Prenatal immune activation alters the adult neural epigenome but can be partly stabilised by a n-3 polyunsaturated fatty acid diet

Abstract

An unstable epigenome is implicated in the pathophysiology of neurodevelopmental disorders such as schizophrenia and autism. This is important because the epigenome is potentially modifiable. We have previously reported that adult offspring exposed to maternal immune activation (MIA) prenatally have significant global DNA hypomethylation in the hypothalamus. However, what genes had altered methylation state, their functional effects on gene expression and whether these changes can be moderated, have not been addressed. In this study, we used next-generation sequencing (NGS) for methylome profiling in a MIA rodent model of neurodevelopmental disorders. We assessed whether differentially methylated regions (DMRs) affected the chromatin state by mapping known DNase I hypersensitivity sites (DHSs), and selected overlapping genes to confirm a functional effect of MIA on gene expression using qPCR. Finally, we tested whether methylation differences elicited by MIA could be limited by post-natal dietary (omega) n-3 polyunsaturated fatty acid (PUFA) supplementation. These experiments were conducted using hypothalamic brain tissue from 12-week-old offspring of mice injected with viral analogue PolyI:C on gestation day 9 of pregnancy or saline on gestation day 9. Half of the animals from each group were fed a diet enriched with n-3 PUFA from weaning (MIA group, n = 12 units, n = 39 mice; Control group, n = 12 units, n = 38 mice). The results confirmed our previous finding that adult offspring exposed to MIA prenatally had significant global DNA hypomethylation. Furthermore, genes linked to synaptic plasticity were over-represented among differentially methylated genes following MIA. More than 80% of MIA-induced hypomethylated sites, including those affecting chromatin state and MECP2 binding, were stabilised by the n-3 PUFA intervention. MIA resulted in increased expression of two of the 'top five' genes identified from an integrated analysis of DMRs, DHSs and MECP2 binding sites, namely Abat (t = 2.46, p < 0.02) and Gnas9 (t = 2.96, p < 0.01), although these changes were not stabilised by dietary intervention. Thus, prenatal MIA exposure impacts upon the epigenomic regulation of gene pathways linked to neurodevelopmental conditions; and many of the changes can be attenuated by a low-cost dietary intervention.

Fig. 1. DNA hypomethylation in hypothalamus of MIA-exposed offspring.

Percent methylation is shown on the y-axis; error bars are standard error of the mean (SEM). *p < 0.05; **p < 0.005; p values calculated from a stratified t-test between a maternal immune activation and control groups given a control diet and b MIA groups given a n-3 PUFA or control diet. Treatment groups shown on the x-axis: n-3 POL—Omega3 PolyI:C, n-6 POL—Omega6 PolyI:C, n-3 SAL—Omega3 Saline, and n-6 SAL—Omega6 Saline

Fig. 2. Genome-wide methylation differences in the hypothalamus.

a MIA-induced DMRs plotted against chromosome locations, with Track1: mouse genome with cytobands. Track2: connectors between the gene names to the chromosome location. Track3: differentially methylated genes connected to the locations. Track4: red dots indicating hypermethylated sites and blue dots indicating hypomethylation. Track5: CpG site distance to the gene; red indicates after transcription start site (TSS) and blue indicates before TSS. Track6: negative log q-values for the corresponding CpG sites at that genomic location. Track8: comparison of Δmethylation distribution detected from MIA effect and omega-3 effect. Y axis indicates percentage Δmethylation values, x axis represents CpG sites from 165 stabilised genes. b Quantile–Quantile plot for MIA-induced differentially methylated CpG sites, ‘observed’ p values are plotted on the y axis against ‘expected’ values on the x axis. The red line indicates the distribution of p value under the null progression. Control group comparison is expected to follow the null and is plotted to check the level of ‘noise’. c Quantile–Quantile plot for differentially methylated CpG sites following dietary intervention. Observed p values are plotted observed on y axis against expected values on x axis. The red line indicates the distribution of p value under the null progression. Control group comparison is expected to follow the null and is plotted to check the level of ‘noise’. d Methylation profile of genes affected by MIA after n-3 intervention. (1) 165 genes stabilised by n-3 PUFA dietary intervention (Orange); (2) 40 genes affected by MIA but not affected by n-3 PUFA (Blue) and (3) not altered by MIA but altered by n-3 PUFA diet (Green)

Fig. 3. Effect of MIA-associated DMRs in gene transcription.

a Overlap between the DMRs, DHSs and MECP2 binding sites. b Abat and Gnas9 mRNA quantification using RT-qPCR. Quantities are calculated against a standard curve and values are normalised to 18S RNA. Error bars are standard error of the mean (SEM). *p < 0.05; p values calculated from a stratified t-test between MIA and control groups given a control diet