Specification and epigenomic resetting of the pig germline exhibit conservation with the human lineage

Abstract

Investigations of the human germline and programming are challenging because of limited access to embryonic material. However, the pig as a model may provide insights into transcriptional network and epigenetic reprogramming applicable to both species. Here we show that, during the pre- and early migratory stages, pig primordial germ cells (PGCs) initiate large-scale epigenomic reprogramming, including DNA demethylation involving TET-mediated hydroxylation and, potentially, base excision repair (BER). There is also macroH2A1 depletion and increased H3K27me3 as well as X chromosome reactivation (XCR) in females. Concomitantly, there is dampening of glycolytic metabolism genes and re-expression of some pluripotency genes like those in preimplantation embryos. We identified evolutionarily young transposable elements and gene coding regions resistant to DNA demethylation in acutely hypomethylated gonadal PGCs, with potential for transgenerational epigenetic inheritance. Detailed insights into the pig germline will likely contribute significantly to advances in human germline biology, including in vitro gametogenesis.

Keywords: DNA demethylation; X-chromosome reactivation; epigenetic resetting; escapees; germ cells; pig; single-cell RNA-seq; transgenerational inheritance.

Graphical abstract

Figure 1

Transcriptional profile of pPGCs and comparison with hPGCs (A) Bright-field top view image of a pig embryo (left) and diagrammatic representation (right) showing key structures. Bottom image: IF staining of a midline sagittal section of an E14 embryo, showing a PGC cluster (white arrow) in the caudal end. Scale bar, 20 μm. (B) UHC clustering of all expressed genes. A subset of selected marker genes was used for the heatmap. Color scale unit, log-transformed transcripts per kilobase million (TPM). (C) t-SNE showing relationships between E11 Epi cells, E14 somatic cells, and E14 and E31 PGCs. (D) UMAP plot showing integration of cyPGCs (E13-55) and pPGCs (E14-31) and somatic cells. (E) UMAP plot showing integration of hPGCs (weeks 4–7) and pPGCs (week 5). (F) Expression profiles of pluripotency genes and lineage markers in pPGCs, hPGCs, cyPGCs, and somatic cells. Wk4-7: Weeks 4–7; cy ePGCs: early cyPGCs (E13–E20); Cy lPGCs: late cyPGCs (E36–E55); cy gast soma: cy gastrulating cells (E13–E20); CS7: Carnegie stage 7; S, soma; P, PGCs. (G) Schematic highlighting species differences in expression of key PGC genes. See also Figures S1 and S2 and Tables S1 and S4.

Figure 2

Active DNA demethylation in pre-migratory pPGC. (A) IF staining for 5hmC and 5mC in a E14 PGC cluster (yellow dashed lines). PGCs are marked by SOX17, Sda/GM2, and Nanog. Scale bar, 20 μm (B) 5hmC and 5mC levels determined by LC-MS/MS. Methylation levels are indicated relative to total levels of deoxyguanine (dG). The p values are based on combined ANOVA and Holm’s post hoc test. Data points indicate biological replicates. (C) Expression of epigenetic modifiers for DNA methylation/demethylation and BER pathway components in E11 Epis, E14 somata, and E14 and E31 PGCs. (D) IF staining for UNG and PARP. The yellow circle marks PGCs. Scale bar, 20 μm. (E) Expression heatmap of epigenetic modifiers differentially expressed in pPGCs, hPGCs, and cyPGCs compared with somata. Cy ePGC, early cyPGC (E13–E20); Cy lPGC, late cyPGC (E36–E55); Cy Gast soma, cynomolgus monkey gastrulating cells (E13–E20). Gray color in the heatmap indicates not available. Z scores of log-transformed matrices were used. Because different expression units are used for each species, values in the color scale are replaced by HIGH and LOW. See also Figure S3.

Figure 3

Histone remodeling in pre-migratory, early migratory, and gonadal pPGCs (A) H3K27me3 IF in pPGCs. Yellow dashed lines mark PGCs. Scale bar, 20 μm. (B) Quantification of H3K27me3 in PGCs. The red line indicates the median value. Significance was determined by Mann-Whitney U test. (C) Violin plot showing expression of H2AFY. (D) IF of macroH2A1. Scale bar, 20 μm. PGCs are shown by yellow dashed lines. See also Figure S3.

Figure 4

XC reactivation in pre-migratory pPGCs (A) IF staining for H3K27me3. Xi-associated H3K27me3 is detected in somatic cells (arrows). The yellow dashed circle marks PGCs. Scale bar, 20 μm. (B) Expression of KDM6A in E14 cells. The p value was determined by Mann-Whitney U test. (C) Expression of XIST in E14 somatic cells and E14 and E31 PGCs. F, female; M, male. (D) Female-to-male expression ratio of XC genes versus autosomes (chr1, chr2, and chr3) in E14 somatic cells, E14 PGCs, and E31 PGCs. (E) Median female-to male-expression ratio across XC. The p values (^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001) were determined by pairwise Wilcoxon test. Presumed XCI is indicated in green. (F) Biallelically detected SNPs on XC genes. Each dot represents one biallelically detected SNP. x axis, sum of reads (RPM) that are mapped to the reference alleles; y axis, sum of reads (reads per kilobase million [RPM]) that are mapped to the alternative alleles. (G) Number of biallelically expressed genes. The p value was determined by Kruskal-Wallis test followed by Dunn’s test. See also Figure S4.

Figure 5

PBAT reveals the basal level of methylation in gonadal pPGCs (A) CpG methylation levels in 1-kb genomic tiles of week 5 (E35) female and male pPGCs and gonadal somatic cells. Black points indicate the median. (B) Averaged CpG methylation level profiles of all genes from 5 kb upstream (−) of transcription start sites (TSSs) through scaled gene bodies to 5 kb downstream (+) of transcription end sites (TESs). Different y axes are used for pPGCs and somatic cells because of the extremely low level of methylation in pPGCs. (C) Violin plots showing CpG methylation levels in different genomic features. (D) CpG methylation levels of imprinted regions in pPGCs and somata. (E) Top: proportion of demethylated loci (meth−) and demethylation-resistant loci (meth+) in week 5 pPGCs, week 7–9 hPGCs, and E13.5 mPGCs (the number of meth+ and meth− 800-nt genomic tiles are indicated in the pie chart). Bottom: CpG methylation levels of meth− and escapees (meth+) in three species. White dots indicate the median, and black bars indicate the interquartile range. (F) Distribution of TE families that overlap with TE-rich escapees in week 5 pPGCs. Enrichment scores (ESs) of more than 2 for all Tes are shown, except for those marked by am asterisk, which had a score below 1. An ES above 2 and p < 0.001 (determined by Fisher’s test) indicates that the TE family is more frequent than what would be expected by chance. (G) Examples of TE-rich escapee loci overlapping with L1_SS, L1-2_SSc, and Pre0_SS. See also Figure S5 and Table S5.

Figure 6

Common and unique features in DNA demethylation escapees between mouse, human, and pig (A) Distribution of TE-poor (<10% overlap with TEs) and TE-rich (≥10% overlap) escapees in week 5 pPGCs, week 7–9 hPGCs, and E13.5 mPGCs. The number of escapees (n) was determined by methylation level (at least 30% in human and 15% in pig and mouse). (B) Overlap of syntenic TE-poor escapees among pig, human, and mouse. Escapee regions in pig (205) and mouse (208) were lifted over to compare with syntenic regions in the human genome. (C) Overlap of homologous TE-poor escapee genes among pig, human, and mouse. (D) The TE-poor escapee regions within SORCS2 and PLCH2 are conserved between human and pig, whereas a pig-specific escapee is identified within FTO. (E) Diagram of the events in the pig germline. SP, germline specification; OxPhos, oxidative phosphorylation. A dashed line indicates expected DNA synthesis. See also Figure S6.