Genetic gradual reduction of OGT activity unveils the essential role of O-GlcNAc in the mouse embryo
Fig 2
Maternal inheritance of a severe Ogt mutation causes preimplantation sub-lethality and perturbs retrotransposons silencing.
(A) Number and frequency of genotypes for the progeny produced by the mating of female heterozygous OgtT931A/+ with WT males. The left pie chart shows the theoretical Mendelian ratios for an X-linked mutation, the middle and right pie charts show the observed distribution of genotypes at the blastocyst stage and at weaning, respectively. P-values were computed using the chi-squared test for independence. (B) Experimental design for the study of blastocysts with maternal inheritance of a mutation at Ogt-T931. Oocytes from females OgtT931A/+ and seemingly Ogt+/+ female littermates were fertilized in vitro with WT sperm and the embryos grown ex vivo to the blastocyst stage (E4), when they were collected for single embryo mRNA-Seq. For the embryos produced by OgtT931A/+ females, the sex and presence of the T931A mutation was determined by PCR genotyping of the cDNA and Sanger sequencing. For the rest of the embryos, the T931del allele was identified a posteriori by manual inspection of the RNA-Seq reads, and the embryos were genotyped based on sequencing reads mapping to Ogt. The WT genotype was assigned based on the absence of reads bearing a mutation at T931. The total number of embryos analyzed per genotype after the in silico filtering steps (S5 Table) is indicated. Note that the absence of T931del-containing reads cannot formally exclude the presence of this allele within the group of heterozygous females. (C) Transcriptomes of individual male blastocysts produced by the experiment in (B) in the space defined by PC2 and PC3 of their principal component analysis (PCA), which result in a better separation based on embryonic genotype than PC1 and PC2 (shown in S2B Fig). The variance explained by each PC is shown in parentheses. (D) MA-plot from DESeq2 differential expression analysis of single copy genes in OgtT931del/Y versus WT male blastocysts from all maternal genotypes (N = 10 embryos per embryo genotype). All genes with mean DESeq2-normalized gene counts > 10, adj. p-value < 0.05, any log2FC (DEGs) are colored, and their number is indicated. Dashed lines show log2FC = ±0.5. Genes standing out (and with abs(log2FC) ≥ 0.2) are labeled. (E) Gene set enrichment analysis (GSEA) of gene expression change in OgtT931del/Y versus WT male blastocysts. Among significant Biological Process (BP) and Cellular Component (CC) gene ontology (GO) terms, the first 20 based on Normalized Enrichment Score (NES) are shown. Terms are ordered based on gene ratio. The size of dots is proportional to the number of total genes of a GO term. Gene ratio = fraction of total genes of the GO term which are concordantly changing in mutant embryos. (F) Top: expression dynamics of OgtT931del/Y DEGs belonging to the indicated GO terms throughout preimplantation development. The two biological replicates per stage were averaged. For each gene, the TPM value in the E3.5 blastocyst is the highest TPM value between the values in E3.5 ICM and E3.5 trophectoderm. The mean among all genes is drawn, as well as the 95% confidence interval, computed using basic nonparametric bootstrap (R function ‘mean.cl.boot’). Y-axis ticks are in log2 scale. TPM: Transcript Per Million. Bottom: enrichment analysis of up- and downregulated OgtT931del/Y DEGs in the genes changing in unperturbed embryos between the 8-cell stage and the trophectoderm of the E3.5 blastocyst (ranked by -log10(adj. p-value)*sign(log2FC) from E3.5TE-vs-8-cell comparison). Gene ratio = fraction of total up-/downregulated DEGs which are up-downregulated between the two preimplantation stages. Top and bottom: mRNA-Seq data for unperturbed embryos are from GSE66582 and GSE76505. (G) Heatmap of the expression (in Fragments Per Kilobase Per Million, FPKM) of the main families of retrotransposons in the single male blastocysts produced in (B). Values are scaled by rows. Retrotransposon families are clustered based on Euclidean distance and coloured by class. P-values were computed using unpaired Wilcoxon rank sum exact test of FPKM values, and they are indicated when < 0.05. (H) MA-plot from DESeq2 differential expression analysis of retrotransposons in OgtT931del/Y versus WT male blastocysts (N = 9 WT, 10 OgtT931del/Y embryos). All repeats with mean DESeq2-normalized gene counts > 10, adj. p-value < 0.1, any log2FC are colored by their class. Repeats standing out are labeled.