Re: longevity
Sun, 21 Nov 1999 16:06:25 EST

In a message dated 11/20/99 6:42:40 PM Pacific Standard Time, writes:

> (J. R. Molloy)
> Sender:
> Reply-to:
> To:
> Curt Adams wrote,
>>Longer lifespan with no disadvantages would be selected for, roughly to the
>>point where only a negligeable number of mice would survive that long due to
>>other causes (a lot less than 200 years, I'd think). If the effects of the
>>gene are more complex (only works if a particular set is in the organism,
>>useless or harmful otherwise) then selection is less reliable, hence weaker
>>and the effect "gives out" earlier.

>When the effect gives out, do the descendents of long-lived mice return to
>their former (and shorter) lifespans?

Sorry, I didn't phrase that right. If selection is weaker due to complicated genetics,
the lifespan increase selection will be able to establish is shorter. If you introduced
hypothetical improvements with simple genetics and no disadvantages, maybe mice in the wild would end up with a lifespan of four years (I'm making up the numbers)
If the mutations had complicated genetics, maybe the best selection would be able
to do is 3 years.

>>Of course longer lifespan may have costs. In the Rose lab, the long-lived
>>strains take longer to develop and lay fewer eggs in the first part of life.

> So, shorter life cycle strains would outperform long-lived strains?

With these particular strains. Research has shown that some "short-lifespan" alleles
carry compensating advantages (for the genes, not necessarily the organism) while
others serve no purpose and have eluded selection. Theory suggests (although I'm not
aware of any experimental tests) that genes with important and straightfoward "short-lifespan" effects essentially only come attached to compensating benefits.

The best area of "natural" effects to look at, by this theory, are short-lifespan alleles
which help the interests of the gene rather than the organism - mostly fertility
and sexual selection genes. It may well be possible to find genes that, say, increase
the growth of secretory breast or prostate cells. Conking those genes will reduce
fertility (but we don't care much; we can deal with such problems) but they'll cut
your chance of breast or prostate cancer. Since the genes are evolutionarily advantageous, they will exist even if easy to knock out and so be good targets for medicine.

We can also look at adaptations for our ancestral environement that are no longer
appropriate. Genes for high LDL cholesterol and adult-onset diabetes seem to be
starvation defenses. Those should also be amenable to intervention, with no cost
as long as we keep the supermarkets stocked.

Our best bet for the costless problems (whatever they are) is to look at things that
would be hard for evolution to do, so gene knockouts, drug alterations and simple
mutations are poor choices. We would want to bring in novel systems, from design
or from other organisms. An example (from Caleb Finch) would be whale anti-cancer
mechanisms. Whales have a per-cell cancer chance that is orders of magnitude less
than yours. Presumably they have at least one additional layer of defense. If we
could figure out what it was, we could take out a stem cell type, splice in the cancer
defense, wipe out the cells in the body, and then insert the "improved" cells. Voila,
90% reduction in leukemia. If we're really lucky, we might be able to improve cell
senescence at the same time.