Paul Brookes wrote:
> Surely if man can develop such technology on a short timescale
> then nature would have done so already.
This is indeed the argument everyone uses, and I think it's quite
persuasive, but wrong. Animals have a second barrier to transfer
of genes during evolution, namely the disparity of genetic codes;
that barrier is probably even more insurmountable by evolution than
the hydrophobicity one (judging from the plant successes I noted).
But humans have got round it easily, because we've got a tool that
nature hasn't, i.e. site-directed mutagenesis. So I agree that we
will probably only get around the problem of hydrophobicity by use
of a tool hugely different from what evolution has had available,
but I don't agree that such tools are inconceivable. In fact, the
second half of my Tibtech paper is just such a proposal, concerning
using inteins to reform the correct protein after import of a much
less hydrophobic precursor. If you don't know what inteins are,
you're very far from alone; see http://www.neb.com/neb/inteins.html
> Any chance of a tibtech article preprint?
I'm told I will receive my proofs as a PDF file tomorrow; I'll send
it along.
> assuming its possible that flux control exists between the
> individual sub-units of a protein but that's another story.
Hm... and not one I've ever thought about. No comment :-)
> Is it just that these proteins have to be on mtDNA so they're readily
> up/down-regulated in response to mitochondria-specific stimuli?
I think that such considerations very probably make gene transfers
somewhat harder than they would be otherwise, but only somewhat: they
aren't the sort of all-out showstoppers that the hydrophobicity and
genetic code issues are. Otherwise we wouldn't see such variability
of gene content in plants.
Aubrey de Grey
