Hello, Peter! Thanks for your comments.
>In the same issue of Nature, there is a companion article to this one,
>commenting on it: Linda Partridge & Nicholas H. Barton, "Evolving
>evolvability", Nature 407 (2000), 457-458. But the following comments
>are my own.
I've also read Partridge and Barton's comment, but it seems to me that they
missed the whole point of the prion paper. For instance, they tend to rule
out selection for evolvability because it has to be a form of group
selection, which is not necessarily the case for two reasons. First of all,
clade selection is different from group selection; also, evolvability may
provide individual advantages, which is what the prion paper just showed.
[...]
>The prion system, like a mutation in any translation factor, may
>interfere with many genes at once. What makes the prion system a very
>special case is that a prion is autocatalytically produced from its
>"normal" cellular conformer. The yeast prion [PSI+] is a
>self-aggregating conformer of the "natural", non-aggregating translation
>termination factor Sup35. Both forms have the same primary sequence, but
>different tertiary structures. Even one single molecule of the prion
>conformation catalyzes the conformational change of Sup35 molecules into
>the prion form, such that all Sup35 in the cell is transformed into
>prions, which aggregate, disrupting the normal translation termination.
>Within the given cell line, this "epigenetic" change is heritable. But
>it can be reversed by counterselection.
But isn't it the germ line heritability what really matters? Isn't this
just a kind of molecular-level phenotypic plasticity? If it is, what is so
revolutionary about it to be published in Nature?
[...]
>Both the pseudogene and the prion systems, however, can only activate
>something that already exists. If this is to be a new functionality, it
>must have emerged by a mutational random walk in the cryptic state,
>hidden from natural selection. Can we expect really novel
>functionalities to emerge by such random processes extending over
>several mutations, but without intermediate selection?
I totally agree with you here. It seems to me that, over time, there should
be an exponential decrease in the probability that this potentially new
function could arise. Maybe this is the reason why the rate of gene
duplications seems to be surprisingly high (at least to me) in some cases
(sometimes in the order of 10^-3 per generation!!). Eventually this
pseudogene should be deleted, otherwise genome size would grow
exponentially, together with the accompanying costs. By the way, do you
have any idea of how this overflow of new pseudogenes could be detected and
removed from the genome?
Best
Marcio
This archive was generated by hypermail 2b29 : Sat Oct 21 2000 - 19:51:24 EDT