Zrsr2 and functional U12-dependent spliceosome are necessary for follicular development

Abstract

ZRSR2 is a splicing factor involved in recognition of 3'-intron splice sites that is frequently mutated in myeloid malignancies and several tumors; however, the role of mutations of Zrsr2 in other tissues has not been analyzed. To explore the biological role of ZRSR2, we generated three Zrsr2 mutant mouse lines. All Zrsr2 mutant lines exhibited blood cell anomalies, and in two lines, oogenesis was blocked at the secondary follicle stage. RNA-seq of Zrsr2 ^mu secondary follicles showed aberrations in gene expression and showed altered alternative splicing (AS) events involving enrichment of U12-type intron retention (IR), supporting the functional Zrsr2 action in minor spliceosomes. IR events were preferentially associated with centriole replication, protein phosphorylation, and DNA damage checkpoint. Notably, we found alterations in AS events of 50 meiotic genes. These results indicate that ZRSR2 mutations alter splicing mainly in U12-type introns, which may affect peripheral blood cells, and impede oogenesis and female fertility.

Keywords: Biological sciences; Molecular biology; Transcriptomics.

Graphical abstract

Figure 1

Diagrams of Zrsr2 gene structure and mutations generated in this study and their expression (A) Diagram of the Zrsr2 gene with its functional domains: ZF1 and ZF2 (zinc finger domains 1 and 2), RRM (RNA recognition motif), and RS (arginine-serine rich). The red lightning indicates the target region for each different mutation. The protein sizes and structures produced by the different mutations (Zrsr2^muA, Zrsr2^muB, and Zrsr2^muC) are represented in the lower panels, shaded in pink. The position of the primers used for Figure 1C (2-3F and 7R) and Figure 1C (9F and 11R) are indicated. (B) Western blot analysis of ZRSR2 protein expression in spleen of WT and the three mutant mice, using ACTB as a loading control. The positions of molecular mass markers are indicated on the left. Blue arrows indicate ZRSR2 mutant protein produced by different alternative splicing events. (C) Expression level of Zrsr2 mRNA in ovaries from 3-month-old females of the three mutant lines and WT mice determined by RT-qPCR. Biological triplicate data for qPCR are presented as mean ± SEM. Bars with different superscripts differ significantly (p < 0.05). (D) Non-quantitative RT-PCR analysis of Zrsr2 in WT and Zrsr2^muA, Zrsr2^muB, and Zrsr2^muC mice. The right panel shows a diagram of the affected exons in Zrsr2^muC mice. Isoforms were confirmed by sequencing.

Figure 2

Ovarian morphology and histology and follicle count in WT and Zrsr2 mutant mice (A) WT and Zrsr2 mutant ovaries from the three lines at 5-, 8-, and 11-week-old mice. Scale bar, 200 μm. (B) Hematoxylin and eosin-stained sections of WT, Zrsr2^muA, Zrsr2^muB, and Zrsr2^muC ovaries from 5- and 8-week-old mice. Scale bar, 200 μm. (C) Bar graph representing the ovary length of WT, Zrsr2^muA, Zrsr2^muB, and Zrsr2^muC ovaries from 5-, 8- and 11-week-old mice (n = 5 per line). Biological triplicate data for qPCR are presented as mean ± SEM. Bars with different superscripts differ significantly (p < 0.05). (two-way ANOVA followed by Tukey’s post hoc test). (D) Primary, secondary, and antral follicles count in dissected ovaries from 8- to 10-week-old females from the WT and the three Zrsr2^mu lines (n = 5 per line) injected with pregnant mare’s serum gonadotropin (PMSG). Biological triplicate data for qPCR are presented as mean ± SEM. Bars with different superscripts differ significantly (p < 0.05) (two-way ANOVA followed by Tukey’s post hoc test).

Figure 3

Peripheral blood cell counts Boxplot showing the most representative parameters of each cell type series (white blood cells, red blood cells, and platelets). WBC, white blood cells; LYM, lymphocytes; MON, monocytes; GRAN, granulocytes; RBC, red blood cells; MCV, erythrocyte volume; MCH, hemoglobin content; RDW, variation coefficient; PLT, platelets; MPV, mean platelet volume; PDW, heterogeneity; PCT, percentage of blood occupied by platelets. Bars with different superscripts differ significantly (p < 0.05) (one-way ANOVA followed by Tukey’s post hoc test).

Figure 4

Zrsr2 expression profile and RNA-seq analysis of WT and Zrsr2^muC follicles (A) Relative expression levels of Zrsr2 in developing follicles; biological triplicate results presented as the mean ± SEM. Bars with different superscripts differ significantly (p < 0.05 by two-way ANOVA followed by Tukey’s post hoc test). (B) Pie charts show the proportion of up-regulated and down-regulated genes as well as the cellular components of the DEGs. (C) Volcano plot of differentially expressed genes between Zrsr2^muC and wild-type samples, in which genes with an false discovery rate value under 0.01 and a fold change over 2 are represented in red. The genes with the greatest fold change and/or significant difference are indicated. Four genes of calcium-binding protein are highlighted surrounded by blue. (D) Bar plot of the number of genes belonging to the three main gene types in the set of DEGs. (E) Number of DEGs that are expressed preferentially in the oocyte, in the granulosa cells or both (oocyte expression data: GEO-GSE111687, granulosa cells expression data: GEO-GSE158218). Missing genes in those datasets are indicated as “Non-classified.” (F) Significantly enriched molecular functions terms in down-regulated transcripts in the Zrsr2^muC follicles. (G) Significantly enriched biological process terms in down-regulated transcripts in the Zrsr2^muC follicles. (H) RNA-seq coverage plot of Zrsr2 gene (top panel) representing the WT data (green) and the Zrsr2^muC data (red). The lower panel shows a zoom of the last exons of Zrsr2, showing the coverage and the splice junctions (shown in blue) of WT and Zrsr2^muC data.

Figure 5

RNA-seq analysis of alternative splicing in Zrsr2^muC follicles (A) Distribution of categories of alternative splicing (AS) events differing in Zrsr2^muC secondary follicles versus WT. Percentages of each class of event in vast-tools (VastDB annotation) and in Zrsr2^muC follicles are indicated. 3SS, alternative 3′ splice sites; A5SS, alternative 5′ splice sites; ES, exon skipping; MIC, alternative micro cassette exon ≤15 nucleotides; IR, intron retention. (B) Enrichment of differentially spliced events in Zrsr2^muC secondary follicles versus WT with respect to the total number of events annotated in the mouse genome. MIC values are chopped for clarity purposes. (C) Violin plots of AS events differing in Zrsr2^muC secondary follicles versus WT. (D) Differences in intron retention events detected in Zrsr2^mu compared with WT follicles (measured as delta percent-spliced-in [DPSI] ratio of intron read counts in Zrsr2^muC versus WT for different intron categories, as indicated). (E) Distributions of the locations of affected U2-type introns in Zrsr2^muC follicles relative to U12-type introns present in the same gene. (F) Percentage of events altered in Zrsr2^muC compared with WT follicles that were also detected in Zrsr1^mu testis (Horiuchi et al., 2018). DEG, differentially expressed genes; Other AS: 3SS, 5SS, ES, and MIC; IR (U2), U2-type introns that showed IR; IR (U12), U12- and U2-type introns located in genes with U12 that showed IR. The table on the right shows the genes with common IR (U12) events between the two experiments. (G) Significantly enriched GO terms and KEGG pathways in MIGs showing retention in U12- or U2-type introns in genes containing U12-type introns in Zrsr2^muC. (H) Significantly enriched KEGG pathways in MIGs showing retention in up-regulated U2 introns in Zrsr2^muC