questions about mitochondria

Aubrey de Grey ag24 at
Thu Aug 10 08:57:25 EST 2000

Iuval Clejan wrote:

> Why is everyone insisting that there is a selection mechanism operating
> prior to meiosis? Sorry Aubrey but I have not found support for this in
> the references you provided (maybe I didn't understand).

The cellular selection that I and others (eg Nature 400:125, Nature 403:
500) have discussed occurs during ovulation.  It is driven (in my model,
at least) by prior, intracellular selection for any mutant mtDNA that a
given oocyte may contain during the many years while it is quiescent
and non-dividing -- similar to what happens in muscle.  The thing that
happens much earlier is a mtDNA population bottleneck; that process is
not selection in and of itself, but it amplifies the power of the later
selection process by reducing the possibility that there will be some
but not many mutant mtDNA genomes in the oocyte at the time of ovulation.

> Maybe something similar occurs in replicating
> oocytes to what occurs in replicating yeast with ERCs (what is the
> biochemical reason for the asymmetry in that case?)

Still unclear, I believe.  One hypothesis is that the extrusion of
material into the daughter cell is a rather carefully regulated process
such that things typically only get into the daughter cell by an active
process, and ERC's are not actively taken there so they predominantly
stay in the mother.

> 3. What proteins that are involved in mitochondrial welfare are coded
> for by the nucleus? Anything for repair of mtDNA? Replication of mtDNA/
> transcription of mtDNA?

All of the above.  The only proteins that are encoded in the mtDNA are
11 subunits of the respiratory chain and two of the ATP synthase.

> Why is tellomerase anti-apopototic? Is it
> possible that senescent cells stop expressing some of the genes
> necessary for mitochondrial welfare?

The only theory I know of (see Zhang et al, Genes Dev. 13:2388) is that
telomerase specifically inhibits apoptosis caused by aneuploidy.  When 
telomerase is absent and telomeres get critically short, chromosomes
get joined end-to-end (Robertsonian fusions) and aneuploidy results.
However, the possibility remains very open that telomere shortening
and consequent gene-expression changes could induce apoptosis in other
ways too.  Eugenia Wang and others found some time ago that fibroblasts
become *less* apoptosis-prone when they approach replicative senescence,
but others have disputed that finding and it's now known that other cell
types show the reverse.

> I am thinking that one of the main
> differences between mitos and aerobic bacteria is that mitos are
> symbiotic with the nucleus, whereas bacteria are rugged individualists.
> So what if mitos are mutating at a high rate, so are bacteria, as long
> as they keep reproducing they will survive.

Ah, but that's because the mutant bacteria stop reproducing [so well].
Since the mtDNA encodes none at all of the machinery for mitochondrial
biogenesis, mutations aren't necessarily selected against.

> One thing that might keep
> them from reproducing is a nuclear clock (e.g. tellomere shortening
> followed by activation of P53 followed by apoptotic signals) This
> scenario would favor mitos in post mitotic cells, if other mechanisms
> didn't come into play.

You've lost me here.  Favour mutant mitos in post mitotic cells?

Jim Cummins wrote:

> In sharp contrast to the over fifty pathogenic mutations described
> in mtDNA, only one disease-causing mutation has been identified in a
> nuclear gene, that encoding the flavoprotein subunit of complex II
> (Bourgeron et al., 1995).  This situation will certainly change

And indeed, see last week's Science 289:782 which identifies the adenine
nucleotide carrier as the gene mutated in autosomal dominant progressive
external ophthalmoplegia.  The study is excellent, though spoiled by a
bizarre discussion of a very nearly impossible mechanism and the virtual
non-reference to a perfectly simple mechanism proposed last year for the
same phenotype in ANT1 knockout mice (Esposito et al, PNAS 96:4820).
The most obvious hypothesis for why the mutation is dominant, namely that
the protein acts as a dimer so is quite likely to be haploinsufficient,
is also not discussed.

Aubrey de Grey

More information about the Ageing mailing list