Caloric restriction and mitochondrial fatty acid composition

Paul S. Brookes. brookes at
Tue Jan 26 12:45:52 EST 1999

I didn't want to go down this path but have now realised that I've opened
up a can of worms regarding mitochondrial polyunsaturation, proton leak,
and superoxide generation.

Now intuition and a lot of experiments tell me that proton leak correlates
with polyunsaturation.  Skulachev reckons proton leak is a sink for redox
energy to prevent superoxide formation (the jury's out on that one),   So
in theory, high proton leak (and polyunsaturation) = low mito' superoxide

Proton leak correlates POSITIVELY with superoxide generation.  This is true
from a number of different model systems - pigeons vs. rats (rat mito's
have more polyunsaturation and make more superoxide).    Ageing (old
animals have greater proton leak and also greater mito' superoxide
generation).  Body mass - smaller animals mito's have more polyunsat', make
more superoxide, have greater proton leak.

So, the question is really whether any of these things are causative of
each other - does polyunsaturation cause proton leak?  Does a high proton
leak cause less superoxide generation?  Is mitochondrial superoxide
generation important for ageing at all?

It's possible that superoxide generation itself causes proton leak via
lipid peroxidation/membrane damage, as peroxidised lipid membranes are more
proton leaky.  So a polyunsaturated membrane may be more leaky because it
gets peroxidised more easily.  This would explain why leak and superoxide
generation correlate.

My current answer is that probably all three (polyunsat', leak, supox'
gen') are regulated by something upstream that also governs metabolic rate,
and that only changing one (e.g. dietary modification of polyunsaturation)
may not have any effect on the others.

I disagree that because there isn't much 22:6n3 doesn't mean its not
important.  What about cholesterol which makes up a tiny fraction of the
membrane but has a massive effect on its physical properties.  I don't
think there's a single parameter such as "18:2/20:4 ratio" that can be
applied to fully describe a membrane composition.   The problem is that
lipid compositions are expressed as relative Mole %, i.e. the total of all
the values adds up to 100%, so a drop in 18:2 will always cause a rise in
20:4, even if 20:4 level doesn't actually change.   Far better to use a
more generic parameter like unsaturation index, or better still to look at
each lipid individually.

Enough rambling for now

Dr. Paul S. Brookes.         (brookes at
UAB Department of Pathology,  G004 Volker Hall
1670 University Blvd., Birmingham AL 35294 USA
Tel (001) 205 934 1915  Fax (001) 205 934 1775

More information about the Ageing mailing list