The RNA ligase RNA terminal phosphate cyclase B regulates mRNA alternative splicing and is required for mouse oocyte development and maintenance

During the development of growing oocytes into fully grown oocytes, an increase in oocyte volume is coupled with active transcription events (Pangas and Rajkovic, 2006). Pre-mRNA is processed to remove introns and splice exons via alternative splicing (AS) (Darnell, 2013; Ding et al., 2014). Through AS, multiple transcripts and protein variants can be transcribed and translated from a single gene, which significantly enriches the diversity of transcriptomes and proteomes (Braunschweig et al., 2013; Stamm et al., 2005). Common AS events include intron retention (IR), skipped exons (SE), mutually exclusive exons (MXE), and alternative 5′ and 3′ splice sites (A5SS and A3SS). Recent studies have shown that intron retention (IR) occurs in many mRNAs and is a frequent event in plants and humans (Ner-Gaon et al., 2004; Reddy et al., 2013). IR is recognized as a key mechanism in the regulation of gene expression. Specific transcripts with IR are expressed in specific cell types or situations, and therefore control developmental transitions or cellular stress responses (Boothby et al., 2013; Braunschweig et al., 2014; Shalgi et al., 2014; Wong et al., 2013).

During mRNA retained intron splicing, two half-molecules with 2′,3′-cyclic phosphate and 5′-OH ends are considered unconventional products (Englert et al., 2012). There are two known pathways for rectifying these two ends. Classically, ligation occurs via a multienzyme pathway. Classical ligases catalyze the ligation of 5′-P and 3′-OH ends with the help of ATP and Mg²⁺ (Desai et al., 2013). Before sealing, 2′,3′-cyclic phosphate is hydrolyzed to 3′-OH by phosphodiesterase, and 5′-OH is phosphorylated by polynucleotide kinase to generate 5′-phosphate (5′-P). The second pathway depends on RTCB. The role of RTCB as an RNA ligase was recently confirmed. It directly ligates the 2′,3′-cyclic phosphate and 5′-OH terminal half molecules (Desai et al., 2013; Popow et al., 2011; Tanaka et al., 2011a,b; Tanaka and Shuman, 2011). With the help of GTP and Mn²⁺ ions, RTCB can treat 2′,3′-cyclic phosphate and 5′-OH ends in one step to form a phosphodiester bond (Chakravarty and Shuman, 2012; Chakravarty et al., 2012; Desai and Raines, 2012; Englert et al., 2012). RTCB plays a key role in tRNA maturation and splicing of Xbp1 mRNA (Tanaka et al., 2011b), and can catalyze the reconnection of short 16S rRNA produced by stress (Maughan and Shuman, 2016). Xbp1 is the only confirmed RTCB target mRNA (Lu et al., 2014). RTCB regulates the unfolded protein response (UPR) through the unconventional splicing of Xbp1 mRNA in the cytoplasm (Shinya et al., 2011). RTCB also plays a key role in regulating tRNA splicing after intron removal (Tanaka et al., 2011b). According to a previous study, ∼5% of tRNA genes in mice contain introns (Wu and Hopper, 2014). tRNA introns are positioned downstream of the anticodon and are smaller than mRNA introns. After intron removal, these pre-tRNAs require ligation to form mature tRNAs. RTCB is responsible for catalyzing the ligation of these tRNAs.

Immature oocytes of mice are arrested at the germinal vesicle (GV) stage in growing ovarian follicles unless they are stimulated by a gonadotrophin signal or released from an inhibitory follicular environment. At this stage, GV oocytes can be classified based on their chromatin configuration in non-surrounded nucleolus (NSN) oocytes and surrounded nucleolus (SN) oocytes. NSN oocytes have low developmental competence and their chromatin does not form a Hoechst-positive rim around the nucleolus. In contrast, the chromatin of SN oocytes forms a Hoechst-positive rim around the nucleolus, and transcription in SN oocytes is silenced (Tan et al., 2009). However, many of the synthesized mRNAs are stored in the ooplasm. The dormant mRNAs drive meiosis and early embryonic development. The key events of meiosis mainly include germinal vesicle breakdown (GVBD) and the following two consecutive M phases (MI and MII) (Jones, 2011).

Splicing factors in oocytes, such as SRSF3 and BCAS2, have been shown to play an important role in oocyte maturation and the maintenance of transcriptome integrity (Do et al., 2018; Zhang et al., 2022). Without RTCB, the lifespan of Caenorhabditis elegans decreases, germ cells fail to mature into oocytes, and the animals are rendered completely sterile (Kosmaczewski et al., 2014). We hypothesized that this new mechanism of rapid gene regulation via RTCB may also exist in mouse oocytes. To test this hypothesis, we specifically knocked out the Rtcb gene in oocytes as early as the primordial follicle stage using Gdf9-Cre. Our results show that RTCB is responsible for regulating mRNA splicing after intron removal during oocyte meiosis. The absence of RTCB resulted in the accumulation of a series of intron-containing transcripts, mainly including DNA methylation and DNA damage repair-related transcripts. As a result, the levels of these maternal proteins decreased. Follicular development, meiosis recovery and early embryonic development were impeded. In summary, our results confirm that RTCB-regulated alternative splicing plays an important role in female reproduction.

First, we detected Rtcb expression in various tissues and organs by quantitative RT-PCR. The results showed that Rtcb was highly expressed in female germ cells (Fig. 1A). In addition, Rtcb mRNA levels remained relatively stable in oocytes and increased drastically after the two-cell stage (Fig. 1B). These results suggest that RTCB may play an important role in oocyte development and embryo preimplantation. Thereafter, we detected RTCB levels by western blotting (WB) during oocyte meiosis. The results showed that RTCB protein levels at the GVBD stage were higher than in the germinal vesicle (GV), MI and MII stages (Fig. 1C).

To study the in vivo function of RTCB in oocytes, we constructed oocyte-specific Rtcb knockout mice. By crossing Rtcb-floxed mice (Rtcb^fl/fl) with Gdf9-Cre transgenic mice, we deleted the Rtcb gene in oocytes as early as the primordial follicle stage (Fig. 1D). WB and immunofluorescence (IF) results revealed that RTCB protein was completely depleted in GV oocytes (Fig. 1E,G). IF results showed that RTCB was distributed in the nuclei and cytoplasm of the oocytes. In contrast, there was no RTCB signal in Rtcb-deleted oocytes (Fig. 1E). A subcellular fractionation assay of 293T cells also proved the nuclear and cytoplasmic localization of RTCB (Fig. 1F). This result is consistent with previous reports in HeLa cells (Jurkin et al., 2014). Thereafter, we conducted a fertility test for 7 months by mating Rtcb^fl/fl female mice (hereafter referred to as wild type) and Rtcb^fl/fl;Gdf9-Cre female mice (hereafter referred to as Rtcb^oo−/−) with wild-type males. The Rtcb^oo−/− female mice were completely sterile (Fig. 1H). These results demonstrate that RTCB in oocytes is required for female fertility in mice.

To study the effect of Rtcb deletion on follicle development in the ovary, we collected mouse ovaries at various stages and compared their shapes and sizes. Compared with wild-type mice, the ovaries of 4-week-old Rtcb^oo−/− mice (not stimulated by exogenous gonadotropins) were almost similar in size. However, in Rtcb^oo−/− mice older than 6 weeks, the ovaries were significantly smaller than those of wild-type mice (Fig. 2A,B).

Subsequently, we investigated follicle development at various ages using Hematoxylin and Eosin staining, and immunohistochemistry (IHC). The ovaries of 4-week-old Rtcb^oo−/− mice still contained many primordial, primary, secondary and antral follicles (Fig. 2C). Follicular counting results showed that the ovaries of 6-week-old Rtcb^oo−/− mice (not stimulated by exogenous gonadotropins) contained similar numbers of primordial, primary and bilayer secondary follicles relative to wild-type mice, but fewer secondary and antral follicles with multilayer granulosa cells (Fig. 2D,H). Furthermore, fewer primordial, primary and secondary follicles remained in the ovaries of the 8-week-old mice (Fig. 2E and I). No follicles were found in the ovaries of 10-week-old or 7-month-old knockout mice (Fig. 2F,G).

To investigate why premature ovarian failure occurs, we examined the signal of phosphorylated H2AX (γH2AX, an unrepaired DNA lesion marker) in oocytes in ovarian sections. We found that, in 6-week-old Rtcb^oo−/− mice, more than 50% of oocytes in secondary and antral follicles had obvious DNA damage signals (Fig. 2J,K). Moreover, TdT-mediated dUTP nick-end labeling (TUNEL) showed a higher proportion of apoptotic signals in the follicles of Rtcb^oo−/− mice than in those of wild-type mice (Fig. 2L,M). In summary, these observations suggest that RTCB is essential for ovarian follicle development and that Rtcb deletion in oocytes leads to DNA damage and induced follicle atresia.

Because fully grown GV oocytes were still present in the ovaries of 6-week-old Rtcb^oo−/− mice, we analyzed the oocytes isolated from 6-week-old wild-type and knockout mice treated with pregnant mare serum gonadotropin (PMSG). The diameters of the Rtcb-null oocytes were significantly smaller than those of the wild-type oocytes (Fig. 3A,B). In addition, numerous brown cytoplasmic granules were detected in Rtcb-null oocytes (Fig. 3A). DAPI staining revealed that more than 80% of fully grown oocytes from 6-week-old wild-type mice had a surrounded nucleolus (SN) configuration. However, the proportion of SN configurations was only ∼40% in the Rtcb-null oocytes (Fig. 3C,D). These results indicate that Rtcb deletion inhibited oocyte growth and nucleolar maturation.

To explore the effect of Rtcb deletion on maternal transcript levels, we collected fully grown GV oocytes from 6-week-old wild-type and knockout mice for RNA sequencing (RNA-seq) analyses. Gene expression levels were evaluated by fragments per kilobase of transcript per million mapped reads (FPKM). According to the RNA-seq results, we detected more genes with decreased mRNA copy numbers [1861 when fold change >2, false discovery rate (FDR) <0.05] than those with increased mRNA copy numbers (498 when fold change >2, FDR<0.05) (Fig. 3E). We performed gene ontology (GO) analysis on the decreased transcripts and found that they were mainly enriched in protein transport and DNA damage repair pathway (Fig. 3F), which may partially explain the phenotype of follicular atresia observed in 6-week-old Rtcb^oo−/− mice. As RTCB mainly affects mRNA splicing, we tested the mRNA levels of the retained introns. We detected the level of transcripts containing intron retention (IR) using IRFinder (Middleton et al., 2017). The IR ratio was calculated as the ratio of unspliced exon-intron junction reads to total junction reads (unspliced exon-intron and spliced exon-exon). The IR change was calculated from the IR ratio of knockout samples minus that of wild-type samples (Fig. 3G). According to IRFinder analysis, we detected many retained introns (Fig. 3I and Table S1). When IR change>0.1 and P<0.1, 91 transcripts containing IR were detected (Fig. 3H and Table S2). In addition, we used rMATS software to analyze other AS events (Shen et al., 2014). In addition to IR, many other AS events underwent significant changes because of Rtcb deletion (Fig. 3J and Table S3). GO enrichment analysis of different splicing events showed that RNA splicing was the most affected (Fig. S1). Based on these results, we conclude that RTCB plays a crucial role in RNA splicing in GV oocytes.

Based on previous results, we conclude that Rtcb deletion leads to IR accumulation and decreases many maternal mRNA levels. To determine whether IR accumulation can lead to changes in mRNA levels, we analyzed 91 transcripts with significant IR changes (eight transcripts from two identical genes were merged). The results showed that the expression levels of most transcripts decreased (Fig. 4A). Thereafter, we confirmed the IR of these transcripts by RT-qPCR. MBD1 (methyl-CpG binding domain protein 1), also known as CXXC3, plays an important role in DNA methylation (Jørgensen et al., 2004). DNMT3B (DNA methyltransferase 3β) is responsible for establishing DNA methylation (Chen et al., 2004). TDRD9 is an ATP-binding RNA helicase required for spermatogenesis (Guijo et al., 2018). RAD50 is a component of the MRN complex and plays a vital role in double-strand break (DSB) repair (Iijima et al., 2008; Syed and Tainer, 2018). RT-qPCR results showed that IR occurred in these transcripts and that all three introns of Tdrd9 were retained. In addition, the levels of most transcripts decreased (Fig. 4B). Visual assessment of RT-PCR products through agarose gel electrophoresis showed that IR events occur in these transcripts after RTCB depletion (Fig. 4C). The RNA-seq results also showed a decrease in Dnmt1, Dnmt3a and Dnmt3b mRNA levels (Fig. 4D).

As Rtcb deletion resulted in the IR of DNA methylation-related mRNAs, such as Dnmt3b and Mbd1, we detected changes in the levels of these related proteins. The results showed that the levels of DNMT1, DNMT3A, DNMT3B and RAD50 were decreased (Fig. 4E). In addition, IF staining showed that the levels of 5-methylcytosine (5mC) rather than 5-hydroxymethylcytosine (5hmC) in Rtcb^oo−/− SN oocytes were significantly downregulated compared with those in wild-type oocytes (Fig. 4F,G).

These results demonstrate that by affecting RNA splicing, RTCB regulates the transcription of other maternal epigenetic regulators and DSB repair factors during oocyte growth, and renders developmental competence to oocytes. When Rtcb is deleted, IR occurs in many DNA methylation- and DNA damage repair-related transcripts, and the total levels of these transcripts are downregulated.

As Rtcb deletion led to the accumulation of introns among maternal transcripts, we then investigated whether these retained introns affected the normal progression of oocyte meiosis by measuring the development rate at different time points in oocytes cultured in vitro (Fig. 5A). In vitro culture results showed that Rtcb deletion significantly downregulated the GVBD rate (Fig. 5B). However, Rtcb-null oocytes that were able to resume meiosis showed polar body emission (PBE) rates comparable with those of the wild-type oocytes (Fig. 5C). IF results showed that the absence of RTCB had a minor effect on chromosome alignment and spindle organization in MII oocytes (Fig. 5D,E). We also obtained oocytes from oviducts 16 h after human chorionic gonadotropin (hCG) injection. The Rtcb-null oocytes could be ovulated after hCG injection; however, the number of ovulated oocytes from Rtcb^oo−/− mice was only half that in the control mice (Fig. 5F). In addition, half of the ovulated oocytes were arrested at the germinal vesicle (GV) stage (Fig. 5G,H). Similar to the in vitro results, these ovulated Rtcb-null MII oocytes showed normal chromosomal alignment and spindle organization (Fig. 5I,J). In summary, these results suggest that a lack of RTCB mainly affects meiosis resumption and has a minor effect on the subsequent PBE process.

We examined the effects of Rtcb deletion on the levels of meiotic resumption-related proteins. Cyclin B1 is a key cell cycle-regulating protein that promotes G2/M transition (Adhikari and Liu, 2014). CDC20 is a positive regulator of the anaphase-promoting complex (APC) (Sudakin et al., 2001). However, WB results revealed that the levels of cyclin B1 and CDC20 proteins increased because of Rtcb deletion. Resumption of meiosis requires activation of cyclin-dependent kinase 1 (CDK1), and phosphatase cell division cycle 25 B (CDC25B) is an activator of cyclin-dependent kinases (Lincoln et al., 2002). The dephosphorylation of threonine 14 and tyrosine 15 of CDK1 by CDC25 phosphatases is a key step in the activation of CDK1-cyclin B protein kinase (Morgan, 1995). We found that the levels of CDC25B and p-T161-CDK1 proteins were lower than those in the wild-type oocytes (Fig. 6A).

Phosphorylated lamin A/C (p^S22lamin A/C) is a direct substrate of active phosphorylated CDK1 (Heald and McKeon, 1990). Therefore, we examined changes in the inhibitory phosphorylation of CDK1 (p^T14/Y15CDK1) and p^S22lamin A/C. In wild-type GV oocytes, p^T14/Y15CDK1 levels decreased and p^S22lamin A/C levels increased when GVBD was about to occur (GV oocytes cultured for 2 h). However, in the Rtcb-null oocytes cultured in the same manner, the p^T14/Y15CDK1 levels remained high and lamin A/C was not successfully phosphorylated (Fig. 6B-E). This observation indicates that the decline in CDC25B leads to the persistent phosphorylation of CDK1 on threonine 14 and tyrosine 15, and the failure of CDK1 phosphorylation on threonine 161. Consequently, lamin phosphorylation and GVBD failed to occur.

To further test our hypothesis, we overexpressed continuously activated CDK1 (CDK1^T14A,Y15F) in wild-type and Rtcb-null oocytes at the germinal vesicle (GV) stage (Fig. 6F). The GVBD rate of Rtcb-null oocytes overexpressing CDK1^T14A,Y15F was ∼70% (Fig. 6G). This result indicates that RTCB plays an important role in maintaining the levels of activated CDK1 proteins.

Thereafter, we superovulated 6-week-old wild-type and Rtcb^oo−/− females and mated them with wild-type males. The diameters of the zygotes obtained from Rtcb^oo−/− mice were significantly smaller than those of the wild-type mice, and brown cytoplasmic granules aggregated in maternal Rtcb-deleted (hereafter Rtcb^♀−/♂+) zygotes (Fig. 7A). IF results showed many abnormal or multiple pronuclei in Rtcb^♀−/♂+ zygotes (Fig. 7B). The proportion of Rtcb^♀−/♂+ zygotes containing normal pronuclei was significantly lower than that of Rtcb^♀+/♂+ zygotes (Fig. 7C). Subsequently, zygotes were cultured in KSOM medium. Although most Rtcb^♀+/♂+ zygotes developed to the two-cell stage, ∼70% of Rtcb^♀−/♂+ zygotes underwent cell cycle arrest at the one-cell stage. However, fewer than 20% of Rtcb^♀−/♂+ zygotes developed into blastocysts (Fig. 7D,E). These results indicate that maternal RTCB is important for early embryonic development. However, as Rtcb-null oocytes suffer from maturation obstacles, subsequent defects in early embryonic development may be secondary effects caused by defective oocytes.

To detect the global translation level, we incubated wild-type and Rtcb^♀−/♂+ embryos at the one- and two-cell stages with L-homopropargylglycine (HPG), an analog of methionine that can be examined using the Click-iT cell reaction kit (Life Technologies) (Zhang et al., 2019). At the one-cell stage, the level of the HPG signal was not significantly different between the wild-type and Rtcb^♀−/♂+ zygotes. However, Rtcb^♀−/♂+ two-cell embryos showed weaker HPG signals than the wild-type embryos (Fig. 7F,G).

To evaluate whether zygotic genome activation (ZGA) was suppressed in arrested Rtcb^♀−/♂+ embryos, we labeled newly synthesized RNAs with 5-ethynyl uridine (EU) added to the culture medium 2 h before fixation of the one-cell stage. EU signals were lower in Rtcb^♀−/♂+ embryos than in Rtcb^♀+/♂+ (Fig. 7H,I). To confirm that ZGA was suppressed in the arrested Rtcb^♀−/♂+ embryos, we examined the expression of six ZGA-related transcripts [MuERV-L (Erv4), Npl, Guca1a, Zscan4c, Eif1a and Zscan4b] using quantitative RT-PCR at the one- and two-cell stages. The levels of these transcripts should increase during ZGA in Rtcb^♀+/♂+ embryos; however, their transcriptional activation was significantly suppressed in Rtcb^♀−/♂+ embryos (Fig. 7J).

In summary, because of the lack of RTCB protein in oocytes, ZGA is suppressed and the global protein translation level is downregulated, leading to embryonic development arrest. These results indicate that maternal RTCB proteins are important for ZGA and early embryonic development.

Among the various AS events, intron retention has been frequently reported in various species. mRNA isoforms that contain introns are not useless. They are temporarily preserved in the cytoplasm and then spliced to synthesize functional proteins during developmental transitions or the cellular stress response. The ligation of these RNAs after intron removal is regulated by the RNA ligase RTCB. According to previous studies, IR occurs in many physiological processes, including the unfolded protein response, tRNA and 16S rRNA splicing (Maughan and Shuman, 2016; Shinya et al., 2011; Tanaka et al., 2011b). However, the role of IR in oocyte meiosis is still vague.

Many maternal mRNAs and proteins stored in the ooplasm of GV stage-arrested oocytes drive oocyte meiosis and early embryonic development. Our study revealed a previously unreported function for RTCB in mouse oocyte maturation. RTCB conditional knockout mice exhibited a phenotype of female infertility, suggesting the important role of RTCB-mediated mRNA splicing in female fertility. The female infertility phenotype can be attributed to multiple factors.

In contrast, the lack of RTCB caused premature ovarian failure in female mice after 6 weeks of age, and we detected the apoptosis signal of granulosa cells on ovarian sections at this stage; however, the ovaries of 4-week-old Rtcb^oo−/− mice still contained many primordial, primary, secondary and antral follicles. However, regardless of in vivo superovulation or in vitro culture, most GV oocytes from 6-week-old RTCB knockout mice have difficulty resuming meiosis. This result was not unexpected because Rtcb deletion resulted in the accumulation of many IR-containing mRNAs in the GV stage, including DNA methylation (Mbd1, Dnmt3b) and DNA damage repair-related mRNA (Rad50). These results imply that RTCB may be involved in regulating follicle activation, except for the first wave. Combined with changes in the oocyte transcriptome, premature ovarian failure may be a secondary effect caused by the decline in transcripts enriched in protein transport and DNA damage repair pathways.

IR accumulation of these transcripts leads to changes in protein levels. RAD50 is crucial for DNA damage repair, and decreased RAD50 protein levels may lead to GVBD failure (Inagaki et al., 2016). MBD1 and DNMT3B are crucial for regulating DNA methylation (Hendrich and Bird, 1998; Nowialis et al., 2019). The decreased levels of related proteins result in a series of epigenetic changes that impair the developmental potential of the oocyte and embryo. Although the oocytes that resumed meiosis were able to extrude the polar body and had normal chromosome alignment and spindle assembly, most of the MII oocytes from Rtcb^oo−/− mice had difficulty with fertilization and ZGA was suppressed. Defects occur during subsequent fertilization and early embryonic development. This result illustrates that RTCB may be dispensable for oocyte maturation after the resumption of meiosis. However, epigenetic changes could affect the developmental potential of oocytes, and the defects become evident at the zygote stage; therefore, the process of ZGA cannot be completed. GVBD failure, premature ovarian failure and ZGA defects caused by epigenetic changes in GV stage oocytes lead to infertility in female mice. These phenotypes also illustrate that RTCB-regulated IR splicing is important for female reproduction.

In addition, we investigated the differences between AS types after Rtcb deletion in oocytes. Surprisingly, many AS events were observed, including SEs. However, no relevant studies have examined the correlation between RTCB and exon skipping. GO enrichment analysis showed that the relevant transcripts were mainly enriched in the category of RNA processing, which differed from the transcripts found in IRFinder. These results indicate that RTCB may act through other mechanisms in oocytes.

Our study found that Rtcb deletion caused IR accumulation and significantly downregulated the global levels of these IR-containing transcripts. In addition, this has not yet been reported. Different mRNA degradation pathway may cause decreased levels of IR-containing transcripts, which may be because they cannot synthesize proteins and are degraded by nonsense-mediated mRNA decay, as the intron retention of these transcripts affects their open reading frame. However, the decrease in the levels of other IR-free transcripts remains unclear.

Knocking out the Rtcb gene in mature mouse B cells leads to structural defects within the endoplasmic reticulum and insufficient antibody secretion (Jurkin et al., 2014). However, our study found that Rtcb deletion in oocytes led to obstacles in the meiotic process, premature ovarian failure and loss of fertility. Xbp1 mRNA was the only confirmed target mRNA of RTCB. The different phenotypes in germ cells show that Rtcb affects tRNA and Xbp1 mRNA, potentially has other substrates, such as Rad50 and Dnmt3b, and has an impact on fertility (Fig. 8).

Our study examined the role of Rtcb across the life of the oocyte and transition to embryo. IR-containing mRNAs are spliced appropriately through RTCB to synthesize proteins and perform their functions. This mechanism can regulate oocyte meiosis and the subsequent early embryonic development.

Rtcb^fl/fl mice were engineered as described previously (Li et al., 2021) and provided by Prof. Nai-Zheng Ding from the College of Life Science, Shandong Normal University, Jinan City, China. To construct oocyte-specific knockout Rtcb mice, Rtcb^fl/fl mice were bred with the Gdf9-Cre mouse line to excise loxP-flanked exon 4, thus generating Rtcb-conditional-knockout mice.

All mice were bred under specific pathogen-free (SPF) conditions in a stable environment, including 50-70% humidity, 20-22°C, a 12 h light/12 h dark cycle, and food and water provided ad libitum. All the mice had a C57BL/6J genetic background. The experimental procedures were approved by the Zhejiang University Institutional Animal Care and Research Committee (approval ZJU20210252 to H.-Y.F.) and mouse care and use were conducted following the relevant guidelines and regulations of Zhejiang University. The primers used for genotyping are listed in Table S4. The age, sex and size of each sample are shown in the related figure legends.

Three-week-old female mice were intraperitoneally primed with 5 IU of pregnant mare serum gonadotropin (PMSG) and fully grown oocytes were obtained in M2 medium (M7167; Sigma-Aldrich) 44 h later. These oocytes were cultured in mini-drops of M16 medium (M7292; Sigma-Aldrich) covered with mineral oil (M5310; Sigma-Aldrich) and then incubated at 37°C in a 5% CO₂ atmosphere.

HEK 293T cells were obtained from ATCC and were recently authenticated and tested for contamination. Cells were cultured in DMEM plus 10% fetal bovine serum and 1% penicillin-streptomycin solution at 37°C in a humidified 5% CO₂ incubator.

Three-week-old female mice were injected with 5 IU PMSG. These mice were injected with 5 IU of human chorionic gonadotropin (hCG) 44 h later. After an additional 16 h, oocyte-cumulus complexes were separated from the oviducts. Individual oocytes were obtained via digestion with hyaluronidase (Sigma-Aldrich). To acquire early embryos, 12-week-old wild-type males were mated with female mice. Vaginal plugs indicate successful mating. Embryos were obtained from oviducts at the indicated time points after hCG injection.

Continuously activated CDK1 was constructed by mutating T14 and Y15 of human CDK1 to alanine and phenylalanine, respectively. The corresponding cDNA was then subcloned into a FLAG-tagged expression vector (pDEST). The expression vector was linearized with HindIII and the linearized RNA was transcribed in vitro using the SP6 mMESSAGE mMACHINE kit (Invitrogen, AM1450). Phenol/chloroform extraction and ethanol precipitation were performed for RNA extraction and purification, respectively. The transcribed RNAs were in vitro polyadenylated using a poly (A) tailing kit (AM1350; Invitrogen).

All microinjections were performed using an Eppendorf TransferMan NK2 micromanipulator. Fully grown oocytes were cultured in M2 medium containing 2 μM milrinone to suppress spontaneous GVBD. Approximately 5-10 pl of 500 μg/ml mRNA was microinjected into the cytoplasm of the oocytes. As a negative control, 5-10 pl of 500 μg/ml transcribed RNA from the empty vector (pDEST-FLAG) was microinjected into the control oocytes. The oocytes were cultured in M16 medium containing 2 μM milrinone at 37°C in a 5% CO₂ atmosphere.

Total RNA was extracted from oocytes using an RNeasy Mini Kit (QIAGEN) according to the manufacturer's instructions. Reverse transcription was performed using a Superscript RT kit (Bio-Rad). Quantitative RT-PCR was conducted using Power SYBR Green PCR Master Mix (Applied Biosystems, Life Technologies) with an ABI 7500 Real-Time PCR system (Applied Biosystems). The primers used are listed in Table S4.

Oocytes were fixed with 4% paraformaldehyde (PFA) for 30 min and permeabilized for 20 min with 0.3% Triton X-100 in PBS. Antibody staining was performed according to previously described protocols (Zhang et al., 2020). The antibodies used in these experiments are listed in Table S5. Imaging was performed using a Zeiss LSM710 confocal microscope. Quantitative analysis of the fluorescence signals was performed using the ImageJ software.

Embryos were cultured in KSOM medium with 100 μM 5-ethynyl uridine (EU) for 2 h. Fixation, permeabilization and staining were performed using the Click-iT RNA Alexa Fluor 488 Imaging Kit (ThermoFisher, 48 C10329) according to the manufacturer's protocol. The embryos were imaged using a Zeiss LSM710 confocal microscope.

Oocytes were incubated in M2 medium primed with 100 mM L-homopropargylglycine (HPG) for 1 h. These oocytes were then fixed for 30 min in 4% PFA. HPG signals were examined using a Click-iT HPG Alexa Fluor Protein Synthesis Assay Kit (Life Technologies). The mean HPG signal was measured and quantified using ImageJ software.

Ovaries were harvested and fixed in formalin overnight, processed and embedded in paraffin wax, according to standard protocols. The ovaries were serially sectioned at 5 μm and stained with Hematoxylin and Eosin. IHC was performed according to standard protocols. All the antibodies and their related dilution factors are listed in Table S5.

Oocytes were lysed in a loading buffer containing β-mercaptoethanol and then heated at 95°C for 10 min. Immunoblotting and SDS-polyacrylamide gel electrophoresis were performed following standard procedures using the Mini-PROTEAN Tetra Cell System (BioRad). All the antibodies and their related dilution factors are listed in Table S5.

Fully grown GV oocytes were collected from 6-week-old wild-type and Rtcb^fl/fl;Gdf9-Cre female mice after PMSG (10 fully grown GV oocytes per sample). Ten oocytes with equal amounts of external RNA control transcripts from the External RNA Controls Consortium (ERCC) were subjected to each RNA-seq library construction. Total RNA was extracted using the RNeasy micro kit (Qiagen) following the manufacturer's instructions after picking the cells into 350 μl of lysis buffer supplied in the RNeasy micro kit (Qiagen). Libraries were prepared using a NEBNext Ultra RNA Library Prep Kit (Illumina). The libraries were sequenced on an Illumina platform with 150 bp paired-end reads. We used a script to filter out low-quality reads, and the clean reads were mapped to the mouse genome mm10 with TopHat (version 2.0.6). Gene expression levels were calculated and normalized to FPKM using cuffquant and cuffnorm. In RNA-seq data normalization, sequencing reads were mapped to an ERCC reference to obtain the percentage of ERCC reads in total reads. Gene expression levels were then normalized by multiplying the raw FPKM values by a normalization factor (normalization factor=percentage of ERCC in wild-type oocytes/percentage of ERCC in the sample). FPKM and mRNA copy numbers represent gene expression levels. Only the expressed genes (FPKM>1 in at least one sample) were considered in all analyses. A two-tailed Student's t-test was used to generate statistically significant values.

Intron retention (IR) was analyzed using IRFinder software. The IR ratio was calculated as the ratio of unspliced exon-intron junction reads to total junction reads (unspliced exon-intron and spliced exon-exon). The IR change was calculated from the IR ratio of knockout samples minus that of wild-type samples. According to IRFinder, only these transcripts (IR change>0.1) exhibited obvious IR accumulation in all the analyses. Changes in IR in genes are shown in Table S1, and the IR number of the filter is shown in Table S2.

Using replicate multivariate analysis of transcript splicing (rMATS) (Shen et al., 2014), we detected five different AS events, including SE, A5SS, A3SS, MXE and retained intron (RI) (Shen et al., 2014). Various AS events were considered only when FDR<0.05. AS events are listed in Table S3.

Results are presented as mean±standard error of the mean (s.e.m.). Most experiments comprised at least three independent samples and were repeated at least three times. The results for the two experimental groups were compared using two-tailed unpaired Student's t-tests. *P<0.05, **P<0.01, ***P<0.001. At P<0.05, results were considered statistically significant. n.s. indicates a non-significant result.

We thank the Life Sciences Institute core facilities at Zhejiang University for their technical assistance.

The peer review history is available online at https://journals.biologists.com/dev/lookup/doi/10.1242/dev.200497.reviewer-comments.pdf.