The basis of life depends upon the complex interplay of information stored in the epigenome, genome, and cellular machinery. This is thought to be the software and biological hardware. However, whether a breakdown in the software or hardware causes aging is not yet known. In the 1950s, Szilard and Medawar independently proposed that aging results from the loss of genetic information caused by DNA damage. The double-stranded DNA break (DSB) is most commonly linked to aging. However, in recent years, the assumption that mutations play an important role in aging has been questioned.
Study: Loss of epigenetic information as a cause of mammalian aging Image Credit: picmedical / Shutterstock
Chromatin structures and transcriptional networks are known to specify cell identity during development which directs cells into metaphorical valleys in the Waddington landscape. Cells must retain their identity through the preservation of epigenetic information and a state of low Shannon entropy for the maintenance of optimal function. Yeast studies in the 1990s have reported that a loss of epigenetic information compared to genetics can cause aging. Few other studies also confirmed that epigenetic changes are not just a biomarker but a cause of aging in yeasts.
Epigenetic changes associated with aging include changes in DNA methylation (DNAme) patterns, H3K27me3, H3K9me3, and H3K9me3. Many epigenetic changes have been observed to follow a specific pattern. However, the reason for changes in the mammalian epigenome is not yet known. A few clues can be obtained from yeast, where DSB is a significant factor whose repair requires epigenetic regulators Esa1, Gcn5, Rpd3, Hst1, and Sir2. As per the ‘‘RCM’’ hypothesis and ‘Information Theory of Aging’’, aging in eukaryotes occurs due to the loss of epigenetic information and transcriptional networks in response to cellular damage such as a crash injury or a DSB.
A new study in the journal Cell aimed to determine whether epigenetic changes are a cause of mammalian aging.
About the study
The study involved a system that comprised the fusion of the I-PpoI (an endonuclease from Physarum polycephalum) gene to the C terminus of tamoxifen (TAM)-regulated mutant estrogen receptor domain gene (ERT2), a TAM-regulated Cre recombinase gene (Cre-ERT2) upstream of a ubiquitin promoter, and a transcriptional loxP-STOP-loxP cassette. Transgenic mice with heterozygous Cre-ERT2 and ERT2-I-PpoI were termed as inducible changes to the epigenome or ICE mice, while Cre, IPpoI, and Wild type (WT) were negative controls.
Western blot analysis was carried out using mouse embryonic fibroblast (MEF) cells and tissue samples, while southern blot was carried out using genomic DNA. Surveyor assay was carried out by amplifying I-PpoI target regions from genomic DNA followed by metabolic labeling of MEFs. After that, protein synthesis and DSB were quantified. Respiration exchange ratio (RER), carbon dioxide production (VCO2), oxygen consumption (VO2), ambulatory activity, and food consumption were measured using indirect calorimetry.
Monochrome multiplex quantitative PCR followed by the assessment of the frailty index (FI) was carried out. Lens opacity scoring was performed along with micro CT scanning, quantification of optic nerve axons, quantification of subepidermal thickness, immunohistochemistry for mouse skin, and brain immunohistochemistry, as well as the measurement of mtDNA and ATP. Thereafter, several tests were carried out, including the contextual fear conditioning test, Barnes maze test, treadmill test, grip strength test, and the measurement of lactate and ambulatory activity.
Measurement of gait patterns took place using forced walking on a treadmill. Capillary density, cytochrome oxidase (COX) staining, and electron microscopy were carried out using isolated muscles from mice. Quantification of podocyte density occurred, followed by an analysis of the glomerular injury and parietal epithelial cell to mesenchymal transition. Following this, 5-Ethynyl-2’-deoxyuridine (EdU) staining, imaging, and microscopy for kidney, immunocytochemistry, and senescence-associated b-galactosidase (SA-b-Gal) assay were utilized. Small molecule-driven neuronal reprogramming was performed, followed by quantitative real-time PCR for transcription of repetitive elements and determination of mutation frequency of 28SrDNA.
Furthermore, the production and transduction of adeno-associated viruses took place, followed by RNA-seq and retinal ganglion cells (RGC) sorting. Chromatin immunoprecipitation followed by DNA sequencing (ChIP-seq), Hi-C, HiChIP assay, and whole-genome sequencing was also performed. In addition, epigenetic clocks were analyzed for fibroblasts, muscles, and blood. Finally, histone mass spectrometry was performed.
The results indicated that HA-I-PpoI was detected in nuclei of ICE cells following the addition of TAM but not in control cells. The number of serine-139-phosphorylated H2AX (gH2AX) foci which is a marker of DSB, was observed to reach a 4-fold background in 24 hours in ICE cells. No changes in cell-cycle profile, senescence, apoptosis, mutation frequency, overall translation efficiency, or RNA levels were observed during and after I-PpoI induction. ICE cells were reported to be about 1.5-fold older as compared to Cre control cells.
After post-treatment, ICE cells were more vulnerable to DNA-damaging agents than controls. There was also an increase in indicators of cellular senescence in post-treated cells. The ICE system was observed to induce specific cuts in vivo but did not result in mutations or immediate deleterious effects. Additionally, 10 months post I-PpoI induction, ICE mice were observed to show signs of aging. Also, signs of skin, brain, muscle, and kidney aging were quite apparent in 10-month post-treated ICE mice.
Furthermore, the epigenetic aging rate was reported to be about 50 percent faster than in Cre control mice. Higher amounts of H3K122ac while lower amounts of H3K56ac and H3K27ac were observed in treated ICE cells. Moreover, the effects of DSB on the expression of homeobox (Hox) developmental transcription factor genes were observed not to depend on the location of the DNA break. Additionally, DSB breaks altered spatial chromatin contacts.
Post-treated ICE fibroblasts were reported to lose their ability to maintain their cellular identity and differentiate into other cell types. Regions with lower H3K27ac in post-treated ICE mice were enriched for muscle tissue signatures, while regions with higher H3K27ac were enriched for immune cell enhancers. Finally, Yamanaka factors Oct4, Sox2, Klf4, and Myc (OSKM) were expressed to reverse age-associated changes in post-treated ICE mice.
Therefore, the current study demonstrates that DSB results in a loss of epigenetic information, further accelerating aging and age-related diseases. Mammals, however, seem to possess a backup copy of youthful epigenetic information that can help restore old tissue functions.
The study has certain limitations. First, the study did not determine which chromatin factors were relocalized. Second, it did not involve analysis of in vivo chromatin contacts. Third, epigenomic analyses were not performed at the single-cell level. Fourth, induction of ICE did not take place in a tissue-specific manner. Fifth, some of the effects in the study might be due to the cutting of the rDNA locus.