In utero exposure to chlordecone affects histone modifications and activates LINE-1 in cord blood

Abstract

Environmental factors can induce detrimental consequences into adulthood life. In this study, we examined the epigenetic effects induced by in utero chlordecone (CD) exposure on human male cord blood as well as in blood-derived Ke-37 cell line. Genome-wide analysis of histone H3K4me3 distribution revealed that genes related to chromosome segregation, chromatin organization, and cell cycle have altered occupancy in their promoters. The affected regions were enriched in ESR1, SP family, and IKZF1 binding motifs. We also observed a global reduction in H3K9me3, markedly in repeated sequences of the genome. Decrease in H3K9me3 after CD exposure correlates with decreased methylation in LINE-1 promoters and telomere length extension. These observations on human cord blood were assessed in the Ke-37 human cell line. H3K4me3 and the expression of genes related to immune response, DNA repair, and chromatin organization, which were affected in human cord blood were also altered in CD-exposed Ke-37 cells. Our data suggest that developmental exposure to CD leads to profound changes in histone modification patterns and affects the processes controlled by them in human cord blood.

Figure S1.. Schematic presentation of the experiments.

In left panel, we describe the experiments carried out in umbilical cord blood samples; in right, Ke-37 cell line experiments are described.

Figure 1.. H3K4me3 and H3K9me3 histone marks were decreased in CD-exposed umbilical cord blood cells.

Representative images of structurally preserved nuclei stained with anti-H4ac (upper panel), anti-H3K9me3 (middle panel), or anti-H3K4me3 (lower panel) antibodies. Pictures were taken with fixed exposure time at ×63 objective. Corresponding quantitative analysis of the nuclear specific signal is presented as relative fluorescence compared with control ± SD. *P < 0.05, Wilcoxon–Mann–Whitney test, n = 8 for each group (H4ac), n = 10 for each group (H3K9me3), n = 8 for control, and n = 10 for CD (H3K4me3).

Figure S2.. Pairwise Spearman’s correlation analysis.

The correlation analysis was performed using R application.

Figure S3.. Scatter plot of H3K4me3 peaks.

Peaks with increased and decreased H3K4me3 signals are shown in red and blue, respectively. Peaks with less than 1.2-fold change are shown in gray.

Figure 2.. The occupancy of H3K4me3 is globally decreased in CD-exposed umbilical cord blood cells.

(A) H3K4me3 occupancy has decreased in the promoter of BRCA1. In contrast, it has increased in the vicinity of proto-oncogene RET. The image represents the averaged signal of six replicates. Dashed boxes indicate the position of strongest changes. (B) Most differential peaks are located near TSS ± 2 kb. (C) Differential peaks were enriched in a G-rich (upper panel) and a T-rich (low panel) motif. The parts of the motifs were similar to the sequences of SP family and IRF3 binding sites, respectively. (D) Several biological functions are overrepresented within the genes located in the neighborhood of the differential genes. Gene Ontology clusters are sorted by P-values. Adjusted P-values (hypergeometric test, FDR) are indicated at the end of each bar of the histogram. (E) Differential peaks are enriched in the genes regulated by JARID2 and ESR1 proteins.

Figure S4.. Most H3K4me3 peaks are located in the vicinity of the TSS.

(A, B) The whole set of H3K4me3 peaks identified in cord blood and (B) differential H3K4me3 peaks identified in CD-exposed newborns have preferential TSS distribution.

Figure 3.. ESR1 binding was altered in CD-exposed cells.

(A) Representative images of z-stack sums obtained from structurally preserved nuclei immunostained with ESR1 antibody. Images were taken with fixed exposure time at x63 objective. ESR1 (green) is localized in the nucleus (visualized by DAPI staining—blue). (B) Quantitative analysis of ESR1 signal in nuclei was performed using ImageJ software. Data are presented as relative fluorescence compared with control ± SD. *P < 0.05, Wilcoxon–Mann–Whitney test, and n = 8 for each group. (C) Representation of the conservative ESR1 binding site within the KMT2E gene. The binding site is highlighted in green. (D) ESR1 binding at the promoter of target genes was assessed by ChIP-qPCR. Values were normalized to the RPLP0 region. Control samples are in blue and CD-treated samples are in red. The data were averaged and presented as normalized ChIP fold change compared with control ± SD. *P < 0.05, **P < 0.01, Wilcoxon–Mann–Whitney test, n = 7 for control, and n = 8 for CD.

Figure 4.. H3K9me3 occupancy and telomere length changes in CD-exposed umbilical cord blood.

(A) H3K9me3 occupancy at the promoter of target genes and satellite was assessed by ChIP-qPCR. Values were normalized to the RPLP0 region. Control samples are in blue, and CD-treated samples are in red. The data were averaged and presented as normalized ChIP fold change compared with control ± SD. *P < 0.05, **P < 0.01, Wilcoxon–Mann–Whitney test, and n = 6 for each group. (B) Telomere length was assessed by qRT-PCR and normalized to the qPCR value of a single-copy gene (RPLP0) to determine the T/S ratio. **P < 0.01, Wilcoxon–Mann–Whitney test, n = 16 for control, and n = 18 for CD. (C) Histone occupancy of H3K4me3 at the promoter of genes involved in telomere maintenance function. Control samples are in blue and CD-treated samples are in red. Peak intensities were extracted from ChIP-seq data and were plotted as fold change compared with control ± SD. All plotted regions displayed statistical difference in H3K4me3 occupancy (moderated t test, Limma package), n = 6 for each group.

Figure S5.. H3K4me3 occupancy has increased in umbilical cord blood of newborn exposed to CD.

(A) H3K4me3 occupancy at the promoter of TERT in CD and control group visualized by Integrative Genomics Viewer genomic viewer. (B) Normalized H3K4me3 signal calculated at the promoter of TERT tends to increase.

Figure S6.. Telomere length was not affected in Ke-37 CD-exposed cells.

Telomere length was assessed by qRT-PCR, and values were normalized to a single-copy gene (RPLP0) to determine the T/S ratio, n = 8 for each group.

Figure 5.. Exposure to low dose of CD leads to DNA methylation reduction and increased H3K4me3 at the UID-34 L1 in umbilical cord blood cells and increased expression of LINE-1 in Ke-37 cells.

(A) Methylation analysis was performed by qPCR analysis using the intact cord blood genomic DNA or digested by methylation-sensitive restriction enzyme from control samples (in blue) and CD-treated samples (in red). Data are presented as relative methylation compared with control ± SD. *P < 0.05, Wilcoxon–Mann–Whitney test, and n = 8 for each group. Below the graph, the sequence of the L1.12 promoter with CpG positions is shown. (B) H3K4me3 occupancy at the promoter of UID-34 L1. The image represents the averaged signal of six replicates. Dashed boxes indicate the position of the strongest changes. CpGs positions are indicated. (C) H3K4me3 occupancy at CpG3 of the L1 promoter. Values were normalized to the RPLP0 region. The data were averaged and presented as normalized ChIP fold change compared with control ± SD. P = 0.1, Wilcoxon–Mann–Whitney test, and n = 7 for each group. (D) The expressions of LINE-1 ORF1 and ORF2 in Ke-37 cells were assessed by qRT-PCR following exposure to 100 nM. (E) The expressions of LINE-1 ORF1 and ORF2 at 300 nM of CD. The expression of target genes was normalized to the expression of the housekeeping RPL37A gene. The data were averaged and presented as normalized expression fold change compared with control ± SD. **P < 0.01, ***P < 0.001, Wilcoxon–Mann–Whitney test, n = 18 for control, n = 20 for CD (100 nM), and n = 6 for each group (300 nM).

Figure 6.. Gene expression changes were associated with H3K4me3 occupancy alterations in CD-exposed Ke-37 cells.

(A, B) Gene expression in Ke-37 cells was assessed by qRT-PCR after exposure to 100 nM (A) or 300 nM (B) of CD. The expression of target genes was normalized to the expression of the housekeeping RPL37A gene. The data were averaged and presented as normalized expression fold change compared with control ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, Wilcoxon–Mann–Whitney test, n = 18 for control, n = 20 for CD (100 nM), and n = 6 each group (300 nM). (C, D) H3K4me3 occupancy at the promoter of target genes in Ke-37 cells was assessed by ChIP-qPCR after exposure to 100 (C) or 300 nM (D) of CD. Values were normalized to the RPLP0 region. The data were averaged and presented as normalized ChIP fold change compared with control ± SD. *P < 0.05, **P < 0.01, Wilcoxon–Mann–Whitney test, n = 7 for each group (100 nM), n = 5 for control, and n = 6 for CD (300 nM).

Figure S7.. H3K4me3 but not H3K9me3 histone marks were decreased in Ke-37 cells after exposure to 300 nM of CD.

Representative images of structurally preserved nuclei stained with anti-H3K4me3 (upper panel) or anti-H3K9me3 (lower panel) antibodies. Pictures were taken with fixed exposure time at ×63 objective. For each mark, quantitative analysis of the nuclear specific signal is presented as relative fluorescence compared with control ± SD. *P < 0.05, Wilcoxon–Mann–Whitney test, n = 5 for each group (H3K4me3), and n = 6 for each group (H3K9me3).

Figure S8.. Chlordecone concentration measured in umbilical cord blood.

The values of CD in umbilical cord blood were presented in Excel graph as ng of CD per ml of blood.

Figure S9.. H3K4me3 occupancy is unaffected at nucleated blood cell–specific markers.

The cell-specific marker H3K4me3 occupancies were shown for CD4 and CD8A (T lymphocytes), CD14 (monocytes), CD56/NCAM1 (natural killer cells), CD19 (B lymphocytes), CD71/TFRC (nucleated red blood cells), and CD15/FUT4 (granulocytes). The images represent the averaged signal of six replicates for control (blue) and CD group (red).