Transgenerational epigenetic effects imposed by neonicotinoid thiacloprid exposure

Abstract

Neonicotinoids are a widely used class of insecticides that are being applied in agricultural fields. We examined the capacity of a neonicotinoid, thiacloprid (thia), to induce transgenerational effects in male mice. Pregnant outbred Swiss female mice were exposed to thia at embryonic days E6.5-E15.5 using different doses. Testis sections were used for morphology analysis, ELISAs for testosterone level analysis, RT-qPCR and RNA-seq for gene expression analysis, MEDIP-seq and MEDIP-qPCR techniques for DNA methylation analysis, and Western blot for a protein analysis. The number of meiotic double-strand breaks and the number of incomplete synapsed chromosomes were higher in the thia 6-treated group of F3 males. Genome-wide analysis of DNA methylation in spermatozoa revealed that differentially methylated regions were found in all three generations at the promoters of germ cell reprogramming responsive genes and many superenhancers that are normally active in embryonic stem cells, testis, and brain. DNA methylation changes induced by thia exposure during embryonic period are preserved through several generations at important master regulator regions.

Figure S1.. Schematic presentation of experiments.

Pregnant outbred Swiss mice were treated from E6.5 to E15.5, with thia doses of 0.6, and 6 mg/kg/day, and control mice “0” received only vehicle (oil). F1 and F3 mice were euthanized at embryonic day 15, postnatal day 35, and at the age of 2 mo. Schema of breeding described in “Mouse treatment and dissection” section of Materials and Methods. For spermatozoa DNA methylation analysis, F2 mice were also euthanized.

Figure S2.. Body and organ weights in 2-mo-old F1 and F3 males.

Male progeny of mice were euthanized at the age of 2 mo, and the murine organs were dissected and weighed. (A, B) Body weights in F1 (A, B) in F3. (C, D) Testis weights in F1 (C) and in F3 (D). (E, F) Epididymis weight in F1 (E) and (F) in F3. Testis and epididymis weights were normalized to body weights. F1: n = 23 dose “0,” n = 17 dose “0.6,” n = 42 dose “6”; F3: n = 23, dose “0,” n = 27 dose “0.6,” n = 35 dose “6,” *P < 0.05, Kruskal–Wallis test. All plots on the figure represent averaged values ± SD.

Figure 1.. Effects of ancestral thiacloprid exposure on the reproductive system.

(A) Spermatozoa numbers in F3 males. Spermatozoa were counted in the epididymis of 2-mo-old F3 male mice, the data were averaged and compared with the control, n = 10 dose “0,” n = 7 dose “0.6,” n = 7 dose “6,” *P < 0.05, Kruskal–Wallis test. (B, C) Testosterone levels (B) in F1 and (C) in F3 blood serum of 2-mo-old males, F1, n = 20 dose “0,” n = 23 dose “6”; F3, n = 20 dose “0,” n = 14 dose “6,” *P < 0.05, nonparametric Mann–Whitney test. (D) A representative image of paraffin 35-d-old F3 male testis section immunostained by γH2AX in control (top) and in the thia 6-derived (bottom) mice, bar is 50 µΜ, images were taken using 20x objectives, the last images are 63x magnification images, z, zygotene cells, p, pachytene cells, bar is 20 µΜ. (E) Quantitative analysis of cells in seminiferous tubules in 35-d-old testis of F3 males, n = 6 dose “0,” n = 6 dose “6,” *P < 0.05, nonparametric Mann–Whitney test. (F) A representative image of paraffin 35-d-old testis section immunostained by PCNA in F3 control (top) and in the F3 thia 6-derived (bottom) mice. PCNA protein presents in leptotene–zygotene and pachytene–diplotene spermatocytes in mouse (white arrow), bar is 50 µΜ. The last images are 63x magnification images, white arrows indicate a PCNA-positive cells, bar is 20 µΜ. (G) Quantitative analysis of PCNA-positive cells in seminiferous tubules in 35-d-old testis F3, n = 4 dose “0,” n = 4 dose “6,” *P < 0.05, nonparametric Mann–Whitney test. (H) Quantitative RT‒qPCR analysis of germ cell population markers in 35-d-old testis F3 males, n = 8 dose “0,” n = 8 dose “6.” *P < 0.05, **P < 0.01, nonparametric Mann–Whitney test. All plots are averaged values ± SD, SG, spermatogonia.

Figure S3.. Testis morphology in F1 and F3 males.

Mice were euthanized at 35 d of age. The testes were fixed in Bouin’s solution and embedded in paraffin. Paraffin sections were stained with H&E and cells were counted in stage six to seven tubules. (A) Representative images of testes from mice derived from treatment doses of 0, 0.6 and 6 mg/kg/day, bar is 100 µΜ. (B) Number of cells per tubule in F3. (C) Lumen size in F3 males, F3, n = 6 dose “0,” n = 7 dose “0.6,” n = 6 dose “6,” Kruskal–Wallis test. All plots on the figure represent an averaged values ± SD.

Figure 2.. Meiotic defects in exposed mice.

(A) Representative image of 35-d-old F3 male testis cell spreads immunostained against central, SYCP1 (red) and lateral, SYCP3 (green) components of the synaptonemal complex, white circle indicates sex chromosomes. (B) Quantitative analysis of the synapsing defects in 35-d-old testis, FS, fully synapsed, NC, not complete, MC, multiple connections, n = 4 dose “0,” n = 4 dose “6,” non-parametric Mann-Whitney test, bar is 5 µΜ. (C) Representative image of pachytene stage immunostained by DMC1 (red) and SYCP3 (green) in control (top) and thia 6-derived (bottom) 35-d-old testis cell spreads, bar is 5 µΜ. (D) Quantitative analysis of DMC1 foci in F3 35 d-old generation males, sex and autosome (Aut.) chromosome (Chr.) foci were counted independently and compared with control samples, n = 4 dose “0,” n = 4 dose “6,” *P < 0.05, nonparametric Mann–Whitney test. All plots on the figure represent an averaged value ± SD.

Figure 3.. The effects of ancestral thiacloprid exposure on H3K9me3 levels and retroelement transcription activity.

(A) Representative image of 35-d-old F3 male testis spreads immunostained against H3K9me3 (red) or SYCP3 (green), bar is 5 µΜ. (B) Quantitative analysis of H3K9me3 intensity in F3 generation males, n = 5 dose “0,” n = 4 dose “6,” *P < 0.05 nonparametric Mann–Whitney test. (C) Representative image of H3K9me3 Western blot of testis tissue, c1-c5 are control, t1-t5 are thia -derived samples. (D) Quantitative analysis of Western blot, n = 5 dose “0,” n = 5 dose “6,” **P < 0.01 nonparametric Mann–Whitney test. (E) Quantitative ChIP‒qPCR analysis of H3K9me3 in 35-d-old testis of F3 males, n = 5 dose “0,” n = 5 dose “6,” *P < 0.05, **P < 0.01 nonparametric Mann–Whitney test. (F) The analysis of retroelement expression in 35-d-old testis of F3 males by RT–qPCR, n = 7 dose “0,” n = 7 dose “6.” All plots on the figure represent an averaged values ± SD.

Figure S4.. Uncut Western blot images.

Histones were extracted from testis of 35-d-old F3 male mice, n = 5 dose “0,” n = 5 dose “6.” (A, B) H3K9me3 WB image, (B) Ponceau Red-stained membrane.

Figure S5.. The analysis of RNA-seq data in F1 and F3 embryonic testis.

(A, B, C) Principal component analysis (PCA) plot, (B) a sample-to-sample dispersion heatmap, and (C) MA-plot in F1 RNA-seq. (D, E, F) PCA plot, (E) a sample-to-sample dispersion heatmap and (F) MA-plot in F3; n = 3 dose “0,” n = 3 dose “6” in both F1 and F3 generations. Plots were generated by Deseq2 package which uses Wald test for statistical significance assessment. PCA plots show clusters of samples based on their similarity. MA plots represent log fold-change versus mean expression between control and exposed group, sample-to-sample dispersion heatmaps visualize the variations between replicates.

Figure 4.. RNA-seq analysis in embryonic E15.5 F1 and F3 testes.

(A) Volcano plot of differentially expressed genes in F1 testis, down-regulated genes are in blue and up-regulated are in red, n = 3 dose “0,” n = 3 dose “6.” (B) Functional annotation of the down-regulated (blue) or up-regulated (red) differentially expressed genes in F1 embryonic testis were done using DAVID, Fisher’s exact test was adopted in DAVID to measure the gene-enrichment in annotation terms, the genes were sorted by P-value in F1. (C) Volcano plot of differentially expressed genes in F3, n = 3 for dose “0” and dose “6.” (D) The down-regulated (blue) and up-regulated genes (red) were annotated separately by using DAVID. The genes were sorted by adjusted P-value in F3. The bars represent the number of genes in each Gene Ontology group. (E) Quantitative RT‒qPCR analysis of DEGs in embryonic F3 testis, the RNA-seq fold change (FC) values are indicated under the graph. RT–qPCR plots represents an averaged values ± SD, n = 5 dose “0,” n = 4 dose “6.” *P < 0.05, nonparametric Mann–Whitney test.

Figure S6.. A ratio heatmap in F1 and F3 of peaks located near the genes.

MEDIP sequencing reads were mapped against a reference mm10 genome, and the resulting .bam files were normalized and combined for each group. The bed files were generated from the resulting bam files and were used for the analysis by EaSeq, the signal was calculated and sorted by Log2FC.

Figure S7.. MEDIP-seq analysis in spermatozoa of F1, F2, and F3.

(A, B, C, D, E, F, G, H, I, J, K, L) A coverage map in (A, B) in control, (C, D) and 6 mg/kg/day thia F1, (E, F) in control F2 and (G, H) in 6 mg/kg/day thia F2, (I, J) in control F3 and (K, L) in F3 6 mg/kg/day thia. The X-axis represents the 50,000 bp surrounding each region, which was segmented into 400 bins and smoothed for 1 bin. Plots were generated by EaSeq.

Figure S8.. The analysis of MEDIP-seq data in F1, F2, and in F3.

(A, B, C) Principal component analysis plot (A) in F1, (B) in F2, and (C) in F3. Principal component analysis plots show a cluster of samples based on their similarity. Control samples are shown in blue and thia -derived are shown in red. (D, E, F) Volcano plots (D) in F1, (E) in F2, and (F) in F3; F1: n = 3 dose “0,” n = 3 dose “6”; F2: n = 3 dose “0,” n = 4 dose “6”; F3: n = 3 dose “0” n = 4 dose “6.” The regions with decreased DNA methylation regions are shown in blue, regions with increased DNA methylation are shown in red, and nonsignificant regions are in grey.

Figure S9.. Pie plot presentation of the location of DMRs.

(A, B, C) The location of DMR (A) in F1, (B) in F2, and (C) in F3. Plots were generated by ChIPseeker using the coordinates of the differentially methylated regions and reference annotation file mm10.refGene.gtf.

Figure 5.. Genome-wide DNA methylation analysis in spermatozoa of the 2-mo-old F1, F2, and F3 mice.

(A) A representative image of mapped reads. The differentially methylated peaks near Dnmt3a, Sox9, and Gsc are shown in red dashed boxes. Plots from control samples are shown in blue, plots from thia-derived samples are shown in red. Each control and treatment groups contained minimum three replicates. The sequencing reads were mapped to the reference mm10 genome, normalized, and converted to bedgraph files which were visualized in IGV, the signal intensity is shown in brackets, and the differential peaks are marked by dashed box. (B) Venn plot represents common genes located in DMRs in F1, F2, and F3. (C) Functional annotation of genes located in common DMRs, bars sorted by adjusted P-values, and each bar represents the number of genes in each group. (D, E, F) Functional annotation “biological process” of genes located in DMRs (D) in F1, (E) in (F2) and (F) in F3 was performed by GREAT. Each bar represents –log 10 (binominal P-value) calculated by GREAT. (G, H) MEDIP-qPCR analysis in F1 and (H) in F3 E15 embryonic testis. Right plots are MEDIP–qPCR analysis of spermatogenesis and SE genes from spermatozoa of 2-mo-old males. F1, n = 4 dose “0,” n = 4 dose “6,” F3, n = 4 dose “0,” n = 4 dose “6.” The right plots represent the analysis of MEDIP-qPCR analysis of retroelements, F1: n = 6, dose “0,” n = 6 dose “6”; F3: n = 4 dose “0” and n = 4 dose “6.” MEDIP-qPCR plots are averaged MEDIP values ± SD, *P < 0.05, **P < 0.01 nonparametric Mann–Whitney test.

Figure S10.. Plots of DMRs at the Ddx4 and Sycp2 GRR gene.

Control plots are shown in blue; plots from thia-derived mice are in red. Each group contained minimum three replicates in each generation.

Figure 6.. DNA methylation of SE and GRR regions changed in spermatozoa.

(A) At the vicinity of regions including superenhancers active in ES cells (Mescd1, Nanog). Each plot represents the normalized sequencing reads, the signal was calculated for each sample, and the statistically significant difference was calculated by the Limma test. Overlaid biological replicate plots from the control group are shown in blue, and overlaid biological replicate plots from the treatment group are shown in red. Signal intensity is shown in the brackets. (B, C) MEDIP–qPCR analysis in embryonic E15.5 testis (B) in F1 and (C) in F3, F1 and F3 n = 6 dose “0,” n = 6 dose “6.” MEDIP–qPCR plots represent averaged values of qPCR ± SD, *P < 0.05, **P < 0.01, nonparametric Mann–Whitney test. (D, E) Differentially expressed genes in RNA-seq have DMRs in MEDIP-seq data (D) in F1 and (E) in F3. The genes were sorted by P-value (in F1) or by adjusted P-value (in F3).

Figure S11.. Gene expression analysis in 35-d-old testes.

(A, B) RT–qPCR in F1 and (B) RT–qPCR in F3. In F1 and F3, n = 8 dose “0,” n = 8 dose “6,” *P < 0.05, nonparametric Mann-Whitney test. All plots on the figure represent averaged values ± SD.

Figure S12.. Gene expression analysis in the cerebellum of 2-mo-old adult mice.

(A, B) RT-qPCR analysis in F1, (B) in F3; F1 and F3 n = 7 dose “0,” n = 7 dose “6,” *P < 0.05, **P < 0.01, nonparametric Mann–Whitney test. All plots on the figure represent an averaged values ± SD.