Steve LaBonne and Mitochondrial genetic codes

Erich Schwarz schwarze.ccomail at starbase1.caltech.edu
Mon Aug 28 00:45:08 EST 1995


HPYockey wrote:

> [X] asks for molecular biology types. Guess who responds: Nobody
> but Steve LaBonne.

    Well, most of us either don't surf the Net, or do browse Usenet but
are totally swamped with our own work.

    I'm sure there's a way to rigorously settle this, but it would involve
not only reading your book but also reading about 10 other books on
evolutionary analysis and then writing one of my own in response to you. 
Which might be a worthy enterprise for Mark Siddall or Joe Felsenstein,
but would be completely peripheral to my own never-cleared stack of
immediate obligations.

--Erich Schwarz


Postscript:

*Warning!*  Pontification follows:

    My own *opinion*, if anyone cares, is that mitochondria are a bad
place on which to base arguments about the genetic code, because they do
not have particularly normal genomes.  In mammals, and possibly many
organisms, mitochondrial chromosomes are given a sinecure -- with only
about 15 proteins that they must encode.  They consequently can probably
tolerate considerably more fluctuation of their genetic code than even
_Mycoplasma_ (encoding ~800 proteins) could ever do.  At the same time,
they are probably under considerable pressure to minimize genome size, and
also probably have somewhat lower fidelity of DNA replication than either
eubacteria or eukaryotes do.  Finally, they encode their own rRNAs and
tRNAs, unlike viruses which must depend entirely on the translational
machinery of a host cell.  Add this all up over many years, and you've got
a plausible setting for *late* alteration of mitochondrial codons.

    If phylogenetic analysis were to show that codon alterations were
derived characters that arose in the late Precambrian, the argument would
be over; I leave it to others to describe whether such an analysis is yet
available, and what it says.

    More generally, it's a risky proposition to argue from first
principles, without a large dose of mere empiricism, when talking about
biology.  This was first pointed out by Descartes and amplified by Max
Delbruck.  It was more recently pointed out by somebody on this newsgroup
whose name I forget, who said something like: "Would anybody have
predicted the dog's nest of proteins in eukaryotic transcription, on the
basis of physics?"


--E.S.



More information about the Mol-evol mailing list