Epigenetics of Aging, and Prospects for Rejuvenation

Last week, I attended the tail end of a Keystone conference on Epigenetic Regulation of Aging, followed by a one-day brainstorming session to kick off a project called GILGA-mesh, intended to take this bull by the horns. Though the subjects of the two days were virtually identical, the approach and attitudes of the scientists in attendance set very different tones.  Both days featured smart, creative and careful scientists, but they saw the same material through different frameworks.  Sometimes philosophy makes a difference.

For readers who know me less well, I should introduce my perspective: I believe that aging is an evolved epigenetic program.  When we are young and growing, particular genes are turned on and off with exquisite timing to determine the growth and development of bones, muscles, and organs.  When we are old, the program continues, more slowly and more diffusely, but inexorably nonetheless.  Genes are turned on that destroy us with inflammation and cell senescence and auto-immunity and programmed cell death, while the systems that protect us from pathogens and from free radical damage are gradually shut down.  Evolution has left nothing to chance.

[I first wrote an academic paper about this idea in 2013, excited by a paper by Aviv  Johnson on methylation, but unaware that Tom Rando had written on the same lines the previous year. Jeff Bowles had hinted at similar ideas in a paper more than a decade earlier. Soon the field was broken wide open by the work of a bio-statistician.
Steve Horvath ran a computer analysis on thousands of genes as they are  expressed in young and old humans, and produced an “epigenetic clock” that could accurately report how old a person using measurementis of methylation in 353 DNA sites in particular.]



Epigenetics is a new science in the 21st century.  All the cells in one body have the same DNA (pretty much), but differernt genes are “expressed” (translated into proteins) in different tissues and at different times, and this is what controls the body’s metabolism.  In fact, only 2% of our DNA is genes, and 98% determines how the DNA is folded and spooled, opened and closed at particular times and places, and this in turn controls gene expression.  We are 2% genetic and 98% epigenetic.

There is a language called the “genetic code” which determines how genes are translated into proteins.  It was decoded by Francis Crick and others in the 1950s.  It is as simple as it can be, and is completely understood.  There is another language, the “epigenetic code” that determines gene expression.  It is anything-but-simple, with a convoluted and self-referential syntax that we are just beginning to understand.  The epigenetic code starts with signals embedded in the DNA that serve as “start” and “stop” codons. The stretch in between comprises a piece of a gene, a kind of Gutenberg movable type that is transcribed from the chromosome and then spliced and combined to form functional RNAs and proteins. The complicated part of the epigenetic code is implemented as a pattern of methyl and acetyl groups.  These are little chemical decorations that attach to the DNA and to the “histones” (spools around which DNA is wound up in the cell nucleus for safe storage).  The methyl and acetyl groups are continually being attached and removed according to instructions that come from within the cell and other instructions that are passed through the blood. It is the methyl and acetyl groups that determine how the DNA is folded and spooled, which effectively turns particular genes on and off as needed.

The part of the epigenetic code on which we have the best handle at present is called “methylation of CpG islands”.  Long stretches of DNA have CGCGCGCG… on one strand, complemented by GCGCGCGC… on the other.  Often the C’s in this region get an extra methyl group, turning from cytosine to 5-methylcytosine.  Then this stretch becomes a “repressor region,” a signal to NOT express the adjacent gene.

DNA methylation can be persistent, turning a gene off for decades at a time.  When a cell divides and its DNA is copied, the methylation pattern can be copied with it.  This accounts for some of the persistence of epigenetics, and the way gene expression can be inherited across generations.

DNA methylation has been appreciated for 30 years, but two recent developments make the subject attractive and accessible to research.  (1) There is now a simple lab/computer technique for reading the methylation pattern from DNA.  It relies on commercially available, automated machinery for PCR to sequence a full genome before and after chemical modification of the methylated C’s.  (2) There is now a simple lab/computer technique for changing the methylation state of any chosen target site in the DNA.  It is based on CRISPR technology that is taking genetics labs by storm the last two years.

from Hannum et al., 2013, Genome-wide Methylation Profiles

from Hannum et al., 2013, Genome-wide Methylation Profiles

Epigenetics and aging

Three years ago, Horvath demonstrated that there are specific patterns of methylation associated with particular ages of the body.  It’s not just that the fresh, clear pattern of youthful gene expression becomes muddied and random with age—although there is some of that.  But it’s also true that some genes that are active in youth become inactive as we get older and (especially) that other genes that were suppressed in youth become activated in old age.  What Horvath’s paper says is, “show me  methylation pattern of a person’s cells, and I can tell you how old s/he is.”


Is epigenetics a cause or effect of aging?

The correlation between aging and epigenetic status is established beyond dispute.  But what does it mean?  This is the big question.  Most researchers think of the body as programmed by evolution to be as strong and healthy as possible.  So, when different genes are expressed in old age, they find it natural to assume that the body is protecting itself in response to damage that it has suffered over the years.  We express different genes when we are older because we need different genes when we are older.  This was the predominant attitude at the first conference (where I was present just for the last day).

The other possible interpretation is my own, and it has become common among those who are closest to the field of epigenetics.  It is that epigenetic changes with age are means of self-destruction.  The body is programmed to die, and its suicide plan is laid out in the form of transcribing an unhealthy combination of genes.  This idea flies in the face of traditional evolutionary theory.  (How could natural selection prefer a genome that destroys itself and cuts off its own reproduction?)  Nevertheless, the evidence for this hypothesis is robust.  The genes that are turned on don’t protect the body—quite the opposite.  Genes for inflammation are dialed up.  Genes for the body’s defense against free radicals are dialed down.  Cell turnover is dialed down.  DNA repair is dialed down. The mechanisms of programmed cell death (apoptosis) are strengthened in healthy cells, at the same time that they are perversely weakened in cells that are a threat to the body, like infected cells and cancer cells.


How will we determine who is right?

In my opinion, the existing evidence heavily favors the hypothesis that aging is caused by epigenetic changes, rather than the other way around.  When we look at the kinds of changes that occur, they seem to be pouring fuel on the fire, not putting it out.  Protective genes are turned off and inflammatory genes are turned up.  I also think that
parabiosis experiments provide a strong clue.  Three researcher groups (at Stanford Harvard, Berkeley) have shown that injecting blood plasma from a young mouse into an old mouse makes the old mouse healthier, and relieves some problems associated with age.  The blood plasma contains no cells—only signal molecules that are the product of gene expression.  This is powerful evidence that youthful gene expression is supporting a strong and youthful body, and (conversely) that the kind of gene expression that characterizes old age is not doing the body any good.

But the ultimate experiment will be to re-program gene expression in an old mouse and see if there is a rejuvenating effect.


My proposal

As of now, the GILGA-Mesh project is dominated by numbers geeks (like me) who practice the “Google approach” to bioinformatics.  Huge databases of gene expression are screened for epigenetic candidates that seem to be well-correlated with good outcomes.  I think what we need is an infusion of biolochemists who understand something about the body’s signaling networks, and can orient us toward “upstream” and “downstream” molecules.  Here’s my proposed program:

  1. Repeat Horvath’s (human) analysis for mice.  In other words, identify several hundred places where methylation is different in young and old mice.
  2. Determine which genes are associated with these regions.  (Map needed for this should already be available.)
  3. Look at the set of genes and identify transcription factors. These are likely to be “upstream”, in that they control other genes.
  4. Start with old mice.  Use CRISPR to change the methylation status in a handful of promoter regions that control transcription factors, making them match the methylation status of young mice.
  5. Measure metabolic functions to see if the old mice are more healthy or less after these procedures.  Look particularly for changes in inflammation, propensity for cancer, and especially life span.

If this experiment goes as I expect, we will be ready for rejuvenation experiments in humans.


How does the body know how old it is?

Even further upstream, is there a central master clock that dictates the body’s epigenetic expression, and thereby determines our biological age?  Logically, it seems that the body would need an accurate clock to time the events of growth and development.  Evolution likes to re-use the parts she has created, and it would not surprise me if the developmental clock morphs into an aging clock.

have reasoned that there are two possibilities.  It may be that there is a timekeeper, probably in the neuro-endocrine regions of the brain, that controls the processes of development and aging.  This possibility is supported by works of Kasper Daniel Hansen and Claudia Cavadas.  If this pans out, it would present the handiest target for true rejuvenation in humans.  But it also may be that epigenetic expression itself is a kind of clock that is diffused through the body.  Today’s gene expression includes transcription factors that control tomorrow’s gene expression, and so epigenetic state may be a feedback loop, or self-contained clock.  This may also be a target for rejuvenation, but a little accessible, harder to address or to tinker with.

Random notes—other things I learned last week

I was tickled to find how many members of the GILGA-mesh team already support the
programmed aging perspective that I have advocated.  I was particularly gratified to receive encouragement from Caleb Finch, a grand old man of the field who wrote the
encyclopedia of aging in 1990, and continues a very active research program today.

From Finch, I learned that infections in childhood and even in the womb can have a serious effect on diseases of old age, decades after the fact.  He hypothesizes a lifelong burden of inflammation.  Evidence includes an elevated incidence of heart disease for the cohort born just after the influenza epidemic of 1918.

I was chagrined to learn that air pollution, especially particulate matter, is associated with increased risk of dementia.  This poses a personal dilemma for me, as I plan to spend the summer at the lab of Meng-qiu Dong in Beijing.

I learned that hospital errors are the third leading cause of death in the US, accounting for about 10% of all deaths, about the same number as smoking.  Maybe you already read that in the New York Times.


74 thoughts on “Epigenetics of Aging, and Prospects for Rejuvenation

  1. I wonder if you can throw any light on why human females and some whales are the only mammals that experience menopause? Just curious.

  2. Hi Josh,

    Thanks for your latest article. Your proposed second alternative that an aging clock could be the epigenetic expression itself, seems less likely in light of the results of the parabiosis experiments from my perspective, your thoughts?

      • Hi Josh,

        To my recollection of the parabiosis studies, I do not remember hearing anything about the soluble factors causing a one time permanent reset of the aging clock , doesn’t the youthful state of the old animal require persistent connection to a young animal to prevent the old animal from reverting to an old phenotype, quickly?

        If you are talking about the epigenetic state of the neuroendocrine system as constituting the clock, that seems feasible. Kind of a unification of the two mechanisms?

        • Hi Josh,
          Another element that I think is in support of the neuroendocrine clock of aging model can be found in
          semelparous animals such as the pacific salmon and the a pacific octopus the aging mechanisms that dominate as death mechanisms strongly appears to be neuroendocrine. Removal of the optic gland in the octopus delays aging in the species. Considering the genetic distance between these species, it appears that a neuroendocrine model has been conserved in evolution.

  3. Epigenetics is a tool that a species can use to adapt organisms to its environment quickly, via say, through an epigenetic response to the stress of starvation, or in response to increased or decreased predation rates that up or down regulates the rate of maturation and as a result the associated rate of aging. This is much faster process than using randomly derived mutations in the gene pool that then must be selected over generational time to change attributes in the population as a whole.

    How much of the aging clock is genetic and how much is epigenetic? It would be interesting to compare the epigenetic constitution of the possums that Steven Austad studied on Sapelo Island to their 50 percent shorter lived relation on the mainland. Do you happen to know if Steven’s lab has already done this work or not?

    Do you

    • Let’s not conflate “epigenetics” with “epigenetic inheritance”. Epigenetics is the science of gene expression. Epigenetic inheritance is the transmission of traits via extra-genetic means.

      • Hi Josh,

        The “Do you” at the bottom of my post was not intended, it was an artifact of how I was starting my post until I changed my mind. It was past the bottom of my editing window and I did not see it was there before I submitted my post. Once posted I did not know how to remove it.

        What I was getting at was, just as DNA base pair sequences can be inherited to define the attributes of future generations, methylation or acetylation can occur or be removed which can then persist through the germ line and effect future generations.

  4. Hello Josh…

    What do you mean I “hinted at” the idea that DNA methylation controlled aging….it was plainly stated as if it was a fact throughout almost half the paper which was 40 some odd pages! (The Evolution of Aging-A new approach to an Old Problem of Biology- Med Hyp Sep 9 1998)

    Anyway- bats also experience menopause

  5. Hi Josh,

    I believe DNA methylation is downstream of histone methylation. I think this because there are only 3 or 4 DNA methylation enzymes known and they are all pretty blunt. On the other hand there are quite a few histone modification enzymes. You can perform much more sophisticated logic with more enzymes and more targets.
    Unfortunately histone modification seems to be white spot on the map. I dont know if there exists a technique through which the histone modification map of the genomes of different cells could be thoroughly investigated. I think until we get such tool we are in the dark. RNA expression profiles maybe most useful until then.

  6. Josh, I agree with your theory of programmed aging in broad strokes, and your theory of aging’s origin in the evolutionary pressure applied to groups is compelling.

    The problem I find is that, if a single individual turns off programmed aging (partly or wholly) that individual would be expected to obtain great and enduring reproductive success. The fact that this *eventually* leads to a population crash that destroys the group is too late / too slow a mechanism to stop individuals from “cheating” on the aging program to obtain a little extra individual success in the short term.

    If you agree this problem is real, the question is: what mechanism prevents cheating? Are we looking for a system of cooperation and verification similar to evolved altruism in some animals? Some signalling connected to sexual reproduction, whereby those that don’t age can’t reproduce (could be a social or even biochemical mechanism)?

    I think that if we really understood this aspect of how aging evolved, it could help guide the research into the epigenetic aging clock. We would have some clues about whether to look for possible upstream drivers of egigenetic aging.

      • Hello Jeff, so I went ahead and read your book. There was some interesting stuff in there, but I didn’t see an answer to the question I posed.

        To restate: let’s just stipulate that programmed aging has evolved somehow. What keeps it stable against individuals that turn off the aging program for a short-term advantage?

        Josh’s theory seems to be that turning off aging leads to ecosystem destabilization and population crashes. However, if that’s the only mechanism in play, it’s too slow to prevent a “cheater” from gaining reproductive success in the short term. We should see a recurring, in fact near-constant pattern of populations of organisms becoming unstable as age-defying “cheaters” come to dominate the group and cause wild population swings.

        Likewise with your theory of predation driving diversity. We should see a recurring pattern of species failure as longer-lived “cheater” individuals dominate a gene pool, diversity is lost, and the species fails to adapt. While we do see species fail all the time, very possibly from lack of diversity, I don’t think there’s evidence of that being preceded by lengthening lifespans (however if there was, that would be a huge win for your theory!).

        • This is a difficult question, a worthy challenge. I have been thinking about it for many years and here’s the best I’ve come up with:

          Runaway reproduction without aging and genetically programmed birth control lead to population extinctions. Many, many times cheaters arise, their progeny come to dominate the population, and eventually bring it down through demographic instability. After a great number of cycles, nature “learns” to keep these cheater genes tucked away where they are not so accessible to mutation, and the cheaters arise less often. My take on Antagonistic Pleiotropy is that it is an evolved adaptation for the purpose of keeping preventing mutations from re-exploring dead ends.

          Still, cheaters still arise today from time to time. The Rocky Mountain Locust plague of the late 19th century may have been an example. Several times, they made a dust bowl of the American West before driving themselves to extinction.

          – Josh

          • Josh, when I read your comment, I had just finished writing *precisely* this same defense of your theory in my notes. And it looks like Jeff Bowles has said substantially the same thing: basically the multiple “clocks” of programmed aging have evolved so that they are difficult to defeat by random mutation.

            Here’s what I don’t like about this idea: we keep finding startlingly simple interventions that extend healthspan and/or lifespan. I’m thinking telomerase expression and culling senescent cells in mice, and the CR / sirtuins / mTOR pathway in several mammals.

            These all seem relatively easy to achieve by mutation, which brings me back to expecting the “cheater” phenomenon to be extremely widespread and obvious, even if it’s doomed to fail within several generations.

            Maybe these healthspan enhancements we’ve discovered have subtle drawbacks, such that in the wild we have a situation of antagonist pleiotropy that we don’t see in a controlled lab. But then, your own take on CR is that it makes creates more robust so they survive famine!

            I don’t mean to bash either (excellent) theory, but my intuition is that there’s some more immediate selection pressure. Something to do with the greater value of sexually reproduced progeny as compared to a long-lived self (which is tantamount to an asexual clone). Still noodling on this.

          • Don’t apologize, Charles – you are right to dwell on this point because it is the weakest link in our understanding.

            Telomerase is locked up very tight–we don’t understand how, but apparently, it takes the special machinery of meiosis to unravel the part of the genome where telomerase lives.

            Culling senescent cells is a more interesting example – why doesn’t the body do it? All I can say is that this is part of a general deterioration in the regulation of apoptosis that has broad consequences for cancer and for sarcopenia as well.

            – Josh

          • Good point about telomerase and its complex regulatory network. But then, I would make the analogy to sophisticated software copy protection schemes, that always end up circumvented by some brilliant Latvian teenager who makes a 2-byte modification to the compiled binary 🙂

            Also, aside from point mutations, isn’t it possible that a splice error could introduce a second copy of the telomerase gene into a different regulatory context?

            And there are ashwagandha, silymarin, and maybe cyclo-astragenol, all naturally occurring molecules that seem to upregulate telomerase. And there’s whatever BioViva did to their CEO in Columbia, probably a modified CRISPR technique applied to extracted hematopoietic stem cells, and still well within the range of what could be envisioned as a naturally occurring adaptation.

            Finally, just a couple of weeks ago, nicotinamide riboside joined the exclusive club of compounds showing life extension in mammals:

            To quote Dune, “we’re finding these sabotage devices too easily” 🙂

            Re: senescence and cancer: when we get through the “low hanging fruit” of extending telomeres, reversing age-related methylation, killing senescent cells and re-training the immune system to attack cancers (manually as now or via thymus reinvigoration), I wonder what obstacle we’ll see next? Cancers arising at increasing frequency until there’s just no functioning cells left?

            My suspicion is that programmed aging is real, but wear-and-tear aging is also real, just at a much longer timeframe. Once we turn off the self-sabotage of programmed aging, we will still have the problem of long-term “information coherence” in the organism, which is analogous to a computation being performed on an unreliable substrate – the lower the error rate you need, the more redundancy and associated energy costs you have.

            Of course there’s a next level of possible interventions for this slowly accumulating damage – but the point is, it’s much easier to explain programmed aging if evolution is choosing the optimal lifespan between, say, 13 and 300, rather than between 13 and infinity.

          • I like Charles’s thinking and approach-Charles why don’t you join the party and work on a comprehensive paper on aging!

          • Thank you Jeff 🙂 I will definitely throw my hat into the ring as soon as I have a hat to throw! I certainly don’t intend to be “that guy” who critiques other’s work and puts forward nothing of his own.

          • Hi Charles,

            your points are really thought provoking, but – non scientific – opinion is that these you mention are not really low hanging fruits.
            Actually I dont know about any research that shows that whether telomerase replacement has any life extension effects. It may have I just havent yet read about it.
            mTor/CR has obvious drawbacks in the wild – anyone on CR will be just predated by microorganisms or carnivores, because slower muscles and dumped immune system.
            Reversing DNA methylation – to my knowlede no one has actually done this and yes we have only 4 DNA methyltransferases, but I suspect the regulation of DNA methylation is in the much more diverse histone modification . At least manipulation of DNA methyltransferases is lethal/ increases cancer incidence so there must be a much finer regulation layer above it.

            Anyways, the only known pathway mTOR/CR, etc works only to the extent of 30-40% life extension in mice but we know that there are rodents that live 1000% more than mice (naked mole rat ).

            My proposal is to take the naked mole rat genome and try to replace the mouse genome with it iteratively. First try chromosome by chromosome or if it produces unviable offsprings then replace by chromosome crossover regions.
            At some point we should be able to get viable offsprings with different mouse/naked mole rat fraction in their genome and thus see if there are specific regions which extend lifespan.

          • Telomerase gene therapy extends lifespan in mice by 24% (healthspan too):

            CRISPR/Cas9 can be used to edit methylation sites:
            There are other techniques too:

            CR in the wild is obviously a bad idea, but resveratrol, pterostilbene and other compounds seem to activate this same pathway. There’s been no successful life extension in mammals through these compounds, even though they seem to make positive metabolic changes (even in humans):
            It may be that we never develop a CR mimetic that extends lifespan for mammals, but my intuition is that we will.

            To clarify “low-hanging fruit”, I mean techniques that are available now through about 10 years, and I don’t mean FDA approved – just feasible.

            The telomerase technique that worked with mice has already be applied to a human:

            I like your naked mole rat idea, but Josh’s proposed experiment, if it worked, would be immediately applicable to already-living humans.

          • Hello there

            the important experiments with mice and their telomeres. I believe, were done by Depinho where he knocked out their telomerase gene. And watched what happened. Telomere shortening was constant but had no effect on the mice in the first threee generations, but finaly in the 4th, the mice started losing hair and gettgin hair graying and became sterile I think.
            mice have telomeres that are 4X ;longer than humans and thus I see them as a deactivated aging system in mice that will only kick in once they evolve much longer lifespans..maybe at year 10 or so??

            The experiment where elongation of the telomeres increased mouse lifespan I think was in some sort of special type of mouse that had been bred to show diseases at a faster pace or something like that

          • Hi Charles,

            thanks a lot for taking your time to reply. I think I have read all these papers before, somehow my mind dismissed those findings with telomere expression in mice. I just found 1o-2o % increase too low. But the effect of telomere elongation on lifespan is there, true.
            About methylation via CRISPR, I knew this was possible however with aging we are facing hundreds or thousands of differentially methylated sites the majority is – thought to be enthropy driven – hypomethylation – thats not a low hanging fruit in my understanding for both random cheater mutations and therapeutic interventions. Also for the DNA methylation to take any effect on ageing and not just be some random consequence of an upstream process, I think DNA methylation should have some influence on transcription profiles. Steve Horvath first tried to go for age related changes in expression profiles but he found none – this is very strange isnt it?
            Moreover I think DNA methylation is too blunt to be the primary driver in aging. There are only 3-4 enzymes and I think at least KO or knock down experiments have been conducted for all of them with bad results. I believe there is something upstream DNA methylation – I suspect its histone modification. There are much much more histone modifier enzymes then DNA methyltransferases, so a more sophisticated expression program could be constructed with them.

          • The telomerase result I cited was with normal mice. There were some earlier studies involving progeroid or otherwise special mice; the study I linked cites those earlier studies as reasons they studied normal mice.

            Given that methylation is the difference between pluripotent cells and specialized tissues, they cannot possibly have the same transcription profile. I think whatever you were reading might have been a rather narrow finding – citing it would help clear this up.

            Likewise I’d be interest in those links on the effect of knockouts for methylation enzymes. Again there, if all means of methylation were actually disabled, one would expect a ball of undifferentiated cells that never develops into a functioning organism, so that finding might have been rather narrow as well.

          • Here is a description of Depinho’s first study of telomerase knockout “normal” mice….It sounds like they do pretty well without it for quite some time…….

            BOSTON — Mice lacking a gene for making telomeres — chromosomal elements with a conjectured but controversial role in aging and cancer — were found to go gray, lose hair faster, and recover less easily from the stress of surgery and chemotherapy than normal animals. They also developed tumors more often and died earlier, a team of Dana-Farber Cancer Institute (DFCI) researchers report in the March 5 Cell.

            “Until now no one has shown the phenotypic effect of telomere loss on the aging organism,” says Lenhard Rudolph, research associate in medicine at DFCI and Harvard Medical School (HMS), and lead author of the study. Telomeres — nubs of protein and nucleic acids poised at either end of DNA strands — are produced during embryonic development. They crumble away as cells mature and divide, suggesting telomere erosion might somehow control aging. But the process had been studied only in cultured cells.

            By knocking out one of the genes for a telomere-making complex, telomerase, Ron DePinho, the American Cancer Society professor at DFCI, and his colleagues created successive generations of mice, each with shorter telomeres. Surprisingly, symptoms were not observed in the telomere-deprived newborns. For example, third generation mice displayed symptoms only after they had reached the age of 18 months. Even the most telomere-deficient sixth generation mice had to go through a waiting period, albeit shorter, before symptoms appeared.

            In this and other respects, the mouse findings confirm but, intriguingly, confound expectations about the role of telomeres in aging and cancer. While telomere shortening is clearly important, it is not sufficient to bring about the depredations of age. “Something else in the aging organism cooperates with telomere dysfunction to compromise fitness,” says DePinho, who is also a professor of medicine (genetics) at HMS. What else that is remains unclear. “That’s one of the central questions in biology? What is the basis for organismal aging? Is it free radical damage that accumulates? Is it general mitochondrial dysfunction? Or are there other mechanisms?”

            The experiments also demonstrate the widely held belief that telomeres are associated with cancer — but not in the way researchers were expecting. Previously, tumor cells had been observed to display an excess of telomerase activity suggesting telomerase might be involved in the cell becoming immortal. But the new findings indicate a lack of telomerase can also lead to cancer. “These experiments have been to a certain extent mind-bending. No one would have anticipated that you could have gotten an increase in cancer from a lack of telomerase,” DePinho says.

            However, there is a possible explanation. Normally, telomeres keep chromosomal tips from unraveling or sticking to other chromosomes. Deprived of these protective shields, chromosomes might fuse, break, or undergo other changes that could lead to the loss or gain of genes and, in turn, to cancer. “So the studies teach us that if we lose telomerase function in normal cells, that might also set the stage for the initiation of cancer in aged individuals,” DePinho says.

            One of the goals fueling telomerase research is the hope that it might lead to new cancer treatments. Anti-telomerase compounds are being developed that could prevent aspiring cancer cells from making new telomeres and thus dividing endlessly. Rather than dashing such hopes, DePinho believes the real lesson of the new experiments is that cancer cells can turn all kinds of situations to their advantage. Some cancer cells may require telomerase while others thrive in its absence.

            “We have learned that there are many roads to cancer. But if we throw up good roadblocks we can diminish the rate of cancer formation. So it’s a numbers game,” he says. “In a situation where many aspiring cells are moving towards cancer, by knocking out telomerase we may have a reasonable chance of decreasing the ratio of tumor formation,” he says.

            Earlier research had shown DePinho that the telomere story was not a simple one. For example, soon after cancer cells were shown to express telomerase, DePinho and his colleagues found that mice lacking the complete telomere-making complex could develop cancers. But how often they did so in comparison to normal mice was not known. Nor had anyone looked systematically at the effects of telomere loss on aging animals.

            Previous observations had shown that while sixth generation knockout mice show some slight abnormalities even at birth, third generation newborns appear perfectly normal. Tracking the third generation mice, Rudolph found that 18-month-olds display an unusually high incidence of skin ulcerations in addition to graying hair and hair loss — signs that organs with high proliferative rates were slowing down or undergoing cell death.

            More significantly, when exposed to a blood-cell depleting agent, normal adults showed signs of recovering their blood cell supply after about a week whereas most of the older knockouts did not repopulate. In fact, many died. The knockouts also took longer to heal from minor wounds. Sixth generation adults ran into the same trouble, but sooner. Interestingly, young animals — normal and knockout — did not appear to be affected by the interventions.

            Knowing more about such stress responses could help explain why surgery and chemotherapy take a greater toll on older people, the researchers say. Similarly, the finding that many telomere-deficient adult mice developed tumors could yield insight into why people are more likely to develop cancers as they age.

            But just as telomere loss may mean different things to aspiring cancer cells, it may yield several different outcomes in other kinds of diseased cells. “The consequences of the loss of telomere function, and especially whether or not it facilitates or inhibits cancer, may be context dependent,” DePinho says. “The bottomline is that it is really complex. Telomeres are definitely important — they impact on life span, organismal fitness, and cancer incidence. But what this story tells us is that there’s tremendous complexity.”

            This work was funded by the National Institutes of Health.

          • Just to avoid confusion: Jeff is citing a different study. The study I cited shows lifespan and healthspan extension from telomerase gene therapy in C57BL6 mice, the most common strain of lab mouse:


            These are not telomerase KO mice or progeroid mice. Just normal mice.

          • Hi Jeff,

            I think the classic mTert KO experiments does not disprove the beneficial effects of telomere elongation. Because the mouse germline might not attrite telomeres very much compared to the rapidly replacing mouse soma.
            Thinking about cheaters, we have a natural experiment at hand: domesticated dogs and cats.
            At least in the case of dogs, despite having 10.000 years of divergent evolution the more stable environment provided by humans did have little effect on their lifespan. The same is true for post paleolithic humans, for whom the agriculture meant drastically better survival chances into old age, yet we have not witnessed much expansion of lifespan in the past 10 thousand years for settled humans despite the growing average lifespan.
            Somehow cheating in the wild is not that easy.

          • Hi there

            i have always wondered why cats have such long lifespans for their little body size while dogs don’t

            The answer i came up with is that dogs are there to protect us and thus are at high risk of being killed in predator encounters (defending us)

            while cats just hide around our houses all day and hide when there is danger… so even being smaller than dogs they have evolved life spans longer than dogs

          • Hi Jeff,

            my guess is that neither cats nor dogs are really endangered by predators they are both pretty close to being top predators. However cats avoid conflicts more than dogs and they can climb trees which is an additional layer of safety.

            Thinking about the epigentic clock further. My take on DNA methylation changes are that there is a generic, random demethylation of the genome and there is a targeted methylation of specific sites in the genome (mostly bivalent chromatin).
            What if the driver for the hypomethylation process is indeed wear and tear?
            If an organelle is damaged, enzymes can repair/reconstruct it. If an enzyme is damaged, the ribosomes will reconstruct it. If the ribosomes are damaged they are reconstructed from RNA. RNA is continuously replenished from DNA template. Single stranded DNA damage can be repaired using the other strand as template. Double strand breaks can be repaired by homologous end joining. But how can DNA methylation damage be repaired? There is no template! DNA methylation carries information from the past about the differentiation and potency fate of the cell.
            My guess is that evolution did not need a method to fix DNA methylation damage. DNA metthylation patterns are reestabilished at fertilization and early embrional development (together with telomere length) and there is no reason fixing it in the soma as DNA methylation apparently provides a rather dynamic ageing envelope from weeks to hundred years that fits most niches.
            And indeed we know there are de novo methyltransferases DNMT3a and b that do the resetting of the epigenome in embryonic stage. And we have DNMT1 which methylates the newly synthesized DNA strand in the daughter cells using the old DNA strand as template in somatic cell divisions. Is there an enzyme for DNA methylation damage repair?

          • Hi there

            If you ever get a chance to read the last few chapters of my book “The Unselfish Genome” you will see we are in some agreement that some DNA methylation (and remethylation) appears to be haphazard.My take on it is that the hormonal milieu that the DNA finds itself in leads to general remethylation / stable methylation or demethylation. Antioxidants tend to catalyze remethylation reactions. I also propose that the hormones do not have to bind specific receptors but can also co-localize with stretches of DNA here and there. In the last chapter of my book I look at the various structures of hormones and supplements /herbs that are good for you and note that they have at least one of 3 things in common. They are shaped like a joined base pair of DNA (see steroid hormones vs GC or AT) they have methyl’s attached that can be donated and/or they have =o or -oh’s attached which can serve as antioxidants. So in addition to binding actual hormone receptors I believe this bath of hormones/methyl/donors/antioxidants determines the overall methylation or demethylation propensity of the DNA’s environment.\

            Also there are sme bad things that increase with age, like homocysteine which is a methyl grabber!

        • hi Charles

          Glad you looked over the book.

          I expect there are cheaters arising all over the place all the time ( but evolution has put the brakes on how fast they can “cheat”….If tbhey get to cheat enpough, they then destroy the diversity of the gene pool with respect to allowing a defense to evolving predation to emerge fast enough to avoid everyone gettgin eaten. So usually just as fast as the cheaters arise, they go extinct. Unless they come up with a defense to predation that predators find hard to crack….Then they can become “cheaters” and start evolving longer lifespans.

          I guess your question is why can’t cheaters pop up all of the sudden and live forever? I think evolution has been discouraging that since the beginning of life, and in my first paper on aging published in Sep 98 Med Hyp Jt Bowles I make the case that there are multiple aging systems (at least 6) & likely many more ticking away in your body. Each with a certain fuse length. The shortest fuse has to be overcome first, then the next longest, etc…to extend lifespan. I expect evolution has put up many roadblocks for a cheater to arise in the short term….but in the longer term, cheaters can arise, and this leads to longer avg life span for the species in question.

          So how long would an individual live before it would be considered a cheater. While there are quite a few very old trees and bushes sometimes from 5000 to even 12,000 years old, how do you know these are not the surviving “cheaters”? How do you know the weren’t a lot of the same species that died off at age 200 or 300 or so and now their genes have been replaced??

    • Hi Charles,

      I agree with you that stopping the cheats is a problem with group selection that has been long expressed by Richard Dawkins in his many books. However altruistic traits and other traits that are hard to explain via individual selection are common. As you hinted I think the answer lies in sexual selection.

      I have developed a simulation program to examine if sexual selection can up regulate and down regulate rates of senescence when factors like predation rate vary over time. I have found that they do. Additionally the same simulation demonstrates that when no sexual selection occurs, as is common in corals, sponges, clams, etc, the rate of senescence tends towards zero over multiple generations which is consistent with the negligible senescence of many of these species.

      If you are interested, I can provide the program in its compiled state that will run on a Mac and I can provide you source if you can read “C” code.

      • Hi Kevin, thanks for your thoughts. I am proficient with C and have Macs on hand.

        Before I dive into code though, could you articulate the theory you’re testing further? Are you taking as a premise that signs of aging cause positive sexual selection?

        Feel free to drop to email for a deeper explanation. You can email me at isomorphic.com, account name is just “charles”.

        • @ Charles

          Every ‘cheating’ mechanism you propose, and countless others, have already been devised and dealt with way before the Cambrian explosion, the only part of an endgame of a brutal arms race which is accessible to us. (tho I bet nobody really looked that hard on the fossil record for population crashes, or perhaps some of the earlier “great extinctions” are precisely due to aging still evolving, huh?)

          And “social cooperation”? Wouldn’t that contradict your own doubts? How about the converse? Look around you, how often do you see the ugly ducklings ganging up against the pretty swans? Even in “higher” animals to this day 😉

          • There are compelling explanations for how seeming altruism and other forms of cooperation have evolved, which do not rely on group- or species-level selection.

            I’m not really sure what you mean about cheating having been “dealt with”. If programmed aging evolved principally through group selection, then at any given time there is the possibility of one organism turning off aging for individual lineage benefit (“cheaters”). Cheating of this kind might have caused population crashes in the Cambrian, but that doesn’t remove the need for a mechanism that is preventing cheating *now*.

          • And every single time that happened it was “dealt with”. (get it? : ) simple darwinian selection)
            And every single time *that* happened it reduced exponentially the probability of “the possibility of one organism turning off aging”.

            Eventually, it won’t happen anymore (“naturally” ; ) thus your examples) in the lifetime of the universe, as much as getting the paste back in the tube would.
            And you don’t even need to qualify it with “principally through group selection” because evolvability is the name of the game, it will beat any other evolutionary pressure down to a “primordial soup” any day.

            Now, don’t get me wrong, I do not deny group selection. It’s just that, every time you add a new copying/transmission/reproduction checkpoint, you blunt the power of evolution by raising the costs (and thus demands on ROI) of any adaptable mutation. (have you considered this in your C-mulation?)
            That’s where deniers get it right. On the other hand it doesn’t make it impossible, that’s where they get it wrong. (the investment analogy I just did should make it clear, even for them, tho I’m afraid their jobs depend on it 😉 mwahaha) Otherwise chemistry itself would be impossible (I guess some would deny even that), RNA had to give rise to DNA, then aminos, when we added a huge checkpoint (chemical activation energy in this analogy) with the ribosome, then another big one with unicellular cloning (which is a LOT of work).

            But here’s the rubber, nothing is constant in this universe, once your Ecosystem — our keyword here — gets large enough, the network effect kicks in and your “ROI” goes through the roof. That’s where you should see cheaters arising and, like a mole, invariably being whacked.

            The really interesting question from all this would be how aging’s evolution proceeded when multicellular and, more relevantly, higher/CNS animals arrived. Did it just usurped/co-opted the unicellular mechanisms or was it the other way around? Was it a gradual race to the bottom, as evolvability would suggest, from the starting point of only unicellular aging being enough to grant death by predation to the chronologically old? Or did it happen in fits and starts, as irony would have it, being this the opposite of “cheating” in our context? Your thoughts…

          • Sorry, no, I don’t get it. If programmed aging has evolved, and one mutated lineage turns part of the aging program off, enjoys longer healthspan and lifespan, but then ultimately dies off, this in itself does not reduce the likelihood that the same mutation is going to occur again (let alone reduce the probability exponentially..).

            If this happens many, many, many times over the epochs, then Josh’s theory is that the programmed aging system might evolve to be more “tamper proof”. But we’ve already found multiple, relatively simple ways to extend lifespan and healthspan in mammals, and this seems to argue against any significant tamper proofing.

          • The ultimate way to “cheat” would be to eliminate the need for sex which has already happened in a number of very complicated animals like turkeys, various snakes and lizards. Yet there are very few examples of them throughout the world. They say it would take about 100 various mutations to restore this ability in humans. That seems like a pretty good example of a roadblock. I discuss this also in my book-
            “The Unselfish Genome”

          • > “Sorry, no, I don’t get it.”

            I think I’ve identified why this may be. What are your assumptions regarding the definition of “crashing”?

            It certainly doesn’t, nor can it, mean an ‘atomic’ / point-wise / immediate crash, as soon as said mutation is expressed.

            It can’t because, first, said “ROI” must be made manifest. You mentioned it must have an advantage (“why doesn’t it take over”, IIRC?) well, how else would said mutants manage to crash the entire ecosystem as their progenitors could not?

            This entails time, and many generations, enough to overshoot and then some (a bit more time after the point of no return), this is also exponential which in turn entails there are many early generations with very little impact (~linear) on the ecosystem. This leaves A LOT of room for evolution to keep working on side-effects and meta-mutants to survive the eventual crash. The very overshoot time will provide enough cortisol (or otherwise stress) to make the genome go “all in” on mutations in that whole, now gigantic/diverse, population.

            Much like we’ve been doing with antibiotics, there is bound to have a few survivors, specially in such a relatively much bigger/fast-paced/adaptable world as viewed from unicellular life’s perspective (when most of these “patches” were ‘committed’).

            > “multiple, relatively simple ways ”

            Like I said.. “naturally”. You mentioned SIRTuins for instance, there seems to have been a lot of controversy even calling that into question. But, how much of that would be required to affect the genome? I don’t think mice (or some other, vegetarian, mammal) would be able to munch that many panda-worth volume of grapes. (hey, maybe we should see if pandas have some of those molecules in their diets and check if it makes a difference?)

            We’d have to go through each one, and see what exactly an animal in the wild would have to do to manage to emulate said interventions. It may be simple for human ingenuity, maybe not so simple to evolved instincts to motivate a simple brain enough for it? Specially since most of the mechanisms are at celullar level, like telomerase, hidden in the chromatin, who knows.. Do you have an example you already went through this exercise, that we should discuss in more detail?

            > “this seems to argue against any significant tamper proofing.”

            These things are intertwined redundant systems. The most efficient way to grant no “cheater” will cross both bridges at the same time. We’d have to knock all Holy Grail Guardians simultaneously, to see “radical” results. But yes, it shouldn’t be *that* difficult for the creative and humble species.


            Choose wisely. : )

          • The background of this thread is Josh Mitteldorf’s concept of programmed aging evolving through group selection, where groups that don’t age fail through population swings that become so violent, they can wipe out the whole group:


            I’m not talking about organisms in the wild seeking out tons of grapes (LOL!) just as I’m not suggesting that wild mice might learn to seek out telomerase gene therapy!

            The point is that one can easily imagine a mutation or series of mutations that simply turns on telomerase, or activates the CR pathway without actual calorie restriction, or kills senescent cells a bit more aggressively. Each of these interventions, performed in a lab setting, has been shown to provide significant lifespan/healthspan benefits that would presumably help an organism outcompete peers.

            If the programmed aging system were “tamper-proofed”, I would expect that in order to cause any significant lifespan/healthspan benefits, we would have to manipulate, say, 6 genes at once, and that affecting any one gene on its own would kill the organism. That would plausibly prevent mutations in the wild from turning off aging, since in this example, 6 mutations would have to all occur in a single organism at the same time, and if only 5 occurred, it would be fatal, so there would be no way to accrue the necessary mutations across multiple generations.

            Instead we keep finding that the programmed aging system is relatively simple to manipulate, and, at least in my opinion, we can readily imagine mutations occurring in the wild that are equivalent to the interventions we’ve performed in the lab.

          • Descartes could imagine his disembodied awareness and he thought this proved Dualism. Here’s a guy with (estimated) IQ of almost 200 ^__^

            btw, what’s your definition of “significant”? Again, we’d have to take each of the interventions you suggest and imagine what the animal/population would have to do to *emulate* it by *natural* means.

            Remember, higher animals have ‘higher’ forms of clocks which have co-opted all those mechanisms into a web of interlocked teeter-totters. Haven’t we’ve seen “significant” life extension in *unicellular* organisms? or HeLa cells? Even multicellular but still ‘primitive’ (with an exact body design down to the number of cells): C. Elegans? Maybe here’s the hint… Yea, “tamper-proof” doesn’t mean impossible. Just enough to have kept “cheaters” in check up to this point. There is a mouse which the males go extinct every reproducing season of the year, yet you can keep them alive for years in the lab. Not every mechanism is hard-coded. What idiosyncrasies of their ecosystems are keeping your “cheaters” from rising again?

          • You keep talking about what animals would “do” in the wild.. again, we’re talking about mutants where the programmed aging system is partially disabled due to genetic differences. They don’t do anything, they are.

            For the different healthspan/lifespan effects, read the cited papers – some of the larger effects are 20%+.

          • Sorry for the anthropomorphizing. What I mean is by which pathways would “cheaters” have to mutate to emulate what humans can do to them in the lab?

            I’m not sure 20% is all that significant, it makes it dangerously close to whatever the investment probably is (we’d have to answer the previous question to get a handle on that tho)

      • just some interesting abstracts re telomeres/methylation

        Int J Epidemiol. 2016 Apr 13. pii: dyw041. [Epub ahead of print]
        The epigenetic clock and telomere length are independently associated with chronological age and mortality.
        Marioni RE1, Harris SE2, Shah S3, McRae AF3, von Zglinicki T4, Martin-Ruiz C5, Wray NR6, Visscher PM7, Deary IJ8.
        Author information
        Telomere length and DNA methylation have been proposed as biological clock measures that track chronological age. Whether they change in tandem, or contribute independently to the prediction of chronological age, is not known.
        We address these points using data from two Scottish cohorts: the Lothian Birth Cohorts of 1921 (LBC1921) and 1936 (LBC1936). Telomere length and epigenetic clock estimates from DNA methylation were measured in 920 LBC1936 participants (ages 70, 73 and 76 years) and in 414 LBC1921 participants (ages 79, 87 and 90 years).
        The epigenetic clock changed over time at roughly the same rate as chronological age in both cohorts. Telomere length decreased at 48-67 base pairs per year on average. Weak, non-significant correlations were found between epigenetic clock estimates and telomere length. Telomere length explained 6.6% of the variance in age in LBC1921, the epigenetic clock explained 10.0%, and combined they explained 17.3% (allP< 1 × 10-7). Corresponding figures for the LBC1936 cohort were 14.3%, 11.7% and 19.5% (allP< 1 × 10-12). In a combined cohorts analysis, the respective estimates were 2.8%, 28.5% and 29.5%. Also in a combined cohorts analysis, a one standard deviation increase in baseline epigenetic age was linked to a 22% increased mortality risk (P= 2.6 × 10-4) whereas, in the same model, a one standard deviation increase in baseline telomere length was independently linked to an 11% decreased mortality risk (P= 0.06).
        These results suggest that telomere length and epigenetic clock estimates are independent predictors of chronological age and mortality risk.

        Clin Epigenetics. 2016 Feb 26;8:21. doi: 10.1186/s13148-016-0186-5. eCollection 2016.
        Frailty is associated with the epigenetic clock but not with telomere length in a German cohort.
        Breitling LP1, Saum KU1, Perna L1, Schöttker B2, Holleczek B3, Brenner H2.
        Author information
        The epigenetic clock, in particular epigenetic pre-aging quantified by the so-called DNA methylation age acceleration, has recently been suggested to closely correlate with a variety of disease phenotypes. There remains a dearth of data, however, on its association with telomere length and frailty, which can be considered major correlates of age on the genomic and clinical level, respectively.
        In this cross-sectional observational study on altogether 1820 subjects from two subsets (n = 969 and n = 851; mean ± standard deviation age 62.1 ± 6.5 and 63.0 ± 6.7 years, respectively) of the ESTHER cohort study of the elderly general population in Germany, DNA methylation age was calculated based on a 353 loci predictor previously developed in a large meta-study, and the difference-based epigenetic age acceleration was calculated as predicted methylation age minus chronological age. No correlation of epigenetic age acceleration with telomere length was found in our study (p = 0.63). However, there was an association of DNA methylation age acceleration with a comprehensive frailty measure, such that the accumulated deficits significantly increased with increasing age acceleration. Quantitatively, about half an additional deficit was added per 6 years of methylation age acceleration (p = 0.0004). This association was independent from age, sex, and estimated leukocyte distribution, as well as from a variety of other confounding variables considered.
        The results of the present study suggest that epigenetic age acceleration is correlated with clinically relevant aging-related phenotypes through pathways unrelated to cellular senescence as assessed by telomere length. Innovative approaches like Mendelian randomization will be needed to elucidate whether epigenetic age acceleration indeed plays a causal role for the development of clinical phenotypes.
        Clin Epigenetics. 2015 Oct 7;7:107. doi: 10.1186/s13148-015-0140-y. eCollection 2015.
        Selective increase in subtelomeric DNA methylation: an epigenetic biomarker for malignant glioma.
        Choudhury SR1, Cui Y1, Milton JR2, Li J3, Irudayaraj J1.
        Author information
        Subtelomeric regions dynamically change their epigenetic pattern during development and progression of several malignancies and degenerative disorders. However, DNA methylation of human subtelomeres and their correlation to telomere length (TL) remain undetermined in glioma.
        Herein, we report on the selective changes in subtelomeric DNA methylation at the end of five chromosomes (Chr.) (7q, 8q. 18p, 21q, and XpYp) and ascertain their correlation with TL in patients with glioma. Subtelomeric methylation level was invariably higher in glioma patients compared to the control group, irrespective of their age and tumor grade. In particular, a significant increase in methylation was observed at the subtelomeric CpG sites of Chr. 8q, 21q, and XpYp in tissues, obtained from the brain tumor of glioma patients. In contrast, no significant change in methylation was observed at the subtelomere of Chr. 7q and 18p. Selective changes in the subtelomeric methylation level, however, did not show any significant correlation to the global TL. This observed phenomenon was validated in vitro by inducing demethylation in a glioblastoma cell line (SF-767) using 5-azacytidine (AZA) treatment. AZA treatment caused significant changes in the subtelomeric methylation pattern but did not alter the TL, which supports our hypothesis.
        DNA methylation level dramatically increased at the subtelomere of Chr.8q, 21q, and XpYp in malignant glioma, which could be used as an early epigenetic diagnostic biomarker of the disease. Alterations in subtelomeric methylation, however, have no effects on the TL.

        J Cell Physiol. 2015 Oct;230(10):2337-44. doi: 10.1002/jcp.24980.
        Roles for Histone Acetylation in Regulation of Telomere Elongation and Two-cell State in Mouse ES Cells.
        Dan J1, Yang J1, Liu Y2, Xiao A2, Liu L1.
        Author information
        Mammalian telomeres and subtelomeres are marked by heterochromatic epigenetic modifications, including repressive DNA methylation and histone methylation (e.g., H3K9me3 and H4K20me3). Loss of these epigenetic marks results in increased rates of telomere recombination and elongation. Other than these repressive epigenetic marks, telomeric and subtelomeric H3 and H4 are underacetylated. Yet, whether histone acetylation also regulates telomere length has not been directly addressed. We thought to test the effects of histone acetylation levels on telomere length using histone deacetylase (HDAC) inhibitor (sodium butyrate, NaB) that mediates histone hyperacetylation and histone acetyltransferase (HAT) inhibitor (C646) that mediates histone hypoacetylation. We show that histone hyperacetylation dramatically elongates telomeres in wild-type ES cells, and only slightly elongates telomeres in Terc(-/-) ES cells, suggesting that Terc is involved in histone acetylation-induced telomere elongation. In contrast, histone hypoacetylation shortens telomeres in both wild-type and Terc(-/-) ES cells. Additionally, histone hyperacetylation activates 2-cell (2C) specific genes including Zscan4, which is involved in telomere recombination and elongation, whereas histone hypoacetylation represses Zscan4 and 2C genes. These data suggest that histone acetylation levels affect the heterochromatic state at telomeres and subtelomeres, and regulate gene expression at subtelomeres, linking histone acetylation to telomere length maintenance.

  7. Hi Josh.
    I’m curious as to what evolutionary mechanism you see driving the epigenetics of aging? Also, through-out most of human history our ancestors had a much shorter life span than modern man. What opportunity would there be for a mechanism that would have had to evolve over millions of years to occur, if human life span has been brutally short?

  8. Nice post. As far as I know, there is no direct evidence suggesting a possible impact of telomere lenght on methylation levels in cells, so I would like to know more about this subject. Is there any hint in current literature? Thanks. Hope to hear from you.

  9. The link about Steve Horvath was especially interesting. (Can we read about that brain storming session anywhere btw?)

    What I find most interesting is that the gene expression profiles yielded no results. I simply cannot believe it. If there is difference in the phenotype between old and young cell, and of course we know there is – there must be difference in the expression profile as well – or the whole direction of research you and others follow is dead end. I can hardly think of an aging program that does not show up in transcribed or translated products.
    Maybe they looked at some sparse RNA chip data only and not deep whole transcriptome sequencing?

  10. It’s the 2nd time I see you crediting Francis Crick in this blog, Josh. (and I think I saw Harold do it on the previous comment section as well.. of all people.. ’cause it’s not enough to live forever, we also gotta have all the glory. : ) Just kidding Harold, still love ya.. even if you don’t gimme any credit ^__^)

    I cued up an interesting piece of information for you here: https://youtu.be/j7EBObU5Tjk?t=33m48s

    Symbolism is extremely important btw ; ) and shall only become ever more so.

  11. Hi Josh,

    what if the epigenetic clock is driven by wear and tear? I mean wear and tear of the DNA methylation state? This would be quite logical. DNA methylation state is reset at embryonic development only. Then it is a one way road for most of the cells. They go from totipotency to senescence, from undifferentiated to differentiated state.
    Aging is characterized by a generic, random demethylation of the DNA and methylation at specific sites (bivalent chromatin – genes switched on and off by PRC complex)
    Do we have a repair mechanism if a CpG site becomes demethylated on both strands by accident?
    I know of only single strand repair (DNMT1 – needed for maintaining methylation state at mitosis) and de novo methylation (DNMT3a and b).
    We know organismal aging is correlated with DNA methylation. We also know maximum lifespan is correlated with free radical production by mitochondria when comparing different species.
    What if the two are related? What if the wear and tear hypothesis is correct but not against the DNA or cellular organelles, but the DNA methylation state – which is reset only in embroynic age.
    And what if this is more than sufficient for maintaining a diversity of lifespans across species spanning from weeks to hundred years? I mean its enough to satisfy population dynamics requirements set by the niche the species living in?
    Interesting read on methylation maintainance

  12. To Charles Kendrick

    Yes, I did not doubt your citation on the effect of telomere elongation in mice. I just dismissed it when I had read it earlier and then forgot about it.
    The reason of my dismissal was that like many interventions, I feel it is just palliative, somehow reducing the burden of old age, but does not fundamentally extend lifespan. I think the same about mTOR/IGF1/GH related findings and senolytics, too. As I wrote before I would like to see a mouse that ages like a cat before I believe anything significant had been found.
    Also in my view telomere elongation is really downstream. Adult stem cells are supposed to elongate their telomeres to live and divide for decades.
    Telomere erosion in differentiated cells shouldnt matter as there is ample supply of new cells from stem cells.
    Non dividing cells do age, too, and telomeres should not shorten in non dividing cells. Also telomere length shows a very noisy correlation with aging.

    Yes, you are correct that DNMT KO is lethal very early on.

    I could not really find anything about someone upregulating DNMTs. The general thought is that these are oncogenes because they are upregulated in many cancers. There are cancer drugs targeting DNMTs
    However I would not link increased DNA methylation with cancer, I think its probably about selective methylation of onco suppressors.
    I would be really glad if I could read about research with overexpressing DNMTs.

    General DNA hypomethylation also helps cancer (in my view this is the primary reason for increased cancer incidence with age)

    And when I wrote about transcription profile I was referring to the link cited by Josh about Steve Horvath’s research.
    He was trying to find differences in the transcritption profile of old vs young somatic cells and he found none. This is very strange since we know he discovered a very specific aging pattern in DNA methylation.

    • In this wonderful 2002 paper
      the authors overexpresed DNMT1 by 250% (measured by methyltrasferase assay) in ES cells. The 5mC content hardly raised but there were adverse and embryonic lethal side effects.
      My understanding is that DNMT1 level in ES cells is saturated so overexpression there does not help anything only makes things worse. Would love to see an analysis of methyltransferase activity in old differentiated and adult stem cells vs young and also DNMT overexpression there.

    • I agree with you GaborB most of the research today is towards palliative solutions and not upstream enough. Thank you for the link to Benayoun, Polina, Brunet paper

  13. Hi Josh I truly admire how well you try to give a 3D perspective of a topic without any bias. No wonder you have so many readers and fans. I agree with this post that evidence is clearly leaning towards programmed aging across all living. One of the latest developments which can be cited towards this is argument is the discovery made by Hayashi and his team at Tsukuba University. No difference in DNA damage between a foetus, 12 year old or a 90 Year old. A lot of anti aging research by very important scientists is based on trying to attenuate the symptoms of aging via telomeres/telomeradr or NAD+ or glycine replenishment. This may give small benefits but won’t lead to a cure of aging. We are constantly run by a software right from the time we are in the womb forming our body parts to a fast growing baby to an adolescent hitting puberty to aging and death. It is difficult to believe that a software program runs making big and subtle changes to us only till we reach adulthood and then abruptly stops and then wear and tear becomes the cause of changes. It is so obvious that our entire life cycle is a coded program. Survival of species may be what has patterned this program. Whether it is cancer cells, humans or trees there are various mechanisms adopted to optimize our chances of survival. The day we as a species reach a point where we do not need a bio-chemical program to survive we will discover the cure to aging. Of all the species we are the closest. Fascinating times we live in.
    I am in a position to raise funding from investors for experiments mentioned by Harold Katcher (HPE on humans via plasmaperesis) and what Josh you mentioned in this post because of my background in investment banking and private equity. I have access to HNIs old timers and family offices who would invest to be first in que if the experiments succeed. Will hope that something good emerges out of such research and experiments in our lifetime as it is a shame that a human that accumulates and evolves knowledge over years to then turn into rotting flesh in a short span.

    • If you are willing to help raise funds, here are three projects I would offer as underfunded and very promising:

      1) Telomerase activation strategies. Screening chemicals in vitro, then testing them in vivo.

      2) Plasma transfusions. I would like to see an 80-year-old receive transfusions of blood plasma from 20-year-olds, perhaps once a week for several months. If there are dramatic results, this would attract new research and investment in the field.

      3) Testing known anti-aging treatments in combination in rodents. We know only about action of treatments one-at-a-time. If these 10% and 20% benefits can add up, that would be great. But I think it likely that most combinations won’t work that way. Let’s do a broad search for combinations that synergize. There is a crowd-funded group at MMTP which I am already in touch with, and I have a proposal in need of funding. Because of economies of scale, this would need to be funded with several hundred thousand dollars to produce results, but those results would be enormously valuable. http://joshmitteldorf.scienceblog.com/2015/12/22/we-know-nothing-about-longevity-drug-interactions/

      • Lolz Josh, what took Harold so long to bring this to your attention? I think you should be glad I don’t “turn off the computer and spend time with” monkeys after all, heh? 😉

      • Yes would love to contribute in a research project of mutual interest.
        1. I am not a big fan of Telomeres and Telomerase path to Longevity – the idea is to upstream as much as possible and Telomeres are as downstream as one can get. although I don’t know what you think of BioViva’s claim of lengthening the Telomeres of their CEO. Personally I do not see a cure for aging coming from interventions aimed at the symptoms of aging.
        2. Parabiosis and may be HPE have an intriguing prospect. Any coding needs to transmit it’s code to multiple targets for execution. Parabiosis offers prospects of swapping this message in transit. We can try and mobilize funding for HPE and tinker around the methodologies to see if it works but we will need to provide verification to our investors for which Steve Horvath’s DNA methylation lab would be required.
        3. I agree testing known anti aging interventions in various permutations and combinations may lead to valuable results.
        I am already in the process of funding an anti-aging clinic and lab and one of the key focus areas is to discover and bank combinations that ensure bioavailability. My guess is that 95% of supplements fail because of poor absorption rates. We are only looking at natural compounds or occurring in our body (which need augmenting due to decline with aging). So rapamycin and metformin will not be part of our offered therapies. An average person who has begun to age is not aware of all the safe interventions which have reasonable body of evidence and may allow for mitigation of some of the onslaught of aging – so at this clinic we hope to offer the most comprehensive portfolio of anti aging interventions which would be further tested at our labs for bioavailability, efficacies and safety.
        Would be happy to discuss over email what we can target jointly for funding while keeping in mind that investors will expect program that may lead to a commercial outcome and not just a knowledge building outcome.

  14. Ontogeny recapitulates philology! I asked a doctor friend two decades ago how the code that made us up was changed to produce the various organs in our bodies. He said the DNA was edited and implied that different DNA was in our skin and liver. Hail epigenome! You have mentioned that you think in numbers and code. I am thinking we are going to have added difficulty because the epigenome is not a single code overlaying all of our DNA but different code with different aging sequences within the different types of cells in our body. Am I right?

  15. I was at a party this last weekend and talked to a man with Parkinson’s Disease. Since there is only a week link to genetics could this be a result of corruption of the epigenome?

    I talked to a woman who said she was allergic to whey protein. She said her doctor told her that the reason she was allergic was because she is Swiss and has consumed a huge amount of dairy her whole life. The doctor told her that if a person was to consume several carrots a day that by the time they were well into adulthood they too would generate an allergy to carrots. I told her that this did not fit the allergy model. Are the whey and gluten intolerant suffering epigenetic or microbiome problems?

  16. In support of Josh Mitteldorf proposed program:
    Even a non-specific action on DNA methylation (experiments with 5-azacytidine – inhibitor of DNA methyltransferase enzyme, Dnmt1), http://www.pnas.org/content/113/16/E2306.abstract allows to reprogram the cells in vitro into safe tissue-regenerative multipotent stem cells. Another DNA methyltransferase inhibitor, RG108 significantly induces the expression of TERT by blocking methylation at the TERT promoter region. Under RG108 treatment in human bone marrow mesenchymal stromal cells, the anti-senescence genes TERT, bFGF,VEGF, and ANG were increased, whereas the senescence-related genes ATM, p21, and p53 were decreased. Number of senescence-associated β-galactosidase-positive cells was significantly decreased. (http://onlinelibrary.wiley.com/doi/10.1002/bab.1393/full?elq_mid=6940&elq_cid=1845997)

  17. It would be interesting to find evolutionarily conserved regions of epigenetic clock CpG methylation – that are changing in the same way in humans, mice, rats, dogs, cats, cows, etc. during aging.
    Furthermore, in addition to the plan.of Josh Mitteldorf
    It would be nice to try to start the epigenetic reprogramming of human blood cells by separating them from the plasma, and then transfer these blood cells to the young donor plasma (in parallel with the forced by dCas9-activator transcription activation of genes that were working in the young blood and dCas9-TET1 selective demethylation).

Leave a Reply

Your email address will not be published. Required fields are marked *