Why we age, and why we might not have to…
A deep dive with Dr. David Sinclair about his new book and the state of longevity science
Extending lives has been a dream since humans first evolved enough consciousness to reflect on their inevitable deaths. Could it, might it, finally be the case that after so many false promises we are finally on the doorstep of substantial life extension increases?
David Sinclair, a well-known and respected professor of genetics with Harvard Medical School, and co-director of the Paul F. Glenn Center for the Biology of Aging, believes that we are getting close. Sinclair runs a lab at Harvard Medical School and also at the University of New South Wales in his native Sydney, Australia, supervising in total more than 30 researchers.
From this vantage point at the top of his field he tells us, with a charming Australian accent: yes, we are on the verge of truly remarkable extensions in lifespan and healthspan.
Various other researchers have had a message that was similarly optimistic to Sinclair’s, but very few have had his combination of optimism and academic pedigree. You may not agree with his projections for the future of human lifespan but you certainly can’t ignore him or his optimism.
Of course, no one knows with any certainty what human lifespan will be in 2050, 2100 or even in 2030. But Sinclair suggests that we have a good chance of increasing life expectancy for humans in advanced nations to about 113 by 2070, up from about 78 currently, using already available medical knowledge and therapeutics. That’s one of his more conservative projections.
Here’s a zinger from his new book, Lifespan: Why We Age — And Why We Don’t Have To: “It is not at all extravagant to expect that someday living to 150 will be standard. And if the Information Theory of Aging is sound, there may be no upward limit; we could potentially reset the epigenome in perpetuity.” (Italics in original). Resetting the epigenome in the manner he mentions, in perpetuity, would mean an effective immortality for those choosing to undergo such treatments — if we want it.
I and many other people want to believe these projections. I’d like the option to extend my own life and those of whom I love indefinitely. Choices are good. But we have a long way to go before we’ll know if Sinclair’s speculations are accurate.
His book contains far more, however, than speculation about the state of aging 50–100 years from now. Most of his book focuses on the science of what is available now for extending lifespan and healthspan, including diet, exercise, sleep, and other lifestyle choices. He also discusses somewhat extensively, without being comprehensive, therapeutics available today that may extend lifespan in humans, including resveratrol, metformin, rapamycin and NAD boosters like nicotinamide riboside (NR) and nicotinamide mononucleotide (NMN).
His basic thesis, which forms the core of his new “Information Theory of Aging,” is that all life forms include a “survival circuit” that evolved to help organisms through tough times like drought and low or no food. When hard times activate this survival circuit our bodies kick into ultra-efficient mode. And by stimulating various aspects of the survival circuit the right amount — not too much and not too little — we can trick the body into living longer and healthier without actually putting our health at risk.
Additionally, it is epigenetic changes, not genetic mutations, that lead to aging and death, Sinclair argues in his book. If we can undo epigenetic changes we can undo aging. We can effectively induce our body to slow down or turn back the clock on epigenetic changes, through lifestyle changes and therapeutics. But it will require not-yet-available medical interventions like epigenetic reprogramming to achieve substantial gains in human lifespan.
These are big claims and we examine some of the basis for these claims in the below interview.
We conducted this interview via email during the fall of 2019.
You write optimistically in your new book that we are on the verge of major changes to human lifespan. You identify partial cellular (epigenetic) reprogramming as perhaps the most compelling research for these conclusions. Why does this research make you so optimistic?
For the past 20 years, the field has been working on fasting mimetics. That has gone well and there are drugs in development. New research indicates the actual age of cells can be reset by reprogramming, that there is a backup copy of the original state of the epigenome. This makes me optimistic that one day aging can not only be delayed but aspects can be reversed.
What other areas of research do you consider to be the most promising for extending human lifespan and healthspan?
Senolytics are super promising. Agents that suppress retrotransposons. Activators of autophagy. And rapalogs that mimic low amino acid conditions.
You suggest that eliminating senescent cells may lead to significant improvements in the elderly for various pathologies. Are senolytics an effective treatment if through such treatment these cells are eliminated rather than rejuvenated or rehabilitated back to normal functioning? Can senolytics be used effectively multiple times if we are eliminating a certain number of senescent cells each time we do such treatments?
Reprogramming could be used to prevent senescence but when senescent cells nevertheless accumulate, then you’d presumably need to bring in the senolytics. Presumably there is a limit to how many times you can eliminate senescent cells or reprogram the body. We don’t know that number yet.
Is it risky to be using senolytics without knowing the long-term effects of killing senescent cells? Could stem cell therapy be used effectively to replace cells killed with senolytics and thus avoid long-term loss of tissue function through elimination of too many senescent cells?
Killing senescent cells doesn’t seem to hurt old mice unless there are massive amounts of senescent cells in the body, such as when they have a large tumor. That said, this is something we are continuing to study so that we can better understand any risks involved and the long-term effects.
How do we know what the youthful epigenome should look like in humans for each cell type? Or do we not need to know as long as we apply Yamanaka factors (the genes that roll back the epigenetic clock) the appropriate amount of time?
We looked for a reversal of the DNA methylation clock and restoration of a youthful gene expression pattern. We also look for a restoration of tissue function.
So we don’t need to know the specific youthful epigenome patterns in each tissue type as long as we can test for a youthful gene expression pattern and/or tissue function for each cell and tissue type?
Yes, but it’s always better to know aging has been mitigated and by how much. We currently need to make clocks for each tissue but we are also learning how to make universal clocks that work on any tissue.
Will reprogramming be possible for nervous system tissues? Could issues arise with respect to loss of memories or other neural functions from turning back the epigenetic clock in these tissues?
We think so because we have been able to in mouse neurons. We don’t know the effect on memory but are testing that.
You don’t discuss telomere therapy very much in your book, which has received a lot of attention over the years. Is this because you don’t see it as particularly promising?
Telomeres have received less attention lately (by scientists in the field) because telomeres aren’t as accurate as the methylation clock and they don’t always affect aging in animal studies.
Since epigenetic reprogramming seems to return telomeres to a more youthful state (longer telomeres) as well as changing epigenetic markers like methylation, how do we determine the most likely mechanism(s) for the observed youthful gene expression and/or tissue function that results from reprogramming (Lu, et al. 2019 preprint)? Could it be telomere lengthening that is the more direct mechanism for these observed outcomes? Or is it pretty obvious that epigenetic changes are the mechanism?
Nerve cells aren’t dividing so it seems unlikely that lengthening telomeres (to allow for more cell division) is the reason for the restored eyesight we see [in mice]. And when we accelerate aging in our mice, telomere length isn’t appreciably affected, yet they get older.
Isn’t it true that shorter telomeres can greatly affect healthy gene expression and epigenetic regulation independently of cell division issues that occur when telomeres become excessively short? (E.g. Mukherjee, et al. 2018; see also my interview with Michael Fossel discussing telomere attrition as an analog process of steadily declining cell health). So isn’t it possible in this case that telomere dynamics are playing a role?
I believe so. When we studied yeast sirtuins, we saw effects of telomeres on epigenetic regulation and vice versa.
Amano et al. 2019 (you are a coauthor) describes how “telomere shortening in livers of telomerase knockout mice leads to a p53-dependent repression of all seven sirtuins,” and also highlights gaps in our knowledge of telomere and sirtuin dynamics: “While both telomeres and sirtuins are independently implicated in aging and disease, how they are interconnected in driving disorders is not well understood.” Does this data support the single cause approach (all aging is caused by epigenetic dysregulation) that you advocate with your Information Theory of Aging?
Telomere length and sirtuin-regulation of the epigenome are intimately linked. Sirtuins protect telomeres and telomere length controls sirtuin activity. In yeast, Sir2 performs this function and in humans, SIRT1 and SIRT6 do. When telomeres shorten, the sirtuins are freed to go bind somewhere else in the genome and when telomeres trigger a DNA damage response they also affect sirtuin activity. What this means is that telomeres can affect the epigenome in profound ways even before they cause cells to senesce.
I, like a great many people, are very excited that you are so optimistic about the potential for significant human healthspan and lifespan improvements before long. However, isn’t it pretty easy to be critical of your optimism given that we have no completed phase 3 clinical trials (and very few even in phase 2) for any therapeutics at this juncture? Won’t it be two or three decades until we can know if we’ll have effective and safe therapeutics of the kind you suggest becoming commercially available, let alone available and cheap for the mass market?
There are already two generic drugs sold on the market that fit the criteria of a medicine that slows aging (as I say in the book). No need to wait “three decades.”
Are you referring to resveratrol and metformin, which you and some of your family and colleagues have been taking, and are available and affordable now? Won’t we need decades of track record in multiple trials and long-term studies to establish with any certainty that these drugs do achieve significant added years to human lifespan? And to be sure that there are no unexpected side effects from long-term use? Ditto with new therapeutics as they become commercialized.
Yes. We need more clinical trials. Unfortunately, they aren’t easy and they take many years and millions of dollars. But my esteemed colleagues and I in the aging field are doing what we can with the resources we have.
Turning to theories of aging, you outline in your book, for the first time as a formal theory, what you call the Information Theory of Aging. In your theory, aging is the result solely of loss of epigenetic regulation, which leads to undesired gene expression and the lack of some genes being expressed that should be expressed. What kinds of tests could falsify the ITA?
If you accelerate loss of epigenetic information in an animal or human and there is no impact on aging or only a few aspects of aging happen then it is not a universal theory. Like all good theories it is easily disproved.
You describe group selection theory in your book as “dead wrong” (p. 10), but how else can we reasonably explain the occurrence of aging and death far sooner than we know nature could achieve, including in humans, (all else equal, evolutionary theory predicts that organisms will live and reproduce as long as possible)? Since nature has evolved a number of organisms with negligible senescence, there seems to be no fundamental natural road block for the evolution of negligible senescence in humans and other mammals, and yet there aren’t any such mammals (excepting perhaps naked mole rats). Are you familiar with Josh Mitteldorf’s Demographic Theory of Aging (based on group selection theory) and his responses to critics of group selection theory (e.g. Chapter 3 and 4 of his book Cracking the Aging Code)?
There isn’t strong pressure to build long lived bodies. Evolution only selects for bodies that survive long enough to ensure genes are passes on, and little more than that.
Similarly, if calorie restriction, exercise, and other types of hormesis are sirtuin activators (p. 91), leading to longer healthspan and lifespan, why hasn’t natural selection found other ways to achieve sustained sirtuin activity in times of normal caloric intake or caloric abundance?
Building a long-lasting body (or soma) is non-adaptive if you are unlikely to survive for long. Instead, the species evolves to put more resources into reproduction.
You argue in your book that epigenetic changes cause aging, whereas it seems many other scientists remain agnostic whether epigenetic markers are a cause or a consequence of aging. Morgan Levine stated in my recent interview with her: “I don’t think there is evidence to suggest that altering the epigenome directly will be beneficial. If some of these changes are effects (read-outs) of aging, then they are not the correct points of intervention. Further, many of the changes may be compensatory, and thus making an old cell epigenetically young but leaving it in an aged organismal environment could be detrimental and possibly contribute to neoplastic transformations.” A 2019 preprint (Lu, et al.) of yours, which you highlight as important in demonstrating reprogramming in mice retinal cells, states: “In mammals, progressive DNA methylation changes serve as an epigenetic clock, but whether they are merely an effect or a driver of ageing is not known.” At the end of the paper you and your colleagues conclude: “altered DNA methylation patterns may not just [be] a measure of age but participants in ageing.” What are the key data that have convinced you that epigenetic changes are indeed a cause and not a consequence of aging?
I’m never “convinced,” which is why I use the word “may.” The necessity of the Tet enzymes that remove DNA methylation for rejuvenation is suggestive.
But you argue in your book, stating it multiple times, that aging is the result of a single cause: epigenetic changes (for example, p. 84 discusses building a “single dam-at the source” to prevent aging). So you are arguing that epigenetic changes are causal and not mere consequences of aging elsewhere in the cell, but also that they are the only cause of aging. Is this your position, or the softer position stated in Lu et al.?
Epigenetic changes are causal.
Kane and Sinclair 2019, another recent paper of yours, summarizes the epigenetic theory of aging (without labeling it the Information Theory of Aging, as you do in your book) and concludes, in part: “Epigenetic changes may contribute to many of the hallmarks of aging including cellular senescence and mitochondrial dysfunction.” This seems to be a softer assertion than in your book, which argues strongly that all hallmarks are indeed caused by epigenetic dysregulation. Is this difference in emphasis a result of your personal conviction that aging is monocausal and epigenetic, and others are slowly catching up to this realization?
I have put forth an Information Theory of Aging. Like every other scientist on the planet, I don’t know if this hypothesis is correct. Fundamental hypotheses of merit need to stand up to millions of experiments, many of which could weaken or disprove the hypothesis. As I wrote in my book, most hypotheses are eventually replaced. I will spend many years trying to disprove this theory. If we can’t disprove it, we will be more convinced of its validity.
You state (p. 171) that we don’t know how the mechanisms behind reprogramming with OSK “know” which epigenetic markers to remove. How do we make progress on this key part of the puzzle? Is there a vintage marker of some sort in each epigenetic tag that allows TET to select which vintage to remove and thus what epigenetic age to return to?
I’m working hard on this problem. I don’t want to give away all my ideas.
Last, what are the biggest societal changes that we can expect when people can reasonably hope to live to 150 years or longer?
I don’t like this number being thrown about casually. It is far in the future. We have more pressing near-term changes to be discussing.
But you bring up this figure in your book and also discuss possible impacts of such increases in lifespan. You spend the last third of your book looking at social and ethical issues surrounding lifespan extension. I agree that this is an extremely important part of the discussion.
The biggest change will be in how people view their lives. If you are 50 and you have another 50 healthy years to go, then you will live your life very differently. You will have the time for multiple careers, see far more of the world, spend more time with healthy great-grandparents. There will be other big changes to retirement age, social security and healthcare costs, political movements, population growth — some good things, some bad things — but longer lives are coming.