Functional annotation of the human brain methylome identifies tissue-specific epigenetic variation across brain and blood

Abstract

Background: Dynamic changes to the epigenome play a critical role in establishing and maintaining cellular phenotype during differentiation, but little is known about the normal methylomic differences that occur between functionally distinct areas of the brain. We characterized intra- and inter-individual methylomic variation across whole blood and multiple regions of the brain from multiple donors.

Results: Distinct tissue-specific patterns of DNA methylation were identified, with a highly significant over-representation of tissue-specific differentially methylated regions (TS-DMRs) observed at intragenic CpG islands and low CG density promoters. A large proportion of TS-DMRs were located near genes that are differentially expressed across brain regions. TS-DMRs were significantly enriched near genes involved in functional pathways related to neurodevelopment and neuronal differentiation, including BDNF, BMP4, CACNA1A, CACA1AF, EOMES, NGFR, NUMBL, PCDH9, SLIT1, SLITRK1 and SHANK3. Although between-tissue variation in DNA methylation was found to greatly exceed between-individual differences within any one tissue, we found that some inter-individual variation was reflected across brain and blood, indicating that peripheral tissues may have some utility in epidemiological studies of complex neurobiological phenotypes.

Conclusions: This study reinforces the importance of DNA methylation in regulating cellular phenotype across tissues, and highlights genomic patterns of epigenetic variation across functionally distinct regions of the brain, providing a resource for the epigenetics and neuroscience research communities.

Figure 1

Methylomic profiling across multiple brain areas and blood from a cohort of individuals highlights clear tissue-specific differences in DNA methylation. (a, b) DNA methylation was calculated from ultra-deep MeDIP-seq data using 500 bp bins across the genome, and the relationship between tissues determined by Pearson correlation (a) and unsupervised hierarchical clustering (b). BA, Brodmann area; Ent Ctx, entorhinal cortex; STG, superior temporal gyrus; Vis Ctx, visual cortex.

Figure 2

Although DNA methylation at CGIs is relatively conserved across tissues, intragenic CGIs are dramatically over-represented and promoter CGIs under-represented in the most tissue-variable CGIs. (a) Average DNA methylation values calculated by MEDIPS from MeDIP-seq data for all annotated gene features: CGIs (yellow), CGI shores (blue), gene promoters (red) and CDSs (green). DNA methylation is lower in promoter CGIs compared to intragenic, 3' UTR and intergenic CGIs. CGI shores are characterized by higher DNA methylation than CGIs, with less location-dependent variation. Promoter DNA methylation shows a strong inverse correlation with GC density, with LCPs showing a higher average level of DNA methylation than CDSs. Error bars represent standard error of the mean. (b) Although TS-DMRs are distributed across all feature types, there are marked differences in the between-tissue correlation of DNA methylation across each of the broad feature categories we examined, with CGIs being more correlated across cortex, cerebellum and blood than CGI shores or CDSs. (c, d) There is a highly significant enrichment of intragenic CGIs (P = 2 × 10^-102) in analyses of CGI DMRs differentiating blood, cortex and cerebellum (c), and an even more dramatic enrichment (P = 1 × 10^-246) in comparisons between cortex and cerebellum (d). EXP, expected; OBS, observed.

Figure 3

Verification and replication of MeDIP-seq data for three top-ranked CGI DMRs. (a, b) Tissue-specific DNA methylation across an intragenic CGI in the JMJD2B/KDM4B gene. (a) MeDIP-seq analysis shows this region is hypermethylated in blood DNA compared to cortex and cerebellum (the red bar depicts the region subsequently analyzed by bisulfite pyrosequencing). (b) Pyrosequencing data for this region in an extended sample set confirm significant tissue-specific methylation patterns (P = 2 × 10^-8). (c, d) Tissue-specific DNA methylation across a CGI in the promoter of the EOMES gene. (c) MeDIP-seq analysis shows this region is hypermethylated in cerebellum DNA compared to cortex and blood. (d) Pyrosequencing data for this region in an extended sample set confirm significant tissue-specific methylation patterns (P = 2 × 10^-5). (e, f) Tissue-specific DNA methylation across an intragenic CGI in the BDNF gene. (e) MeDIP-seq analysis shows this region is hypermethylated in blood DNA compared to cortex and cerebellum from the same individuals. (f) Pyrosequencing data for this region in an extended sample set confirm significant tissue-specific methylation patterns (P = 4 × 10^-9). Error bars represent standard error of the mean.

Figure 4

Weighted gene co-methylation network analysis of DNA methylation at intragenic CGIs. (a) Dendrograms produced by average linkage hierarchical clustering of intragenic CGIs on the basis of topological overlap. Modules of co-methylated loci were assigned colors as indicated by the horizontal bar beneath each dendrogram. The 'blue' module was strongly negatively co-methylated (r² = -0.98, P = 4 × 10^-5) in cortex. (b) IPA on the genes associated with the blue module highlighted a network involved in nervous system development and function.

Figure 5

DNA methylation across LCPs is strongly associated with tissue type. (a) MEDIPS scores across both HCPs and LCPs can be used to accurately cluster samples by tissue type, but the strength of clustering, indicated by Pearson dissimilarity on the y-axis, is much higher in LCPs. (b, c) This pattern is reflected in three-factor PCA plots (b) and correlation analyses (c), with LCPs demonstrating stronger tissue-specific patterns of DNA methylation than HCPs.

Figure 6

Between-individual variation in DNA methylation is often correlated between blood and brain. (a) Between-individual variation in DNA methylation is highest in blood and lowest in the cortex. All tissues show a similarly significant (ANOVA P < 0.001) distribution of variability across features, with the greatest between-individual variation occurring in non-promoter CGIs. Scores represent the mean difference in normalized MeDIP-seq read density between individual 1 and 2 for each of the feature categories. Error bars represent standard error of the mean. Features 1 to 4 = promoter, intragenic, 3', and intergenic CGIs; features 4 to 8 = promoter, intragenic, 3', and intergenic CGI shores; feature 9 = CDS; features 10 to 12 = HCPs, ICPs, and LCPs. (b) Between-individual differences in DNA methylation observed in blood are significantly (P < 0.001) correlated with differences observed in the cerebellum (correlation = 0.76) and cortex (correlation = 0.66) from the same individuals. Scores represent the mean difference in normalized MeDIP-seq read density between individual 1 and 2 for each of the quantified features.