Epigenetic modifications induce premature aging in mammals
Figure 1 – Inducible changes to the epigenome alter aging in mammals, but it is a reversible process [Image created with BioRender.com]
Abstract
An increase in entropy is an hallmark of aging, leading organisms to lose genetic and epigenetic information. Here, a system called ‘‘ICE’’ (inducible changes to the epigenome) to cause non-mutagenic double strand breaks (DSBs) is applied on mice. The consequent relocalization of chromatin-modifying proteins shows that faithful DNA repair accelerates aging at physiological, cognitive, and molecular levels, involving also alterations in the epigenome, cellular exdifferentiation, and progression of the epigenetic clock. The expression of a subset of Yamanaka factors, OSK, can reverse these changes and operate a rejuvenation process, as the Information Theory of Aging suggests.
Review
Introduction
All the processes of the living cells are driven by a proper usage (through the epigenome) of the information stored in the DNA (genome). If cellular machinery is viewed as the biological hardware of the cell, the genome and epigenome would be the software. Aging is here conceptualized as a “software problem” affecting epigenetic information causing changes in cellular identity and function over time, which might be restored from an existing backup copy through reprogramming. The question remains whether aging results from a breakdown in cellular machinery, within the genome and epigenome, or both.
To maintain proper functions, cells need to preserve their epigenetic information, keeping a low state of Shannon entropy. However, over time, every organism have to face an increase in entropy, which causes a loss of genetic and epigenetic information.
Historically, the prevailing view was that aging resulted from the loss of genetic information due to DNA damage, with double strand breaks (DSBs) being the most common. It is now clear that old cells often have fewer mutations, and organisms with higher mutation rates do not necessarily exhibit premature aging; because of that, loss of epigenetic information, rather than genetic information, could be the primary cause of aging. Studies in yeast have already demonstrated that epigenetic changes, like relocalization of Sir complex or histone modifications, are not just biomarkers but causative factors of aging. Similar effects on epigenetic landscape are present also in inferior multicellular organisms.
Here, the new hypothesis about the “relocalization of chromatin modifiers” (RCM) is presented in mammals [1], suggesting that aging in eukaryotes is due to the loss of transcriptional networks and epigenetic information over time, driven by a conserved mechanism evolved to respond to cellular damage. This hypothesis is based on another theory proposed by the Sinclair group called ” The Information Theory of Aging” [2], which states that epigenetic reprogramming can improve the function of damaged and aged tissues, catalyzing age reversal.
The ICE system
To determine whether epigenetic changes are a cause, rather than merely an effect of mammalian aging, the ICE system was designed [1]. This system induces double strand breaks (DSBs) without causing mutations, using a specific endonuclease called I-PpoI. This enzyme recognizes and cuts the DNA in 20 canonical sites in non-coding genomic regions, causing DSBs. I-PpoI is an optimal choice because of its low rate of mutation.
The ICE system is a transgene containing a fusion of the I-PpoI gene to an HA-tagged tamoxifen-regulated mutant estrogen receptor (HA-ERT2), a GFP as a reporter gene and a transcriptional loxP-STOP-loxP cassette (Figure 2). A second transgene consists in a tamoxifen-regulated Cre recombinase gene (Cre-ER) upstream of a ubiquitin promoter for whole-body expression. Heterozygous ICE mice were generated by crossing I-PpoI STOP/+ mice to CreERT2/+ mice harboring a single ERT2 fused to Cre recombinase that is induced whole body. Wild type (WT) mice, I-PpoI/+ mice and CreERT2/+ mice are used as negative controls. When tamoxifen (TAM) is administered through the diet, a minimal intranuclear I-PpoI expression is obtained and a consequent enzymatic activity is detected inducing DSBs. Analysis showed that, 12 months after the treatment, no alterations in protein synthesis, mutation frequency, or changes in copy number are present.
Figure 2 – The ICE system with a TAM-inducible I-PpoI endonuclease [Image created with BioRender.com]
Physiological and cognitive aging
The ICE system is applied on ICE mice. After 12 months from the induction, ICE induces signs of aging such as alopecia, pigment loss, reduced body weight, decreased food intake, and lower respiratory exchange ratio. The frailty index of 12-month-old ICE mice matched with 24-month-old WT mice. Additionally, ICE mice exhibit aging in the central nervous system as in the 24-month-old WT mice, having reduced movement in the dark phase, loss of coordination, and impairments in short and long-term memory. Besides, it is detected an increased activation of astrocytes and microglia, leading to a hyper-inflammatory environment. Muscle aging is also evident in ICE mice, with less muscle mass, reduced endurance, and a lower capillary-to-fiber ratio compared to WT mice.
Significantly, there are fewer cytochrome oxidase-positive myofibers and an increase in the transcription of retrotransposons, indicating a loss of silencing at repetitive elements. Eventually, the gene expression in skeletal muscle of 12-month-old ICE mice is compared to the 24-month-old WT showing a similar profile, particularly in Cdkn1a, Myl4, Nlrc5 and Mrpl55 genes that are known to be involved in muscle aging.
Epigenetic aging
The epigenetic clock is evaluated using age-associated CpGs, identified with RRBS. The correlation between epigenetic age and biological age in ICE mice shows that this group ages up to 50% faster than the control group. It is now clear that some histone modifications are related to advancing aging, like a general decrease of H3K27 and H3K56 acetylation; in fact, lower levels of these modifications are found in ICE mice. Particularly, the erosion of H3K27ac landscape is interesting and previously observed in aging; as also shown by the ATAC seq, after I-PpoI treatment, regions with higher accessibility lost this modification, whereas regions with lower accessibility gained it.
Gene ontology (GO) analysis, performed on genes with significant increase in H3K27ac, showed that most of the processes are involved in development. Indeed, Hoxa genes groups present significant alterations of H3K27ac, H3K56ac, H3K27me3 and in mRNA levels. To rule out that the relocalization of chromatin modifiers is specifically due to local I-PpoI cuts, another endonuclease (I-SceI) is tested on MEF cells. This experiment demonstrated that the effect of DSBs on Hoxa expression does not depend on where the DNA breaks occur. Furthermore, new H3K27ac-associated chromatin contacts were observed with active enhancers in adjacent TADs, proving for the first time that faithful DNA repair can impair spatial chromatin contacts, function and insulation.
GO analysis reveals that ICE cells experience a loss of cellular identity. Fewer levels of repressory histone modifications are present in genes involved in neuronal development, which force ICE cells to ex-differentiate and shifting toward a neuronal lineage. In the skeletal muscle of ICE mice, ChIP-seq demonstrates that super-enhancer regions of immune cells are enriched, and muscle cells start to express MHC groups II genes and Nfkbid gene. Hence, due to DSBs, cells tend to re-differentiate and they contribute to generate an inflammatory state which is consistent with premature aging.
OSK-mediated rejuvenation
The cyclic expression of Yamanaka factors (Oct4, Sox2, Klf4, and Myc, or OSKM) suggests that cells possess a backup copy of youthful epigenetic information, which can restore cellular identity without inducing pluripotency [3]. Indeed, after whole-body OSK induction via adeno-associated virus (AAV) in post-treated ICE mice, the epigenetic age of OSK-treated ICE mice reverses up to 57%. A rejuvenation of aging markers in the kidney and muscle can also be observed together to a rescue of age-associated mRNA changes (such as Hmgb, Chaf1, Hoxa, and H3/H4 genes). Specifically, levels of H3K9me3 in the kidney and H3K36me2 in the muscle are restored to a young-like state.
Conclusion and future perspectives
Even though the ICE system proves to be a valid method to study aging in mammals, it is worth mentioning that the article was questioned on Cells about I-PpoI genotoxicity and cell elimination, the use of proper negative controls and the functional rejuvenation by OSK [4]. However, Yang et al. provided a response and carried out the corrections, adding further informations that are all present in the final article on Cells [5].
The non-mutagenic repair of DSBs induces an erosion of the epigenetic landscape, which is consistent with the RCM theory; this could explain why ageing proceeds through a predictable series of molecular and physiological changes, even though DNA damage can occur anywhere in the genome. The fact that lifespan in mammals is correlated with DSBs, and not with other types of damage, perhaps indicates that only severe threats to cell survival sufficiently disrupt the epigenome to cause ageing. In fact, RCM may have evolved to allow cells to survive despite the abundance of DSBs that occur during rapid DNA replication in microorganisms and embryos, promoting the survival in young cells but having a negative long-term effect causing aging. How the epigenetic clock is affected by DSBs is yet unknown, but one possibility is that DSBs cause the relocalization of TET enzymes and DNA methyltransferases. Particularly, other histone modifications need to be analyzed in the future; consistent with H3 modifications discussed in this paper, other studies show that the shortage of H3 in the chromatin structure provides more chromatin accessibility and increase RNA Polymerase II elongation speed, which is positively correlated with aging [6]. DSBs can trigger a sequence of signals to initiate DNA repair, creating new spatial chromatin contacts; thus, the epigenome reorganization can be the result of these signals spreading through chromatin. This can partially explain why even relatively few DSBs can cause a rearrangement of the entire epigenome, but this topic will need further investigations. It is important to emphasize how ICE can be used to faster the epigenetic age in mammals so that they resemble aged tissues, providing a new time-effective tool to study human aging.
References
- Yang, J. H., Hayano, M., Griffin, P. T., Amorim, J. A., Bonkowski, M. S., Apostolides, J. K., … & Sinclair, D. A. (2023). Loss of epigenetic information as a cause of mammalian aging. Cell, 186(2), 305-326.
- Lu, Y. R., Tian, X., & Sinclair, D. A. (2023). The information theory of aging. Nature Aging, 3(12), 1486-1499.
- Ocampo, A., Reddy, P., Martinez-Redondo, P., Platero-Luengo, A., Hatanaka, F., Hishida, T., … & Belmonte, J. C. I. (2016). In vivo amelioration of age-associated hallmarks by partial reprogramming. Cell, 167(7), 1719-1733.
- Timmons, J. A., & Brenner, C. (2024). The information theory of aging has not been tested. Cell, 187(5), 1101-1102.
- Yang, J. H., Hayano, M., Rajman, L. A., & Sinclair, D. A. (2024). Response to: The information theory of aging has not been tested. Cell, 187(5), 1103-1105.
- Debès, C., Papadakis, A., Grönke, S., Karalay, Ö., Tain, L. S., Mizi, A., … & Beyer, A. (2023). Ageing-associated changes in transcriptional elongation influence longevity. Nature, 616(7958), 814-821.