Epigenetic Drift Is a Determinant of Mammalian Lifespan

Abstract

The epigenome, which controls cell identity and function, is not maintained with 100% fidelity in somatic animal cells. Errors in the maintenance of the epigenome lead to epigenetic drift, an important hallmark of aging. Numerous studies have described DNA methylation clocks that correlate epigenetic drift with increasing age. The question of how significant a role epigenetic drift plays in creating the phenotypes associated with aging remains open. A recent study describes a new DNA methylation clock that can be slowed by caloric restriction (CR) in a way that correlates with the degree of lifespan and healthspan extension conferred by CR, suggesting that epigenetic drift itself is a determinant of mammalian lifespan. Genetic transplantation using genomic editing of DNA methylation homeostatic genes from long-lived to short-lived species is one way to potentially demonstrate a causative role for DNA methylation. Whether the DNA methylation clock be reset to youthful state, eliminating the effects of epigenetic drift without requiring a pluripotent cell intermediate is a critical question with profound implications for the development of aging therapeutics. Methods that transiently erase the DNA methylation pattern of somatic cells may be developed that reset this aging hallmark with potentially profound effects on lifespan, if DNA methylation-based epigenetic drift really plays a primary role in aging.

Introduction

The epigenome, which controls cell identity and function, is not maintained with 100% fidelity in somatic animal cells. Errors in the maintenance of the epigenome lead to epigenetic drift,^1,2 an important hallmark of aging.³ Because changes in the epigenome can result in changes in gene expression,⁴ and such changes in turn can alter which and how many proteins are made by a cell, it is evident that epigenetic drift could significantly impact homeostasis over time resulting in phenotypes characteristic of aging.

Determining the relative contribution of each hallmark of aging is of great importance to understanding the mechanisms that underlie aging with the potential to increase healthspan and eventually lifespan. Maintenance of homeostasis, growth control and stress response at the cellular level are key mechanisms that are conserved in animals. Numerous key genes, such as Daf-2/insulin receptors, and related pathways when modified or mutated can extend lifespan in organisms ranging from invertebrates, such as worms (Caenorhabditis elegans) and fruit flies (Drosophila), to mammals such as mice, rats and monkeys.⁵ But characterizing the nature and contribution of the specific underlying molecular and cellular processes beyond the general idea that damage or entropy drives increasing dysfunction with age is critical to understanding aging. Unlike simple machines, biological systems possess varying degrees of regenerative/rejuvenative capacity that counteracts myriad forms of damage. The extent to which an animal's somatic cells can counter intrinsic or extrinsic entropic forces largely determines its lifespan. In the extreme, some organisms such as hydra appear immortal, in that they maintain a constant mortality with age.⁶ The mechanism involves nonaging stem cells capable of replacing all cells in the organism. The differentiated cells in the hydra themselves likely age, but specialized mechanisms to maintain stem cells indefinitely must exist at least in hydra. Hydra have solved the problems of extrinsic damage, due in part to DNA and other biomolecule damage via radiation and the environment, and intrinsic damage due in part to problematic cell divisions and maintenance of cell state via the epigenome.

In animals with a finite lifespan (the vast majority of known species!), epigenetic drift occurs,⁴ however its relative importance compared to other hallmarks of aging such as cell senescence is a matter of debate. A simple prediction is that the most important hallmarks of aging will most closely correlate with chronological age in any specific organism and that differences between lifespans of organisms will also correlate quantitatively with the specific hallmark.

Patterns of DNA methylation at deoxycytosines in CpGs, a critical epigenetic modification at least in vertebrates, have been observed to change with age in relatively characteristic ways. These changes are especially significant in CpG islands (CGIs), clusters of CpGs that localize to promoter regions in mammals. CGIs tend to be unmethylated, while overall most CpGs which are localized to non-CGIs are methylated.⁷ Various aging “clocks” based on sequencing of cells and tissues from young and old mice, rats, monkeys, humans and other mammals have been reported.^8

–12 These clocks identify sets of specific DNA methylation changes that when statistically analyzed as a set, can be used to predict the chronological and perhaps even the biological age of an organism with accuracy in humans of the order of ±3.6 years for the Horvath clock⁹ and 3.3 weeks in the mouse.¹³ In addition, there are numerous reports that patterns of methylation of DNA can predict all-cause mortality in humans.^14

–18 Correlations of increased epigenetic drift in DNA methylation patterns have been reported for cognitive and physical functioning,¹⁴ obesity,¹⁹ menopause,²⁰ osteoarthritis,²¹ neurodegenerative diseases including Alzheimer's disease,²² Huntington's disease,²³ and Parkinson's disease²⁴ and Down syndrome,²⁵ lung cancer,²⁶ HIV infection,²⁷ while centenarians^18,28 exhibit a more youthful less altered pattern.

Interestingly these DNA methylation clocks in general did not correlate with transcriptional changes in the set of genes comprising the clock, which is surprising because DNA hypermethylation at promoters tends to repress transcription, while DNA hypomethylation tends to activate RNA transcription in mammals and probably all vertebrates. Transcriptional changes over time have been associated with aging.^6,7 Moreover, the changes in the clock genes are statistical, there is mosaicism at the individual cell level,^29,30 which would suggest that adult organisms may carry a mixture of young, middle-aged and old cells at any point in their lifetime, with the mixture tending toward more old cells with increasing age. Not surprisingly, at least one of the more well-studied DNA methylation clocks is closely linked to cell proliferation, the greater number of cell divisions, the older the cell appears.⁹ Moreover, a DNA methylation based mitotic clock for stem cells has been established that connects extent of cell proliferation with cancer.³¹ Thus, diseases characterized by physical damage to tissues that provokes a repair response involving cell division exhibit an aged methylation phenotype. The processes of DNA methylation based epigenetic drift appears independent of replicative and cellular senescence in that a correlation of altered DNA methylation with cell passage persists in cell culture, but not with senescence status.³² However, it seems likely that the role of proinflammatory signals from senescent cells on bystanders is likely to accelerate epigenetic drift in the bystanders.

Clearly, changes in methylation of DNA are potentially good biomarkers for aging in mammals, but are they actually one of the fundamental mechanisms that cause loss of homeostasis that is observed with aging? One complicating problem is that methylation of DNA at CpG plays a less critical role in invertebrate gene expression versus mammals. For example, in well-studied Drosophila, levels of CpG are very low and the regulatory role is unclear, and in C. elegans, CpG is essentially not present at all. However, epigenetic drift has been reported for C. elegans at the level of histone modifications H3K27me3,^33,34 and Drosophila ³⁵ and associated with “transcriptional drift”,³⁶ increasing number of errors in the pattern of transcribed genes. There is speculation from work describing extensive methylation of DNA in pacific oysters that CpG DNA methylation may in fact, date to the common ancestor of eukaryotes,³⁷ although its differential role in influencing gene expression versus stabilizing the genome and suppressing transposable elements may vary greatly through evolutionary time.³⁸ In vertebrates and mammals methylation of DNA at CpG modulates gene expression and plays a significant role in maintaining cell identity, as is clear from experiments in which treatment with 5-azacytidine, a DNA (cytosine-5)-methyltransferase (DNMT) inhibitor that hypomethylates DNA, can transform fibroblasts into myocytes, chondrocytes and adipocytes.³⁹ Clearly, alteration of methylation of DNA with aging could have profound effects on cell state and function.

Epigenetic Drift Is a Determinant of Mammalian Lifespan

In work that extends our understanding of the possible role that DNA methylation-based epigenetic drift plays in determining lifespan, Maegawa et al.⁴⁰ show that the rate of epigenetic drift correlates with lifespan among mice, rhesus monkeys and humans. Furthermore, caloric restriction (CR), which extends lifespan in a large number of organisms, in monkeys and mice resulted patterns of methylation of DNA in blood cells significantly younger than ad libitum (AL) controls. The authors suggest in a somewhat circular fashion that CR slows aging in part through slowing epigenetic drift and that epigenetic drift may be a key determinant of mammalian lifespan.⁴⁰ Actually, more sophisticated experiments in which maintenance of methylation of DNA was altered directly by genetic means or drugs in the context of CR, would be needed to actually prove these hypotheses. A potentially transformative experiment would be to use CRISPR gene editing to substitute key DNA methylation metabolism genes DNMT1, DNMT2, DNMT3A, DNMT3B, TET1, TET2 and TET3 from a long-lived species, such as humans into short-lived mice, and examine the effect on genetic drift and lifespan.

In Maegawa et al.'s experiments, unlike much of the previous work that generated DNA methylation clocks the Digital Restriction Enzyme Analysis of Methylation (DREAM) sequencing method was used.⁴⁰ In this method DNA is digested by two restriction enzymes. One only cuts CCCGGG without a methylated CpG, leaving a blunt ended fragment cut in the middle of the recognition sequence, then another enzyme is used that will cut CCCGGG regardless of C(m)pG leaving a 5′ overhand creating a methylation signature for next generation sequencing.⁴¹ This method is very sensitive, but will miss many methylated CpGs not in the recognition sequence. Analyzing 19 mice from 0.3 to 2.8 years, 16 monkeys from 0.8 to 30 years and 16 humans from 0 (cord blood) to 86 years, using unsupervised hierarchical clustering, a standard statistical technique, found that for the subset of genes in unmethylated CGI sites (less than 5%) methylation went from 2% ± 0.1% in the young to 18% ± 5% in the old mice (p = 0.03), 2% ± 0.3% to 22% ± 3% in young versus old monkeys (p = 0.002) and 3% ± 0.5% to 20% ± 4% in newborn versus old humans (p = 0.009). The reverse pattern was seen for non-CGI sites with the most methylation (>90%): DNA methylation went from 94% ± 0.4% (young mice) to 78% ± 4% (old mice) (p = 0.003) and similarly from 94% ± 0.3% to 73% ± 4% in young versus old monkeys (p = 0.007) and from 93% ± 1% in newborns to 74% ± 2% in newborns to old humans (p < 0.001). So which genes were involved? Ingenuity pathway analysis of hypermethylated genes showed enrichment for genes involved in development, signaling, growth, cell maintenance and in cancer and cardiovascular disease. Pathways tended to be conserved across species. They then repeated this analysis for granulocytes, T-cells and CD34⁺ cells to determine which of the analyzed genes might have cell-type specific methylation patterns, and were able to show that variation in cell numbers of these cells could not explain the variability, although they should have done a more comprehensive study looking at neutrophils, monocytes, eosinophils, and B-cells as well. Then focusing on promoter regions, Maegawa et al. find that the genes involved in “methylation drift” are conserved across all three species.⁴⁰ But keep in mind the potential artifacts from requiring the CCCGGG sequence to be present to be detected in any particular gene, which would occur 1 out of 4096 random DNA sequences.

How did the DREAM DNA methylation analysis compare to other studies using different techniques? There was overlap, but the DREAM method detected a much higher number of drifted genes. What about correlation with gene expression, which was generally not observed in previous reports? Using DREAM, Maegawa et al. found substantial correlation with gene expression. Genes with increased expression with age were significantly demethylated, and the smaller number of genes with reduced expression with age had increased DNA methylation. In order to confirm the data, another method of detecting DNA methylation was used: bisulfite pyrosequencing, in which treatment with bisulfite converts cytosine, but not methylated cytosine to uracil; uracil then sequences as a thymidine for 34 candidate mouse genes, 36 candidate monkey genes and 16 human genes. Clear age-related patterns emerged, but with the caveat that some genes showed significant DNA methylation changes by bisulfite pyrosequencing that were undetected by DREAM, which is consistent with the more thorough capability of the former. The authors suggest that their DREAM analysis may underestimate the extent of conservation and alterations among species.⁴⁰

To determine a rate of age-related hypermethylation, Maegawa et al. considered 10 conserved hypermethylated genes and plotted the data against age to calculate drift rates of 4.1% ± 1.2% per year in mice, 0.34% ± 0.14% per year in monkeys, and 0.10% ± 0.02% per year in humans. The numbers were similar considering all tested hypermethylated genes regardless of conservation: 5.1% ± 0.4% per year in mice, 0.47% ± 0.02% per year in monkeys, 0.09% ± 0.01% per year in humans. This data is consistent with the idea that epigenetic drift is a determinant of maximum lifespan, but is only correlative. There was an inverse correlation in a log-log plot of methylation rate with maximum lifespan (Fig. 1).⁴⁰ In effect, Maegawa et al. have built a new DNA methylation aging clock. Interestingly some monkeys appear to have a methylation rate significantly lower than some humans, and there is a fourfold difference in rates methylation rate for humans which may be due to noise, or to the possibility that the model needs refinement.

FIG. 1.

Maximum lifespan is correlated with methylation rate of change per year (adapted from Maegawa et al.⁴⁰). Color images available online at www.liebertpub.com/rej

Because CR is known to extend lifespan in mice, and monkeys, the effect of CR on DNA methylation drift was examined using DREAM and pyrosequencing on blood cells in old mice (2.7–3.2 years) subjected to 40% CR from 0.3 years and on old monkeys (22–30 years) subjected to 30% CR from 7 to 14 years old. CR slowed DNA methylation drift for both the mice and monkeys. The average methylation of all 24 genes was 26% ± 2% in AL old mice and 17% ± 0.7% in CR old mice and average methylation was 27% ± 0.7% in AL monkeys compared to 24% ± 0.9% in CR monkeys. The 12 CR mice had an average chronological age of 2.8 years and a “methylation age” of 0.8 years (3.5-fold improvement in epigenetic drift rate), and the 18 CR monkeys had an average chronologic age of 27 years with the predicted methylation age of 20 years.⁴⁰ The clock may need calibration, as 2.8 year old mice are reporting 0.8 years, are unlikely to live another 2 years or more as would be predicted from the clock. The decidedly smaller effect for monkeys may be due to a combination of less CR, starting CR later, or primates having less potential to improve epigenetic drift given inherently lower drift.

To determine tissue specificity of epigenetic drift, 12 hypermethylated and 3 hypomethylated genes were studied. Most of these genes showed age-associated DNA methylation drift in most tissues. Interestingly, as might be predicted from cell proliferation rates, kidney and liver showed less age-related hypermethylation, while the large intestine showed even larger drift than blood. On the other hand, blood, spleen, and kidney showed similar hypomethylation drift, while liver, small, and large intestine and bone marrow had reduced drift.⁴⁰ Since hypomethylation tends to be occurring at non-CGI locations it suggests that mechanisms other than proliferation rate may actually underlie this kind of epigenetic drift.

Given the hypothesis that telomere shortening is a potential biomarker for aging, as well as promoter of aging, quantitative PCR was used to examine telomere shortening in mice, monkeys and humans. Interestingly, only a small amount of telomere shortening was found with age, on which CR treatment had no measurable effect, in agreement with a previous report in monkeys,⁴² suggesting that telomere length is a poor biomarker for biological age.

As the authors point out, because their data was cross-sectional, the experiments could be improved by carrying out a longitudinal study which is likely to provide better and more accurate data.⁴⁰ Larger data sets and greater use of pyrosequencing across the whole genome on pure populations of cells will eventually yield far better data and models.

Maegawa et al. hypothesize that the most reproducible changes they observe are in stem or progenitor cells, given that differentiated cells might be lost upon cell death, which may be an oversimplification as some differentiated cells in the immune system, liver, kidneys etc. can be quiescent for long periods, and that differentiation does not automatically imply a postmitotic state.⁴⁰ However, their hypothesis is strongly supported by emerging data. Epigenetic drift may cause loss of stem cell plasticity by aging and a key regulator of lifespan³⁰ Stem cell dysfunction and exhaustion have been linked to aging-associated dysfunction.^43
–45 Given that epigenetic programs may be established and reset during the DNA replication to some extent, it is plausible that DNA methylation drift derives from accumulated random epigenetic errors during stem cell proliferation. They suggest that species-specific differences in rates of methylation drift relate to differing rates of stem cell turnover. This is plausible, although in need of supporting data.

We suggest that there are other intriguing possibilities. The fidelity of epigenetic maintenance in cell proliferation may be different between species. For example, the error rate of methylation maintenance has not been exhaustively studied. A rate of 10e⁻⁴ to 10e⁻⁵ loss/gains per division was observed for CpGs of the APRT gene in cultured mice kidney cells.⁴⁶ Small differences in the epigenetic maintenance rate per cell division could translate into large differences in the epigenetic drift rate per year. Another possibility is that there is a close link between DNA damage and repair capability and epigenetic drift. Cells from species with enhanced DNA repair capacity, which is associated with longevity,⁴⁷ are likely to better maintain chromatin structure and patterns of DNA methylation after constant damage from environmental radiation. Consistent with such a hypothesis, CR or fasting enhances DNA repair in mammals and protects from DNA damage from chemotherapy and radiation damage in mice and humans,⁴⁸ and CR helps to delay accelerated aging in a DNA repair deficient mouse model.⁴⁹ Perhaps the same mechanisms that protect from DNA damage, help maintain the methylome.

Because previous studies have shown that chronic inflammation, which is known to shorten lifespan, also accelerates methylation drift,⁵⁰ and because so many diseases with an inflammatory component accelerate DNA methylation clocks it is tempting to speculate that increased numbers of senescent cells promoting inflammation may very well accelerate epigenetic drift, and that epigenetic drift may lead to cell stress that promotes cell senescence, a vicious cycle. For example, inhibition of DNA methyltransferases by 5-azacytidine or siRNA induces senescence in mesenchymal stem cells (MSCs) by altering DNA methylation and histones associated with promoters.⁵¹ Could increased DNA methylation also trigger senescence? In support of the latter idea a stochastically acquired DNA methylation signature for senescence is acquired in subpopulations of MSCs over time.⁵²

Postmitotic differentiated cells such as neurons may be subject to epigenetic drift with aging as well. For example, the epigenome in postmitotic Purkinje cells is reported to be dynamic during development⁵³ and neuronal activity has been reported to modify patterns of DNA methylation.⁵⁴ Moreover, it has been suggested that at least some of the aging-associated DNA methylation changes observed with aging in the brain derive from neurons⁵⁵ and DNA methylation changes in synaptic genes in the prefrontal cortex associate with aging.⁵⁶ DNA damage from radiation and other external insults would be predicted to play a key role in postmitotic cells. In a recent report, epigenetic drift associated with aging was hypothesized to play a role in altering autophagy in adult neurons in C. elegans.⁵⁷ Autophagy, typically a beneficial homeostatic process that extends lifespan and counters dysfunctions of aging, was transformed in older worms into a destabilizing process that shortens lifespan because autophagosomes lose their ability to resolve correctly with age. Worm lifespan could actually be extended 30% by inhibiting autophagy initiating factors such as bec-1 in adult worms, preventing the buildup of toxic autophagosomes.⁵⁷ That potentially restricted and as yet uncharacterized changes in gene expression that occur during aging could so profoundly alter a homeostasis maintaining process, suggests that there are many ways even modestly altered gene expression can be amplified into pathological phenotypes. One caveat is that CpG DNA methylation plays no role in epigenetic drift in C. elegans, which lacks methylated CpG. However, epigenetics is far more complex than just DNA methylation. For example, it is known that a global decrease in histone modification H3K27me3 occurs in aging worms,³³ although this aging-associated epigenetic change does not appear conserved in mouse satellite cells⁵⁸ or killifish.⁵⁹

Medical Implications

Although the existence of epigenetic drift and DNA methylation clocks is established, more work is needed using single cells and complete genome coverage to fully assess how DNA methylation comprehensively changes with aging. Furthermore, such studies need to assess the relative amounts of hydroxymethylcytosine as well as methylcytosine.⁶⁰ There is evidence that hydroxymethylcytosine plays a role in nucleosome positioning with distinctive roles in neurons and primordial germ cells.⁶¹ To date, clocks of aging based on methylation of DNA have been derived from studies that ignore this potential complicating distinction, because the most frequently used techniques to detect DNA methylation report hydroxymethylcytosine as cytosine.

Clocks of epigenetic drift based on patterns of DNA methylation provide a powerful molecular biomarker for testing the effects of therapeutics on aging. When performed on cells from the same individual animal or human being, it should be possible to obtain an accurate assessment of at least one molecular hallmark of aging that corresponds well to biological age.

CR has been known for a long time as a means to slow aging in mammals, so it is gratifying to observe that alterations in methylation of DNA reflect a magnitude of change consistent with known effects on lifespan for mice and monkeys.⁶² That the effect was reduced for monkeys, suggests that longitudinal studies in humans may be very beneficial in determining just how much potential CR has to extend the lifespan of humans. The recent CALERIE trial showed CR slowed measures of physiological aging in humans.⁶³ It is quite possible that the potential individual benefits from CR will vary greatly. Similarly, the potential benefits from treatment with a CR mimetic such as rapamycin⁶⁴ will vary. Moreover there is the caveat that CR mimetics sometimes oppose the effects of CR. For example, although CR generally inhibits mTOR, CR actually activates mTOR in adult gut stem cells stimulating their proliferation,⁶⁵ while rapamycin always inhibits mTOR, which would block beneficial gut stem cell expansion. As for healthspan, the effects of CR on cell physiology in humans and animal systems are under intense study. There are points of potential divergence between potential increased lifespan, healthspan and negative tradeoffs due to metabolic alterations. For example, CR will, by necessity, reduce muscle mass. An interesting report suggests that CR can improve hematopoietic stem cell (HSC) function by increasing cell numbers and preventing a skewing toward myeloid lineages, but at the same time inhibits differentiation of HSC into lymphoid lineages and blocks proliferation of lymphoid progenitors, effectively impairing immune function.⁶⁶ So, it does a person little good if she is biologically younger than her untreated peers, but more prone to mortality. However, other reports suggest that lifelong CR had no effect on immune function.⁶⁷

Can the epigenetic clock be reset?⁶⁸ Reprogramming a differentiated cell to a pluripotent state and then redifferentiating cells to the original cell type is the only known way to reverse epigenetic drift and reset the epigenome.⁶⁹ An outstanding question is does partial reprogramming via transient induction of IPS inducing factors actually reset the epigenetic state of a cell.⁷⁰ The answer to this question should be forthcoming with profound implications for the development of antiaging therapeutics. A more modest approach, transdifferentiating one cell type (fibroblasts) into another cell type (neurons) appears to retain an epigenetic aging signature of the parental cell.⁷¹ An interesting possibility is that DNA methylation inhibitors such as 5-azacytidine could be used during the transdifferentiation process to erase the aging-associated hypermethylation, and rejuvenate the cells. Potentially, cells may be cis-differentiated into their original phenotype or transdifferentiated into a key stem cell type to rejuvenate their epigenome and/or perhaps reconstitute a stem cell compartment. Determining which drugs or genetic manipulations can erase DNA methylation to change cell state,⁷² and then whether the correct DNA methylation pattern can be recovered subsequently will have profound impact on this strategy.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Martin

. Epigenetic drift in aging identical twins. Proc Natl Acad Sci U S A, 2005; 102:10413–10414.

Fraga

, Ballestar

, Paz

, Ropero

, Setien

, Ballestar

, et al. Epigenetic differences arise during the lifetime of monozygotic twins. Proc Natl Acad Sci U S A, 2005; 102:10604–10609.

López-Otín

, Blasco

, Partridge

, Serrano

, Kroemer

. The Hallmarks of Aging. Cell, 2013; 153:1194–1217.

Booth

, Brunet

. The aging epigenome. Mol Cell, 2016; 62:728–744.

Kenyon

. The genetics of ageing. Nature, 2010; 464:504–512.

Schaible

, Scheuerlein

, Dańko

, Gampe

, Martínez

, Vaupel

. Constant mortality and fertility over age in Hydra. Proc Natl Acad Sci U S A, 2015; 112:15701–15706.

Bird

. DNA methylation patterns and epigenetic memory. Genes Dev, 2002; 16:6–21.

Hannum

, Guinney

, Zhao

, Zhang

, Hughes

, Sadda

, et al. Genome-wide methylation profiles reveal quantitative views of human aging rates. Mol Cell, 2013; 49:359–367.

Horvath

. DNA methylation age of human tissues and cell types. Genome Biol, 2013; 14:3156.

10.

De Paoli-Iseppi

, Deagle

, McMahon

, Hindell

, Dickinson

, Jarman

. Measuring animal age with DNA methylation: From humans to wild animals. Front Genet, 2017; 8:106.

11.

Bocklandt

, Lin

, Sehl

, Sánchez

, Sinsheimer

, Horvath

, et al. Epigenetic predictor of age. PLoS One, 2011; 6:e14821.

12.

Garagnani

, Bacalini

, Pirazzini

, Gori

, Giuliani

, Mari

, et al. Methylation of ELOVL2 gene as a new epigenetic marker of age. Aging Cell, 2012; 11:1132–1134.

13.

Stubbs

, Bonder

, Stark

A-K

, Krueger

, von Meyenn

, Stegle

, et al. Multi-tissue DNA methylation age predictor in mouse. Genome Biol, 2017; 18:68.

14.

Marioni

, Shah

, McRae

, Chen

, Colicino

, Harris

, et al. DNA methylation age of blood predicts all-cause mortality in later life. Genome Biol, 2015; 16:25.

15.

Christiansen

, Lenart

, Tan

, Vaupel

, Aviv

, McGue

, et al. DNA methylation age is associated with mortality in a longitudinal Danish twin study. Aging Cell, 2016; 15:149–154.

16.

Perna

, Zhang

, Mons

, Holleczek

, Saum

K-U

, Brenner

. Epigenetic age acceleration predicts cancer, cardiovascular, and all-cause mortality in a German case cohort. Clin Epigenetics, 2016; 8:64.

17.

Chen

, Marioni

, Colicino

, Peters

, Ward-Caviness

, Tsai

P-C

, et al. DNA methylation-based measures of biological age: Meta-analysis predicting time to death. Aging, 2016; 8:1844–1859.

18.

Horvath

, Pirazzini

, Bacalini

, Gentilini

, Di Blasio

, Delledonne

, et al. Decreased epigenetic age of PBMCs from Italian semi-supercentenarians and their offspring. Aging, 2015; 7:1159–1170.

19.

Horvath

, Erhart

, Brosch

, Ammerpohl

, von Schönfels

, Ahrens

, et al. Obesity accelerates epigenetic aging of human liver. Proc Natl Acad Sci U S A, 2014; 111:15538–15543.

20.

Levine

, Lu

, Chen

, Hernandez

, Singleton

, Ferrucci

, et al. Menopause accelerates biological aging. Proc Natl Acad Sci U S A, 2016; 113:9327–9332.

21.

Vidal

, Lopez-Golan

, Rego-Perez

, Horvath

, Blanco

, Riancho

, et al. Specific increase of methylation age in osteoarthritis cartilage. Osteoarthritis Cartilage, 2016; 24:S63.

22.

Levine

, Lu

, Bennett

, Horvath

. Epigenetic age of the pre-frontal cortex is associated with neuritic plaques, amyloid load, and Alzheimer's disease related cognitive functioning. Aging, 2015; 7:1198–1211.

23.

Horvath

, Langfelder

, Kwak

, Aaronson

, Rosinski

, Vogt

, et al. Huntington's disease accelerates epigenetic aging of human brain and disrupts DNA methylation levels. Aging, 2016; 8:1485–1504.

24.

Horvath

, Ritz

. Increased epigenetic age and granulocyte counts in the blood of Parkinson's disease patients. Aging, 2015; 7:1130–1142.

25.

Horvath

, Garagnani

, Bacalini

, Pirazzini

, Salvioli

, Gentilini

, et al. Accelerated epigenetic aging in Down syndrome. Aging Cell, 2015; 14:491–495.

26.

Levine

, Hosgood

, Chen

, Absher

, Assimes

, Horvath

. DNA methylation age of blood predicts future onset of lung cancer in the women's health initiative. Aging, 2015; 7:690–700.

27.

Horvath

, Levine

. HIV-1 infection accelerates age according to the epigenetic clock. J Infect Dis, 2015; 212:1563–1573.

28.

Horvath

, Mah

, Lu

, Woo

, Choi

O-W

, Jasinska

, et al. The cerebellum ages slowly according to the epigenetic clock. Aging, 2015; 7:294–306.

29.

Bahar

, Hartmann

, Rodriguez

, Denny

, Busuttil

, Dolle

MET

, et al. Increased cell-to-cell variation in gene expression in ageing mouse heart. Nature, 2006; 441:1011–1014.

30.

Issa

J-P

. Aging and epigenetic drift: A vicious cycle. J Clin Invest, 2014; 124:24–29.

31.

Yang

, Wong

, Kuh

, Paul

, Rakyan

, Leslie

, et al. Correlation of an epigenetic mitotic clock with cancer risk. Genome Biol, 2016; 17:205.

32.

Lowe

, Horvath

, Raj

. Epigenetic clock analyses of cellular senescence and ageing. Oncotarget, 2016; 7:8524–8531.

33.

Maures

, Greer

, Hauswirth

, Brunet

. H3K27 demethylase UTX-1 regulates C. elegans lifespan in a germline-independent, insulin-dependent, manner. Aging Cell, 2011; 10:980–990.

34.

, Ebata

, Alipanahiramandi

, Lee

. Two SET domain containing genes link epigenetic changes and aging in Caenorhabditis elegans . Aging Cell, 2012; 11:315–325.

35.

Wood

, Hillenmeyer

, Lawrence

, Chang

, Hosier

, Lightfoot

, et al. Chromatin remodeling in the aging genome of Drosophila. Aging Cell, 2010; 9:971–978.

36.

Rangaraju

, Solis

, Thompson

, Gomez-Amaro

, Kurian

, Encalada

, et al. Suppression of transcriptional drift extends C. elegans lifespan by postponing the onset of mortality. Elife, 2015; 4:e08833.

37.

Wang

, Li

, Lian

, Li

, Jin

, Cai

, et al. Genome-wide and single-base resolution DNA methylomes of the Pacific oyster Crassostrea gigas provide insight into the evolution of invertebrate CpG methylation. BMC Genomics, 2014; 15:1119.

38.

Benayoun

, Pollina

, Brunet

. Epigenetic regulation of ageing: Linking environmental inputs to genomic stability. Nat Rev Mol Cell Biol, 2015; 16:593–610.

39.

Taylor

, Jones

. Multiple new phenotypes induced in 10T1/2 and 3T3 cells treated with 5-azacytidine. Cell, 1979; 17:771–779.

40.

Maegawa

, Lu

, Tahara

, Lee

, Madzo

, Liang

, et al. Caloric restriction delays age-related methylation drift. Nat Commun, 2017; 8:539.

41.

Jelinek

, Madzo

. DREAM: A simple method for DNA methylation profiling by high-throughput sequencing. Methods Mol Biol, 2016; 1465:111–127.

42.

Smith

, Mattison

, Desmond

, Gardner

, Kimura

, Roth

, et al. Telomere dynamics in rhesus monkeys: No apparent effect of caloric restriction. J Gerontol A Biol Sci Med Sci, 2011; 66A:1163–1168.

43.

Beerman

, Bock

, Garrison

, Smith

, Gu

, Meissner

, et al. Proliferation-Dependent alterations of the DNA methylation landscape underlie hematopoietic stem cell aging. Cell Stem Cell, 2013; 12:413–425.

44.

Schultz

, Sinclair

. When stem cells grow old: Phenotypes and mechanisms of stem cell aging. Development, 2016; 143:3–14.

45.

, Lee

, Wagers

. Stem cell aging: Mechanisms, regulators and therapeutic opportunities. Nat Med, 2014; 20:870–880.

46.

Rose

, Yates

, Simpson

, Tischfield

, Stambrook

, Turker

. Biallelic methylation and silencing of mouse Aprt in normal kidney cells. Cancer Res, 2000; 60:3404–3408.

47.

MacRae

, Croken

, Calder

, Aliper

, Milholland

, White

, et al. DNA repair in species with extreme lifespan differences. Aging, 2015; 7:1171–1182.

48.

Dorff

, Groshen

, Garcia

, Shah

, Tsao-Wei

, Pham

, et al. Safety and feasibility of fasting in combination with platinum-based chemotherapy. BMC Cancer, 2016; 16:360.

49.

Vermeij

, Dollé

MET

, Reiling

, Jaarsma

, Payan-Gomez

, Bombardieri

, et al. Diet restriction delays accelerated aging and genomic stress in DNA repair deficient mice. Nature, 2016; 537:427–431.

50.

Issa

J-PJ

, Ahuja

, Toyota

, Bronner

, Brentnall

. Accelerated age-related CpG island methylation in ulcerative colitis. Cancer Res, 2001; 61:3573–3577.

51.

A-Y

, Jung

J-W

, Lee

, Kim

H-S

, Kang

K-S

. DNA methyltransferase controls stem cell aging by regulating BMI1 and EZH2 through MicroRNAs. PLoS One, 2011; 6:e19503.

52.

Franzen

, Zirkel

, Blake

, Rath

, Benes

, Papantonis

, et al. Senescence‐associated DNA methylation is stochastically acquired in subpopulations of mesenchymal stem cells. Aging Cell, 2017; 16:183–191.

53.

Zhou

, Resendiz

, Lo

C-L

, Chen

. Cell-wide DNA de-methylation and re-methylation of Purkinje neurons in the developing cerebellum. PLoS One, 2016; 11:e0162063.

54.

Guo

, Ma

, Mo

, Ball

, Jang

M-H

, Bonaguidi

, et al. Neuronal activity modifies DNA methylation landscape in the adult brain. Nat Neurosci, 2011; 14:1345–1351.

55.

, Ebrahimi

, Wang

S-C

, Cortese

, Kaminsky

, Gottesman

, et al. Epigenetic assimilation in the aging human brain. Genome Biol, 2016; 17:76.

56.

Ianov

, Riva

, Kumar

, Foster

. DNA Methylation of synaptic genes in the prefrontal cortex is associated with aging and age-related cognitive impairment. Front Aging Neurosci, 2017; 9:249.

57.

Wilhelm

, Byrne

, Medina

, Kolundžić

, Geisinger

, Hajduskova

, et al. Neuronal inhibition of the autophagy nucleation complex extends life span in post-reproductive C. elegans. Genes Dev, 2017; 31:1561–1572.

58.

Liu

, Cheung

, Charville

, Hurgo

BMC

, Leavitt

, Shih

, et al. Chromatin modifications as determinants of muscle stem cell quiescence and chronological aging. Cell Rep, 2013; 4:189–204.

59.

Baumgart

, Groth

, Priebe

, Savino

, Testa

, Dix

, et al. RNA-seq of the aging brain in the short-lived fish N. furzeri–conserved pathways and novel genes associated with neurogenesis. Aging Cell, 2014; 13:965–974.

60.

Song

C-X

, Szulwach

, Fu

, Dai

, Yi

, Li

, et al. Selective chemical labeling reveals the genome-wide distribution of 5-hydroxymethylcytosine. Nat Biotechnol, 2011; 29:68–72.

61.

Teif

, Beshnova

, Vainshtein

, Marth

, Mallm

J-P

, Höfer

, et al. Nucleosome repositioning links DNA (de)methylation and differential CTCF binding during stem cell development. Genome Res, 2014; 24:1285–1295.

62.

Mattison

, Colman

, Beasley

, Allison

, Kemnitz

, Roth

, et al. Caloric restriction improves health and survival of rhesus monkeys. Nat Commun, 2017; 8:14063.

63.

Belsky

, Huffman

, Pieper

, Shalev

, Kraus

. Change in the rate of biological aging in response to caloric restriction: CALERIE biobank analysis. J Gerontol A Biol Sci Med Sci, 2017. DOI:10.1093/gerona/glx096

64.

Gillespie

, Pickering

, Eskiw

. Better living through chemistry: Caloric restriction (CR) and CR mimetics alter genome function to promote increased health and lifespan. Front Genet, 2016; 7:142.

65.

Igarashi

, Guarente

. mTORC1 and SIRT1 cooperate to foster expansion of gut adult stem cells during calorie restriction. Cell, 2016; 166:436–450.

66.

Tang

, Tao

, Chen

, Koliesnik

, Calmes

, Hoerr

, et al. Dietary restriction improves repopulation but impairs lymphoid differentiation capacity of hematopoietic stem cells in early aging. J Exp Med, 2016; 213:535–553.

67.

Tomiyama

, Milush

, Lin

, Flynn

, Kapahi

, Verdin

, et al. Long-term calorie restriction in humans is not associated with indices of delayed immunologic aging: A descriptive study. Nutr Healthy Aging, 2017; 4:147–156.

68.

Rando

, Chang

. Aging, rejuvenation, and epigenetic reprogramming: Resetting the aging clock. Cell, 2012; 148:46–57.

69.

Okita

, Ichisaka

, Yamanaka

. Generation of germline-competent induced pluripotent stem cells. Nature, 2007;advanced online publication. doi:10.1038/nature05934.

70.

Ocampo

, Reddy

, Martinez-Redondo

, Platero-Luengo

, Hatanaka

, Hishida

, et al. In vivo amelioration of age-associated hallmarks by partial reprogramming. Cell, 2016; 167:1719–1733.e12.

71.

Mertens

, Paquola

ACM

, Ku

, Hatch

, Böhnke

, Ladjevardi

, et al. Directly reprogrammed human neurons retain aging-associated transcriptomic signatures and reveal age-related nucleocytoplasmic defects. Cell Stem Cell, 2015; 17:705–718.

72.

Lewis

, Brewster

, Tian

, Shi

. Direct reprogramming facilitated by small molecules. J Stem Cell Transplant Biol, 2015; 1:103.