Evolution Rate (Was Re: Creationist Garbage)

Eric Cabot ecec at kimbark.uchicago.edu
Fri Oct 15 14:02:25 EST 1993

In article <MAILQUEUE-101.931015075712.713 at micro.uct.ac.za> ed at micro.uct.ac.za writes:
>No species is "more evolved
>> " than any other. That's where Ed Rybicki makes his mistake too. 
>For about the last time......
>And RNA viruses ARE more evolved than anything else - they CAN 
>evolve a HELL of a lot faster than anything else (10exp5 or so), 
>thanks to more inaccurate RNA-dep RNA pol.  Of course, that doesn't 
>mean they DO, all the time; just that they CAN, when selected.
>  ____________________________________________________________________
> | Ed Rybicki, PhD             |                                      |

I'd like to propose that we be a little more careful when we use the term
"selected" in the context of the rate of evolution of "retro-replicating"
genetic elements (if you will forgive that term :-)

I presume that you mean that if they replicate that they have enhanced fitness
since they differentially increase their representation in genomes and/or
in populations. The problem is that you'd expect selection to 
act against many if not most of those variants created by error-prone 
  I can envision two scenarios, however, where that might not necessarily
 be the case. One is a spontaneous burst
of transposition or replication that produces an enormous number of defective
elements that amplify just once. That' probably a pretty short-term gain.

    The other  situation would be a case where
a few competent elements can provide, *in trans* necessary functions
to replication-defective related elements.  I have some simulation results
which suggest that such a system will invariably run down
as  more and more competent elements produce copies going into the defective,
albiet mobilizable class of elements (defectors? ;->)

 I'm basing most of my reasoning on retrotransposons but I think that one
can easily make similar arguments concerning real retroviruses. (Although
things are made even more complicated by having to consider evolution during
passage from cell to cell as well as passage from host to host, where what 
is really at issue with retrotransposons is germline transmission). True, a
virus like HIV is quite variable but, although I haven't reviewed much of the
available data, I'm pretty certain that the enzyme coding pol domain experiences
considerably more selective constraint than the hypervariable env domain.
I think that one might even be able to construct a case for positive selection
for changes in the env domain if it provides a way of exploiting new cell
types or evading immunological response.

So my point is that there are several levels of natural selection involved
in the evolution of reverse-transcriptase dependent replicating elements
so it's not particularly informative to issue statements that imply that 
"retro-replicators" evolve faster when selected.

Eric Cabot
<despite the .sig below, please respond to cabot at trog.mbb.sfu.ca as
 I give this account a 99% probability of being extinct by next week.>
    Eric Cabot                         |       "Non Nobis Nati Solum" 
ecec at midway.uchicago.edu               | 

More information about the Mol-evol mailing list