From: Aubrey de Grey (ag24@gen.cam.ac.uk)
Date: Wed May 21 2003 - 16:13:21 MDT
Rafal Smigrodzki wrote:
> ### Do you have a hypothesis as to what this difference might be? What
> are the postulated non-degradable molecular species?
Quite a few are known; luckily, one of my prospective collaborators on
this work, Keri Carpenter, has been at the forefront of research on
the chemical composition of atheroma for many years. Mostly they are
oxidation products of cholesterol, such as 7-ketocholesterol.
> > A key question is, are there any other comparably legitimate
> > hypotheses?
>
> ### The dynamic equilibrium hypothesis.
I don't think you ever explained how that hypothesis can be reconciled
with the appearance of fatty streaks so young. That's my main problem
with the hypothesis -- the advance of lipid deposition seems to be a
continuous, lifelong process.
> ### I found the following: Homma S. Ishii T. Tsugane S. Hirose N.
> Different effects of hypertension and hypercholesterolemia on the
> natural history of aortic atherosclerosis by the stage of intimal
> lesions. [Journal Article] Atherosclerosis. 128(1):85-95, 1997 Jan 3.
> Apparently even at around age 79 you still find people with no intimal
> lesions. How did they escape their accumulation?
Thanks for this reference. I would need to consult someone like Keri
to be sure, but I think the explanation is just random sampling of a
few sites within the aorta rather than examination of the whole aorta.
According to this paper's Fig 2, 2/3 of those in their 20s had no foam
cells; this is ostensibly at odds with the consensus (stated in, for
example, Jake Lusis's review in Nature 407:233) that they are usually
present in the first decade.
> ### I don't understand. The mutants take over mainly as a result of
> stochastic processes, with no clear mitochondrial proliferative
> response to partial loss of function. There is a substantial functional
> reserve of mitochondria, and the nucleus seems to maintain a constant
> mass of genomes (at least in some experimental paradigms), so there is
> no mitochondrial hyperproliferation until very late in the process, if
> at all. Since the stochastic hypothesis makes no reference to the cause
> of hyperproliferation, merely predicting the emergence of
> COX-negativity, it is compatible with either hyperproliferation before
> COX negativity, after, or none at all, if the cell simply decides not
> to upregulate mitos. These two processes are most likely largely
> independent, over large range of heteroplasmy levels.
No, Chinnery's model *does* specify the cause of hyperproliferation.
He says that the cell maintains the same number of *functional* mt
genomes. Hence, he concludes that the random fluctuation of wild-type
versus mutant genomes will push the number of mutant genomes up along
with the wild-type when the wild-type randomly diminish, but won't
push it down again when the wild-type randomly exceed the necessary
number. Whereas in real life, just as you say, no clear mitochondrial
proliferative response to partial loss of function is seen.
He has a new paper out about this in J Theor Biol which explains and
extends the model somewhat.
> ### In the CNS there are neurons with enormous differences in cellular
> mass, from tiny interneurons to the giant alpha motor neurons. The
> giant ones do have a very active metabolism. If indeed single
> deleterious mutations had a significant replicative advantage, the
> large neurons would die much faster than the interneurons, which is
> not the case.
Um, hang on -- there are at least three things wrong with this logic.
First, neurons die of lots of things, not just OXPHOSlessness. Second,
they don't necessarily die even of OXPHOSlessness, because if they did
then we wouldn't see COX-negative ones accumulating with age. And third,
most brain regions (of whatever size) have far fewer COX-negative neurons
than the substantia nigra. It could be that this is because they resist
apoptosis in the substantia nigra more than elsewhere, but chances are
that it's mutation rate, because of the H2O2 production associated with
dopamine synthesis.
> Also, we know that the mutation level in single neurons in elderly
> adults is in the range of 50 to 800 diverse aminoacid-changing
> mutations per million bases (according to Simon and Beal), meaning that
> more than ninety percent of genomes are mutated, and without doubt many
> of them are deleterious. This leaves little place for a preferential
> amplification of low psi or otherwise impaired mitos.
Ah, not so fast. Are you referring to Hum Mol Genet 11(2):133? I'm
keeping an open mind about that study until the appearance of a ms in
preparation that that article cites, but it's not clear to me that
they made a compelling case for the reliability of their quantitation.
For example, they made no allowance for the isolation of mtDNA that
was in the process of being degraded because of damage (see Suter and
Richter, Biochemistry 38(1):459). The acid test is the single-cell
analysis being done so effectively by Khrapko's group, who have now
started serious work on neurons (see Mech Ageing Dev 123(8):891 as a
starting point). It's possible that only deletions clonally expand
and point mutations don't, but I doubt it.
Aubrey de Grey
This archive was generated by hypermail 2.1.5 : Wed May 21 2003 - 16:25:52 MDT