Methylome-wide association study of whole blood DNA in the Norfolk Island isolate identifies robust loci associated with age

A methylation panel of 10 CpGs for accurate age inference via stepwise conditional epigenome-wide association study

Estimating individual age from DNA methylation at age associated CpG sites may provide key information facilitating forensic investigations. Systematic marker screening and feature selection play a critical role in ensuring the performance of the final prediction model. In the discovery stage, we screened for 811876 CpGs from whole blood of 2664 Chinese individuals ranging from 18 to 83 years of age based on a stepwise conditional epigenome-wide association study (SCEWAS). The SCEWAS identified 28 CpGs showing genome-wide significant and independent effects. Further restricting this panel to 10 most informative CpGs showed a tolerable loss of information. A linear model consisting of these 10 CpGs could explain 93% of the age variance (R2 = 0.93) in the training set (n = 2664). In an independent test set of Chinese individuals (n = 648), this model also provided highly accurate predictions (R2 = 0.85, mean absolute deviation, MAD = 3.20 years). The model was additionally validated in a public dataset of multiple ancestral origins (86 Europeans, 14 Asians, and 273 Africans) and the prediction accuracy reduced significantly (R2 = 0.85, MAD = 6.21 years), as might be expected due to different genomic backgrounds, sample sizes, and age ranges. Our 10 CpG model also outperformed the recently proposed 9-CpG model constructed in 390 Chinese males (R2 = 0.79 in test set). We also demonstrated that our SCEWAS approach outperformed the traditional EWAS and the elastic net approach in obtaining a small set of most age informative CpGs. Overall, our systematic genome-wide feature selection identified a small panel of 10 CpGs for accurate age estimation with high potential in forensic applications.

Epigenome-Wide Association Study of Asthma Exacerbations in Europeans

BACKGROUND

Asthma exacerbations (AEs) represent the major contributor to the global asthma burden. Although genetic and environmental factors have been associated with AEs, the role of epigenetics remains uncovered.

OBJECTIVE

This study aimed to identify whole blood DNA methylation (DNAm) markers associated with AEs in Europeans.

METHODS

DNAm was assessed in 406 blood samples from Spanish individuals using the Infinium MethylationEPIC microarray (Illumina). An epigenome-wide association study was conducted to test the association of DNAm with AEs at differentially methylated positions, regions, and epigenetic modules. CpGs suggestively associated with AEs (false discovery rate [FDR] < 0.1) were followed up for replication in 222 European individuals, and the genome-wide significance (p < 9 × 10-8) was declared after meta-analyzing the discovery and replication samples. Additional assessment was performed using nasal tissue DNAm data from 155 Spanish individuals. The effects of genetic variation on DNAm were assessed through cis-methylation quantitative trait loci (meQTL) analysis. Enrichment analyses of previous EWAS signals were conducted.

RESULTS

Four CpGs were associated with AEs, and two were replicated and reached genomic significance in the meta-analysis (annotated to ZBTB16 and BAIAP2). Of those, CpG cg25345365 (ZBTB16) was cross-tissue validated in nasal epithelium (p= 0.003) and associated with five independent meQTLs (FDR < 0.05). Additionally, four differentially methylated regions and one module were significantly associated with AEs. Enrichment analyses revealed an overrepresentation of prior epigenetic associations with prenatal and environmental exposures, immune-mediated diseases, and mortality.

CONCLUSIONS

DNAm in whole blood and nasal samples may contribute to AEs in Europeans, capturing genetic and environmental risk factors.

An epigenome-wide study of selenium status and DNA methylation in the Strong Heart Study

BACKGROUND

Selenium (Se) is an essential nutrient linked to adverse health endpoints at low and high levels. The mechanisms behind these relationships remain unclear and there is a need to further understand the epigenetic impacts of Se and their relationship to disease. We investigated the association between urinary Se levels and DNA methylation (DNAm) in the Strong Heart Study (SHS), a prospective study of cardiovascular disease (CVD) among American Indians adults.

METHODS

Selenium concentrations were measured in urine (collected in 1989-1991) using inductively coupled plasma mass spectrometry among 1,357 participants free of CVD and diabetes. DNAm in whole blood was measured cross-sectionally using the Illumina MethylationEPIC BeadChip (850 K) Array. We used epigenome-wide robust linear regressions and elastic net to identify differentially methylated cytosine-guanine dinucleotide (CpG) sites associated with urinary Se levels.

RESULTS

The mean (standard deviation) urinary Se concentration was 51.8 (25.1) μg/g creatinine. Across 788,368 CpG sites, five differentially methylated positions (DMP) (hypermethylated: cg00163554, cg18212762, cg11270656, and hypomethylated: cg25194720, cg00886293) were significantly associated with Se in linear regressions after accounting for multiple comparisons (false discovery rate p-value: 0.10). The top hypermethylated DMP (cg00163554) was annotated to the Disco Interacting Protein 2 Homolog C (DIP2C) gene, which relates to transcription factor binding. Elastic net models selected 425 hypo- and hyper-methylated DMPs associated with urinary Se, including three sites (cg00163554 [DIP2C], cg18212762 [MAP4K2], cg11270656 [GPIHBP1]) identified in linear regressions.

CONCLUSIONS

Urinary Se was associated with minimal changes in DNAm in adults from American Indian communities across the Southwest and the Great Plains in the United States, suggesting that other mechanisms may be driving health impacts. Future analyses should explore other mechanistic biomarkers in human populations, determine these relationships prospectively, and investigate the potential role of differentially methylated sites with disease endpoints.

Epigenetic associations with neonatal age in infants born very preterm, particularly among genes involved in neurodevelopment

The time from conception through the first year of life is the most dynamic period in human development. This time period is particularly important for infants born very preterm (< 30 weeks gestation; VPT), as they experience a significant disruption in the normal developmental trajectories and are at heightened risk of experiencing developmental impairments and delays. Variations in the epigenetic landscape during this period may reflect this disruption and shed light on the interrelationships between aging, maturation, and the epigenome. We evaluated how gestational age (GA) and age since conception in neonates [post-menstrual age (PMA)], were related to DNA methylation in buccal cells collected at NICU discharge from VPT infants (n = 538). After adjusting for confounders and applying Bonferroni correction, we identified 2,366 individual CpGs associated with GA and 14,979 individual CpGs associated with PMA, as well as multiple differentially methylated regions. Pathway enrichment analysis identified pathways involved in axonogenesis and regulation of neuron projection development, among many other growth and developmental pathways (FDR q < 0.001). Our findings align with prior work, and also identify numerous novel associations, suggesting that genes important in growth and development, particularly neurodevelopment, are subject to substantial epigenetic changes during early development among children born VPT.

Omics feature selection with the extended SIS R package: identification of a body mass index epigenetic multimarker in the Strong Heart Study

The statistical analysis of omics data poses a great computational challenge given their ultra-high-dimensional nature and frequent between-features correlation. In this work, we extended the iterative sure independence screening (ISIS) algorithm by pairing ISIS with elastic-net (Enet) and 2 versions of adaptive elastic-net (adaptive elastic-net (AEnet) and multistep adaptive elastic-net (MSAEnet)) to efficiently improve feature selection and effect estimation in omics research. We subsequently used genome-wide human blood DNA methylation data from American Indian participants in the Strong Heart Study (n = 2235 participants; measured in 1989-1991) to compare the performance (predictive accuracy, coefficient estimation, and computational efficiency) of ISIS-paired regularization methods with that of a bayesian shrinkage and traditional linear regression to identify an epigenomic multimarker of body mass index (BMI). ISIS-AEnet outperformed the other methods in prediction. In biological pathway enrichment analysis of genes annotated to BMI-related differentially methylated positions, ISIS-AEnet captured most of the enriched pathways in common for at least 2 of all the evaluated methods. ISIS-AEnet can favor biological discovery because it identifies the most robust biological pathways while achieving an optimal balance between bias and efficient feature selection. In the extended SIS R package, we also implemented ISIS paired with Cox and logistic regression for time-to-event and binary endpoints, respectively, and a bootstrap approach for the estimation of regression coefficients.

Exploring the Functional Basis of Epigenetic Aging in Relation to Body Fat Phenotypes in the Norfolk Island Cohort

DNA methylation is an epigenetic factor that is modifiable and can change over a lifespan. While many studies have identified methylation sites (CpGs) related to aging, the relationship of these to gene function and age-related disease phenotypes remains unclear. This research explores this question by testing for the conjoint association of age-related CpGs with gene expression and the relation of these to body fat phenotypes. The study included blood-based gene transcripts and intragenic CpG methylation data from Illumina 450 K arrays in 74 healthy adults from the Norfolk Island population. First, a series of regression analyses were performed to detect associations between gene transcript level and intragenic CpGs and their conjoint relationship with age. Second, we explored how these age-related expression CpGs (eCpGs) correlated with obesity-related phenotypes, including body fat percentage, body mass index, and waist-to-hip ratio. We identified 35 age-related eCpGs associated with age. Of these, ten eCpGs were associated with at least one body fat phenotype. Collagen Type XI Alpha 2 Chain (COL11A2), Complement C1s (C1s), and four and a half LIM domains 2 (FHL2) genes were among the most significant genes with multiple eCpGs associated with both age and multiple body fat phenotypes. The COL11A2 gene contributes to the correct assembly of the extracellular matrix in maintaining the healthy structural arrangement of various components, with the C1s gene part of complement systems functioning in inflammation. Moreover, FHL2 expression was upregulated under hypermethylation in both blood and adipose tissue with aging. These results suggest new targets for future studies and require further validation to confirm the specific function of these genes on body fat regulation.

Combined Methylation and Transcriptome Analysis of Liver Injury of Nonalcoholic Fatty Liver Disease Induced by High Alcohol-Producing Klebsiella pneumoniae

It has been known that high alcohol-producing Klebsiella pneumoniae (HiAlc Kpn) is one of causative agents of nonalcoholic fatty liver disease (NAFLD). However, how HiAlc Kpn promotes liver injury remains unclear. Recent findings suggest that DNA methylation might associate with the pathogenesis of NAFLD. Herein, the role of DNA methylation in HiAlc Kpn-induced liver injury was investigated. Murine models of NAFLD were established in C57BL/6N wild-type mice by gavaging HiAlc Kpn for 8 weeks. The liver injury was assessed based on the liver histopathology and biochemical indicators. In addition, DNA methylation in hepatic tissue was assessed by using dot bolt of 5-mC. RNA sequencing analysis and whole-genome bisulfite sequencing (WGBS) analysis were also performed. HiAlc Kpn significantly increased the activity of aspartate transaminase (AST), alanine transaminase (ALT), triglycerides (TGs), and glutathione (GSH), while hypomethylation was associated with liver injury in the experimental mice induced by HiAlc Kpn. The GO and KEGG pathway enrichment analysis of the transcriptome revealed that HiAlc Kpn induced fat metabolic disorders and DNA damage. The conjoint analysis of methylome and transcriptome showed that hypomethylation regulated related gene expression in signal pathways of lipid formation and circadian rhythm, including Rorα and Arntl1genes, which may be the dominant cause of NAFLD induced by HiAlc Kpn. Data suggest that DNA hypomethylation might play an important role in liver injury of NAFLD induced by HiAlc Kpn. Which possibly provides a new sight for understanding the mechanisms of NAFLD and selecting the potential therapeutic targets. IMPORTANCE High alcohol-producing Klebsiella pneumoniae (HiAlc Kpn) is one of causative agents of nonalcoholic fatty liver disease (NAFLD) and could induce liver damage. DNA methylation, as a common epigenetic form following contact with an etiologic agent and pathogenesis, can affect chromosome stability and transcription. We conjointly analyzed DNA methylation and transcriptome levels in the established murine models to explore the potential mechanisms for further understanding the role of DNA methylation in the liver damage of HiAlc Kpn-induced NAFLD. The analysis of the DNA methylation landscape contributes to our understanding of the entire disease process, which might be crucial in developing treatment strategies.

Defining Specific Cell States of MPTP-Induced Parkinson's Disease by Single-Nucleus RNA Sequencing

Parkinson's disease (PD) is a neurodegenerative disease with an impairment of movement execution that is related to age and genetic and environmental factors. 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) is a neurotoxin widely used to induce PD models, but the effect of MPTP on the cells and genes of PD has not been fully elucidated. By single-nucleus RNA sequencing, we uncovered the PD-specific cells and revealed the changes in their cellular states, including astrocytosis and endothelial cells' absence, as well as a cluster of medium spiny neuron cells unique to PD. Furthermore, trajectory analysis of astrocyte and endothelial cell populations predicted candidate target gene sets that might be associated with PD. Notably, the detailed regulatory roles of astrocyte-specific transcription factors Dbx2 and Sox13 in PD were revealed in our work. Finally, we characterized the cell-cell communications of PD-specific cells and found that the overall communication strength was enhanced in PD compared with a matched control, especially the signaling pathways of NRXN and NEGR. Our work provides an overview of the changes in cellular states of the MPTP-induced mouse brain.

Lipidomics profiling of biological aging in American Indians: the Strong Heart Family Study

Telomeres shorten with age and shorter leukocyte telomere length (LTL) has been associated with various age-related diseases. Thus, LTL has been considered a biomarker of biological aging. Dyslipidemia is an established risk factor for most age-related metabolic disorders. However, little is known about the relationship between LTL and dyslipidemia. Lipidomics is a new biochemical technique that can simultaneously identify and quantify hundreds to thousands of small molecular lipid species. In a large population comprising 1843 well-characterized American Indians in the Strong Heart Family Study, we examined the lipidomic profile of biological aging assessed by LTL. Briefly, LTL was quantified by qPCR. Fasting plasma lipids were quantified by untargeted liquid chromatography-mass spectrometry. Lipids associated with LTL were identified by elastic net modeling. Of 1542 molecular lipids identified (518 known, 1024 unknown), 174 lipids (36 knowns) were significantly associated with LTL, independent of chronological age, sex, BMI, hypertension, diabetes status, smoking status, bulk HDL-C, and LDL-C. These findings suggest that altered lipid metabolism is associated with biological aging and provide novel insights that may enhance our understanding of the relationship between dyslipidemia, biological aging, and age-related diseases in American Indians.

Counteracting aged DNA methylation states to combat ageing and age-related diseases

DNA methylation (DNAm) overwrites information about multiple extrinsic factors on the genome. Age is one of these factors. Age causes characteristic DNAm changes that are thought to be not only major drivers of normal ageing but also precursors to diseases, cancer being one of these. Although there is still much to learn about the relationship between ageing, age-related diseases and DNAm, we now know how to interpret some of the effects caused by age in the form of changes in methylation marks at specific loci. In fact, these changes form the basis of the so called "epigenetic clocks", which translate the genomic methylation profile into an "epigenetic age". Epigenetic age does not only estimate chronological age but can also predict the risk of chronic diseases and mortality. Epigenetic age is believed to be one of the most accurate metrics of biological age. Initial evidence has recently been gathered pointing to the possibility that the rate of epigenetic ageing can be slowed down or even reversed. In this review, we discuss some of the most relevant advances in this field. Expected outcome is that this approach can provide insights into how to preserve health and reduce the impact of ageing diseases in humans.

An epigenome-wide study of DNA methylation profiles and lung function among American Indians in the Strong Heart Study

BACKGROUND

Epigenetic modifications, including DNA methylation (DNAm), are often related to environmental exposures, and are increasingly recognized as key processes in the pathogenesis of chronic lung disease. American Indian communities have a high burden of lung disease compared to the national average. The objective of this study was to investigate the association of DNAm and lung function in the Strong Heart Study (SHS). We conducted a cross-sectional study of American Indian adults, 45-74 years of age who participated in the SHS. DNAm was measured using the Illumina Infinium Human MethylationEPIC platform at baseline (1989-1991). Lung function was measured via spirometry, including forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC), at visit 2 (1993-1995). Airflow limitation was defined as FEV1 < 70% predicted and FEV1/FVC < 0.7, restriction was defined as FEV1/FVC > 0.7 and FVC < 80% predicted, and normal spirometry was defined as FEV1/FVC > 0.7, FEV1 > 70% predicted, FVC > 80% predicted. We used elastic-net models to select relevant CpGs for lung function and spirometry-defined lung disease. We also conducted bioinformatic analyses to evaluate the biological plausibility of the findings.

RESULTS

Among 1677 participants, 21.2% had spirometry-defined airflow limitation and 13.6% had spirometry-defined restrictive pattern lung function. Elastic-net models selected 1118 Differentially Methylated Positions (DMPs) as predictors of airflow limitation and 1385 for restrictive pattern lung function. A total of 12 DMPs overlapped between airflow limitation and restrictive pattern. EGFR, MAPK1 and PRPF8 genes were the most connected nodes in the protein-protein interaction network. Many of the DMPs targeted genes with biological roles related to lung function such as protein kinases.

CONCLUSION

We found multiple differentially methylated CpG sites associated with chronic lung disease. These signals could contribute to better understand molecular mechanisms involved in lung disease, as assessed systemically, as well as to identify patterns that could be useful for diagnostic purposes. Further experimental and longitudinal studies are needed to assess whether DNA methylation has a causal role in lung disease.

Genetic loci and metabolic states associated with murine epigenetic aging

Changes in DNA methylation (DNAm) are linked to aging. Here, we profile highly conserved CpGs in 339 predominantly female mice belonging to the BXD family for which we have deep longevity and genomic data. We use a 'pan-mammalian' microarray that provides a common platform for assaying the methylome across mammalian clades. We computed epigenetic clocks and tested associations with DNAm entropy, diet, weight, metabolic traits, and genetic variation. We describe the multifactorial variance of methylation at these CpGs and show that high-fat diet augments the age-related changes. Entropy increases with age. The progression to disorder, particularly at CpGs that gain methylation over time, was predictive of genotype-dependent life expectancy. The longer-lived BXD strains had comparatively lower entropy at a given age. We identified two genetic loci that modulate epigenetic age acceleration (EAA): one on chromosome (Chr) 11 that encompasses the Erbb2/Her2 oncogenic region, and the other on Chr19 that contains a cytochrome P450 cluster. Both loci harbor genes associated with EAA in humans, including STXBP4, NKX2-3, and CUTC. Transcriptome and proteome analyses revealed correlations with oxidation-reduction, metabolic, and immune response pathways. Our results highlight concordant loci for EAA in humans and mice, and demonstrate a tight coupling between the metabolic state and epigenetic aging.

Multi-Omics Integration in a Twin Cohort and Predictive Modeling of Blood Pressure Values

Abnormal blood pressure is strongly associated with risk of high-prevalence diseases, making the study of blood pressure a major public health challenge. Although biological mechanisms underlying hypertension at the single omic level have been discovered, multi-omics integrative analyses using continuous variations in blood pressure values remain limited. We used a multi-omics regression-based method, called sparse multi-block partial least square, for integrative, explanatory, and predictive interests in study of systolic and diastolic blood pressure values. Various datasets were obtained from the Finnish Twin Cohort for up to 444 twins. Blocks of omics-including transcriptomic, methylation, metabolomic-data as well as polygenic risk scores and clinical data were integrated into the modeling and supported by cross-validation. The predictive contribution of each omics block when predicting blood pressure values was investigated using external participants from the Young Finns Study. In addition to revealing interesting inter-omics associations, we found that each block of omics heterogeneously improved the predictions of blood pressure values once the multi-omics data were integrated. The modeling revealed a plurality of clinical, transcriptomic, and metabolomic factors consistent with the literature and that play a leading role in explaining unit variations in blood pressure. These findings demonstrate (1) the robustness of our integrative method to harness results obtained by single omics discriminant analyses, and (2) the added value of predictive and exploratory gains of a multi-omics approach in studies of complex phenotypes such as blood pressure.

Capturing SNP Association across the NK Receptor and HLA Gene Regions in Multiple Sclerosis by Targeted Penalised Regression Models

Conventional genome-wide association studies (GWASs) of complex traits, such as Multiple Sclerosis (MS), are reliant on per-SNP p-values and are therefore heavily burdened by multiple testing correction. Thus, in order to detect more subtle alterations, ever increasing sample sizes are required, while ignoring potentially valuable information that is readily available in existing datasets. To overcome this, we used penalised regression incorporating elastic net with a stability selection method by iterative subsampling to detect the potential interaction of loci with MS risk. Through re-analysis of the ANZgene dataset (1617 cases and 1988 controls) and an IMSGC dataset as a replication cohort (1313 cases and 1458 controls), we identified new association signals for MS predisposition, including SNPs above and below conventional significance thresholds while targeting two natural killer receptor loci and the well-established HLA loci. For example, rs2844482 (98.1% iterations), otherwise ignored by conventional statistics (p = 0.673) in the same dataset, was independently strongly associated with MS in another GWAS that required more than 40 times the number of cases (~45 K). Further comparison of our hits to those present in a large-scale meta-analysis, confirmed that the majority of SNPs identified by the elastic net model reached conventional statistical GWAS thresholds (p < 5 × 10−8) in this much larger dataset. Moreover, we found that gene variants involved in oxidative stress, in addition to innate immunity, were associated with MS. Overall, this study highlights the benefit of using more advanced statistical methods to (re-)analyse subtle genetic variation among loci that have a biological basis for their contribution to disease risk.

Combining current knowledge on DNA methylation-based age estimation towards the development of a superior forensic DNA intelligence tool

The estimation of chronological age from biological fluids has been an important quest for forensic scientists worldwide, with recent approaches exploiting the variability of DNA methylation patterns with age in order to develop the next generation of forensic 'DNA intelligence' tools for this application. Drawing from the conclusions of previous work utilising massively parallel sequencing (MPS) for this analysis, this work introduces a DNA methylation-based age estimation method for blood that exhibits the best combination of prediction accuracy and sensitivity reported to date. Statistical evaluation of markers from 51 studies using microarray data from over 4000 individuals, followed by validation using in-house generated MPS data, revealed a final set of 11 markers with the greatest potential for accurate age estimation from minimal DNA material. Utilising an algorithm based on support vector machines, the proposed model achieved an average error (MAE) of 3.3 years, with this level of accuracy retained down to 5 ng of starting DNA input (~ 1 ng PCR input). The accuracy of the model was retained (MAE = 3.8 years) in a separate test set of 88 samples of Spanish origin, while predictions for donors of greater forensic interest (< 55 years of age) displayed even higher accuracy (MAE = 2.6 years). Finally, no sex-related bias was observed for this model, while there were also no signs of variation observed between control and disease-associated populations for schizophrenia, rheumatoid arthritis, frontal temporal dementia and progressive supranuclear palsy in microarray data relating to the 11 markers.

Multi-phenotype genome-wide association studies of the Norfolk Island isolate implicate pleiotropic loci involved in chronic kidney disease

Chronic kidney disease (CKD) is a persistent impairment of kidney function. Genome-wide association studies (GWAS) have revealed multiple genetic loci associated with CKD susceptibility but the complete genetic basis is not yet clear. Since CKD shares risk factors with cardiovascular diseases and diabetes, there may be pleiotropic loci at play but may go undetected when using single phenotype GWAS. Here, we used multi-phenotype GWAS in the Norfolk Island isolate (n = 380) to identify new loci associated with CKD. We performed a principal components analysis on different combinations of 29 quantitative traits to extract principal components (PCs) representative of multiple correlated phenotypes. GWAS of a PC derived from glomerular filtration rate, serum creatinine, and serum urea identified a suggestive peak (pmin = 1.67 × 10-7) that mapped to KCNIP4. Inclusion of other secondary CKD measurements with these three kidney function traits identified the KCNIP4 locus with GWAS significance (pmin = 1.59 × 10-9). Finally, we identified a group of two SNPs with increased minor allele frequencies as potential functional variants. With the use of genetic isolate and the PCA-based multi-phenotype GWAS approach, we have revealed a potential pleotropic effect locus for CKD. Further studies are required to assess functional relevance of this locus.

DNA methylation and histone variants in aging and cancer

Aging-related diseases such as cancer can be traced to the accumulation of molecular disorder including increased DNA mutations and epigenetic drift. We provide a comprehensive review of recent results in mice and humans on modifications of DNA methylation and histone variants during aging and in cancer. Accumulated errors in DNA methylation maintenance lead to global decreases in DNA methylation with relaxed repression of repeated DNA and focal hypermethylation blocking the expression of tumor suppressor genes. Epigenetic clocks based on quantifying levels of DNA methylation at specific genomic sites is proving to be a valuable metric for estimating the biological age of individuals. Histone variants have specialized functions in transcriptional regulation and genome stability. Their concentration tends to increase in aged post-mitotic chromatin, but their effects in cancer are mainly determined by their specialized functions. Our increased understanding of epigenetic regulation and their modifications during aging has motivated interventions to delay or reverse epigenetic modifications using the epigenetic clocks as a rapid readout for efficacity. Similarly, the knowledge of epigenetic modifications in cancer is suggesting new approaches to target these modifications for cancer therapy.

Ageing-associated changes in DNA methylation in X and Y chromosomes

Background

Ageing displays clear sexual dimorphism, evident in both morbidity and mortality. Ageing is also associated with changes in DNA methylation, but very little focus has been on the sex chromosomes, potential biological contributors to the observed sexual dimorphism. Here, we sought to identify DNA methylation changes associated with ageing in the Y and X chromosomes, by utilizing datasets available in data repositories, comprising in total of 1240 males and 1191 females, aged 14–92 years.

Results

In total, we identified 46 age-associated CpG sites in the male Y, 1327 age-associated CpG sites in the male X, and 325 age-associated CpG sites in the female X. The X chromosomal age-associated CpGs showed significant overlap between females and males, with 122 CpGs identified as age-associated in both sexes. Age-associated X chromosomal CpGs in both sexes were enriched in CpG islands and depleted from gene bodies and showed no strong trend towards hypermethylation nor hypomethylation. In contrast, the Y chromosomal age-associated CpGs were enriched in gene bodies, and showed a clear trend towards hypermethylation with age.

Conclusions

Significant overlap in X chromosomal age-associated CpGs identified in males and females and their shared features suggest that despite the uneven chromosomal dosage, differences in ageing-associated DNA methylation changes in the X chromosome are unlikely to be a major contributor of sex dimorphism in ageing. While age-associated CpGs showed good replication across datasets in the present study, only a limited set of previously reported age-associated CpGs were replicated. One contributor to the limited overlap are differences in the age range of individuals included in each data set. Further study is needed to identify biologically significant age-associated CpGs in the sex chromosomes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13072-021-00407-6.

A genome-wide methylation study of body fat traits in the Norfolk Island isolate

BACKGROUND AND AIMS

Natural variation in body fat is explained by both genetic and environmental effects. Epigenetic mechanisms such as DNA methylation can mediate these effects causing changes in gene expression leading to onset of obesity. Studies of genetic isolates have the potential to provide new epigenetic insights with advantages such as reduced genetic diversity and environmental exposures.

METHODS AND RESULTS

This was an exploratory study of genome-wide DNA methylation in relation to body fat traits in 47 healthy adults from the genetic isolate of Norfolk Island. Quantitative body fat traits (body fat percentage, body mass index, hip circumference, waist circumference, waist-hip-ratio and weight) were carefully measured. DNA methylation data was obtained from peripheral blood using Illumina 450K arrays. Multi-trait analysis was performed using Principal Component Analysis (PCA). CpG by trait association testing was performed using stepwise linear regressions. Two components were identified that explained approximately 89% of the phenotypic variance. In total, 5 differential methylated positions (DMPs) were identified at genome-wide significance (P≤ 2.4 × 10^-7), which mapped to GOT2-CDH8, LYSMD3, HIBADH, ADGRD1 and EBF4 genes. Gene set enrichment analysis of 848 genes containing suggestive DMPs (P≤ 1.0 × 10^-4) implicated the Cadherin (28 genes, P_adj = 6.76 × 10^-7) and Wnt signaling pathways (38 genes, P_adj = 7.78 × 10^-6).

CONCLUSION

This study provides new insights into the epigenetically influenced genes and pathways underlying body fat variation in a healthy cohort and provides targets for consideration in future studies of obesity risk.

Instability of Variable-selection Algorithms Used to Identify True Predictors of an Outcome in Intermediate-dimension Epidemiologic Studies

BACKGROUND

Machine-learning algorithms are increasingly used in epidemiology to identify true predictors of a health outcome when many potential predictors are measured. However, these algorithms can provide different outputs when repeatedly applied to the same dataset, which can compromise research reproducibility. We aimed to illustrate that commonly used algorithms are unstable and, using the example of Least Absolute Shrinkage and Selection Operator (LASSO), that stabilization method choice is crucial.

METHODS

In a simulation study, we tested the stability and performance of widely used machine-learning algorithms (LASSO, Elastic-Net, and Deletion-Substitution-Addition [DSA]). We then assessed the effectiveness of six methods to stabilize LASSO and their impact on performance. We assumed that a linear combination of factors drawn from a simulated set of 173 quantitative variables assessed in 1,301 subjects influenced to varying extents a continuous health outcome. We assessed model stability, sensitivity, and false discovery proportion.

RESULTS

All tested algorithms were unstable. For LASSO, stabilization methods improved stability without ensuring perfect stability, a finding confirmed by application to an exposome study. Stabilization methods also affected performance. Specifically, stabilization based on hyperparameter optimization, frequently implemented in epidemiology, increased the false discovery proportion dramatically when predictors explained a low share of outcome variability. In contrast, stabilization based on stability selection procedure often decreased the false discovery proportion, while sometimes simultaneously lowering sensitivity.

CONCLUSIONS

Machine-learning methods instability should concern epidemiologists relying on them for variable selection, as stabilizing a model can impact its performance. For LASSO, stabilization methods based on stability selection procedure (rather than addressing prediction stability) should be preferred to identify true predictors.

DNA methylation and cancer incidence: lymphatic-hematopoietic versus solid cancers in the Strong Heart Study

BACKGROUND

Epigenetic alterations may contribute to early detection of cancer. We evaluated the association of blood DNA methylation with lymphatic-hematopoietic cancers and, for comparison, with solid cancers. We also evaluated the predictive ability of DNA methylation for lymphatic-hematopoietic cancers.

METHODS

Blood DNA methylation was measured using the Illumina Infinium methylationEPIC array in 2324 Strong Heart Study participants (41.4% men, mean age 56 years). 788,368 CpG sites were available for differential DNA methylation analysis for lymphatic-hematopoietic, solid and overall cancers using elastic-net and Cox regression models. We conducted replication in an independent population: the Framingham Heart Study. We also analyzed differential variability and conducted bioinformatic analyses to assess for potential biological mechanisms.

RESULTS

Over a follow-up of up to 28 years (mean 15), we identified 41 lymphatic-hematopoietic and 394 solid cancer cases. A total of 126 CpGs for lymphatic-hematopoietic cancers, 396 for solid cancers, and 414 for overall cancers were selected as predictors by the elastic-net model. For lymphatic-hematopoietic cancers, the predictive ability (C index) increased from 0.58 to 0.87 when adding these 126 CpGs to the risk factor model in the discovery set. The association was replicated with hazard ratios in the same direction in 28 CpGs in the Framingham Heart Study. When considering the association of variability, rather than mean differences, we found 432 differentially variable regions for lymphatic-hematopoietic cancers.

CONCLUSIONS

This study suggests that differential methylation and differential variability in blood DNA methylation are associated with lymphatic-hematopoietic cancer risk. DNA methylation data may contribute to early detection of lymphatic-hematopoietic cancers.

DNA methylation in blood-Potential to provide new insights into cell biology

Epigenetics plays a fundamental role in cellular development and differentiation; epigenetic mechanisms, such as DNA methylation, are involved in gene regulation and the exquisite nuance of expression changes seen in the journey from pluripotency to final differentiation. Thus, DNA methylation as a marker of cell identify has the potential to reveal new insights into cell biology. We mined publicly available DNA methylation data with a machine-learning approach to identify differentially methylated loci between different white blood cell types. We then interrogated the DNA methylation and mRNA expression of candidate loci in CD4⁺, CD8⁺, CD14⁺, CD19⁺ and CD56⁺ fractions from 12 additional, independent healthy individuals (6 male, 6 female). ‘Classic’ immune cell markers such as CD8 and CD19 showed expected methylation/expression associations fitting with established dogma that hypermethylation is associated with the repression of gene expression. We also observed large differential methylation at loci which are not established immune cell markers; some of these loci showed inverse correlations between methylation and mRNA expression (such as PARK2, DCP2). Furthermore, we validated these observations further in publicly available DNA methylation and RNA sequencing datasets. Our results highlight the value of mining publicly available data, the utility of DNA methylation as a discriminatory marker and the potential value of DNA methylation to provide additional insights into cell biology and developmental processes.

The signature of HBV-related liver disease in peripheral blood mononuclear cell DNA methylation

Background

Hepatitis B virus (HBV)-related liver disease induces liver damage by hepatic immune and inflammatory response. The association between aberrant peripheral blood mononuclear cell (PBMC) DNA methylation and progression of liver disease and fibrosis remains unclear.

Results

Here we applied Infinium 450 K BeadChip investigating PBMC genome-wide methylation profiling of 48 HBV-related liver disease patients including 24 chronic hepatitis B (CHB), 14 compensated liver cirrhosis (LC), and 10 decompensated liver cirrhosis (DLC). In total, there were 7888 differentially methylated CpG sites (36.06% hypermethylation, 63.94% hypomethylation) correlate with liver disease progression. LC was difficult to be diagnosed, intermediating between CHB and DLC. We used least absolute shrinkage and selection operator (LASSO)-logistic regression method to perform a LC predictive model. The predicted probability (P) of having LC was estimated by the combined model: P = 1/(1 − e^−x), where X = 11.52 − 2.82 × (if AST within the normal range − 0.19 × (percent methylation of cg05650055) − 0.21 × (percent methylation of cg17149911 ). Pyrosequencing validation and confusion matrix analysis was used for internal testing, area under receiver operating characteristic curve (AUROC) of model was 0.917 (95% CI, 0.80–0.977). On the fibrosis progress, there were 1705 genes in LC compared with CHB, whose differentially methylated CpG sites loading within the “promoter” regions (including TSS1500, TSS200, 5′UTR, and the 1st exon of genes) subject into the enrichment analysis using Ingenuity Pathway Analysis (IPA). There were 113 enriched immune-related pathways indicated that HBV-related liver fibrosis progression caused epigenetic reprogramming of the immune and inflammatory response.

Conclusions

These data support idea that development of HBV-related chronic liver disease is linked with robust and broad alteration of methylation in peripheral immune system. CpG methylation sites serve as relevant biomarker candidates to monitor and diagnose LC, providing new insight into the immune mechanisms understanding the progression of HBV-related liver fibrosis and cirrhosis.

Locus-Specific Differential DNA Methylation and Urinary Arsenic: An Epigenome-Wide Association Study in Blood among Adults with Low-to-Moderate Arsenic Exposure

BACKGROUND

Chronic exposure to arsenic (As), a human toxicant and carcinogen, remains a global public health problem. Health risks persist after As exposure has ended, suggesting epigenetic dysregulation as a mechanistic link between exposure and health outcomes.

OBJECTIVES

We investigated the association between total urinary As and locus-specific DNA methylation in the Strong Heart Study, a cohort of American Indian adults with low-to-moderate As exposure [total urinary As, mean  ( ± SD )  μ g / g creatinine: 11.7 (10.6)].

METHODS

DNA methylation was measured in 2,325 participants using the Illumina MethylationEPIC array. We implemented linear models to test differentially methylated positions (DMPs) and the DMRcate method to identify regions (DMRs) and conducted gene ontology enrichment analysis. Models were adjusted for estimated cell type proportions, age, sex, body mass index, smoking, education, estimated glomerular filtration rate, and study center. Arsenic was measured in urine as the sum of inorganic and methylated species.

RESULTS

In adjusted models, methylation at 20 CpGs was associated with urinary As after false discovery rate (FDR) correction (FDR <   0.05 ). After Bonferroni correction, 5 CpGs remained associated with total urinary As (p Bonferroni < 0.05 ), located in SLC7A11, ANKS3, LINGO3, CSNK1D, ADAMTSL4. We identified one DMR on chromosome 11 (chr11:2,322,050-2,323,247), annotated to C11orf2; TSPAN32 genes.

DISCUSSION

This is one of the first epigenome-wide association studies to investigate As exposure and locus-specific DNA methylation using the Illumina MethylationEPIC array and the largest epigenome-wide study of As exposure. The top DMP was located in SLC7A11A, a gene involved in cystine/glutamate transport and the biosynthesis of glutathione, an antioxidant that may protect against As-induced oxidative stress. Additional DMPs were located in genes associated with tumor development and glucose metabolism. Further research is needed, including research in more diverse populations, to investigate whether As-related DNA methylation signatures are associated with gene expression or may serve as biomarkers of disease development. https://doi.org/10.1289/EHP6263.

Deary I, Liu Q, Liu S, Elosua R, DeMeo DL, Ordovás JM. Epigenomic Assessment of Cardiovascular Disease Risk and Interactions With Traditional Risk Metrics

Background Epigenome-wide association studies for cardiometabolic risk factors have discovered multiple loci associated with incident cardiovascular disease (CVD). However, few studies have sought to directly optimize a predictor of CVD risk. Furthermore, it is challenging to train multivariate models across multiple studies in the presence of study- or batch effects. Methods and Results Here, we analyzed existing DNA methylation data collected using the Illumina HumanMethylation450 microarray to create a predictor of CVD risk across 3 cohorts: Women's Health Initiative, Framingham Heart Study Offspring Cohort, and Lothian Birth Cohorts. We trained Cox proportional hazards-based elastic net regressions for incident CVD separately in each cohort and used a recently introduced cross-study learning approach to integrate these individual scores into an ensemble predictor. The methylation-based risk score was associated with CVD time-to-event in a held-out fraction of the Framingham data set (hazard ratio per SD=1.28, 95% CI, 1.10-1.50) and predicted myocardial infarction status in the independent REGICOR (Girona Heart Registry) data set (odds ratio per SD=2.14, 95% CI, 1.58-2.89). These associations remained after adjustment for traditional cardiovascular risk factors and were similar to those from elastic net models trained on a directly merged data set. Additionally, we investigated interactions between the methylation-based risk score and both genetic and biochemical CVD risk, showing preliminary evidence of an enhanced performance in those with less traditional risk factor elevation. Conclusions This investigation provides proof-of-concept for a genome-wide, CVD-specific epigenomic risk score and suggests that DNA methylation data may enable the discovery of high-risk individuals who would be missed by alternative risk metrics.

Candesartan Neuroprotection in Rat Primary Neurons Negatively Correlates with Aging and Senescence: a Transcriptomic Analysis

Preclinical experiments and clinical trials demonstrated that angiotensin II AT₁ receptor overactivity associates with aging and cellular senescence and that AT₁ receptor blockers (ARBs) protect from age-related brain disorders. In a primary neuronal culture submitted to glutamate excitotoxicity, gene set enrichment analysis (GSEA) revealed expression of several hundred genes altered by glutamate and normalized by candesartan correlated with changes in expression in Alzheimer's patient's hippocampus. To further establish whether our data correlated with gene expression alterations associated with aging and senescence, we compared our global transcriptional data with additional published datasets, including alterations in gene expression in the neocortex and cerebellum of old mice, human frontal cortex after age of 40, gene alterations in the Werner syndrome, rodent caloric restriction, Ras and oncogene-induced senescence in fibroblasts, and to tissues besides the brain such as the muscle and kidney. The most significant and enriched pathways associated with aging and senescence were positively correlated with alterations in gene expression in glutamate-injured neurons and, conversely, negatively correlated when the injured neurons were treated with candesartan. Our results involve multiple genes and pathways, including CAV1, CCND1, CDKN1A, CHEK1, ICAM1, IL-1B, IL-6, MAPK14, PTGS2, SERPINE1, and TP53, encoding proteins associated with aging and senescence hallmarks, such as inflammation, oxidative stress, cell cycle and mitochondrial function alterations, insulin resistance, genomic instability including telomere shortening and DNA damage, and the senescent-associated secretory phenotype. Our results demonstrate that AT₁ receptor blockade ameliorates central mechanisms of aging and senescence. Using ARBs for prevention and treatment of age-related disorders has important translational value.

Transcriptome analysis identifies a robust gene expression program in the mouse intestinal epithelium on aging

The intestinal epithelium undergoes constant regeneration driven by intestinal stem cells. How old age affects the transcriptome in this highly dynamic tissue is an important, but poorly explored question. Using transcriptomics on sorted intestinal stem cells and adult enterocytes, we identified candidate genes, which change expression on aging. Further validation of these on intestinal epithelium of multiple middle-aged versus old-aged mice highlighted the consistent up-regulation of the expression of the gene encoding chemokine receptor Ccr2, a mediator of inflammation and several disease processes. We observed also increased expression of Strc, coding for stereocilin, and dramatically decreased expression of Rps4l, coding for a ribosome subunit. Ccr2 and Rps4l are located close to the telomeric regions of chromosome 9 and 6, respectively. As only few genes were differentially expressed and we did not observe significant protein level changes of identified ageing markers, our analysis highlights the overall robustness of murine intestinal epithelium gene expression to old age.

Age-dependent methylation in epigenetic clock CpGs is associated with G-quadruplex, co-transcriptionally formed RNA structures and tentative splice sites

Horvath's epigenetic clock consists of 353 CpGs whose methylation levels can accurately predict the age of individuals. Using bioinformatics analysis, we investigated the conformation, energy characteristics and presence of tentative splice sites of the sequences surrounding the epigenetic clock CpGs, in relation to the median methylation changes in different ages, the presence of CpG islands and their position in genes. Common characteristics in the 100 nt sequences surrounding the epigenetic clock CpGs are G-quadruplexes and/or tentative splice site motifs. Median methylation increases significantly in sequences which adopt less stable structures during transcription. Methylation is higher when CpGs overlap with G-quadruplexes than when they precede them. Median methylation in epigenetic clock CpGs is higher in sequences expressed as single products rather than in multiple products and those containing single donors and multiple acceptors. Age-related methylation variation is significant in sequences without G-quadruplexes, particularly those producing low stability nascent RNA and those with splice sites. CpGs in sequences close to transcription start sites and those which are possibly never expressed (hypothetical proteins) undergo similar extent of age-related median methylation decrease and increase. Preservation of methylation is observed in CpG islands without G-quadruplexes, contrary to CpGs far from CpG islands (open sea). Sequences containing G-quadruplexes and RNA pseudoknots, determining the recognition by H3K27 histone methyltransferase, are hypomethylated. The presented structural DNA and co-transcriptional RNA analysis of epigenetic clock sequences, foreshadows the association of age-related methylation changes with the principle biological processes of DNA and histone methylation, splicing and chromatin silencing.

Elastic net regularized regression for time-series analysis of plasma metabolome stability under sub-optimal freezing condition

In this paper, the stability of the plasma metabolome at −20 °C for up to 30 days was evaluated using liquid chromatography-high resolution mass spectrometric metabolomics analysis. To follow the time-series deterioration of the plasma metabolome, the use of an elastic net regularized regression model for the prediction of storage time at −20 °C based on the plasma metabolomic profile, and the selection and ranking of metabolites with high temporal changes was demonstrated using the glmnet package in R. Out of 1229 (positive mode) and 1483 (negative mode) metabolite features, the elastic net model extracted 32 metabolites of interest in both positive and negative modes. L-gamma-glutamyl-L-(iso)leucine (tentative identification) was found to have the highest time-dependent change and significantly increased proportionally to the storage time of plasma at −20 °C (R² = 0.6378 [positive mode], R² = 0.7893 [negative mode], p-value < 0.00001). Based on the temporal profiles of the extracted metabolites by the model, results show only minimal deterioration of the plasma metabolome at −20 °C up to 1 month. However, majority of the changes appeared at around 12–15 days of storage. This allows scientists to better plan logistics and storage strategies for samples obtained from low-resource settings, where −80 °C storage is not guaranteed.

Machine learning for epigenetics and future medical applications

Understanding epigenetic processes holds immense promise for medical applications. Advances in Machine Learning (ML) are critical to realize this promise. Previous studies used epigenetic data sets associated with the germline transmission of epigenetic transgenerational inheritance of disease and novel ML approaches to predict genome-wide locations of critical epimutations. A combination of Active Learning (ACL) and Imbalanced Class Learning (ICL) was used to address past problems with ML to develop a more efficient feature selection process and address the imbalance problem in all genomic data sets. The power of this novel ML approach and our ability to predict epigenetic phenomena and associated disease is suggested. The current approach requires extensive computation of features over the genome. A promising new approach is to introduce Deep Learning (DL) for the generation and simultaneous computation of novel genomic features tuned to the classification task. This approach can be used with any genomic or biological data set applied to medicine. The application of molecular epigenetic data in advanced machine learning analysis to medicine is the focus of this review.