Supplementary Figure 1: Characteristics of nascent strand methylation.

(a) Simplified schematic of BrdU incorporation into the nascent strands of asynchronously replicating cells (adapted from Hansen, R. S. et al. Proc Natl Acad Sci U S A 107, 139–144, 2010). Based on estimated fork speeds of ~ 0.035 nucleotides/second (Shirodkar, A. V. et al. Blood 121, 3531–3540, 2013) we expect to label ~ 100 kb of single stranded DNA for each lagging or leading strand at every replication fork during a 1 hour pulse. (b) Genome browser tracks of read distribution for Repli-seq data from ENCODE (dark blue tracks, BG02 ESCs) and from our nascent DNA data using Repli-BS (light purple, HUES64 ESCs). All samples were auto-scaled to display the maximum range in the data track. An inverted “V” shows the typical pattern for Repli-seq data where read density at the peak corresponds to early replicating domains from which DNA is replicated bi-directionally to generate segregated peaks across S-phase. (c) Methylation levels for individual CpGs show more CpGs with intermediate methylation status in nascent DNA compared to bulk. (d) Methylation levels for matched nascent and bulk CpGs separated into early (S1 + S2; n = 10,351,838), mid (S3 + S4; n = 9,099,483) or late (S5 + S6; n = 3,937,510) S-phase fractions. White dots: median. (e) For CpGs located within specific genomic features, the ratio (nascent to bulk methylation) is shown on the x-axis and bulk mean methylation on the y-axis. CpG islands (CGI), high CpG-dense promoters (HCP), low CpG-dense promoters (LCP), long terminal repeats (LTR), long interspersed nuclear elements (LINE) and short interspersed nuclear elements (SINE). The shaded area highlights the small difference between genomic features.(f) Gene body methylation levels for genes classified as not expressed (FPKM < 1), expressed (FPKM >10) or highly expressed (FPKM > 100). Both bulk and nascent methylation levels are not affected by gene expression level. Bold line: median; box displays interquartile range and whiskers extend to the most extreme data point that is no more than 1.5 times the interquartile range. (g) Boxplots show methylation levels for CpGs with methylation >0.8 in bulk DNA. CpGs were grouped according to the CpG density of their surrounding genomic regions. Bold black lines: median; box displays interquartile range and whiskers extend to the most extreme data point that is no more than 1.5 times the interquartile range. Pie charts (right) show the proportion of CpGs within high CpG dense regions that overlap with promoters (defined as 1 kb upstream to 1 kb downstream of transcription start site), genomic repeats and biologically determined CGIs. (h) Representative CGI with high methylation levels. (i) H3K27me3 (red) and EZH2 ChIP-seq data (blue, top, ENCODE data for human ESC H1 cells) are displayed for a region on chromosome 15 as an example of a site where nascent DNA shows highly reduced methylation compared to bulk (shaded in grey). (j) Boxplot display methylation levels for bulk WT ESCs and DKO (DNMT3A^−/− and DNMT3B^−/−) cells for all CpGs (left) and for only CpGs located in methylated EZH2 sites (right). While globally, only a ~ 10% reduction in methylation is observed in DKO cells, EZH2 sites show almost complete loss of methylation in the DKO line indicating their dependence on DNMT3A and 3B for methylation. Bold line: median; box displays interquartile range and whiskers extend to the most extreme data point that is no more than 1.5 times the interquartile range. Note that as only ~ 10,000 CpGs are located within these particular EZH2-eniched sites and their functional relevance is unkown. (k) Mean CpG (left) and CpA (right) nascent (S1) and bulk methylation in ESCs. (l) For CpAs located within specific genomic features, the ratio (nascent to bulk methylation) is shown on the x-axis with bulk mean methylation on the y-axis for features as defined in e.