Chromosome-wide DNA methylation analysis predicts human tissue-specific X inactivation

Abstract

X-chromosome inactivation (XCI) results in the differential marking of the active and inactive X with epigenetic modifications including DNA methylation. Consistent with the previous studies showing that CpG island-containing promoters of genes subject to XCI are approximately 50% methylated in females and unmethylated in males while genes which escape XCI are unmethylated in both sexes; our chromosome-wide (Methylated DNA ImmunoPrecipitation) and promoter-targeted methylation analyses (Illumina Infinium HumanMethylation27 array) showed the largest methylation difference (D = 0.12, p < 2.2 E-16) between male and female blood at X-linked CpG islands promoters. We used the methylation differences between males and females to predict XCI statuses in blood and found that 81% had the same XCI status as previously determined using expression data. Most genes (83%) showed the same XCI status across tissues (blood, fetal: muscle, kidney and nerual); however, the methylation of a subset of genes predicted different XCI statuses in different tissues. Using previously published expression data the effect of transcription on gene-body methylation was investigated and while X-linked introns of highly expressed genes were more methylated than the introns of lowly expressed genes, exonic methylation did not differ based on expression level. We conclude that the XCI status predicted using methylation of X-linked promoters with CpG islands was usually the same as determined by expression analysis and that 12% of X-linked genes examined show tissue-specific XCI whereby a gene has a different XCI status in at least one of the four tissues examined.

Fig. 1

Promoter methylation analysed in blood (female: n = 59 and male: n = 36) revealed X chromosome sex-specific methylation differences as well as differences based on CpG density. Probes were divided by CpG density (LC black, IC gray, HC white) and classified as unmethylated (0–15% methylated), intermediate (15–60% methylated) or methylated (60–100% methylated) in males and females. a Methylation levels in males and females were significantly different (p value < 0.0001, Mann–Whitney test) across all X-linked probes. The majority of HC and IC promoter probes (n = 560) on the X chromosome were unmethylated in males and intermediate in females. X-linked LCs probes (n = 217) were mostly methylated regardless of sex. b Methylation levels in males and females were not significantly different (p value = 0.2779, Mann–Whitney test) across autosomal probes. The majority of promoter probes on chromosomes 20, 21 and 22 were unmethylated. Probes in HC and ICs (n = 1088) were mostly unmethylated whereas LC (n = 448) probes were mostly methylated. Males and females showed no differences in their methylation classes

Fig. 2

The XCI status predicted using methylation in blood corresponds with previously determined XCI status. a The XCI status (subject black, variable escape diagonal stripes, escape white, unclassifiable gray, conflicts dotted) of 372 X-linked genes with probes in HC and IC islands was predicted using methylation. The percentage of the total X-linked genes with probes in HC and IC islands is given for each predicted XCI status. In blood, the majority (81%) of genes are predicted to be subject to XCI. b The XCI status previously determined by Carrel and Willard (2005) in somatic cell hybrids (subject black, variable escape diagonal stripes, escape white) for those genes predicted by methylation to be subject to XCI (black bar in a and top pie chart) and those genes predicted to escape XCI (white bar in a and bottom pie chart). c–d Q-RT-PCR confirmation in somatic cell hybrids of predicted XCI status based on methylation. The expression level of two genes (TSR2 and ZRSR2) in four somatic cell hybrids containing a human Xi (white t75-2maz34-4a, t48-1a-1DAZ4A, t86-B1maz1b-3a and t11-4Aaz5), two somatic cell hybrids containing a human Xa (light gray AHA-11aB1 and t60-12) and a control female cell line (dark gray GM7350) were compared to confirm that methylation could predict XCI status. Test genes (TSR2 and ZRSR2) were normalized against a gene known to escape XCI (ZFX). Error bars represent the positive and negative error between three replicate PCRs. c TSR2 was unmethylated in male blood and intermediate in female blood and was predicted to be subject to XCI. d ZRSR2 was unmethylated in male and female blood and was predicted to escape XCI

Fig. 3

Most genes show the same predicted XCI status in all tissues examined while 12% of genes show tissue-specific XCI. a Male and female methylation was used to predict XCI status (as outlined in Supplementary Figure S1) of genes with probes in HC and IC islands in fetal muscle (black, female n = 6; male n = 4), fetal neural tissue (gray, female n = 6; male n = 5) and fetal kidney (white, female n = 6; male n = 5). b The combined predicted XCI status in all four tissues examined (blood, fetal muscle, fetal neural and fetal kidney). The majority of genes showed the same XCI status (subject black, variable escape diagonal stripes, escape white, unclassifiable gray, conflicts dotted) in all tissues. 6% of genes were unable to predict an XCI status in at least one tissue but predicted same XCI status in all other tissues (horizontal stripes). 12% of genes showed tissue-specific methylation differences which resulted in at least one tissue having a different predicted XCI from the other tissues (dark gray)

Fig. 4

Genes predicted to show tissue-specific XCI based on methylation from the Illumina Infinium HumanMethylation27 array as found across the X chromosome. The genomic locations of genes which showed the same predicted XCI status in all tissues examined (escape green, subject red) are shown to the left of the X chromosome ideogram. On the right are the genomic locations of genes in which at least one tissue had a predicted XCI status different from the other tissues. The predicted XCI status (subject red, variable escape purple, escape green, unclassifiable gray, conflict yellow) in each tissue examined (blood, fetal neural, fetal muscle and fetal kidney) is shown along with the names of all genes which show tissue-specific XCI

Fig. 5

Methylation histograms reveal X-linked HC promoters show the largest X-linked sex-specific methylation difference. The average male and average female methylation from probes representing four different genomic elements was used to create methylation histograms by determining the frequency at which probes were at a specific level of methylation (20 bins from 0 to 1.0 methylated). Female methylation frequencies are shown as dotted lines and males as solid lines with methylation frequencies from the X chromosome displayed on the upper row and the autosomal average from chromosomes 20, 21 and 22 on the bottom. The percentage of the total chromosomal DNA each element represents is given for the X chromosome and the autosomes (chromosomes 20, 21 and 22). Significance was calculated comparing the distribution of average male and average female methylation using the Kolmogorov–Smirnov test. When p values were greater than 0.0001 they were not significant, however, p values between 0.0001 and 1.0 E−10 (asterisk) and p values < 1.0 E−10 (double asterisk) were considered significantly different. a–b Promoter elements (the 1 kb up and downstream of all TSS) showed differences in methylation frequencies based on CpG density. HC promoters (a) showed males were hypomethylated compared to females on the X chromosome but not the autosomes. LC promoters (b) (contained neither an HC nor IC island) showed no sex-specific methylation difference on either the X chromosome or the autosomes. c–d Non-promoter elements tended to be methylated on the male and female X chromosome and intermediately methylated on the autosomes in both intragenic (c) and intergenic (d) regions

Fig. 6

X-linked introns but not exons show differences in methylation based on expression level. The average male and average female methylation from probes representing four different genomic elements was used to create methylation histograms by determining the frequency of probes at a specific level of methylation (20 bins from 0 to 1.0 methylated). Female methylation frequencies are shown as dotted lines and males as solid lines with methylation frequencies from the X chromosome displayed on the upper row and the autosomal average from chromosomes 20, 21 and 22 on the bottom. X-linked and autosomal (chromosomes 20, 21 and 22) genes were separated based on expression (determined in Su et al. 2002) and the top (light gray) and bottom (dark gray) 20% divided into those which correspond to either the exons (a) or introns (b). Significance was calculated comparing the distribution of male and female methylation using the Kolmogorov–Smirnov test. When p values were greater than 0.0001 they were not significant (ns), however, p values between 0.0001 and 1.0 E−10 (asterisk) and p values < 1.0 E−10 (double asterisk) were considered significantly different. While exons (a) were similarly methylated regardless of sex or expression level on both the X chromosome and the autosomes, introns (b) were more methylated in highly expressed X-linked genes than lowly expressed X-linked genes in both sexes. Autosomal introns showed no methylation difference in either sex