UBE2T resolves transcription-replication conflicts and protects common fragile sites in primordial germ cells

Abstract

The proper development of primordial germ cells (PGCs) is an essential prerequisite for gametogenesis and mammalian fertility. The Fanconi anemia (FA) pathway functions in maintaining the development of PGCs. FANCT/UBE2T serves as an E2 ubiquitin-conjugating enzyme that ubiquitylates the FANCD2-FANCI complex to activate the FA pathway, but its role in the development of PGCs is not clear. In this study, we found that Ube2t knockout mice showed defects in PGC proliferation, leading to severe loss of germ cells after birth. Deletion of UBE2T exacerbated DNA damage and triggered the activation of the p53 pathway. We further demonstrated that UBE2T counteracted transcription-replication conflicts by resolving R-loops and stabilizing replication forks, and also protected common fragile sites by resolving R-loops in large genes and promoting mitotic DNA synthesis to maintain the genome stability of PGCs. Overall, these results provide new insights into the function and regulatory mechanisms of the FA pathway ensuring normal development of PGCs.

Keywords: DNA damage response; Fertility; MiDAS; Replication stress.

Fig. 1

Proliferation defects of Ube2t^–/– PGCs lead to profound germ cells loss a H&E and immunohistochemical staining for DDX4 (a germ cell marker, brown) in the ovaries at PD3 and 3 months. Scale bars = 50 μm. b H&E and immunohistochemical staining for DDX4 (brown) of testes at PD3 and 3 months. Scale bars = 50 μm. c, d Representative images (c) and quantification (d) of wild-type and Ube2t^–/– PGCs using alkaline phosphatase staining of E8.5 embryos, E9.5 embryos, and E11.5 genital ridges. Scale bars = 200 μm. Data from individual embryos are presented as dots and mean ± SD are presented, n = 8/9/8/8/6/6 embryos, ns: not significant, ***P < 0.001. e, f Representative images (e) and percentages (f) of the cell cycle distribution of PGCs in E11.5 wild-type and Ube2t^–/– genital ridges. G1-phase cells, cyclin B1 negative; S-phase cells, EdU positive; G2-phase cells, cyclin B1 strongly positive in cytoplasm; M-phase cells, cyclin B1 accumulation in the nucleus. STELLA-positive indicates PGC, and arrowheads indicate G2-phase PGCs. n = 4 embryos per genotype, ***P < 0.001. Scale bar = 50 μm. Data from individual embryos are presented as dots and mean ± SD are presented

Fig. 2

Loss of UBE2T leads to DNA damage accumulation and p53 signaling activation in PGCs a Immunofluorescence staining for FANCD2 (red), EdU (white), and STELLA (green) to assess FA pathway activation in E11.5 wild-type and Ube2t^–/– genital ridges. The nuclei of S-phase PGCs and S-phase somatic cells are circled out. Arrowheads indicate FANCD2 foci in PGCs. Scale bar = 10 μm. b Representative images of the neutral comet assay for E11.5 wild-type and Ube2t^–/– PGCs. DNA was stained with Hoechst 33342 (white). Scale bar = 50 μm. c Quantification of the OTM in the PGC neutral comet assay. At least 200 PGCs were included per group, ***P < 0.001. d, e Representative images (d) and percentages (e) of 53BP1 foci (green) in PGCs (STELLA-positive) in E11.5 wild-type and Ube2t^–/– genital ridges. The nuclei of PGCs are circled out. Arrowheads indicate 53BP1 foci in PGCs. n = 5 embryos per genotype, ***P < 0.001. Scale bar = 10 μm. f, g Representative images (f) and percentages (g) for p-p53–positive (red) PGCs (STELLA-positive) in E11.5 wild-type and Ube2t^–/– genital ridges. n = 5 embryos per genotype, ***P < 0.001. Scale bar = 10 μm. h, i Representative images (h) and quantification (i) of PGCs (STELLA-positive) in E11.5 genital ridges of the indicated genotypes. n = 6 embryos per genotype, *P < 0.05. Scale bar = 200 μm. Data from individual cells (c) or embryos (e, g, i) are presented as dots, mean values (c) or mean ± SD (e, g, i) are presented

Fig. 3

Loss of UBE2T leads to R-loop and TRC accumulation in PGCs a, b Representative images (a) and quantification (b) of TRCs in PGCs (STELLA-positive) in E11.5 wild-type and Ube2t^–/– genital ridges. PLA (Pol II + PCNA) foci (red) indicate TRCs. At least 200 PGCs were included per genotype, ***P < 0.001. Scale bar = 10 μm. c, d Representative images (c) and quantification (d) of R-loops in PGCs (STELLA-positive) in E11.5 wild-type and Ube2t^–/– genital ridges. The mean S9.6 intensity (red) in the nucleus after subtracting the nucleolar signal indicates the level of R-loops. At least 200 PGCs were included per genotype, ***P < 0.001. Scale bar = 10 μm. Data from individual cells are presented as dots and the mean values are presented (b, d)

Fig. 4

Loss of UBE2T contributes to R-loop and TRC accumulation in MEFs a, b Representative images (a) and quantification (b) of R-loops of wild-type and Ube2t^–/– MEFs overexpressing control (green) or GFP-RNaseH1 (GFP-RNH1, green) adenovirus. The mean S9.6 intensity (red) in the nucleus (except the nucleolar signal that is positive for Fibrillarin staining) indicates the number of R-loops. At least 200 MEFs were included per group, ***P < 0.001. Scale bar = 10 μm. c, d Representative images (c) and quantification (d) of TRCs in wild-type and Ube2t^–/– MEFs overexpressing control (green) or GFP-RNH1 (green) adenovirus. PLA (Pol II + PCNA) foci (red) indicate TRCs. At least 200 MEFs were included per group, ***P < 0.001. Scale bar = 10 μm. (e) Top: The experimental scheme of the DNA fiber assay in MEFs to assess the velocity of RFs. Middle: Representative images of the DNA fiber assay in MEFs. Bottom: CldU track length of the DNA fiber assay in wild-type and Ube2t^–/– MEFs overexpressing control or GFP-RNH1 adenovirus. At least 200 CldU tracks were included per group, ns: not significant, ***P < 0.001. Data from individual cells (b, d) or DNA fibers (E) are presented as dots, means are presented (b, d, e)

Fig. 5

Knockdown of Ube2t causes R-loop accumulation a Annotation of R-loop CUT&Tag peaks with pie charts. The genomic distribution of peaks is shown in the low panel. UTR, untranslated region. b Genomic metaplots of genes with R-loop signals across the 2 kb window around gene bodies. (c) Venn diagrams showing the overlap of R-loop peaks between Ube2t knockdown groups (shUbe2t #1, shUbe2t #2) and control group (shCtrl). d, e Average R-loop read density and heatmap of shUbe2t #1 groups, shCtrl group and no antibody group in the 1 kb window around the center of shUbe2t #1 unique peaks (d) and merged peaks (e). f Genome browser tracks of R-loop coverage at the representative genes Rapgef6 and Taf15 in P19 cells. No antibody (green), control (blue), shUbe2t #1 (orange) and shUbe2t #2 (orange) CUT&Tag data are shown

Fig. 6

Loss of UBE2T leads to increased resection of the nascent DNA in P19 cells and MEFs a, b The experimental scheme, representative images (a) and the CldU/IdU track length (b) in the DNA fiber assay in P19 cells infected with shCtrl, shUbe2t #1, or shUbe2t #2 adenovirus after control, APH (0.2 μM), or HU (0.3 mM) treatment. At least 200 CldU/IdU ratios were included per group, ns: not significant, **P < 0.01, ***P < 0.001. c, d The experimental scheme, representative images (c) and the CldU/IdU track length (d) of the DNA fiber assay in wild-type and Ube2t^–/– MEFs after control, APH (0.2 μM), or HU (0.3 mM) treatment with or without mirin. At least 200 CldU/IdU ratios were included per group. ns: not significant, **P < 0.01, ***P < 0.001. Data from DNA fibers are presented as dots and the mean values are presented (b, d)

Fig. 7

UBE2T is essential for protecting the stability of CFSs a, b Representative images (a) and percentages (b) of 53BP1-NBs in G1-phase PGCs (STELLA-positive) in E11.5 wild-type and Ube2t^–/– genital ridges. Large 53BP1 foci (green) indicate 53BP1-NBs. Cyclin A2-negative (red) and STELLA-positive indicate G1-phase PGCs. The nuclei of G1-phase PGCs are circled out. Arrowheads indicate 53BP1-NBs in PGCs. n = 5 embryos per genotype, ***P < 0.001. Scale bar = 10 μm. c, d Representative images (c) and percentages (d) of binucleated cells with micronuclei in wild-type and Ube2t^–/– MEFs after APH (0.5 μM) or control treatment. DNA was stained with Hoechst 33342 (white), and α-Tubulin (green) indicates the binucleated MEFs. Arrowheads indicate micronuclei in MEFs. Three biological replicates were performed in each group and at least 200 binucleated MEFs were included per replicate, **P < 0.01. Scale bar = 200 μm. e Experimental workflow for analyzing P19 cells in anaphase following UBE2T depletion. f, g Representative images (f) and percentages (g) of UFB-positive P19 anaphase cells following UBE2T depletion. PICH (white) indicates UFBs, and DNA was stained with Hoechst 33342 (blue). Three biological replicates were performed in each group and at least 100 P19 anaphase cells were included per replicate, ***P < 0.001. Scale bar = 10 μm. Data points from individual embryos (b) or groups (d, g) are presented as dots, mean ± SD are presented (b, d, g)

Fig. 8

UBE2T promotes MiDAS by recruiting MUS81 to chromatin a Experimental workflow for analyzing MiDAS in P19 prometaphase cells following UBE2T depletion. b, c Representative immunofluorescence images (b) and quantification (c) of MiDAS foci (labeled with EdU; red) in P19 prometaphase cells following UBE2T deletion. Cyclin B1-positive nuclei indicate cells entering mitosis. DNA was stained with Hoechst 33342 (blue). At least 200 P19 prometaphase cells were included per group, ***P < 0.001. Scale bar = 10 µm. d Quantification of MiDAS foci in wild-type and Ube2t^–/– prometaphase MEFs following APH (0.2 μM) treatment. At least 200 prometaphase MEFs were included per genotype, ***P < 0.001. e Representative western blots of the chromatin-bound fraction of P19 cells at the indicated cell cycle phases for UBE2T, FANCD2, MUS81, POLD3, and Histone H3 (the loading control). Asy, asynchronous; G2, G2 phase; PM, prometaphase; shU#1, shUbe2t #1; shU#2, shUbe2t #2. f Experimental workflow for analyzing metaphase MEFs following APH (0.2 μM) treatment. g Representative images of metaphase chromosomes of wild-type and Ube2t^–/– MEF cells after APH (0.2 μM) treatment. DNA was stained with Hoechst 33342 (white). Arrowheads indicate breaks or gaps in the chromosomes. Magnified images of chromosomes are shown in the red squares. Scale bar = 10 µm. h Quantification of breaks/gaps per metaphase spread from wild-type and Ube2t^–/– MEFs following APH (0.2 μM) treatment. Three biological replicates were performed in each group and at least 100 metaphase MEFs were included per replicate. *P < 0.05. Data from individual cells (c, d) or groups (h) are presented as dots, mean (c, d) or mean ± SD (h) are presented

Fig. 9

UBE2T promotes the resolution of R-loops located at CFSs a Genome browser tracks of R-loop coverage at the large Wwox, Immp2l, and Sec8 loci detected by S9.6 antibody CUT&Tag assay in P19 cells. No antibody (green), control (blue), shUbe2t #1 (orange) and shUbe2t #2 (orange) CUT&Tag data are shown. b Quantification of R-loops by CUT&Tag-qPCR at promoter regions of the large genes Wwox, Immp2l, and Sec8. Data points from individual groups are presented as dots, mean ± SD are presented. Three biological replicates were performed in each group. ns: not significant, *P < 0.05

Fig. 10

The working model of how UBE2T maintains genome stability in PGCs. PGCs encounter endogenous replication stress derived from TRCs and CFSs during rapid division. UBE2T is required to activate the FA pathway to resolve R-loops and stabilize RFs to counteract TRCs. The activated FA pathway also resolves R-loops at large genes and facilitates MiDAS to maintain stability of CFSs. Thus, the FA pathway functions to maintain the genome stability of PGCs