Alcohol exposure alters DNA methylation profiles in mouse embryos at early neurulation

Abstract

Alcohol exposure during development can cause variable neurofacial deficit and growth retardation known as fetal alcohol spectrum disorders (FASD). The mechanism underlying FASD is not fully understood. However, alcohol, which is known to affect methyl donor metabolism, may induce aberrant epigenetic changes contributing to FASD. Using a tightly controlled whole-embryo culture, we investigated the effect of alcohol exposure (88mM) at early embryonic neurulation on genome-wide DNA methylation and gene expression in the C57BL/6 mouse. The DNA methylation landscape around promoter CpG islands at early mouse development was analyzed using MeDIP (methylated DNA immunoprecipitation) coupled with microarray (MeDIP-chip). At early neurulation, genes associated with high CpG promoters (HCP) had a lower ratio of methylation but a greater ratio of expression. Alcohol-induced alterations in DNA methylation were observed, particularly in genes on chromosomes 7, 10, and X; remarkably, a >10 fold increase in the number of genes with increased methylation on chromosomes 10 and X was observed in alcohol-exposed embryos with a neural tube defect phenotype compared to embryos without a neural tube defect. Significant changes in methylation were seen in imprinted genes, genes known to play roles in cell cycle, growth, apoptosis, cancer, and in a large number of genes associated with olfaction. Altered methylation was associated with significant (p<0.01) changes in expression for 84 genes. Sequenom EpiTYPER DNA methylation analysis was used for validation of the MeDIP-chip data. Increased methylation of genes known to play a role in metabolism (Cyp4f13) and decreased methylation of genes associated with development (Nlgn3, Elavl2, Sox21 and Sim1), imprinting (Igf2r) and chromatin (Hist1h3d) was confirmed. In a mouse model for FASD, we show for the first time that alcohol exposure during early neurulation can induce aberrant changes in DNA methylation patterns with associated changes in gene expression, which together may contribute to the observed abnormal fetal development.

Figure 1

Alcohol-induced dysmorphology of embryonic development at E8.25 + 44 hrs. The alcohol-treated embryos showed developmental delay and abnormality. Several examples (in B and C) are demonstrated in abnormality of the heart (H); small forebrain (FB), midbrain (MB), or hindbrain (HB); small eyes (arrow in C); abnormal tail morphology as compared with Control (A). A major dichotomic phenotype is embryos with neural tube closed but with delay and abnormality in development (NTC, B), and neural tube opened (arrowheads, not closed in time; NTO, C) that accompany more severe dysmorphology. An alcohol-treated embryo with neural tube opened (Alcohol-NTO) is shown in (C). Scale bar: 1,000 micron.

Figure 2

Characterization of DNA methylation patterns in different promoter categories. (A) Histogram of the number of promoters in each category. LCP: low-CpG promoter, ICP: intermediate-CpG promoter and HCP: high-CpG promoters (see definition in text). The definition was based on the sequence identity of the promoters from −1,300 to +500 bp from transcription start site. The number indicates the number of genes in each category of all chromosomes. (BI) Shows the percentage of autosomal genes in three promotor categories (CI; HCP, ICP, LCP), and (B2) the percentage of hypermethylated autosomal genes in control (Untreated) and alcohol-treated with phenotype neural tube closed (ALC-NTC) and neural tube opened (ALC-NTO). In HCP/LCP category, there are less hypermethylated (see definition in C) genes in ALC-NTO than in ALC-NTC. (C) Scatter plot of promoter methylation level and promoter CpG content. X-axis denotes the observed-to-expected CpG ratio, and Y-axis represents the average log2 transformation of the methylation signal level through all the probes in −1,300 to +500 bp from transcription start site; LCP (Red), ICP (green), and HCP (blue) promoters. The dash line at y = 0.4 is defined as cutoff for hypermethylation, which represents 1.3-fold of the genome-wide median methylation level. The dots above this line are considered hypermethylated. This scatter distribution pattern is well-conserved across mouse to human.

Figure 3

Correlation between promoter CpG content, DNA methylation levels, and gene expression levels. (A) Percentage of present genes (detection call >=0.5) whose expression levels can be reliably detected in microarray experiments in LCP, ICP and HCP, respectively. A higher percentage of present genes imply an overall higher gene expression profile. It shows that HCP has the highest percentage of gene expression followed by ICP and LCP. (B) Percentage of present genes with different range of DNA methylation levels in LCP, ICP and HCP. In the ICP and HCP, higher methylation correlated with lower gene expression.

Figure 4

Venn diagram denoting the number of genes containing at least two differentially methylated consecutive probes. (A) Increased- and (B) Decreased methylation in two statistical comparisons, control vs. ALC-NTC and control vs. ALC-NTO. There are 147 genes with increased methylation levels in both groups, and 141 with decreased methylation levels by alcohol treatment.

Figure 5

Alcohol-induced DNA methylation changes in neural tube open and close groups. (A) Alcohol affected methylation profiles in high-CpG and low-CpG regions (control compared with ALC-NTO, ALC-NTC combined). The genes with regions of high CpG tend to be increased in methylation, while those of low CpG tend to be decreased in methylation. (B) Number of promoters that are increase- or decrease in methylation in two consecutive probes with high CpG contents. (C) Number of increased- or decreased-methylation promoters in two consecutive probes with low CpG content. In the ALC-NTO group, more increased methylation genes with high CpG content, and more decreased methylation genes with low CpG content, were apparent, in agreement with the more severe neural tube development phenotype in ALC-NTO vs. ALC-NTC.

Figure 6

Number of genes containing two consecutive probes showing increased- and decreased-methylation in each chromosome. (A) Neural tube close group. (B) Neural tube open group. There are decreased- and increased methylation changes of genes in each chromosome. Highly significant changes were seen in Chromosome 7, 10 and X. These chromosomes are also known to be enriched for imprinted or silenced genes.

Figure 7

Sequenom analysis confirming key alcohol-induced DNA methylational changes for (A) Sox21, (B) Nlgn3, (C) Elavl2 and (D) Hist1h3d. For each gene, the upper panel shows the sequenom-detected methylation signal along the genomic DNA region being targeted. The blue circles and red diamonds indicate the % of methylated CpGs in control and alcohol-treated conditions. The lower panel shows the correspondent MeDIP-chip-derived changes; the red dash line between the sequenom and MeDIP-chip data demonstrates the projected genomic regions where the MeDIP-derived DNA fragment can reach (150 bp upstream of the 5'-end and 150 bp downstream of the 3'-end of the two-consecutive probes). The blue bars indicate the positions of probes whose methylational changes are not statistically significant. The red bars denote the fold changes of the two probes whose alcohol-induced methylational changes are statistically significant, based upon the two consecutive analyses.