From: Aubrey de Grey (ag24@gen.cam.ac.uk)
Date: Tue Apr 22 2003 - 09:49:51 MDT
Rafal Smigrodzki wrote:
> ### Do you think that protein aggregation is causative in sporadic AD
> or PD, or that it is involved in the pathomechanism at some later
> stage, perhaps amplifying the damage caused by other factors? Recently
> it turned out that APP has a direct toxic effect on mitochondria, and
> mutant alpha-synuclein has also been implicated in diminished mito
> function.
It's still very unclear what the major pathogenic process is in AD or
PD, but the various aggregates seem to be the only options, because
nothing else has been identified that changes in affected cells (either
with age or faster in the disease than in normal aging). The effect on
mitochondria is interesting, but basically I think that we can be sure
that these aggregates are bad for us, so let's get rid of them, rather
than wasting time worrying about precisely how they're bad for us.
> Also, are protein aggregates in AD and PD really lysosomal, or
> cytoplasmic (or even extracellular)?
In AD, the Abeta aggregates (plaques) are extracellular but can be made
lysosomal by vaccination (Elan's work, etc). Vaccination makes microglia
engulf (endocytose) bits of plaque. It thereby becomes lysosomal. As
for tau (tangles) in AD and alpha-synuclein in PD, the evidence that it
is lysosomal is limited but positive: see eg Neurosci Lett 284:187 or Mol
Cell Neurosci 18:702. It would be very surprising if it weren't, because
the stuff is clearly accumulating to the detriment of the cell. In some
neurodegenerative diseases there seem to be nuclear aggregates, but that
is very much the exception.
> Atheromas are not aggregates of lysosomally indigestible material, so
> improving cellular digestion is not likely to be of help.
Wrong. First, plaques develop from foam cells, which are macrophages
that have endocytosed modified LDL in the artery wall and failed to
digest it all, which is why they became foam cells. Second, in a mature
plaque there is a constant influx of new macrophages that enter it for
the purpose of breaking down the core, and they engulf bits of it, but
they can't break it down so they die and become part of the problem.
> Is there any indication that lipofuscin is indeed causally involved in
> cell death?
The in vitro evidence is now quite compelling -- see recent papers by
Brunk and Terman. In vivo evidence is still rather lacking. Currently
I'm more focused on the age-related diseases (above), for this reason.
> BTW, I agree that the best way of fighting aging is reversing it by
> genetic engineering (once you get it to work). I like your idea of
> transferring mtDNA-encoded function to the nucleus. Any more
> information on this project?
It's going very well in terms of the underlying science but it badly
needs more funding. The most important recent advance is the cloning
of four genes from Chlamydomonas that are mt-coded in animals but
nuclear in Chlamydomonas. These genes have undergone subtle changes to
make the proteins slightly less hydrophobic (hence more importable),
and we should be able to emulate those changes (in all our 13 mt-coded
proteins, not just those four).
Robert Bradbury wrote:
> > BTW, I agree that the best way of fighting aging is reversing it by
> > genetic engineering (once you get it to work). I like your idea of
> > transferring mtDNA-encoded function to the nucleus. Any more
> > information on this project?
> First you have to get genetic engineering *cheap* and *fast) (that I'm
> reasonably sure I know how to do) -- *then* everyone can easily tinker
> with various genomes to get it to "work".
No need here, really -- the hydrophobicity problem can be studied in
vitro quite efficiently enough with existing technology.
> I had the mtDNA-transfer to the nucleus idea probably 7-8 years ago. I
> don't remember whether I came up with it. It may have come from a
> theoretical paper that I read. Aubrey de Grey I think figured it out
> independently and has worked on it a bit more and I think he thinks its a
> bit tricky because the human mtDNA proteins that are currently still in
> the mtDNA are a bit hydrophobic and so transport and getting them across
> the membranes may be difficult.
Indeed, many people have thought of this independently (it's not a very
hard idea to think of!). The first work was in the mid-80s in yeast, by
Nagley's group. Therapeutic relevance was suggested by Lander and Lodish
in 1990. Hoeben was the first to suggest it for aging (in 1993).
> I think he has some possible work-arounds -- do a PubMed search for his
> papers over the last 3-4 years or send him an email.
Trends Biotechnol 18:394 and Antiox Redox Sign 3:1153.
> But the nucleus being a safer place than the mitochondria may be why
> humans such a small mt Genome -- someone (perhaps *you*?) needs to
> write a paper comparing the human, elephant and whale (bowhead or blue)
> mtDNA to each other to see if they "solved" this problem the same way.
No, no point -- elephants and whales have so vastly lower a specific
metabolic rate that they don't need better maintenance machinery than
us in order to live longer. (This is not to say that oxidative damage
is the only thing that matters in aging, of course, only that it affects
everything.) Give me the $50m, I could spend it much better than that...
Also, a genetic code disparity (UGA encoding STOP in the nucleus but
tryptophan in the mitochondria) has been present since animals diverged
from fungi (~1 billion years) and has totally prevented any gene transfer
in that time -- all animals encode exactly the same 13 protein in their
mtDNA. That's a trivial thing to solve by biotechnology but completely
impossible for evolution. (There is no code disparity in plants, which
has allowed Chlamy to do more. Hm, but it didn't do it in order to live
longer.... Evolution is smart, but it has very different goals than us!)
> Alternate ideas I've thought of is intracellular bacteria engineered
> to produce the modified mtProteins or even souped up to produce and
> export ATP in high volumes (essentially recreating the mitochondria
> with lesser free radical generation properties).
Rather harder than modifying 13 genes....
> An alternate
> process might be more robust mtDNA repair -- right now its pretty
> minimal compared to the nucleus (something like 120+ nuclear DNA
> repair proteins).
No - reversal or obviation of damage is what we need, not retardation.
> The short summary is that there are a number of ways to tackle
> this problem. The pain in the rear end problem that we really
> need to solve is what the heck in the mitochondrial transport
> chain is generating the free radicals in the first place.
> Every time I hear a talk on this topic it gives me a headache
> due to the complexity.
Progress is being made here -- we now know roughly where in Complex
I and precisely where in Complex III the superoxide is made. But
making a modified enzyme that makes less (a) is probably very hard
and (b) only retards aging, doesn't reverse it, so we might as well
focus on a proper solution (nuclearising the genes).
Damien Broderick wrote:
> I like the idea of using heavy-duty nuclear maintenance and repair
> mechanisms to keep our mitos nice
It's not the maintenance and repair. The idea is that mtDNA mutations
would carry on happening but they wouldn't matter, because the same
proteins would be coming in from the cytosol.
> but I find it hard to see how this
> would work. You surely can't replace the work of a zillion cytoplasmic
> machines by one or a hundred nuclear-embedded versions. Besides,
> wouldn't that require the helix to be pulled open and naked all the
> time at that locus, which can't be a good look for DNA.
This isn't a problem. Nearly all our mitochondrial proteins are encoded
in the nucleus already. The cytoplasmic protein synthesis machinery
just works really fast compared to the mitochondrial one, so the cell
has no trouble making enough of most of these proteins from just two
copies of the gene. The mt-coded ones aren't present in any greater
abundance than the nuclear-coded ones.
Aubrey de Grey
This archive was generated by hypermail 2.1.5 : Tue Apr 22 2003 - 10:01:13 MDT