Robert Bradbury and Joao Magalhaes on Recent Aging Science

A couple of the regulars were discussing recent advances in the scientific understanding of the aging process on the Extropy-Chat mailing list, and an interesting exchange it was too. I've reproduced it here, with permission. The two principles are Robert Bradbury - who maintains the Aeiveos Research Library and has a very interesting history of work on healthy life extension - and Joao Magalhaes, author of one of the Longevity Meme articles and currently involved in biogerontological studies.

Read on:

Robert Bradbury:

Well, we finally have some real progress on understanding aging.

It looks like SIRT1 (homologue of yeast Sir2), regulates FOXO3 which in turn regulates the enzymes that resist oxidative stress.

Now *before* everyone gets all excited please note that the gene regulation goes against apoptosis (programmed cell death) and for stress resistance (particularly from free radicals). That is probably a reasonable strategy in short lived animals (which include most that scientists work on in labs).

However in long lived larger organisms one does not want to suppress apoptosis (because it will probably lead to an increase in cancer). In long lived species one needs to allow apoptosis or improve the ability of the immune system to recognize and eliminate cancer cells (which people are working on). One also needs to promote stem cell replacement of lost cells (which we have some of but it probably isn't as finely tuned as one would like).

But it is clear that this provides a key piece of the puzzle as to how cells manage the repair/replicate/die decision processes. Now whether the actions of FOXO3 on apoptosis and stress resistance have been split in longer lived organisms (so one has 2 genetic programs under individual controls rather than just a combined genetic program with only 1 control factor) remains to be seen.

Joao Magalhaes:

I don't want to be the skeptic around here but I should remind you that there is NO evidence SIRT1 is in anyway involved in human aging. Yes, in yeast sir2 is involved in cell cycle regulation--which is not the same as aging. Maybe sir2 is involved in aging of C. elegans but results from Drosophila and mice do not suggest any involvement of SIRT1 in aging. Since drosophila and mice are biologically closer to humans than c. elegans and yeast, I'm skeptical that SIRT1 plays a role in human aging.

As for the Forkhead family, these transcription factors are very much involved in development, so it is normal that they affect redox potential and apoptosis. Nevertheless, I wouldn't be surprised if they were involved in mammalian aging since p66 has been associated with the forkhead family. Yet the connection to human aging is not clear because of cancer. After all, yeast, drosophila and c. elegans don't have cancer and mice have much higher cancer incidences. So we must be very careful in extrapolating this sort of data into humans.

Robert Bradbury:

Joao (Hi!), I have no problem with your comments regarding the involvement of SIRT1/sir2 in higher organisms (because I know of no evidence for such involvement as you point out).

But I would offer the idea that it is very very difficult for Nature/evolution to change course. So *if* theapoptosis/stress response pathways were linked to each other very early on in evolution I would propose that it would be difficult for them to become separated. Not impossible mind you -- which is why I'm slowly pushing behind the scenes to get a number of long-lived genomes sequenced -- so we can have the data to figure this out. What I strongly suspect is that there are "patches" on the apoptosis program that may decouple it from the stress response program.

With respect to the cancer incidences -- one has to have an organism that can actually get cancer. Yeast clearly can't and probably C. elegans and Drosophila as well. Cancer is a direct result of a failure of the program of the regulatory processes of cell replication in organisms that have enough cells for this to be important. This probably involves a delicate balance -- in organisms with enough cells you want to kill off those that are replicating out of control. In that case you want to replace those cells so there is presumably a pool of cells biased towards replication (when necessary). In my opinion, it doesn't take too much for that situation to get out of control (which is why cancer causes ~30% of deaths). (IMO)

But good comments. If you would care to expand on the p66 involvement I'd be interested in reading them on/off list. (I know what it is but don't have current knowledge with respect to where it fits into the big picture.)

I would guess the short summary of my previous message is that they now have a strong candidate for the regulation of at least the stress response -- it isn't going to take that long to confirm that or blow it out of the water (even for higher mammals). That is why I called it "real progress". Ultimately, it may not prove to be progress from a biochemical standpoint -- but it is going to open the door somewhat wider towards nailing these pathways down.

Isn't it amazing that we live in an age when we can have these sorts of detailed conversations about the way in which aging works? Not to mention the likelihood that many of the questions brought up in this exchange will be answered within a few years at the current rate of progress. Like many observers, I find the newfound pace of scientific research, powered by bioinformatics and new tools, to be exhilarating.

As Robert Bradbury notes, nailing the biochemical mechanisms of aging, one by one, is the quickest way to move forward. Understanding, rather than blind testing (even massive, parallel blind testing powered by bioinformatics that is reaping so many benefits in modern research) is what will lead us directly to therapies for the aging process.


No, Bradbury is wrong. We know enough about aging. What we do not know enough about is how to repair. We need gene therapies and cell therapies and other types of therapies that fix things.

It is worth trying to develop drugs that will up-regulate repair enzymes and antioxidant enzymes. But the return from doing that is going to be fairly modest. What we really need is the ability to fix the accumulated damage. We need to be able to grow replacement organs, replenish adult stem cell reservoirs, send in genes for xenohydrolases to break down intrecellular junk, and otherwise fix what is broken.

Interventions that raise the life expectancy of short lived species are unlikely to cause the same percentage increase or even anywhere near the same percentage increase in much longer lived species.

Posted by: Randall Parker at February 21st, 2004 11:40 PM

Hmm. I, perhaps in my ignorance, would have said that not being able to create the repairing technologies is a function of not knowing enough about aging.

But I recall you're an advocate of massive parallel testing (of the sort that has become very much more efficient in past years, and looks set to become even more so in the near future), so is that your favored direction for the near future over uncovering mechanisms? Not that researchers can't do both...

I think you and Robert Bradbury are on the same page with regard to your last point there, based on other conversations I've had with him. I let the guys know you commented, and we'll see if they have anything to add.

Founder, Longevity Meme

Posted by: Reason at February 22nd, 2004 12:04 AM

I response to Randall's comments -- I wouldn't say that am "wrong". To resolve aging we have to understand what it is. Then we have to develop therapeutic approaches for specific problems. Randall's comments, particularly with respect to xenohydrolases sound like Aubrey de Grey's suggestions. Nothing wrong with those other than the fact that the immune system will be relatively unhappy with foreign proteins circulating within the body. That is why Alteon has taken the small drug molecule approach to solving part of this problem (protein glycation). I agree that we need to be be able to promote the replenishment of adult stem cells but first the DNA sequencing (or at least genotyping and gene expression technologies) need to come down in cost so one could do that replenishment with the most "perfect" stem cells one can find. It doesn't do you any good to replenish the cells if they are damaged or might become cancerous -- one always has to remember that stem cells are biased towards replication -- too much of that is a big problem. Finally, with regard to growing replacement organs -- I fully support it and believe we will someday have it -- but it is going to require the development of a massive amount of infrastructure to get this. If it isn't developed in a way that it can be used by everyone one runs the risk of a backlash against technologies that only the rich can afford.

Posted by: Robert Bradbury at February 22nd, 2004 4:27 AM

Just a few general remarks (no time for more):

1) Regarding repair technologies, Reason is wrong: nearly all the current barriers to building such technologies are unrelated to what's going on in the body, but are to do with delivering the appropriate cells and genes into the appropriate tissues. In some cases we don't know exactly what genes, but there too the search for what genes to deliver is not usually an examination of our own biology but of the biology of other organisms (e.g. soil bacteria for xenohydrolases, algae for mitcchondrial genes). The only serious exception is the genetic machinery underlying ALT, the telomerase-independent telomere elongation system used by some cancers, which is still woefully unknown.

2) Alteon is using small molecules because they're easier to develop in a whole lot of ways, not just because of immune problems. We will need enzymes to break a range of glycation-derived cross-links that ALT-711 can't, because many of the cross-links found in our tissues are thermodynamically much more stable than the docarbonyl ones that ALT-711 breaks and will not be cleaved (catalytically or otherwise) by any small molecule unreactive enough not to have big side-effects. We have to tolerise ourselves to new proteins for a range of reasons; ways to do this are under intensive exploration.

3) The different behaviour of Sir2 etc in different organisms is a consequence of what organisms need to do to live longer when food is scarce. Animals that don't get cancer don't need to improve their defences against cancer in order to live longer; animals that do, do. A nice example of this was published by the Guarente group recently:

Mammalian SIRT1 Represses Forkhead Transcription Factors
Cell, Vol 116, Issue 4, 5 February 2004
Maria Carla Motta, Nullin Divecha, Madeleine Lemieux, Christopher Kamel, Delin Chen, Wei Gu, Yvette Bultsma, Michael McBurney, and Leonard Guarente

The NAD-dependent deacetylase SIR2 and the forkhead transcription factor DAF-16 regulate lifespan in model organisms, such as yeast and C. elegans. Here we show that the mammalian SIR2 ortholog SIRT1 deacetylates and represses the activity of the forkhead transcription factor Foxo3a and other mammalian forkhead factors. This
regulation appears to be in the opposite direction from the genetic interaction of SIR2 with forkhead in C. elegans. By restraining mammalian forkhead proteins, SIRT1 also reduces forkhead-dependent
apoptosis. The inhibition of forkhead activity by SIRT1 parallels the effect of this deacetylase on the tumor suppressor p53. We speculate how downregulating these two classes of damage-responsive mammalian factors may favor long lifespan under certain environmental conditions, such as calorie restriction.

4) The bad news is that the fact that these pathways are what organisms do naturally in order to live longer when it's evolutionarily advantageous to do so is the real reason why long-lived organisms won't get much out of any manipulations of those pathways (whether CR, CR mimetics, drugs that tweak these genes, whatever). CR and its new genetic approximations are thus a disastrous distraction from the real business of giving organisms substantially greater lifespans than they already have the genetic machinery to exhibit.

Posted by: Aubrey de Grey at February 22nd, 2004 5:28 AM

Meanwhile, the conversation continues on the ExI list.

Joao Magalhaes:

I agree with the argument that the essence of life is very similar amongst eukaryotes. I mean, the way the system is designed is similar between yeast and human cells. Yet unicellular organisms don't need apoptosis and thus linking findings in yeast sir2 to findings in human cells with SIRT1 and apoptosis is dubious.

As for p66, the main interest in it is that knocking out p66 in mice increases their lifespan and may also delay aging. Since p66 is related to apoptosis, and mice have tons of apoptosis, my opinion is that p66 -/- mice live longer because of changes in tissue homeostatsis caused by having lower levels of apoptosis. I very much enjoyed the ideas in:

Warner, H. R., and Sierra, F. (2003). "Models of accelerated ageing can be informative about the molecular mechanisms of ageing and/or age-related pathology." Mech Ageing Dev 124(5):581-587.

Although Warner mentions his ideas in the context of models of accelerated aging, I tend to see the delayed aging witnessed in p66 mice as similar. You can also easily extrapolate these models into Werner's syndrome.

Robert Bradbury:

We seem to be in relative agreement with the similarity of genetic programs in eukaryotes as well as the "all bets are off" perspective when one goes from unicellular or small multicellular (potentially short-lived) organisms to larger longer lived organisms.

So clearly the extrapolation of small, short lived to large, long-lived needs to be resolved. I do not believe that will be easy.

I would grant that less apoptosis (p66 -/-) could easily extend lifespan. The next question I would ask is whether such mice have a higher rate of death from cancer? If so this would imply that short lived animals may be on the same balance as longer lived animals -- increased apoptosis results in increased cell death which leads to aging, lack of apoptosis leads to increased cancer rates that kill you via another path.

This is an important question because it involves whether or not short-lived and long-lived species are on the same cancer vs. aging tradeoff or on different paths.

> Although Warner mentions his ideas in the context of models of accelerated
> aging, I tend to see the delayed aging witnessed in p66 mice as similar.
> You can also easily extrapolate these models into Werner's syndrome.

Be very careful with linking these two (p66 and WS). p66 may be involved in the apoptosis pathway which if successful may prevent the development of cancer but would require a complementary program in stem cell development to provide replacement cells for cells lost due to apoptosis. WS is very different as I believe the WS gene is an exonuclease meaning that it is going to corrupt the DNA code during any repair process.

In short, the decision is:

a) Do I kill cells that have sustained damage -- hoping that new cells derived from stem cells will take over their functions?


b) Do I try to repair cells even though I may be producing a corrupted genome in those cells in the process? (And of course the corrupted genome has greater chances of causing problems in the future.)

I do not believe that nature has developed a good genetic program that resolves these questions. Without the view that genomes are genetic programs with patches on top of patches on top of patches I believe it will be difficult to fully understand them.

Posted by: Reason at February 22nd, 2004 1:16 PM

The discussion continues.

Joao Magalhaes:

Although the p66 mice live longer, there is little information regarding specific age-related pathologies. I don't remember reading anything regarding cancer rates in p66 -/- mice, so I would assume cancer rates do not change but I could be wrong.

As for WS and p66, I think you missed my point. At a molecular level they may have different roles. Yet I was looking at aging from a cell population perspective. At a tissue level, p66 decreases apoptosis, meaning tissues can have cells dividing slower or have less cells dividing to maintain tissue homeostasis. In WS, cells divide slower and less vigorously, meaning there must be more cells dividing to maintain homeostasis. The bottom line is: in p66 you have a higher production of cells while in WS you have a lower production of cells, meaning p66 allows cells to divide less frequently while WS obliges cells to divide more frequently. From this perspective it makes sense that p66 delays aging while WS decreases longevity and accelerates aging.

Just like you can't understand a painting by looking too closely, maybe you have to zoom out from the molecular perspective to understand aging.

Apoptosis is also affected by WS, though I'm not sure whether it's increased or decreased.

A couple of references on p66:

Migliaccio, E., Giorgio, M., Mele, S., Pelicci, G., Reboldi, P., Pandolfi, P. P., Lanfrancone, L., and Pelicci, P. G. (1999). "The p66shc adaptor protein controls oxidative stress response and life span in mammals." Nature 402(6759):309-313.

Nemoto, S., and Finkel, T. (2002). "Redox regulation of forkhead proteins through a p66shc-dependent signaling pathway." Science 295(5564):2450-2452.

And a couple of references on the link between WS and apoptosis:

PS: A good question then would be why mice mutant for WRN do not age faster since their cells divide slower in culture.

Posted by: Reason at February 22nd, 2004 8:15 PM

Thank you for an interesting discussion.
May be you can help, I cannot find any reference: did anybody directly show that calorie restriction downregulates p66?

Posted by: Marina Kovina at May 2nd, 2006 11:31 AM

Post a comment; thoughtful, considered opinions are valued. Comments incorporating ad hominem attacks, advertising, and other forms of inappropriate behavior are likely to be deleted.

Note that there is a comment feed for those who like to keep up with conversations.