eubacterial seuqneces
excpet for Thermotoga. The 16 sRNA dataset on the other hand, includes multiple
representatives of all the major groups of cultivatable eubacteria. I still
think that 16S rRNA is the most useful molecule for eubactieral phylogeny
BECAUSE of the wide taxon sampling. I think we are all aware of the dangers
of constructing phylogenetic trees with small datasets lacking the critical taxa
necessary to answer the question at hand.
>Brown et al. (1994) then go on to compile a list of all Thermotoga genes
>that have been sequenced and they conclude that the majority support
>a close relationship between Thermotoga and Gram-positive bacteria.
Brown et al (1194) say
"We performed several individual BLAST searches using T.maritima
genes to probe the nonredundant database. Table 1 lists, for 11 proteins,
the ranking of species in descending order of similarity to the T.maritima
homolog. The most similiar species to T.maritima were either members
of the Gram-positive (seven genes), cyanobacteria or chloroplasts (two
genes), or the genus Thermus (two genes)."
I maintain that the observed relationship between Thermotoga and
gram-postive bacteria is unresloved.
>Some other Thermotoga genes have subsequently been sequenced (ie. ribosomal
>protein genes; Sanangelatoni et al. (1994)). In most cases the dendrograms
>constructed from these sequences also suggest that Thermotoga is not the
>deepest branch of the bacteria tree but is related to the Gram-positive
>bacteria.
>>It's true that Woese and his "cronies" claim that the 16S RNA sequences
>are the best phylogenetic markers. It's also true that many others
>dispute this claim.
>>>Laurence A. Moran (Larry)
Yes, I share your sceptical view of the world based on 16S rRNA. But
I think it is dangerous to base conclusions on incomplete datasets such
as glutamine synthetase. I welcome other molecular phylogenies of
eubacteria but lets wait until there are more complete datasets before
broad conclusions are made.
And, I would question your statement that glutamine synthetase is more
phylogenetically useful than 16S rRNA. What do you base this statement
on? Taxon sampling? Number of alignable sites?
Dave Edgell
dedgell at ac.dal.ca
