Genome-wide screen for differential DNA methylation associated with neural cell differentiation in mouse

Abstract

Cellular differentiation involves widespread epigenetic reprogramming, including modulation of DNA methylation patterns. Using Differential Methylation Hybridization (DMH) in combination with a custom DMH array containing 51,243 features covering more than 16,000 murine genes, we carried out a genome-wide screen for cell- and tissue-specific differentially methylated regions (tDMRs) in undifferentiated embryonic stem cells (ESCs), in in-vitro induced neural stem cells (NSCs) and 8 differentiated embryonic and adult tissues. Unsupervised clustering of the generated data showed distinct cell- and tissue-specific DNA methylation profiles, revealing 202 significant tDMRs (p<0.005) between ESCs and NSCs and a further 380 tDMRs (p<0.05) between NSCs/ESCs and embryonic brain tissue. We validated these tDMRs using direct bisulfite sequencing (DBS) and methylated DNA immunoprecipitation on chip (MeDIP-chip). Gene ontology (GO) analysis of the genes associated with these tDMRs showed significant (absolute Z score>1.96) enrichment for genes involved in neural differentiation, including, for example, Jag1 and Tcf4. Our results provide robust evidence for the relevance of DNA methylation in early neural development and identify novel marker candidates for neural cell differentiation.

Figure 1. DNA methylation data distribution.

A) DMH feature location across the genome. After array normalization and data processing, the full DMH dataset contained 51,243 features. Thereof, 23,957 features were associated with Transcriptional Start Sites (TSS) of annotated genes, 27,286 were located in distal CG rich areas. B) TSS-associated features (+/−1 kb range) are less methylated with respect to features not associated with TSS. The red line represents the average methylation percentages (Y-axis) in all tissues and cells across DMH features sorted by their distance to the TSS (X-axis). C) Methylation distribution was similar in coding and non-coding regions in features associated to TSS (left panel) and features not associated to TSS (right panel). Distribution of features located in coding and non-coding regions are represented as green and gray shapes, respectively. DMH scores represent the percentage of methylation calculated using enzymatically methylated (100% methylation) and unmethylated (0% methylation) control samples as calibrators. Thus, DMH scores range from 0 to 1, representing 0% and 100% methylation, respectively D) ESCs, NSCs and embryonic brain samples showed bimodal distribution with peaks at 0 (0% methylation) and 1 (100% methylation). While ESCs and NSCs displayed similar distribution with well-differentiated peaks, embryonic brain distribution is shifted towards more intermediate values. Distribution of DMH scores in NSCs, ESCs and embryonic brain are represented as blue, green and red lines, respectively. E) Volcano plot of features showing differential methylation in ESCs and NSCs. 140 regions were highly methylated in NSCs and 62 highly methylated in ESCs. Features were ranked and candidate tDMRs were selected following the criteria detailed in the text. X-axis represents the methylation percentage difference between ESCs and NSCs while Y-axis represents the registered t-statistic for that difference. Red circles highlight candidate tDMRs. Features showing higher methylation in NSCs and ESCs are clustered on the right and left sides, respectively.

Figure 2. TSS-associated tissue-specific differentially methylated regions (tDMRs) in stem cells and embryonic tissues.

202 candidate tDMRs were discovered comparing profiles in NSCs and ESCs. Each row corresponds to a feature in the DMH array while each column corresponds to a sample, i.e. NSCs (n = 2), ESCs (n = 2) embryonic and adult tissues (n = 8). Quantitative methylation analysis results are shown in a color code ranging from yellow (∼0% methylation), over green (∼50% methylation) to dark blue (∼100% methylation).

Figure 3. TSS-associated tDMRs define distinct groups in NSCs, ESCs and embryonic brain.

A) Unsupervised clustering of top-ranked candidates. B) ANOVA of methylation percentage in ESCs, NSCs and embryonic brain defined 6 tDMR groups. 382 candidate tDMRs were selected and ranked. Color code as detailed in Figure 2.

Figure 4. tDMR validation by direct bisulfite sequencing.

A) Averaged methylation values obtained by direct bisulfite sequencing (DBS). 51 candidate regions were studied in embryonic stem cells (ESCs), neural stem cells (NSCs), embryonic spinal cord (EmbSC), embryonic dorsal root ganglia (EmbDRG), adult cerebellum (AdCer), adult spinal cord (AdSC), adult liver (AdLiv) and adult skeletal muscle (AdSM). Color code as detailed in Figure 2. Rows represent each DBS amplicon and columns correspond to the average methylation value per sample type. B) DMH methylation percentage and averaged DBS amplicon methylation values in NSCs and ESCs. 49 matched DMH features/DBS amplicons were studied in the same biological samples. Data are presented in two color-matching matrices for DMH and DBS (left and right matrix, respectively). The numbers in brackets next to each feature ID indicate the number of restriction sites in the respective DMH fragments. Rows represent each DMH feature or DBS amplicon and columns correspond to the average methylation value per sample type. Color code as detailed in Figure 2. C) Correlation analysis of DMH and DBS data. Mean CpG methylation obtained by DBS are shown on the X-axis and DMH scores are showed on the Y-axis. The correlation coefficient was 0.679 for all DMH features/DBS amplicon pairs, while 0.438 for pairs for which the DMH feature contained a single restriction site. Points in circles highlight features with a single restriction site. D) DNA methylation data distribution in DMH features and DBS amplicons. Data distribution was similar with both technologies. The blue line represents the distribution of the DMH scores while the green line the distribution of mean DNA methylation for DBS amplicons.

Figure 5. DMH and MeDIP-chip DNA methylation profiles in NSCs and ESCs.

Both technologies showed similar results for two genes involved in neural differentiation. A) Location of probes covering the investigated regions in the Pou5f1 (upper panel) and Ddah2 (lower panel) genes. Red and green boxes represent areas covered by features in the DMH and MeDIP-chip arrays, respectively. B) DNA methylation values in Pou5f1 and Ddah2 genes in NSCs and ESCs. DNA methylation values obtained with DMH (left panel) and MeDIP-chip (red panel) are comparable. Rows correspond to features in the DMH or MeDIP-chip array and columns correspond to samples (NSCs (n = 2) and ESCs (n = 2)), grouped for DMH and MeDIP-chip respectively. Quantitative methylation values are expressed as methylation percent for DMH and MeDIP-chip respectively and color-coded as in Figure 2.