„Social freezing“ – die männliche Seite Translated title: Social freezing – the male perspective

DNA methylation plays an important role in development of disease and the process of aging. In this study we examine DNA methylation at 476,366 sites throughout the genome of white blood cells from a population cohort (N = 421) ranging in age from 14 to 94 years old. Age affects DNA methylation at almost one third (29%) of the sites (Bonferroni adjusted P-value <0.05), of which 60.5% becomes hypomethylated and 39.5% hypermethylated with increasing age. DNA methylation sites that are located within CpG islands (CGIs) more often become hypermethylated compared to sites outside an island. CpG sites in promoters are more unaffected by age, whereas sites in enhancers more often becomes hypo- or hypermethylated. Hypermethylated sites are overrepresented among genes that are involved in DNA binding, transcription regulation, processes of anatomical structure and developmental process and cortex neuron differentiation (P-value down to P = 9.14*10−67). By contrast, hypomethylated sites are not strongly overrepresented among any biological function or process. Our results indicate that the 23% of the variation in DNA methylation is attributed chronological age, and that hypermethylation is more site-specific than hypomethylation. It appears that the change in DNA methylation partly overlap with regions that change histone modifications with age, indicating an interaction between the two major epigenetic mechanisms. Epigenetic modifications and change in gene expression over time most likely reflects the natural process of aging and variation between individuals might contribute to the development of age-related phenotypes and diseases such as type II diabetes, autoimmune and cardiovascular disease. Show more

Introduction The effects of advanced paternal age have only recently become of interest to the scientific community as a whole. This interest has likely arisen as a result of recent studies that suggest an association with increased incidence of diseases and abnormalities in the offspring of older fathers. Specifically, offspring sired by older fathers have been shown to have increased incidence of neuropsychiatric disorders (autism, bipolar disorder, schizophrenia, etc.) [1]–[3], trinucleotide repeat associated diseases (myotonic dystrophy, spinocerebellar atixia, Huntington's disease, etc.) [4]–[7], as well as some forms of cancer [8]–[11]. Though these are intriguing data, we know very little about the etiology of the increased frequency of diseases in the offspring of older fathers. Among the most likely contributing factors to this phenomenon are epigenetic alterations in the sperm that can be passed on to the offspring. These studies are in striking contrast to the previously held dogma that the mature sperm are responsible only for the safe delivery of the paternal DNA. Intriguingly, with increased investigation has come mounting evidence that the sperm epigenome is not only well suited to facilitate mature gamete function but is also competent to contribute to events in embryonic development. It has been established that even through the dramatic nuclear protein remodeling that occurs in the developing sperm, involving the replacement of histone proteins with protamines, some nucleosomes are retained [12]. Importantly, histones are retained at promoters of important genomic loci for development, suggesting that the sperm epigenome is poised to play a role in embryogenesis [12]. In addition, recent reports suggest that hypomethylated regions with high CpG density also appear to drive nucleosome retention [13]. Similarly, DNA methylation marks in the sperm have been identified that likely contribute to embryonic development as well [12], [14]. These data strongly support the hypothesis that the sperm epigenome is not only well suited to facilitate mature sperm function, but that it also contributes to events beyond fertilization. Looking past fertilization and embryogenesis, sperm appear to contribute to events manifesting later in life. The remarkable claim that sperm, independent of gene mutation, may be capable of affecting phenotype in the offspring was initially proposed as a result of large retrospective epidemiological studies observing changes in the frequency of diseases in the offspring of fathers who were exposed to famine conditions in the early 19th century [15], [16]. Recently, many studies utilizing animal models have discovered similar patterns that comport with the epidemiological data. Specifically, in male animals fed a low protein diet, offspring display altered cholesterol metabolism in hepatic tissue [17]. However, the etiology of this phenomenon is poorly understood. Despite this, there are multiple likely candidates that may drive these effects, such as DNA methylation. Methylation marks at cytosine residues, typically found at cytosine phosphate guanine dinucleotides (CpGs), in the DNA are capable of regulatory control over gene activation or silencing. These roles are dependent on location relative to gene architecture (promoter, exon, intron, etc.). Since these heritable marks are capable of driving changes that may affect phenotype, they represent a possible mechanism to explain the increased disease susceptibility in the offspring of older fathers. Additionally, in both sexes, aging alters DNA methylation marks in most somatic tissues throughout the body. In one of the first large studies to address the question of age-associated methylation alterations, Christensen et al. identified over 300 different CpG loci with age-associated methylation alterations in many tissues [18]. One recent study compared age-associated DNA methylation alterations in blood, brain, kidney and muscle tissue and identified both common and unique methylation alterations between different tissues [19]. Additionally, recent work suggests that DNA methylation can be used to predict the age of an organism based on tissue methylation profiles [20]. This study also supports previous reports which identify global hypomethylation as a hallmark of aging in most somatic tissues [21]. Because of its prevalence in other cell types, age-associated DNA methylation alteration is likely to occur in sperm as well. In further support of this idea is work demonstrating that frequently dividing cells typically have more striking methylation changes associated with age than do cells which divide less often [22]. In this study we have analyzed the age associated sperm DNA methylation alterations that are common among the individuals in our study population to determine the magnitude of sperm DNA methylation changes over time and whether specific regions are consistently altered with age. Results Our study includes 17 sperm donors (of known fertility) that collected an ejaculate in the 1990's. These donors were asked to provide an additional semen sample in 2008, enabling the evaluation of intra-individual changes to sperm DNA methylation with age. These samples are referred to as young (1990's collection) and aged (2008 collection) respectively. The age difference between each collection varied between 9 and 19 years, and the age at first collection (“young” sample) was between 23 and 56 years of age. Table 1 describes the donor demographics within both categories. 10.1371/journal.pgen.1004458.t001 Table 1 Donor demographics. Parameter Young (±SEM) Aged (±SEM) Significance Age 37.7 (±2.12) 50.3 (±2.1) N/A Volume 3.78 (±0.46) 2.85 (±0.45) p = 0.0142 Million/ml 125.4 (±9.16) 145.56 (±15.57) P>0.05 Total count 434.32 (±53.67) 424.67 (±88.69) P>0.05 Total motile 63.38 (±1.64) 61.25 (±4.34) P>0.05 % live 69.08 (±1.47) 61.0 (±3.93) P>0.05 Global methylation analysis To assess global methylation in the samples in question we performed pyrosequencing analysis of long interspersed elements (LINE), a commonly used tool for the analysis of global methylation in many tissues [23], [24]. We identified significant global hypermethylation with age in sperm DNA as previous data from our lab suggests (Figure 1) [25]. Specifically, there was significant hypermethylation with age based on a paired analysis (p = 0.028) or by stratifying the samples by age alone and performing linear regression analysis (p = 0.0062). 10.1371/journal.pgen.1004458.g001 Figure 1 Pyrosequencing results for the LINE-1 global methylation assay. (A) The box plot depicts significantly increased average global methylation with age based on a non-paired t-test of all samples ≤45 (n = 17) years of age vs. all samples >45 (p = 0.001; n = 17). Global methylation was also stratified based only on age at the time of collection for each sample from all 17 donors (a total of 34 samples with each donor represented twice). (B) Linear regression analysis confirmed the significant increases in global sperm DNA methylation with age (p = 0.0062). Array analysis In addition to the global analysis, we performed a high resolution (CpG level) analysis of methylation alterations with age. To perform this we utilized Illumina's Infinium HumanMethylation 450K array. Each sample was hybridized and analyzed on an array and the results were compared to detect changes in methylation that are consistent with age. We utilized a sliding window analysis, coupled with regression analysis (average methylation at identified window versus the age at collection) as an additional filter (any window whose regression p-value was >0.05 was excluded from downstream analysis), to compare changes that are common between paired samples. A Benjamini Hochberg corrected Wilcoxon Signed Rank Test FDR of  = 0.2 (effectively a change in methylation of approximately 10% or greater) was used as our threshold of significance. Raw FDR values have been transformed for visualization in figures and reference in this text ((−10 log10 (q-value FDR)), such that a transformed FDR value of 13 = 0.05, 20 = 0.01, 25 = 0.003, 30 = 0.001, and 40 = 0.0001. With this approach we have identified multiple age-associated intra-individual regional methylation alterations that consistently occur within the same genomic windows in most or all of the donors screened. Specifically, we identified a total of 139 regions that are significantly hypomethylated with age (Log2 ratio ≤−0.2) and 8 regions that are significantly hypermethylated with age (Log2 ratio ≥0.2; Table S1). The average significant window is approximately 887 base pairs in length and contains an average of 5 CpGs with no fewer than 3 in any significant window. Of the 139 hypomethylated regions, 112 are associated with a gene (at either the promoter or the gene body), and of the 8 hypermethylated regions 7 are gene associated. The 8 hypermethylated regions that were found did change in all donor samples, however they did not increase DNA methylation levels beyond 0.1 fraction methylation. In one case we identified 3 significantly hypomethylated windows within a single gene (PTPRN2). Thus there were a total of 110 genes with age-associated hypomethylation. A previous report analyzing multiple somatic tissues suggests that the magnitude of DNA methylation alterations that occurs with age is fairly subtle with an average percent change per year (measured as slope) at a single CpG of approximately 0.05% to 0.15% [19]. Our data, while still subtle, suggest that there may be a stronger effect of age on the methylation alterations in sperm compared with somatic cells. Briefly, in the four tissues screened by Day et al. (blood, brain, kidney and muscle) they identified a total of 8 individual CpGs with a methylation change per year of >0.4% and a single CpG with a yearly change of >0.5%. By comparison, our data have revealed a total of 26 genomic windows (not just individual CpGs) whose average fraction methylation change is >0.4% per year and 13 genomic windows with an average fraction methylation change per year of >0.5% (Figure 2A–B). Specifically in hypermethylated regions, the average fraction methylation change was 0.304% per year (ranging from 0.08% to 0.95% per year). In hypomethylated regions the average fraction methylation change was 0.279% per year (ranging from 0.08% to 0.92% per year). Considering the entire reproductive lifespan of a male, it is not unreasonable to anticipate an average change of 10–12% at these significantly altered regions. Importantly, these alterations all occur in windows with an average initial fraction methylation of  = 45 years of age and 47 samples from individuals 45 years of age is virtually the same, though it is more strongly hypomethylated. In these cases the change is still strikingly significant, but the magnitude of fraction DNA methylation change is minimal. Second, we see a single population in samples collected at 45 years of age. Last, we identified a bimodal distribution in samples collected 45 years of age, is stabilized into a single population (Figure 5). This could be indicative of at least two sperm subpopulations, which are biased to a single, more hypomethylated sperm population with age. In every case the results suggest that all of the alterations we detected with the array are the result of the entire sperm population being altered in similar subtle ways and not a result of a dramatic alteration in a small portion of the sperm population. 10.1371/journal.pgen.1004458.g005 Figure 5 Single molecule analysis reveled 3 distinct alterations that occur with age. (A) DRD4 has only slight alterations associated with age because the young cohort (  = transformed FDR of 40) and an absolute log2 ratio > = 0.2 was used as our threshold for significance. Raw FDR values were transformed for visualization in figures and reference in this text ((−10 log10 (q-value FDR)), such that a transformed FDR value of 13 = 0.05, 20 = 0.01, 25 = 0.003, 30 = 0.001, and 40 = 0.0001, etc. We selected this robust level of significance, as opposed to an FDR of > = 13 (corrected p-value of 0.05), to ensure that we selected only the most striking alterations that are consistently perturbed in most or all of the individuals screened. To confirm the significance of each of the called windows we subjected the mean β-value within the window for each donor (young and aged samples) to a paired t-test. Following this initial filter we additionally subjected each significant window to a regression analysis (age at time of collection versus average methylation within significant windows) to determine the relationship between age and mean methylation within each window. Regression analysis and paired t-tests were performed using STATA 11 software package. A p-value of <0.05 was considered significant for these analyses. Sequencing analysis We performed multiplex sequencing in a replication cohort as a confirmation that the alterations identified in the paired donors via array represent methylation alterations that are common in human sperm with age. First, each donor sample used in the array study was additionally subjected to targeted bisulfite sequencing at loci determined to be most consistently altered based on the window analysis. This step was designed to confirm the array results and to provide greater depth of coverage of the CpGs in the windows of interest. Primers for 21 loci were designed using MethPrimer (Li Lab, UCSF). PCR was performed on samples following sperm DNA bisulfite conversion with EZ-96 DNA Methylation-Gold kit (Zymo Research, Irvine CA). PCR products were purified with QIAquick PCR Purification Kit (Qiagen, Valencia CA) and were pooled for each sample. The pooled products were delivered to the Microarray and Genomic Analysis core facility at the University of Utah for library prep which included shearing of the DNA with a Covaris sonicator to generate products of approximately 300 base pairs, in preparation for 150 bp paired end sequencing, and the addition of sample-specific barcodes for all 34 samples. Multiplex sequencing was then performed on a single lane on the MiSeq platform (Illumina, San Diego CA). Second, 19 sperm DNA samples from an independent, unselected cohort of general population donors who were ≥45 of ages were selected and compared to 47 sperm DNA samples from general population donors who were <25 years of age. These samples underwent the same preparation as described above for multiplex sequencing, though only 15 amplicons were targeted in this study of larger sample size. Average fraction methylation for each window was determined and was subjected to unpaired t tests between the young and aged groups. Single molecule analysis of targeted bisulfite sequencing data Bisulfite sequencing data was aligned against the human reference genome Hg19 using Novoalign. The aligned reads were processed using Novoalign Bisulfite Parser, BisStat and Parse Point Data Context for CG from the USeq package. The binned CpG graphs were generated using a modified version of the Allelic Methylation Detector from the USeq package. In short, all reads were queried for their number of CpGs. A consensus CpG number was then taken based on the highest number of CpGs per read and a minimum of 10% of all aligned reads (approximately 100 reads per region) must cover said number of CpGs. The consensus CpG number then served as the basis for the number of bins per region. Samples that were donated at an age of 45 years or older were coalesced in silico in the “aged donor group”. Conversely, samples younger than 45 years were grouped in the “young donor group”. All reads for the consensus CpG count were summed up based on their age group and then normalized to a 100 reads total. The graphs plotting normalized reads to methylation bins were then generated using the spline function from the R package. GO term/Pathway/disease association analysis GO term Analysis was performed with Gene Ontology Enrichment Analysis and Visualization Tool (GOrilla; cbl-gorilla.cs.technion.ac.il). Pathway and disease association analysis was performed on the Database of Annotation, Visualization, and Integrated Discovery (DAVID; david.abcc.ncifcrf.gov) v6.7. Additional disease association analysis was performed directly on the National Institute of Health's Genetic Association Database (GAD; geneticassociationdb.nih.gov). Additional statistical analyses Fishers exact test was used to determine the differences in frequencies of genes associated with particular diseases between our significant gene group and a background group. This analysis was also used to detect the differences in frequencies of windows that were found in regions of histone retention in the hypomethylation group and the hypermethylation group. Additionally, regression analysis was utilized to determine relationships between age and methylation status at various loci. STATA software package was used to test for significance with a p<0.05 being considered a significant finding. Supporting Information Table S1 Genomic features of significantly altered windows. Represented in this table are the windows of significance that were identified in our study as well as their transformed FDR, log 2 ratio, association to genes, association to known DMR, and CpG Island context. (DOCX) Click here for additional data file. Show more

The accumulation of epigenetic changes was proposed to contribute to the age-related increase in the risk of most common diseases. In this study on 230 monozygotic twin pairs (MZ pairs), aged 18–89 years, we investigated the occurrence of epigenetic changes over the adult lifespan. Using mass spectrometry, we investigated variation in global (LINE1) DNA methylation and in DNA methylation at INS, KCNQ1OT1, IGF2, GNASAS, ABCA1, LEP, and CRH, candidate loci for common diseases. Except for KCNQ1OT1, interindividual variation in locus-specific DNA methylation was larger in old individuals than in young individuals, ranging from 1.2-fold larger at ABCA1 (P = 0.010) to 1.6-fold larger at INS (P = 3.7 × 10−07). Similarly, there was more within-MZ-pair discordance in old as compared with young MZ pairs, except for GNASAS, ranging from an 8% increase in discordance each decade at CRH (P = 8.9 × 10−06) to a 16% increase each decade at LEP (P = 2.0 × 10−08). Still, old MZ pairs with strikingly similar DNA methylation were also observed at these loci. After 10-year follow-up in elderly twins, the variation in DNA methylation showed a similar pattern of change as observed cross-sectionally. The age-related increase in methylation variation was generally attributable to unique environmental factors, except for CRH, for which familial factors may play a more important role. In conclusion, sustained epigenetic differences arise from early adulthood to old age and contribute to an increasing discordance of MZ twins during aging. Show more