Aging effects on DNA methylation modules in human brain and blood tissue

Horvath et al. Genome Biology Aging effects on DNA methylation modules in human brain and blood tissue Steve Horvath 0 2 Yafeng Zhang 2 Peter Langfelder 0 Ren S Kahn 1 Marco PM Boks 1 Kristel van Eijk 1 5 Leonard H van den Berg 4 Roel A Ophoff 2 3 0 Department of Human Genetics, David Geffen School of Medicine, University of California Los Angeles , Los Angeles, CA 90095 , USA 1 Department of Psychiatry, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht , Utrecht , The Netherlands 2 Department of Biostatistics, School of Public Health, University of California Los Angeles , Los Angeles, CA 90095 , USA 3 UCLA Center for Neurobehavioral Genetics, Semel Institute of Neuroscience and Human Behavioral, School of Medicine, University of California Los Angeles , Los Angeles, CA 90095 , USA 4 Department of Neurology, University Medical Center Utrecht , Utrecht , The Netherlands 5 Department of Medical Genetics, University Medical Center Utrecht , Utrecht , The Netherlands Background: Several recent studies reported aging effects on DNA methylation levels of individual CpG dinucleotides. But it is not yet known whether aging-related consensus modules, in the form of clusters of correlated CpG markers, can be found that are present in multiple human tissues. Such a module could facilitate the understanding of aging effects on multiple tissues. Results: We therefore employed weighted correlation network analysis of 2,442 Illumina DNA methylation arrays from brain and blood tissues, which enabled the identification of an age-related co-methylation module. Module preservation analysis confirmed that this module can also be found in diverse independent data sets. Biological evaluation showed that module membership is associated with Polycomb group target occupancy counts, CpG island status and autosomal chromosome location. Functional enrichment analysis revealed that the aging-related consensus module comprises genes that are involved in nervous system development, neuron differentiation and neurogenesis, and that it contains promoter CpGs of genes known to be down-regulated in early Alzheimer's disease. A comparison with a standard, non-module based meta-analysis revealed that selecting CpGs based on module membership leads to significantly increased gene ontology enrichment, thus demonstrating that studying aging effects via consensus network analysis enhances the biological insights gained. Conclusions: Overall, our analysis revealed a robustly defined age-related co-methylation module that is present in multiple human tissues, including blood and brain. We conclude that blood is a promising surrogate for brain tissue when studying the effects of age on DNA methylation profiles. - Background Gene expression (messenger RNA transcript abundance) is modulated by epigenetic factors such as histone modifications, microRNAs, long noncoding RNAs, and DNA methylation. A large body of literature has provided evidence that age has a significant effect on cytosine-5 methylation within CpG dinucleotides [1-4]. A genomewide decrease in DNA methylation has been shown to occur during in vitro aging [5] and in vivo aging [6,7]. Previous studies of aging effects on DNA methylation involved typically adults but recent studies also involved pediatric populations[8] Important insights have been gained regarding what types of genes show promoter hyper- or hypomethylation with age [9-11]. For example, early-life-induced programming that relies on DNA methylation appears to be at a considerable risk to become disrupted during aging [12,13]. Age-associated hypermethylation has been found to preferentially affect loci at CpG islands [14]. Important cancer related genes become hypermethylated during aging, including those encoding the estrogen receptor, insulin growth factor, and E-cadherin, and key developmental genes [9,15,16]. Rakyan et al. [15] showed that aging-associated DNA hypermethylation in blood occurs preferentially at bivalent chromatin domain promoters that are associated with key developmental genes. These genes are frequently hypermethylated in cancers, which points to a mechanistic link between aberrant hypermethylation in cancer and aging. Teschendorff et al. [16] identified a core DNA methylation signature of 589 CpGs that were significantly related to age. Further, the authors showed that Polycomb group protein targets (PCGTs) are far more likely to become methylated with age than non-targets (odds ratio = 5.3, P < 10-10), independently of sex, tissue type, disease state, and methylation platform. The authors identified a subset of 64 PCGTs exhibiting a clear trend toward hypermethylation with age across multiple cell types (blood, ovarian cancer, cervix, mesenchymal stem cells). This is a biologically important insight since gene repression by the PCG protein complex via histone H3 lysine 27 trimethylation (H3K27me3) is required for embryonic stem cell selfrenewal and pluripotency [17,18]. While Teschendorff et al. evaluated the core aging signature in whole blood (WB), solid tissues, lung tissue, and cervix tissue, they did not include brain tissues. In this study, we expand previous studies along multiple directions. First, we study aging effects in brain by evaluating aging effects in human tissue samples of the frontal cortex (FCTX), temporal cortex (TCTX), cerebellum (CRBLM), caudal pons (PONS) [19], prefrontal cortex [20], and mesenchymal stromal cells (Table 1). Second, we contrast aging effects on gene expression levels (mRNA) and DNA methylation levels and in brain and blood tissue. Third, we analyze four novel WB DNA methylation data sets involving n = 752 Dutch subjects. Fourth, we carry out a weighted correlation network Table 1 Description of DNA methylation data sets Consensus 92 WB Consensus 273 WB Consensus 293 WB Dutch controls from ALS study Dutch controls from SZ study Data sets 1 to 10 were used in the consensus network analysis while data sets 11 to 16 were used in the module validation (preservation) analysis. Our novel WB DNA methylation data sets (numbered 1 to 3 and 11) are composed of (n = 92 + 273 + 293 + 94) individuals. The study involved multiple tissues (blood, brain) and different populations (adults and healthy children). Note that the mean age (and age ranges) differ greatly across the studies. ALS, amyotrophic lateral sclerosis; CRBLM, cerebellum; FCTX, frontal cortex; MSC, mesenchymal stromal cell; PONS, caudal pons; SZ, schizophrenia; TCTX, temporal cortex; UKOPS, United Kingdom Ovarian Cancer Population Study; WB, whole blood. analysis (WGCNA) of multiple methylation data sets. We apply the consensus module analysis to ten independent methylation data sets and identify a consensus co-methylation module (referred to as aging module) that contains CpG sites that are hypermethylated with age in multiple human tissues (WB, leukocytes, and different brain regions, including cortex, pons, and cerebellum). We then (...truncated)