Dose dependent impact of epimutagenesis measured in iPSCs exposed to 1, 50, and 100 μM BPS.

a) Ideogram plots displaying chromosomal distribution of genome-wide changes in DNA methylation caused by BPS exposure. b) Mean difference (MD) plots of changes in gene expression following exposure to increasing doses of BPS. Exposure to increasing doses of BPS induced corresponding increases in the numbers of DMCs, DMRs, and DEGs. Blue horizontal lines = hypomethylated DMCs, red horizontal lines = hypermethylated DMCs, black squares = DMRs.

Chromosomal distributions and annotations of BPS-induced epimutations in pluripotent, somatic, and germ cell types.

a) Ideograms illustrating chromosomal locations of DMCs induced by exposure of each cell type to 1 μM BPS. Blue horizontal lines = hypomethylated DMCs, red horizontal lines = hypermethylated DMCs. b) Enrichment plots indicating feature annotations in genomic regions displaying prevalent BPS-induced epimutations in each cell type.

Treatment-specific differentially methylated sites (DMCs) (treated vs. control)

Correlation between cell type-specific expression of estrogen receptors and density of genomic EREs associated with BPS-induced epimutations.

a) Assessment of expression of ERα and ERβ on cell types co-stained for known cell type-specific markers. Somatic cell types express both receptors, pluripotent cells express ERβ but not ERα, and germ cells do not express either estrogen receptor. b) Motif plots displaying the full ERE consensus sequence and the more biologically relevant ERE half-site motifs found to be enriched from ERα ChIP-seq. c,d) Normalized density plots and box plots displaying the frequency of ERE half-sites identified c) within 500 bp of all BPS-induced DMCs genome-wide, or d) within 500 bp of the most enriched categories of BPS-induced DMCs in each cell type (= enhancer regions for somatic and pluripotent cell types and promoter regions in the germ cell type).

Direct comparison of BPS exposure-specific and cell type-specific features between cell types.

a) Assessment Venn diagrams indicating DMCs that are due either to BPS exposure (top, smaller ovals) or inherent cell type-specific differences (bottom, larger ovals). Numbers of apparent endocrine signaling-related DMCs are shown in the light orange arrow, and apparent endocrine signaling-independent DMCs are shown in the dark orange arrows. Enrichment plots indicating feature annotations in genomic regions displaying b) apparent endocrine signaling-related DMCs occurring predominantly in enhancer regions in somatic Sertoli and granulosa cell types or pluripotent cells expressing one or more estrogen receptors, or c) a smaller set of apparent endocrine signaling-independent DMCs occurring predominantly in promoter regions in all four cell types regardless of +/- expression of relevant endocrine receptors. d) Normalized density plots and box plots displaying the frequency of ERE half-sites identified within 500 bp of apparent endocrine signaling-related DMCs occurring predominantly in enhancer regions and apparent endocrine signaling-independent DMCs occurring predominantly in promoters.

Exposure-specific differentially expressed genes*

Correlation between the proximity of DMCs to promoters and dysregulation of gene expression.

a) Proximity plot displaying distances from exposure-specific DMCs to nearest promoter regions. Dotted lines indicate median points of the data for each cell type. b) Correlation plot displaying a negative relationship between the distance from DMCs to nearest promoters and resulting dysregulation of gene expression within each cell type.

Potential involvement of non-canonical estrogen signaling pathways in BPS-induction of epimutations.

Relative expression of genes a) enriched for apparent endocrine signaling-independent promoter-region DMCs or b) dysregulated in PGCLCs which lack expression of estrogen receptors. c) Heatmap of relative expression of estrogen receptor genes (Esr1 and Esr2) and G-coupled protein receptors (Gprc5a, Gpr107, Gprc5b, Gpr161, and Gpr89) in pluripotent, somatic, and germ cell types.

Persistence of BPS-induced epimutations through recapitulation of early germline reprogramming in vitro.

a) Schematic illustrating derivation of PGCLCs from iPSCs in vitro. iPSCs are first induced to form EpiLCs which are then induced to form PGCLCs. iPSCs were exposed to either ethanol + 1 μM BPS or ethanol (carrier) only, then induced to undergo transitions to form EpiLCs and then PGCLCs. b) DNA samples from BPS-exposed or control iPSCs as well as subsequently derived PGCLCs were assessed for DNA methylation epimutations by EM-seq. BPS-treated iPSCs showed 38,105 DMCs and subsequently derived PGCLCs showed 28,169 DMCs. Of those, only 1,417 (3.7%) of the DMCs were conserved from the BPS-exposed iPSCs to the subsequently derived PGCLCs. c) RNA samples from BPS-exposed or control iPSCs and subsequently derived PGCLCs were assessed for global gene expression patterns by RNA-seq. BPS-treated iPSCs showed 1,637 exposure-specific DEGs and subsequently derived PGCLCs showed 1,437 exposure-specific DEGs. Of those, only 138 (8.4%) were conserved from the BPS-exposed iPSCs to the subsequently derived PGCLCs.

Differential expression of potential endocrine-signaling independent DMCs.

Heatmap displaying the relative expression of genes with promoters enriched for apparent endocrine-signaling independent DMCs. The majority of these genes are actively expressed in all cell types and an overall largely similar pattern of expression across all cell types.

Genome-wide annotation of ERE half-sites.

Venn diagram displaying the identification of ERE half-sites localized in known genic and intergenic regions.

Summary of ERE annotations

ICC validation of MF5-9-1 iPSCs.

iPSCs from reprogrammed MEFs were validated for immunolabeling with known pluripotency markers along with negative control.

N2 media components (5 mL)
N2B27 media components (1 L)
N2B27 2i-LIF media components (40 mL)

N2 EpiLC differentiation (∼5 mL)

GK15 (10 mL)

PGCLC differentiation media (10 mL)

TrypLE wash media (75 mL)

ICC control staining of cell type-specific markers.

Validation of immunolabeling for known pluripotent and somatic cell type markers in iPSCs, Sertoli Cells, and Granulosa cells along with negative and positive controls.

FACS sorting for ITGB3/FUT4 enriched primordial germ cell like cells.

a) Gating for cells. b.1-2) Gating for singlet cells. c) Gating for live cells. d.1-2) Single color control ITGB3 positive cells. d.3-4) IgG isotype control. e.1-2) Single color control FUT4 positive cells. f.3-4) IgG isotype control. g) Sorting for PGCLC-enriched ITGB3/FUT4 double positive population (2.18% of total cells)

Relative expression of markers for PGCLC induction from iPSCs.

a) ICC of pluripotency and germ cell marker expression throughout the transition from iPSCs to PGCLCs. b) qRT-PCR of pluripotency, epiblast, & germ cell markers indicating gene expression profiles during induction of PGCLCs from iPSCs. Each gene fold expression is relative to the housekeeping gene Gusb using the ΔCq method. The symbol * indicates that the expression of transcripts in the sample were either non-existent or so low as to be undetectable by qRT-PCR.

Primer designs

Chemical exposure experimental workflow. 1)

Cells are passaged into T-25 flasks with filter caps. 2) Mixed blood gas (carbon dioxide 5%, oxygen 5%, and balance nitrogen) is filtered and bubbled into media to prepare media to be mixed with diluted chemical treatment and added to air-tight cell culture flasks. 3) Media is transferred into glass vials and diluted chemical is added. 4) T-25 filter caps are replaced with air-tight caps with septums and cell media containing chemicals is added via syringe and left for 24 hours. 5) After 24 hours, chemical-containing media is removed and cells are washed with buffer. 6) T-25 air-tight septum caps are replaced with filter caps and cells are cultured for an additional 24-hour “chase” period.

Quality control for Infinium Mouse Methylation BeadChip Array data.

a) Non-linear correction of dye bias removal from sample data. b) Background signal subtraction from samples to limit noise. c) Prediction of correct C57B6 mouse strain from samples included in the study based on built-in controls on Infinium Mouse Methylation BeadChip Array. d) Average CpG probe detection success of 97.58% percent across all samples indicating efficient bisulfite conversion of all samples.

Quality control for RNA-seq data.

RNA-seq quality control data from a one of the three PGCLC replicates exposed to BPS as an exemplar a) Base calls had high quality scores (phred scores > 30) for all bases in reads. b) Reads showed equal distributions of all four bases following the initial adaptor sequence and sufficient base complexity. c) Distribution of GC sequences across reads aligns very closely with the theoretical distribution. d) Duplication plot indicates deduplicated libraries contained ∼67% unique sequences which indicates sufficient library complexity for subsequent downstream data processing.

Quality control metrics for EM-seq data.