feedback requested on ideas(modified)

Aubrey de Grey ag24 at
Fri Feb 23 12:40:37 EST 2001

Iuval Clejan wrote:

> Does this buildup of undegradable junk in lysosomes correlate with the
> time scale for decreased organ function?


> Also, why does it not happen
> in the germline (or does it get selected out)?

Not even selected -- diluted.  Same in all regularly-dividing cells.

> Is it due to the remaining cells being negatively impacted (e.g. having
> to "work harder" or missing signals from their missing neighbors) or is
> it just not having enough cells in an organ decreases its function, in
> which case I would rather classify it as a separate mechanism, or a
> special case of (2).

Both.  Not having enough cells in an organ decreases its function, but
that function may support that same organ as well as the rest of the
body, so the distinction you describe is hard to draw.

> will look up AGEs. Are they degraded or cross linked collagens and
> elastin? I thought extracellular matrix is made up of components
> secreted by cells (mostly fibroblasts).

Cross-linked.  The problem is that the elasticity of the extracellular
matrix is crucial to its function.  In order to have proper elasticity,
the cross-links between collagen or elastin fibres need to be in a regular
pattern (which is how they are laid down), whereas the random cross-links
produced by glycation diminish that elasticity.

> > Hang on.  This list covers only the simple case of mutations that cause
> > modulation of expression of the same protein, in the same cell, that
> > the mutation is in the gene of.
> Not at all. All except possibly (2) are about mutations for genes other
> than the ones for the
> protein being modulated.

4 and 5 also count, don't they?  But sorry -- sure, 1, 3 and 6 are as
you say.  But in fact your point in (4) "a ribosome may be defective and
translation can still proceed, albeit at a lower rate" is the main problem
-- proteins will still work, at variably reduced efficacy, when they have
suffered point mutations.  There are of course awfully many possible
mutations in each gene, and in virtually any gene there will be a range
of reductions in function right the way from tiny to complete loss of
function, depending on which base pair is changed and to what.  Thus,
when you consider expression of a protein, this can be affected by such
mutations not only in ribosomal components but in all the other processes
relevant to protein synthesis -- tRNA synthesis/acylation, transcription,
mRNA splicing, mRNA export from the nucleus, etc -- and also, of course,
in mutations in any of the upstream transcription factors that regulate
the expression of the protein, or in the factors that regulate those
factors' expression, and so on.  When you consider that enzymes may or
may not be up-regulated in expression to compensate for mutations that
somewhat diminish their efficiency, the complications become even bigger.

> I thought things like pH, temperature, oxidative stress, etc., are all
> extremely tightly regulated in vivo on the average. Is there a drift in
> any of these variables (which indeed can influence reaction kinetics in
> continuous ways) with age? I don't think so.

pH and temperature don't change much (pH does go down a tiny bit with
age).  But definitely, oxidative stress (as measured by, for instance,
plasma GSH/GSSG ratio) increases markedly with age.  A good reference
is Samiec et al, Free Rad Biol Med 15:699.

> As far as I know nobody has looked at time averaged mRNA expression.
> What is the time scale for which genes (coding for proteins that are
> involved in decreased organ function with age and associated proteins,
> not for genes associated with daily metabolic changes) are
> transcriptionally active, all else being equal?

Do you mean, how fast does mRNA expression oscillate?  I've no idea, but
it's certain to be very different for different proteins.  But you don't
need to look on purpose for time-averaged expression, because in general
cells will not oscillate synchronously (except for genes regulated by
daylight, for example).  So the average of a lot of cells suffices.

> I didn't think base pair substitution happens in non replicating cells.

See for example Dolle et al, Proc Natl Acad Sci USA 97:8403.

> The scenario you propose seems possible, although I don't know if it
> really happens.  I need to study more on DNA repair. But the concept of
> non-instantaneous repair is the same in non-replicating cells and
> indeed it can happen that damage to more than one repair pathway can
> occur "simultaneously".  Again I must ask for time scales. Is it
> possible with some finite probability to damage only enzymes in two or
> more repair pathways without damaging other DNA which would kill the
> cell? If the time to do this is too long, then there will be time to
> repair the repair enzymes which have been damaged by the ones that have
> not. If the time is too short, then it seems (but may not be)
> improbable that the damage won't be of massive type that destroys alot
> more than just the repair enzymes, and the cell will die.

You're too hung up on repair enzymes.  The scenario I outlined, and many
related ones, will happen with finite probability whether or not any or
all repair pathways are working.

> I thought there must be some way to coordinate
> mitochondrial replication ( in somatic cells, I know this is not the
> case for eggs) with the cell cycle, otherwise there would be too many
> mitochondria.

The replication is balanced by lysosomal destruction (autophagocytosis).

> What regulates their division?

No one knows; the model in my BioEssays paper is a possibility.

> Has it been observed in a
> non dividing somatic cell (or G0/G1/G2 part of the cell cycle in a
> dividing cell) that mitochondria divide?

Oh yes, for many years -- and that they are autophagocytosed.

Aubrey de Grey

More information about the Ageing mailing list