Longevity: The Disposable Soma Hypothesis

The disposable soma theory is an explanation for why ageing happens. As well as answering the "why" question, it tells us quite a lot about how ageing happens — and I think that's really important, because if we understand why something happens in biology, it gives us a feeling for the shape of the processes involved. Ageing is a very complicated process, and trying to understand it purely from a phenomenological basis is hard work.

The essence of the theory is this. If you look at the natural world, animals as a rule don't live terribly long. There are all kinds of hazards out there — they can be killed by a predator, die of starvation, catch an infection, die of cold. So survival in the wild is curtailed by the high level of risk that organisms are exposed to. At the time, I was working on the mechanisms that maintain and repair the body, and some of these are very expensive in terms of metabolism, in terms of the energy required. And I was thinking: if the body isn't going to last very long anyway — because an accident is going to get you — how much should your genes be prepared to invest in its long-term maintenance?

"Soma" is really just Greek for body. It refers to all the parts of the body that are not part of the reproductive lineage — the germ cells. And the idea is that what matters in evolutionary terms is that the germline is maintained well enough to keep the lineage going from generation to generation. Each baby has to be born in tip-top condition, with its clock effectively reset to zero. But for the rest of the body — the soma — all you require is that it remains in good enough shape to last you through the period of time you can expect still to be alive in your wild environment.

Think of a mouse. Small animal, exposed to many dangers. There have been excellent field studies of mouse survival in the wild, and we know that 90% — probably more — of wild mice are dead before their first birthday. Very few make it past 15 or 18 months, tops. So what does a mouse need in terms of bodily maintenance? Enough that it doesn't fall apart too quickly. It's no good falling apart at two or three months. But if you fall apart at 18 or 24 months, it's really not a problem — and more than not being a problem, it may actually free up resources. Rather than spending them on better maintenance than you need, you could use them to do other things that matter more for being competitive in the wild. That, in essence, is the disposable soma theory: you've got to look after your germline, and you've got to look after your soma too — but only just enough.

We're not programmed for death. We're programmed for survival. It was just never a high enough priority in the wild environment to be programmed well enough to last indefinitely.

I'm not an evolutionary biologist, but I'd been introduced to the writings of one of Britain's great evolutionary biologists of the latter 20th century, John Maynard Smith at the University of Sussex. I liked his approach — he was always asking why, why, why. It's like going back to childhood: yes, but why?

This was moonlighting from my work on blood. I'd had a chance encounter with a very eminent geneticist at the National Institute for Medical Research who was working on cellular ageing — trying to understand why ordinary cells of the body can divide only a certain number of times, then stop and seem to give up the ghost and die.

I was also interested in a theory that was quite influential at the time: that cells might age because they make mistakes in building the important macromolecules on which life depends. They make mistakes in DNA — somatic mutations — which could lead to an accumulation of trouble. But the particular idea I was working on was that you'd also make mistakes in protein synthesis. Protein synthesis is not a precise, accurate process; significant fractions of the proteins made in our cells are made with mistakes. If you think about how proteins are synthesised, it requires an incredible amount of molecular discrimination. You have the messenger RNA being translated, the transfer RNAs come along, they plug the codons into the anticodons and carry amino acids on the end, and if all goes well that amino acid gets incorporated into a new protein chain. Almost the most difficult job on the planet is the work done by enzymes called aminoacyl-tRNA synthetases — they have to make sure to put the right amino acid onto the right tRNA.

It had been suggested that if you make mistakes in building proteins, some of those faulty proteins will be part of the machinery that synthesises the next round of proteins, and the next, and the next. So you'd generate a positive feedback, and the whole system would blow up — come to failure. I'd done some mathematical modelling of this which showed it could happen: a runaway "error catastrophe," as it was called. But it was also possible that if you invested enough energy in proofreading these enzymes, or in clearing away your mistakes, you could maintain stability over time. That was the key — you buy stability by investing energy.

And I started thinking, having just been made aware of the distinction between germ cells and somatic cells: if we think about this in almost economic terms, you want to spend a lot on the stability of your germline — but how much do you want to spend on the stability of your soma? Literally, it was sitting in the bath one cold February night in early 1977 that the idea just popped into my head: evolution will prioritise investment in the stability of the germline, and have only limited interest in the stability of the somatic cells. That was the original formulation. But it's a principle that, when you think about it, has quite wide ramifications.

It got me interested in how natural selection acts on repair, regeneration and maintenance programmes generally, and in applying that evolutionary logic to the biology of ageing. It tells you several things. It tells you that ageing and death are not directly programmed — longevity is what's evolutionarily programmed, because your longevity is set by how the genes regulate your maintenance and repair. A very bright graduate student of mine, when I was in Manchester, did a lovely experiment: taking cells from little biopsies from mammals of different lifespans, and showing that cells from the longer-lived mammals were much better at looking after themselves than cells from the short-lived ones. So the theory leads naturally into an explanation of how different species evolved different lifespans, and it makes a clear prediction — cells from long-lived species should be better at maintenance than cells from short-lived ones. We buy our longevity that way. And that has turned out to be the case.

It also tells us that, because faults can arise through a large number of mechanisms, we should be prepared to discover that ageing has multiple causes. Throughout my career there's been a lot of competition between different hypothesised mechanisms — people championing the mutation theory, or oxidative stress from free radicals, or errors in proteins, or this or that. What the disposable soma theory tells us is that it doesn't make sense to think of it as one mechanism versus another, like gangs in a school playground all trying to be dominant. They work together, they synergise, they interact. Some of the most exciting papers I've produced over the last 10 or 15 years have explored precisely this kind of complexity and networking between the different mechanisms of ageing.

A clear prediction is that there's a trade-off between longevity and reproduction. In some experimental systems — nematode worms, fruit flies — you can interfere with and modify the processes that regulate reproduction. And where you do that and don't see the trade-off realised — you change reproduction but see no corresponding change in longevity — people regard that as a challenge to the theory.

I think it's very interesting, but there are reasons to be a little cautious about how we interpret these challenges, because a manipulation you make experimentally won't necessarily play into a trade-off that may be wired into the biology underneath. It's for others to take these ideas forward and see how far they stack up.

Another apparent challenge — more apparent than real, I think — is the observation that calorie restriction can extend the lifespan of experimental animals. On the face of it that's a bit odd: if ageing evolved through competition for resources, why would restricting a mouse's food make it live longer? Back at the end of the eighties it was hypothesised that during a famine — a bad time to reproduce — the organism switches resources in favour of maintenance, in order to wait out the bad times and then pick up and carry on when food returns. That's a hand-waving evolutionary hypothesis.

So 20 years ago another graduate student working with me, Daryl Shanley — now a reader in ageing at Newcastle — tried, as part of his PhD, to test whether that hypothesis could really hold water. The thing about evolutionary ideas is that you can't rely on a verbal articulation; it has to work in terms of how natural selection actually operates. Natural selection is driven by numbers, so you have to translate the hypothesis into numbers to determine whether the strategy you're proposing actually results in an increase in fitness. And it turned out that yes, it could work — certainly for mice and rodents, and for many smaller animals where the food supply is unpredictable. So the apparent challenge to the theory was actually a doorway to a deeper level of understanding: if the disposable soma theory is about how you optimally allocate resources between growth and maintenance, then where there's unpredictability in the environment, some organisms may have evolved the capacity to flex their strategies to suit their circumstances.

You see this in nematode worms, which are widely used as a model for ageing. When worms find themselves somewhere resources are scarce and stress is high, they turn off the developmental process and lock themselves into a highly stress-resistant dispersal form called the dauer larva, whose whole function is to get the hell out of there and find somewhere better. They live quite a long time. A lot of the experiments that first identified genes with a regulatory impact on ageing came from discovering the genes associated with this change of game plan in the worm's development.

A theory isn't there to be proved. If someone offered completely convincing knockdown evidence to totally squash the disposable soma theory, part of me would be absolutely delighted — because that's how science works — and part of me would be a little sad, because I think it has a lot of explanatory power. But I don't feel it's my role to fight off all comers. If you have challenges you'd like to put to me, please go ahead; my job is not to defend it against attack.

It could be. Something I should have said is that this was an evolutionary hypothesis to explain the phenomenon of calorie restriction. It's entirely possible that calorie restriction has explanations that have nothing to do with ageing or evolution, and may not require the disposable soma theory at all — it could be entirely mechanistic. One of the things that happens when you're short of food is that you strip down and recycle molecules: you turn over your proteins more aggressively, the processes of autophagy are really activated. It's entirely conceivable that that would improve bodily health.

The reason for remaining open to the possibility that it's some kind of adaptation is that quite early on it was found that under calorie restriction, mice genuinely seem to up-regulate a lot of their maintenance systems — they do seem to work better. Another feature consistent with the evolutionary hypothesis is that mice should organise themselves to postpone reproduction, and indeed, calorie-restricted mice retain fertility past the age at which normally fed mice would have become reproductively exhausted and sterile. So it's certainly producing effects consistent with what the hypothesis proposes. But one should never be closed to alternative explanations, and it may well be a bit of both.

One thing that inclines me towards thinking it has some genetic, evolutionary component is a very important experiment that caused a stir at the time but somehow got overlooked. People took a large number of different inbred strains of mice — all in the same facility, all exposed to the same regimen of dietary restriction — and found that half the strains showed an increase in lifespan through restriction and half showed a decrease. You saw the whole spectrum: from the most extreme positive effect to the most extreme negative effect, with all the intermediate effects in between, forming a kind of triangular wedge shape. If it were simply the physiological consequence of having less food — greater protein turnover, say — you'd expect it to work the same in all the strains. The fact that it doesn't suggests, to my mind, that there may be a genetic architecture that has evolved for mice to respond to periods of famine and change the game plan in some regulated way. That work needs to be taken further, because it may give us insight into the makeup of that genetic architecture. So I still retain quite an open mind.

Absolutely, yes. These were all the same species — all mice — but different inbred strains. And the impact was not at all consistent from one strain to the next. If it were a purely mechanistic effect of calorie restriction equals longer life, you'd expect a flat line — no slope — with all of them showing the same effect. It's difficult to describe a graph in words, but that's the point.

It's a very exciting time in research on ageing and the mechanisms that determine longevity, but we're still finding our way. We have drugs that influence nutrient sensing — the IGF and growth-factor signalling pathways — which are certainly very interesting. We have drugs like rapamycin, reported to have some impact on longevity in mice. And metformin — as you may know, there's quite a big trial in humans looking at its potential role in modulating the ageing process. Most of our data, for obvious reasons, comes from animal studies; the big metformin trial is really the first to be done in a significant way in humans. Parabiosis is interesting too — we don't yet understand enough about how it works, but it will tell us important things.

The bottom line is that until we have results from the metformin trial, we don't actually have clear evidence for any intervention with a proven impact on the rate of ageing in humans — apart from the things our grandmothers could have told us about: healthy nutrition and exercise. One of the implications of the disposable soma picture is that ageing is malleable. Ageing is not hardwired in the sense that there are genes that actively turn something on to kill us; the length of life is regulated by the efficiency of our maintenance and repair. And we're learning a lot about how the interaction between body and environment affects both our exposure to damage and our capacity to repair. That's where nutrition comes in: a poor diet contains components that add to the burden of damage, whereas a healthy diet — we all talk about the Mediterranean diet, but remember that Japan is the longest-living country in the world, so there are Asian variants too — is associated with longevity. And the evidence that exercise is beneficial for healthy ageing is overwhelming; it's probably the single most important thing we can do if we want a long and healthy life.

So these are things your granny could have told you — but what we can now tell our grannies is that we're beginning to understand how they work. We can demonstrate that what was a matter of belief in previous generations is now firmly founded in science. This is going to be an area of rapid development.

One treatment to add to the list is the emerging science of senolytics. As we get older we accumulate senescent cells in our tissues and organs. This was the system I was working with back in the seventies when I first became interested in ageing — cellular senescence. At that time we believed cells aged and died simply because they accumulated bad stuff. We now know senescence is a much more regulated process; work my colleagues and I did in Newcastle 10 years ago gave insight into how a cell's entry into the senescent state is a regulated response to damage of various kinds. Once cells become senescent they stop dividing, and they often switch into a curious state called the senescence-associated secretory phenotype, or SASP — an inflammatory condition. They sit there in the tissue flashing inflammatory signals everywhere, and they can convert neighbouring cells into senescent cells. It's like wildfire spreading through a forest, igniting the trees around them. The evidence is accumulating that these senescent cells have deleterious consequences as we get older. So the idea has been around for a few years that if you could get rid of them, you might improve health in ageing and maybe even extend lifespan. Proof of principle has come from experiments in mice: if you target senescent cells for removal, you see some improvement in health and in median lifespan — though you don't see much improvement, if any, in the maximum length of life. So it's very much an open question what's changing. In fact, just last month a couple of colleagues and I published a paper looking at cellular senescence from an evolutionary standpoint. Senolysis is another area into which we need to do research — it's quite a hot topic.

You've put your finger on exactly the challenge: the length of human life makes it very difficult to get early results confirming that an intervention will extend it. If you gave the treatment to people already very old — in their eighties — you might get a result sooner, but you might also see no result because their bodies are already too damaged by age. We know from all kinds of drug studies that the earlier you treat, the greater the efficacy. But then, exactly as you said, you'd need a placebo-controlled randomised trial: one group treated, one on placebo, wait half a century, and compare. It's daunting. It's also daunting from an ethical and regulatory point of view, because a drug with potent biological effects — not all of which may be good — would be given to young, fit, healthy people with no underlying condition. That raises real regulatory challenges.

So how do you get around it? One possibility, as you suggested, is a surrogate marker that gives an early indication of whether the drug is working. But we have to recognise that's not the same as proving the treatment extends human life — there's no way around waiting to prove the outcome. What a surrogate would let you do is, for example, abandon trials that show no hope of progress: some earlier marker that lets you say, "good idea, but not working." For that you need a biomarker of age. You'd take a treatment group and a placebo group, and maybe after five years take a blood sample, run the biomarkers, and say, "the treated group is definitely not showing the age-related change we're seeing in the controls." That would give you grounds to believe you were onto something.

It would be lovely to do that with complete reliability, and the hunt is on. Probably the best biomarker for that process is the epigenetic clock, based on epigenetic profiles. The interesting thing is that although it works better as a biomarker of age, we probably have rather little understanding of exactly what it's reading. So it may have enormous promise, but we still need to fill in the gaps. The need for this has been recognised for many years — there was a paper in the Lancet in 1969 arguing for a battery of tests to measure the ageing rate in humans, and the National Institute on Aging in the US ran a big, multi-million-dollar biomarkers-of-ageing programme in the 1990s that produced a lot of data but didn't really come up with anything. That doesn't mean we should be pessimistic now, with the greater knowledge we have — but it's research in progress. What we have to do is develop our capacity to assess biological age at the same time as running these trials.

I'm really optimistic about the potential for ageing science to improve health and quality of life in older age. But the timescale challenge you articulated is a very big one. Remember, with any drug you not only have to determine efficacy, you have to do things like dosage studies — working out the optimum dose and regimen. If your lag time between trying it and getting the answer is decades, it's a real headache.

These are important questions. A lot of the discussion at the moment is about taking existing, already-licensed drugs and repurposing them — attractive partly because you haven't got the big regulatory hurdle of a new drug for which you must prove efficacy and long-term safety. But these aren't likely to be products that earn pharmaceutical companies a huge amount of money. When you think about the lifetime of a patent against the timescale problem — if you patent a new drug and the patent expires before your clinical trials can even be completed — it's not financially attractive. So there has to be serious attention given to how this might work. Existing drugs may deliver the benefits we hope for; or we may need entirely novel drugs.

It also raises a raft of societal and ethical issues. Works of fiction have been written about these scenarios — where someone produces a drug available only to those who can afford it. Brave New World played in that area, and there have been many other books of that kind. So I think we need a serious discussion about how it would happen, and an even more serious discussion about what it is we'd like such drugs to do. We talk about life extension — and yes, it's entirely consistent with the whole progress of modern medicine, which is about preventing people from dying. But is the length of life really the best metric of success here? Might we instead think about treatments that improve health? Within the field there's now a very proper focus on maximising healthspan over and above lifespan — to my mind, absolutely the right priority — but that needs unpacking. There are multiple age-related diseases: do we want to postpone them all, or are there some we'd be more willing to accept than others?

The desire to make progress here is completely understandable. But we need to be clear about the way we're going to do it. We look at the spectacular successes of recent technological advances and slip rather easily into thinking, "gosh, we humans are so clever — look at the internet, look at satellites." I think we need to harness that cleverness to be really incisive about what we're actually trying to do in ageing.

To be honest, I think that's symptomatic of our obsession with immortality, over and above more pressing and urgent goals. It's interesting to read, but I think it remains firmly in the realm of speculation until we show we can do the basics. There have been forceful critical analyses of the idea — the US demographer Jay Olshansky has picked that logic apart. I don't spend a lot of time thinking about it, because it's so far beyond what we're capable of addressing now. It's fun, in the way I still enjoy reading science fiction — but there's so much we should be using our improving understanding of ageing to achieve, obviously important and hopefully achievable, before we could think about escape velocity and immortality, that I think it's self-indulgent to spend a great deal of time on it.

Let's put this in perspective. Over the last 200 years we've doubled the average human lifespan. We live in a world with many unsolved challenges arising from having so many more older people. We see massive social inequality — the social gradient in health and life expectancy. If you live in disadvantaged conditions, you're likely to fall ill with multi-morbidity and to die perhaps 10 or even 15 years before people who are more advantaged. We have proof from what happened in Germany after reunification: life expectancy in the East was significantly lower than in the West, and the improvements in living conditions that came with reunification brought former East Germany up to match the West. We know what the challenges are, and we know what can be done about them, and we should be harnessing everything we can to make those improvements. They're not easy. Some people would say, "yeah, yeah, we know all that — let someone else get on with it; we want to tackle the sexy question of whether we can make people live forever." I'd say the question of how we reduce the social gradient in healthy life expectancy is a really hard problem — so let's work on it. Fairness, equity, and opportunities to ameliorate the conditions that reduce quality of life in older age are very important.

One of the things I did early on at Newcastle was to start the Newcastle 85+ Study, a cohort study of what life was really like for people as they passed 85. It's still ongoing. It started in 2006 with people born in 1921, so this year is the year of their 100th birthdays — and relatively few are still alive. But they've taught us so much. We learnt, for example, that three-quarters of people at age 85 have four or more age-related conditions — yet, in spite of that, four out of five rate their health and quality of life as good, very good or excellent. So we need to understand some of the paradoxes that emerge when we look closely at what life is really like for older people. I love the science, and I love speculating about where it may lead — but I like to believe it will allow slightly more modest achievements than making people live forever: improving quality of life at older ages, and levelling the playing field.

It's an interesting question — and thank you for the Epic of Gilgamesh reference. I'm aware of that story; it's a beautiful example of how the question gets posed. There are other wonderful examples in literature where people face that kind of dilemma. What's interesting is what it tells us about human aspiration. Literature doesn't contain the blueprint for living, but it gives us points to reflect on. It's become a hobby of mine to read books and see how people reflect on ageing and the experience of immortality.

I don't think I can answer the question clearly for myself yet. I'm getting older, but I may not be quite old enough. I've had fun in recent years — ageing has become an interesting topic, and the Nobel Foundation has started to organise dialogues on certain subjects. I've been asked four times now to take part in Nobel dialogues on ageing, where they bring together field experts like myself alongside Nobel laureates — and the laureates aren't all in physiology or medicine; some have won the literature or economics prize. Very often those discussions are so illuminating. I work in a research environment and think of these questions from a research perspective, but I was at an event like this in Madrid a couple of years ago where a Nobel laureate in literature made a very similar point to the one the Epic of Gilgamesh tells us. I find that deeply interesting.

I don't want to encourage people to be thinking all the time about old age and death, but I think we all should think about our life journey. Possibly the one thing that troubles me most about the obsession with the quest for immortality is that it reveals a deep fear of death. And death is going to be with us for quite a long time. I think we all benefit from taking the time, and being bold enough, to think about it — it's a matter that should concern us all. But not all the time; most of the time we should just be getting on with the joy of being alive. There is a certain joy to be found in age as well.

I've benefited from having an interdisciplinary perspective. I did my first degree in maths, then worked across into the life sciences — training first as a biostatistician and later doing a PhD in biology. That interdisciplinary perspective is valuable when you're working on a problem as complex as ageing. Because I came at it that way, I also didn't grow up within an established school of thought. At the time I was taking my early steps, there was really no significant community of researchers on ageing, so almost everything I did I had to discover for myself, to a significant degree. I had mentors — people I took inspiration from — but I wasn't growing up within an established career trajectory. There are pluses and minuses to that, of course, but on balance I think it helped the way my career played out.