From: Rafal Smigrodzki (rafal@smigrodzki.org)
Date: Thu May 01 2003 - 15:34:35 MDT
Aubrey de Grey wrote:
>
>> ### See: Desurmont C. Caillaud JM. Emmanuel F. Benoit P. Fruchart JC.
>> Castro G. Branellec D. Heard JM. Duverger N. Complete atherosclerosis
>> regression after human ApoE gene transfer in ApoE-deficient/nude
>> mice. [Journal Article] Arteriosclerosis, Thrombosis & Vascular
>> Biology. 20(2):435-42, 2000 Feb.
>
> Ah, I wondered whether you were thinking of something like this. That
> sort of experiment "doesn't count", in my view. These animals are so
> high in cholesterol that the recycling system of compounds that are
> normally no problem, such as unmodified cholesterol, is overwhelmed.
> If you make an animal accumulate stuff at a hugely accelerated rate,
> then sure, a lot of what they accumulate will be stuff that a normal
> animal wouldn't have accumulated in the first place because they could
> break it down, so if you then relieve the handicap they will go ahead
> and break it down. What this paper doesn't tell us (I've just read
> the full text) is the key question: do the apoE-/- mice with
> transfected human apoE have less lesion at 199 days old than
> age-matched apoE+/+ mice? I bet they have more -- "complete" in the
> title is misleading, since the text says there was a reduction from
> 220 to 28 mm^2 lesion area, only a factor of eight.
### So you are postulating a major chemical difference between fatty streak
in apoE null mutants and in aged animals. Would it also apply to humans with
hereditary hypercholesterolemia vs. aged humans?
This is a legitimate hypothesis, but I'd need more data before I would
accept it.
-----------------------------------
>
>> ### Genetically determined obesity (e.g leptin deficiency) manifests
>> early in life, and yet we know that maintenance of body weight is a
>> dynamic balancing process, and can be tilted early or late in either
>> direction. You can have fatty streaks appearing early and developing
>> very fast, if your LDL is very high, in hereditary
>> hypercholesterolemia, or they can develop slower, if you have only
>> the Western diet going against you. The balance can be out of kilter
>> because of hereditary, or acquired changes in your genome, or due to
>> non-genetic changes (like formation of insoluble aggregates). So far
>> we have insufficient data to answer the question of the specific
>> contribution of such factors in PD, AD, and, AFAIK, in most other
>> aging-related conditions.
>
> I don't see the relevance of this. My point about fatty streaks in
> childhood is that they appear in the context of a normal diet: we all
> get them in our aortas by the age of ten. I agree fully that changes
> to how much you push a homeostatic system will cause changes in its
> equilibrium state (just like Desurmont's mice), but we're not talking
> about any such changes here.
### Interestingly, there are some lucky seniors who even at age 80 do not
have atherosclerosis. If indeed there was an inexorable accumulation of
indigestible material in most of us, how do they escape it? Would you say
they do not make such supposedly insoluble plaque (hard to imagine,
cholesterol oxidizes the same way in all humans), or do they have superior
lysosomal enzymes?
The equilibrium model, without significant insoluble material, handles this
question better. The healthy seniors have probably a host of beneficial
traits, like better mitochondria, more durable stem cells, more active
antioxidant defenses, a better liver, all adding to the ability not to lay
down plaque, or to dissolve it better than others.
------------------------------
>
>> ### My boss, Davis Parker, likes the idea of clonal expansion, too.
>> Why do you think COX negativity must precede hyperproliferation if
>> this hypothesis is true?
>
> It doesn't need to do so in ALL mechanisms of clonal expansion, by any
> means -- but in Patrick's model, hyperproliferation must precede COX
> negativity, no question. I claim that my model for the mechanism of
> clonal expansion (BioEssays 19:161) is the most consistent with all
> available data ... but then I would, wouldn't I :-)
### I went over Chinnery's articles (Am J Hum Gen, Lancet) and I didn't find
any clear statement that hyperproliferation must precede COX negativity.
Where do you think he says that it does?
Non-specific hyperproliferation in response to a biochemical abnormality
sensed by the cell is not a necessary part of his hypothesis, although it is
added as a secondary process after the clonal expansion to fit with the
existence of MERRF and possibly other conditions characterized by mtDNA
accumulation. In fact, hyperproliferation may well be triggered by COX
negativity. After all, COX negativity is observed once the level of a
mutation (at least in cases of single nucleotide substitutions) exceeds
50-60%, and the proliferation may be a frantic rescue attempt. This would be
more or less unrelated to the normal mechanisms of mtDNA copy maintenance,
which probably involve sensing total mtDNA mass, rather than energy levels.
Talking about your model (and most selection models) - if indeed the mutated
genome has a selective advantage, then cells with more mt genomes would be
at a higher risk of suffering a mutation (this risk goes up as a simple
function of the number of genomes), and once a single mutation is present,
such cell would very quickly shift to 100% mutated genomes, as a result of
selection. This means that large cells would be more likely to be COX
negative than smaller cells. In fact, very large cells, like the ovum, and
the syncytial muscle fiber, with 100 000 mtDNA's, and very small or
quiescent cells, like chondrocytes, with only 1000 to 5000 mtDNA's, would
differ in the accumulation of mutations by an order of magnitude or more. Is
it indeed the case?
The random drift model is much less dependent on the copy number per cell,
and predicts that the COX-negative percentage will vary by a factor 2 even
if copy number is varied by two orders of magnitude, which AFAIK is in
agreement with observations, as limited as they are.
------------------------------------------
>
>> ### If the effective number of mtDNA genomes undergoing replication
>> is smaller than the total number of mitos, then drift can occur
>> pretty fast.
>
> True -- "stem mitochondria" is how Bruce Ames has described this. But
> it really only works if the effective number is 1, because there seem
> to be very few cells with anywhere near a 50/50 ratio of wild-type to
> mutant genomes -- indicating that cells pass through that middle phase
> a lot faster than drift should allow. Evidence for stem mitochondria
> is conspicuous by its absence, so far.
### Chinnery was able to get simulation results consistent with experimental
observations even without stem mitochondria. The model works very well with
anything from 100 to 10 000 mtDNA copies per cell, definitely it is not
necessary to postulate a single proliferating mitochondrion. There are
pretty few data on single cells, especially the data derived by sequencing
of clones rather than direct PCR sequencing, but Chinnery's model indicates
that the number of cells with 50% mutated genomes will be actually pretty
low, on the order of a few percent in aged humans, well agreeing with some
of the crude data available.
This said, without doubt, our understanding of the dynamics of mitochondrial
replication and mutational changes is in its infancy.
Rafal
This archive was generated by hypermail 2.1.5 : Thu May 01 2003 - 12:42:29 MDT