Maternal PRDM10 activates essential genes for oocyte-to-embryo transition

Abstract

PR/SET domain-containing (PRDM) proteins are metazoan-specific transcriptional regulators that play diverse roles in mammalian development and disease. Several members such as PRDM1, PRDM14 and PRDM9, have been implicated in germ cell specification and homoeostasis and are essential to fertility-related processes. Others, such as PRDM14, PRDM15 and PRDM10 play a role in early embryogenesis and embryonic stem cell maintenance. Here, we describe the first PRDM family member with a maternal effect. Absence of maternal Prdm10 results in catastrophic failure of oocyte-to-embryo transition and complete arrest at the 2-cell stage. We describe multiple defects in oocytes, zygotes and 2-cell stage embryos relating to the failure to accumulate PRDM10 target gene transcripts in the egg. Transcriptomic analysis and integration of genome-wide chromatin-binding data reveals new and essential PRDM10 targets, including the cytoskeletal protein encoding gene Septin11. We demonstrate that the failure to express maternal Septin11, in the absence of maternal PRDM10, disrupts Septin-complex assembly at the polar body extrusion site in MII oocytes. Our study sheds light into the essentiality of maternal PRDM10, the requirement of the maternal Septin-complex and the likely evolutionary conservation of this regulatory axis in human female germ cells.

Fig. 1. Prdm10 is a maternal effect gene.

a Expression heatmap (RNA-seq) of Prdm family members in mouse (n = 7, GV1-GV7) and human (n = 14, GV1-GV14) GV (germinal vesical) oocytes. (Prdm7 is primate-specific; Prdm10 is highlighted.) b Expression heatmap (RNA-seq) of Prdm family members in oocytes and early embryos. c RPF (ribosome-protected mRNA fragment) heatmap of Prdm family members in oocytes and early embryos. (FGO [fully grown oocytes], LPI [late prometaphase I oocyte], MII [Metaphase II oocyte], zygotes (PN3/5 [pronuclear stage 3/5]) and preimplantation stages (E2C/L2C [early/late 2-cell stage], 4C/8C [4-cell/8-cell stage], ICM [inner cell mass]). d Expression heatmap (RNA-seq) of Prdm family members in NGO (non-growing oocytes), GO1/2 (growing oocytes) and FGO. e Prdm10 expression in GV oocytes after Zp3-Cre-induced deletion (normalized to Trim28 and shown relative to Ctr expression levels). f Pups/litter born to Ctr and MatKO females mated to wildtype males (n = 15 and n = 8 observed mating plugs for Ctr and MatKO producing mothers, respectively; showing median, upper and lower quartiles; ****, p = 7.9*10⁻¹¹, unpaired, two-tailed, parametric t-test. g In vitro culture of Ctr and MatKO zygotes (E0.5) to the blastocyst stage (E3.5) (scale bar: 100 μm). h Percentage of embryos developing across preimplantation stages to blastocyst in culture (n = 59/106 Ctr or MatKO embryos, respectively). i Representative still images from Supplementary Movie 1 (a) and Supplementary Movie 2 (b) of a Ctr and Prdm10 MatKO zygote, respectively, at various timepoints across first cleavage division. Injected cRNA for H2B-RFP (red) labels chromatin while membrane-targeted GFP (green) labels membranes (scale bar: 50 μm). j Symmetric vs asymmetric blastomere size in Ctr and MatKO 2-cell stage embryos respectively (red arrowheads indicate smaller blastomere, scale bar: 100 μm). k Quantification of skewed blastomere size ratios (****, p = 2.1*10⁻⁵, unpaired, two-tailed, parametric t-test). At least three experimental replicates for all embryo isolations and counts were performed.

Fig. 2. Loss of maternal Prdm10 causes oocyte defects and abnormal zygotic progression.

a Quantification of abnormal spindle occurrence in control (n = 30) and MatKO (n = 29) MII oocytes. In mutants, 14% of MII oocytes showed spindles resembling anaphase, 10% had apolar or multipolar spindles and 7% had misaligned chromosome (b) Exemplary MatKO MII oocytes spindles stained for alpha-tubulin (grey) and DNA (blue). c Spindle width/length ratio in apparently normal MatKO (n = 20) and Ctr (n = 29) MII oocytes (****, p = 7*10⁻⁸, unpaired, two-tailed, parametric t-test) with exemplary images (stained for DNA and tubulin). d Mutant MII oocytes (n = 22) show significantly higher occurrence of aneuploidies compared to controls (n = 24) (****, p = 2.34*10⁻⁹, two-tailed Fisher’s exact test). e Quantification of kinetochores shows predominantly increased genomic content in MatKO vs Ctr oocytes. f Transmission light images of Ctr and MatKO zygotes isolated after natural mating. Spermatozoa in the perivitelline space and/or attached to the zona pellucida of MatKO zygotes indicated with black arrowheads. g Maximal projections of DNA stained Ctr and MatKO zygotes revealing multiple spermatozoa attached to MatKO zygotes (white arrowheads). Dotted, dashed, and solid circles delineate cell borders, pronuclei and polar bodies, respectively. h Quantification of MatKO (n = 57) and Ctr (n = 44) zygotes with overt accumulation of spermatozoa after natural mating. i Cortical granule labelling using FITC-labelled Lens culinaris agglutinin (LCA) in MatKO and Ctr GV or MII oocytes, respectively. j Still images from Supplementary Movie 1 (a) and Supplementary Movie 2 (b) with chromatin labelling (H2B-RFP) revealing multiple, pronuclear-like structures in Prdm10 MatKO zygotes. k Quantification of ‘poly-pronuclei’ phenotype (MatKO n = 38; Ctr n = 65; ****, p < 1*10⁻¹⁵, two-tailed Fisher’s exact test). l Maximal projections zygotes with labelled DNA (DAPI, blue), pronuclei (TRIM28, red) and maternal-derived chromatin (H3K9me3, green). Blue circles indicate paternal pronuclei (H3K9me3-negative), red circles indicate maternal-derived pro-/micronuclei, white circles indicate polar bodies [PB], asterisks indicate abnormally associated spermatozoa, dashed white circles indicate zygotes’ circumference. (all scale bars: 50 μm).

Fig. 3. PRDM10 regulates the maternal transcriptome.

a Volcano plot of RNA-seq data from MatKO and Ctr GV oocytes (grey: FDR [false discovery rate]≤0.05 and baseMean≥10; black: genes FC [fold change]≥2; purple: genes FC ≥ 2 and promoter bound by PRDM10 in ESCs). Genes of interest are highlighted. b Gene ontology analysis overview of regulated genes (FDR≤0.001) (top) and ‘bound&regulated’ genes (FDR≤0.05) (bottom). c Heatmap representation of PRDM10 binding signal (mESCs) of peak centres (+/−1kb) and associated chromatin features of active transcription/DNA accessibility in growing (GO1)(ATAC) and fully grown oocytes (FGO)(ATAC, H3K4me3, H3K36me3). d Proportion of genes significantly (FDR≤0.05) up- or downregulated (FC ≥ 2) amongst PRDM10 bound/unbound fractions. Direct PRDM10 targets are predominantly downregulated. e Heatmap of all deregulated genes (FDR≤0.05, baseMean≥10, FC≥2) ranked from most down- to upregulated with PRDM10 promoter-binding indicated. ‘Bound&regulated’ genes defining GO categories ‘DNA damage response’ (pink) and ‘bacterial invasion of epithelial cells’ (blue) are highlighted. f PRDM10-dependent relative expression, RNA-seq track and promoter-binding peaks of formerly unknown target Sept11. Expression in 2 Ctr and 3 MatKO pools of oocytes is shown, respectively. Three overlayed tracks for PRDM10 ChIP in iKO (inducible knock out) and control mESCs at the Sept11 promoter are shown, respectively (GSE135022). Motif location in respect to peak is indicated (red bar). g PRDM10 binding motif (top) and matching peak-underlying sequence in the Sept11 promoter in mouse with respective fully conserved motifs in humans.

Fig. 4. Septin family gene expression and PRDM10-dependent regulation in early development.

a Heat maps showing mRNA transcription of Septin family genes during oogenesis (RPKM), OET and preimplantation development (FPKM). b Heat map showing RPFs (ribosome protected fragments) of Septins in oocytes and preimplantation embryos (FPKM). c Heat map showing protein intensities of Septin family members in oocytes and preimplantation embryos by LC-MS/MS. Septin members are clustered in functional homology groups. d Predicted Septin hexamer complex based on transcription and proteomic data in Ctr (left) and Prdm10 MatKO oocytes (right). e Expression heatmap from RNA-seq data from mouse Ctr/MatKO oocytes, Ctr/ZygKO 8-cell embryos and Ctr/iKO mESCs, respectively. Septin members are clustered in groups based on structural and functional similarity (column 1). PRDM10 promoter-binding as defined by ChIP-seq in mESCs is indicated (column 2). Only Sept11 is bound by PRDM10 and regulated by PRDM10 (red box/asterisk) in oocytes and mESCs. f Schematic illustration of oocyte activation and staining protocol to detect maternal Septin complex at polar body abscission site. g Immunofluorescence detection of core Septin members 2, 7 and 11 at the polar body abscission site in parthenogenically activated Ctr/MatKO MII oocytes, respectively (DNA blue; SEPTIN2/7/11 green; TUBULIN red). In Ctr parthenotes, SEPTIN2 (n = 22), 7 (n = 13) and 11 (n = 12) were all localized at the cytokinetic ring (white arrowheads) where abscission of the polar body will occur. In MatKO parthenotes only SEPTIN2 (n = 17) was detected at the cytokinetic ring (white arrowheads); SEPTIN11 (n = 12) and 7 (n = 11) were absent (scale bar 50 μm).

Fig. 5. Partial rescue of SEPTIN7 recruitment and 2-cell block by Sept11 cRNA injection.

a Schematic of Sept11-eGFP cRNA injection rescue strategy for SEPTIN7 recruitment: GV oocytes are microinjected, matured to MII oocytes, parthenogenically activated and fixed and stained upon polar body extrusion. b Sept11-eGFP Ctr and MatKO injected and mock injected parthenotes were stained for eGFP (green/arrowhead), SEPTIN7 (red/arrowhead) and DNA (blue), respectively (scale bar 50 μm). Rightmost column shows magnification of the polar body abscission site (scale bar 25 μm). c Schematic diagram for 2-cell block rescue strategy: Zygotes were injected with Sept11-eGFP-cRNA and cultured beyond the 2-cell stage, up to blastocyst formation. d Quantification of in vitro development post-injection: 80% of wildtype embryos developed beyond the 2-cell stage (n = 52/66). Full 2-cell block was observed in mock-injected Prdm10 MatKO embryos (n = 0/12, rescued). Sept11-eGFP-cRNA injection enabled 22% of MatKO embryos to overcome the 2-cell block (n = 11/51, rescued). e Percentual distribution of developmental stage arrests of rescued MatKO embryos overcoming the 2-cell block. f Brightfield images of representative rescue embryos arrested at the 3-cell, 8-cell (2x) and blastocyst stage, respectively (scale bar 50 μm).