ABSTRACT
Many maternal mRNAs are translationally repressed during oocyte development and spatio-temporally activated during early embryogenesis, which is crucial for oocyte and early embryo development. By analyzing maternal mutants of nanog (Mnanog) in zebrafish, we demonstrated that Nanog tightly controls translation of maternal mRNA during oogenesis via transcriptional repression of eukaryotic translation elongation factor 1 alpha 1, like 2 (eef1a1l2). Loss of maternal Nanog led to defects of egg maturation, increased endoplasmic reticulum stress, and an activated unfold protein response, which was caused by elevated translational activity. We further demonstrated that Nanog, as a transcriptional repressor, represses the transcription of eefl1a1l2 by directly binding to the eef1a1l2 promoter in oocytes. More importantly, depletion of eef1a1l2 in nanog mutant females effectively rescued the elevated translational activity in oocytes, oogenesis defects and embryonic defects of Mnanog embryos. Thus, our study demonstrates that maternal Nanog regulates oogenesis and early embryogenesis through translational control of maternal mRNA via a mechanism whereby Nanog acts as a transcriptional repressor to suppress transcription of eef1a1l2.
INTRODUCTION
During oocyte development, many maternal mRNAs are transcribed and accumulated, and temporal and spatial regulation of translational activation or repression of maternal mRNA determines oocyte development, maturation and early embryogenesis. Translation of many maternal mRNAs is repressed during oocyte development (Evans and Hunter, 2005; Piqué et al., 2008; Gosden and Lee, 2010), and maintenance of translational arrest of maternal mRNA is essential for normal oocyte development and maturation (Richter and Lasko, 2011; Yarunin et al., 2011). Failure of translational repression of maternal mRNAs leads to various developmental defects, including apoptosis of oocytes, impaired oocyte maturation and unsuccessful early embryonic development (Kotani et al., 2013; Takahashi et al., 2014; Miao et al., 2017; Petrachkova et al., 2019).
Various mechanisms of translational control have been described using different animal models. In zebrafish, the RNA-binding protein Zar1 binds to zona pellucida (ZP) mRNAs and represses translation of the ZP genes during oogenesis, and loss of Zar1 induces oocyte apoptosis and ovary degeneration (Miao et al., 2017). The RNA-binding protein Ybx1 associates with processing body components and represses global translation activity during early embryogenesis (Sun et al., 2018). In Drosophila, translation of many germ cell-specific mRNAs is repressed by RNA-binding proteins during oocyte maturation. The germline RNAs, oskar and pgc, are targeted by Bruno 1 or Pumilio at the 3′-untranslated region and translationally repressed during oogenesis, and mutations of the RNA-binding sites results in precocious translation and mislocalization of the mRNA (Kim-Ha et al., 1995; Snee et al., 2008; Flora et al., 2018). Me31B mediates translational silencing of both maternal mRNAs during the maternal-to-zygotic transition (MZT) and oocyte-localizing RNAs during transport to the oocyte (Nakamura et al., 2001; Wang et al., 2017). During transport of nanos mRNA to the oocyte, failure of translational repression of nanos mRNA by Smaug results in ectopic translation of nanos and defects of anteroposterior axis formation (Dahanukar and Wharton, 1996; Smibert et al., 1996, 1999). To date, the studies of translational repression in oogenesis have focused mainly on post-transcriptional regulation, in which the translational repressors are mainly RNA-binding proteins. It remains unknown whether there is a general translational repressor that regulates the translation of maternal mRNAs at a global level in oocytes.
Nanog is known for its prominent function as a regulator of pluripotency in embryonic stem cells (Mitsui et al., 2003; Boyer et al., 2005; Loh et al., 2006) and reprogramming of somatic cells to the pluripotent state (Takahashi and Yamanaka, 2006; Silva et al., 2009). In zebrafish, Nanog has been shown to be a transcriptional activator that plays a central role in regulating early embryogenesis. For instance, maternal Nanog mediates endoderm formation through Mxtx2-Nodal signaling (Xu et al., 2012), and is required for both extra-embryonic development (Gagnon et al., 2018) and embryonic architecture formation (Veil et al., 2018). During the MZT, the maternally provided transcription factors Pou5f3, SoxB1 and Nanog open up chromatin in a coordinated manner to initiate zygotic genome activation (ZGA) (Lee et al., 2013; Veil et al., 2019; Pálfy et al., 2020). Our recent study shows that Nanog suppresses the global activation of maternal β-catenin activity to safeguard dorsal-ventral axis formation (He et al., 2020). However, as a strongly maternally expressed gene, the role of Nanog in oogenesis is still unknown.
In this study, we found that the absence of maternal nanog leads to various developmental defects in oocytes and early embryos. Our study demonstrates that global translational activity is greatly enhanced in nanog mutant oocytes and maternal nanog mutant (Mnanog) embryos, as a result of the transcriptional activation of eukaryotic translation elongation factor 1 alpha 1, like 2 (eef1a1l2) during oocyte development and maturation. We further show that maternal depletion of eef1a1l2 significantly rescues the developmental defects of nanog mutant oocytes and early development of Mnanog embryos. Thus, our study reveals a role for Nanog as a general translational repressor through transcriptional repression of eef1a1l2 in zebrafish oogenesis.
RESULTS
Maternal nanog is required for oocyte maturation and early embryonic development
We previously generated a nanog mutant using transcription activator-like effector nuclease (TALEN) technology and addressed its crucial role in regulating dorsal formation by interfering with TCF factors (He et al., 2015, 2020). Here, by analyzing the phenotype of maternal mutant of Mnanog embryos, we found that the Mnanog embryos showed slow epibolic movement, resulting in the accumulation of blastomere cells at the animal pole at gastrulation stage (Fig. 1A), which is similar to the phenotype of the maternal and zygotic mutant of nanog (MZnanog) (He et al., 2020). Comparison of the size of Mnanog and wild-type (WT) embryos at 15 min post-fertilization (mpf) revealed that the Mnanog embryos had significantly smaller chorion diameter and oocyte diameter than did WT embryos (Fig. 1B-D). In addition, we analyzed the activation phenotype of mutant eggs by monitoring cortical granule (CG) exocytosis and cytoplasmic streaming. Fluorescein-conjugated Maclura pomifera lectin was used to label CGs to assess CG exocytosis in activated eggs. Compared with WT eggs, nanog mutant eggs showed many retained CGs (Fig. 1E) at 10 min post-activation (mpa). CellTracker CM-DiI Dye was injected into the yolk of Mnanog and WT embryos to monitor cytoplasmic streaming. In WT embryos, vigorous cytoplasmic movement towards the animal pole was recorded (Movie 1). In contrast, Mnanog embryos showed sluggish cytoplasmic streaming (Movie 2). These results indicate that maternal Nanog is essential for egg activation and early embryonic development.
Moreover, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) revealed apoptotic signals in Balbiani bodies and the cytoplasm of early-stage mutant oocytes, but not in WT oocytes (Fig. S1A). Mitochondria are enriched in the Balbiani body and also present throughout the oocyte cytoplasm (Marlow and Mullins, 2008; Jamieson-Lucy and Mullins, 2019); thus, nanog deficiency induced the mitochondrial apoptosis during oocyte development. Robust active-Caspase3 signals were also detected in Mnanog embryos, but not WT embryos, at 75% epiboly stage (Fig. S1B). These data demonstrate that nanog depletion induces oocyte apoptosis and the death of early embryonic cells.
Through morphological and histological analyses, we found that the gonadosomatic index (GSI; gonad weight/body weight×100%) was significantly reduced in nanog−/− compared with WT females (Fig. 1F,G), indicating defects of oocyte development in nanog mutants. To characterize further the oocyte maturation defects in the nanog mutant, stage IV follicles (follicle-enclosed oocytes) were isolated and treated with (or without) 17α,20β-dihydroxy-4-pregnen-3-one (DHP) to determine the percentage of germinal vesicle breakdown (GVBD) in vitro. After 2 h incubation, the percentage of GVBD in nanog mutant follicles was 6.9% without DHP treatment, which is significantly lower than that in WT follicles (28.2%) (Fig. 1H,I). Even when treated with DHP, the percentage of GVBD in nanog mutant follicles was as low as 12.8%, which is also significantly lower than that in WT follicles (72.4%) (Fig. 1H,I). Moreover, the nanog mutant stage V follicles were less transparent than those of WT (Fig. 1J). During oocyte maturation, the major yolk proteins undergo cleavage and change the appearance of the oocyte from opaque to transparent (Dosch et al., 2004). Therefore, we compared the composition of higher and lower molecular weight yolk proteins (HYP and LYP) from stage III and stage V follicles to evaluate the yolk protein cleavage level. In stage III follicles, the HYP/LYP ratio were even lower in the nanog mutant than in WT, whereas the ratio was greatly decreased in WT stage V follicles than in mutants (Fig. 1K). These results suggested the deficiency of yolk protein cleavage during maturation of nanog mutant oocytes. In addition, we also generated a transgenic line, Tg(CMV:nanog-myc), in a nanog homozygous background. Immunofluorescence staining using anti-Myc antibody showed that Nanog is strongly expressed in early oocytes (Fig. S1C). Moreover, this overexpression of nanog could rescue the early developmental defects of Mnanog (Fig. S1D), demonstrating that oocyte maturation and early embryonic defects were caused by deficiency of Nanog. Therefore, we conclude that loss of maternal nanog leads to pleiotropic defects of oogenesis and early embryonic development.
Loss of maternal nanog elevates the global translation level
In order to understand the molecular mechanism of oogenesis regulation by Nanog, we quantitatively compared the proteomes of nanog mutant eggs with WT using isobaric tags for relative and absolute quantitation (iTRAQ) technology. More than 1600 proteins were identified in eggs from the two genotypes. These proteins were classified using the COG (Clusters of Orthologous Groups of proteins) database, and two of the top categories identified were protein translation-related biological processes (cluster J and O) (Fig. S2A). Comparing the mutant with WT, 67 proteins showed differential expression (P<0.05) (Table S1), and 39 proteins were increased in mutant eggs (Fig. S2B). Gene ontology analysis of the upregulated proteins showed that one of the most significant enriched biological processes was translation elongation factor activity (Fig. S2C). Gene-Concept Network analysis revealed that four elongation factors were enriched (Fig. S2D). These results indicate that global translation activity is elevated in nanog-deficient eggs.
To verify this speculation, we assessed the translation activity of Mnanog embryos at an early developmental stage. The mCherry reporter mRNA was injected into one-cell-stage WT and Mnanog embryos together with the same amount of GFP protein. The injected embryos were imaged under a fluorescence microscope at later stages and fluorescence levels were measured. The GFP protein acted as the loading control for injection. Fluorescence measurement showed that the reporter mRNA translation level was significantly higher in Mnanog embryos (Fig. 2A,B). Western blot analysis confirmed the increase of mCherry reporter translation in Mnanog embryos (Fig. 2C,D). We then treated the stage IV mutant follicles with an eIF4E/eIF4G interaction inhibitor, 4EGI-1, which has been shown to block translation initiation by disruption of the eIF4E/eIF4G association by binding to eIF4E (Moerke et al., 2007), and determined the percentage of GVBD in vitro. After 2 h incubation with 4EGI-1, the percentage of GVBD in nanog mutant follicles was remarkably increased from 10.4% to 28.3% with DHP treatment, and increased from 8.3% to 15.2% without DHP treatment (Fig. S2E,F). Taken together, these results suggest that Nanog is necessary for repressing the global translation of maternal mRNAs during oogenesis and early embryonic development.
Nanog depletion triggers ER stress and the unfolded protein response (UPR)
The ER functions as a crucial machinery for protein synthesis, modification and trafficking in eukaryotic cells. Under ER stress, cells activate the UPR to alleviate ER burden by reducing protein translation, increasing protein degradation and generating additional chaperones to assist protein folding. Therefore, ER stress and the UPR are often associated with aberrant translational derepression (Kaufman, 2002; Miao et al., 2017). The UPR functions through three major pathways, initiated by three ER-localized transmembrane proteins in mammals: protein kinase RNA-like ER kinase (PERK), activating transcription factor 6 (ATF6) and inositol-requiring enzyme 1 (IRE1), to maintain ER homeostasis (Hetz, 2012). Normally, the N termini of these transmembrane ER proteins are held by the ER chaperone Hspa5 (also termed Grp78 or Bip), preventing their aggregation. When misfolded proteins accumulate, Hspa5 releases the proteins, allowing aggregation of these transmembrane signaling proteins, and launching the UPR (Rao and Bredesen, 2004; Shen et al., 2004; Schröder and Kaufman, 2005). Activation of PERK upregulates the expression of CCAAT-enhancer-binding protein homologous protein (CHOP), which induces cell apoptosis and death (Oyadomari and Mori, 2004; Iurlaro and Muñoz-Pinedo, 2016). We detected the transcriptional level and protein level of hspa5 and the CHOP-encoding gene ddit3 in nanog mutant ovary, found that both transcription and translation of hspa5 and ddit3 are increased in nanog−/− ovary (Fig. 2E-G). The TUNEL assay also revealed an obvious apoptosis signal in mutant ovary (Fig. S1A). Owing to the lack of specific antibodies against zebrafish ATF6 and IRE1A, we detected the mRNA expression of atf6 and ire1a (ern1), and discovered that both of atf6 and ire1a expression were increased in nanog mutant ovary (Fig. 2H,I). Given that phosphorylated ribosomal protein S6 (pS6) is considered an indicator of active protein synthesis (Biever et al., 2015; Meyuhas, 2015), we also detected the expression level of total S6 and pS6, and found that the pS6 level was significantly increased in nanog mutant ovary (Fig. 2J,K). Finally, we used PERK inhibitors (GSK2606414 and ISRIB) to treat nanog mutant and WT follicles and determined the occurrence of GVBD. After treatment with the two inhibitors, the GVBD percentage in mutant follicles significantly recovered (Fig. 2L,M). These results demonstrate that loss of nanog triggers ER stress and UPR and thus leads to failure of oocyte development and maturation.
Loss of maternal Nanog upregulates eef1a1l2 transcription level
To find out the molecular mechanism responsible for Nanog regulating translation activity in early embryos, we conducted RNA-sequencing (RNA-seq) analysis to compare nanog mutant eggs and WT eggs. A total of 137 genes were differentially expressed in mutant and WT eggs, with 74 upregulated genes and 63 downregulated genes in mutant eggs. Among the upregulated genes, eukaryotic elongation factor 1 alpha 1, like 2 (eef1a1l2) was one of the most significantly differentially expressed (Fig. 3A). eEF1A1l2 is a major subunit of the translation elongation factor 1 complex (eEF1), which plays a central role in protein synthesis by delivering aminoacyl-tRNAs to the elongating ribosome (Sasikumar et al., 2012). Calculation of reads per kilobase of exon per million reads mapped (RPKM) and visualization of RNA-seq reads mapped to the eef1a1l2 showed significant upregulation of eef1a1l2 in mutant eggs (Fig. S3A,B). Reverse-transcription quantitative PCR (RT-qPCR) further supported the suggestion that the transcription of eef1a1l2 was significantly increased in nanog mutant follicles at five different stages and in early embryonic developmental stages (Fig. 3B, Fig. S3C). In situ hybridization of ovary cryosection also showed increased expression of eef1a1l2 in the nanog mutant (Fig. 3C). In contrast, transgenic overexpression of nanog significantly decreased the high expression levels of eef1a1l2 in mutant oocytes (Fig. 3B,C). These data indicate that nanog deficiency leads to strong transcriptional activation of eef1a1l2 during oocyte development and maturation.
To verify the transcriptional inhibition of eef1a1l2 by Nanog, WT nanog or a constitutive repressor type nanog (Engrailed fusion with Nanog homeodomain, En-nanog) (He et al., 2020) was overexpressed and eef1a1l2 transcription was measured at shield stage. Both in situ hybridization and RT-qPCR analysis showed that the expression of eef1a1l2 was significantly reduced in both nanog and En-nanog overexpressing embryos (Fig. 3D,E), suggesting that Nanog acts as a transcriptional repressor on the regulation of eef1al2. To clarify whether Nanog directly binds to the promoter region of eef1a1l2 to repress its transcription, a chromatin immunoprecipitation (ChIP) assay was conducted. Ovaries of Tg(CMV:nanog-myc) at 6 mpf were dissected and ChIP was performed using an anti-Myc antibody. The precipitated chromatin was then analyzed by PCR using primer pairs that could amplify fragments of eef1a1l2 promoter. A fragment of the rpl5b exon amplified by a specific primer pair was used as control (Belting et al., 2011). As shown in Fig. 3F, the promoter fragment of eef1a1l2 was significantly enriched in the immunoprecipitated sample with no enrichment of the control genomic region rpl5. Thus, this result demonstrates that Nanog directly binds to the promoter of eef1a1l2 to inhibit its transcription. All these data illustrate that Nanog directly inhibits the transcription of eef1a1l2 and that depletion of nanog leads to significantly increased expression of eef1a1l2.
Deficiency of eEF1A1l2 ameliorates impaired oogenesis of nanog mutants
We then generated a homozygous mutant of eef1a1l2 and two types of eef1a1l2 mutants were obtained (Fig. S4A); neither of the two maternal and zygotic mutants of eef1a1l2 showed obvious embryonic defects (Fig. S4B), and we used the ihb99 allele for subsequent study. To determine whether Nanog promotes oocyte development and maturation through suppression of eef1a1l2, we generated a double homozygous mutant of nanog and eef1a1l2 and studied whether depletion of eef1a1l2 could ameliorate the oogenesis defect of nanog mutant. By morphological analysis, we found that the double maternal mutant of nanog and eef1a1l2 (Mnanog,Meef1a1l2) showed increased embryo chorion diameter and oocyte diameter at 15 mpf, compared with Mnanog embryos, whereas the single maternal mutant of eef1a1l2 (Meef1a1l2) showed no difference in embryo chorion diameter and oocyte diameter compared with WT (Fig. 4A-C). The process of CG exocytosis in double mutant eggs (nanog,eef1a1l2) were also comparable to WT at 10 mpa, which showed fewer retained CGs than the nanog mutant (Fig. 4D). Moreover, cytoplasmic streaming labeled by CM-DiI dye in the double maternal mutant embryo (Mnanog, Meef1a1l2) showed vigorous cytoplasmic movement, similar to the WT (compare Movie 1 and Movie 3). These results indicate that depletion of eef1a1l2 rescues the egg activation defect of nanog mutant.
Therefore, we further examined the oocyte development and maturation improvement in the double mutant. Morphologically, The GSI was increased in double mutant (nanog−/−,eef1a1l2−/−) females, compared with nanog−/− females (Fig. 4E,F). The double-mutant stage V follicles were more transparent than the nanog mutant, similar to WT (Fig. 4G). The altered HYP/LYP ratio in stage III and stage V follicles in the double mutant confirmed this conclusion (Fig. 4H). The GVBD percentage was also remarkably increased in double-mutant follicles (Fig. 4I,J). The eef1a1l2 single mutant showed no obvious defects in oogenesis (Fig. 4A-C,E-G,I,J) or embryogenesis (Fig. S4B), so we did not analyze eef1a1l2−/− in subsequent experiments. These results together indicate that Nanog promotes oogenesis by suppressing the expression of eef1a1l2 in the oocyte.
Given the role of eEF1A1l2 in translation elongation, we further investigated the rescue effect of overactivated translation in the double mutant. Previous studies have shown that translation of Cyclin B1 should be repressed in immature oocytes (Vardy and Orr-Weaver, 2007; Kotani et al., 2013; Takahashi et al., 2014), and translation of ZP proteins is also repressed during oogenesis (Miao et al., 2017). Therefore, we detected the translation level of Cyclin B1 and Zp3b in WT, nanog mutant, and nanog and eef1a1l2 double-mutant immature follicles (stage I/II). The result showed that the translation of Cyclin B1 and Zp3b was silent in WT immature oocyte, but the protein levels of Cyclin B1 and Zp3b were significantly increased in the immature oocyte of the nanog mutant, indicating that the translation level is elevated in the nanog mutant. However, this increased translation level could be restored by deletion of eef1a1l2 in the nanog mutant (Fig. S5A-D). In addition, mCherry mRNA reporter and GFP protein were co-injected at the one-cell stage in WT, Mnanog, and Mnanog,Meef1a1l2 embryos. The fluorescence intensity of mCherry was measured at 4 h post-fertilization (hpf). Fluorescence measurement showed that the reporter translation level was significantly reduced in Mnanog,Meef1a1l2 embryos (Fig. S5E,F). These results indicate that Nanog controls the translation of maternal mRNAs by inhibiting the transcription of eef1a1l2.
Depletion of eef1a1l2 alleviates ER stress and UPR in nanog mutant oocytes
Because depletion of eef1a1l2 ameliorates the oogenesis defect of the nanog mutant, we wondered whether eEF1A1l2 depletion alleviates ER stress. We examined the mRNA expression level of ER stress-associated genes, and found that the increased expression of hspa5, ddit3, atf6 and ire1a in nanog mutant ovary were all restored in the nanog and eef1a1l2 double mutant (Fig. 5A-D). IRE1 is a unique RNase that removes an internal 26 nucleotides from X-box binding protein 1 (xbp1) mRNA transcripts in the cytoplasm and activates expression of genes involved in protein folding and degradation; thus, the splicing of xbp1 mRNA has been established as a common indicator of ER stress (Shen et al., 2001; Yoshida et al., 2001; Li et al., 2015). We examined altered splicing of xbp1 mutant and WT ovaries and discovered that the splicing ratio of xbp1 was increased in nanog mutant ovary, and this excessive splicing was reduced in the double-mutant ovary (Fig. 5E,F).
To observe organelle changes that may accompany the response to ER stress, ultrastructure analysis using transmission electron microscopy (TEM) was performed WT, nanog mutant and double mutant stage I oocytes. WT oocytes showed normal morphology of ER, Golgi apparatus and mitochondria (Fig. 5Ga,Gd). However, nanog mutant oocytes exhibited disruption of the Golgi apparatus, including swelling of the Golgi apparatus, dilated and disintegrated vesicles, and collapse of the Golgi complex (Fig. 5Gb, white arrowheads). nanog mutant oocytes also showed incompact and swollen mitochondria (Fig. 5Ge, yellow arrowheads), as well as evident lysosome distribution (Fig. 5Ge, red arrowheads). In contrast, the nanog and eef1a1l2 double mutant the structure of ER, Golgi apparatus and mitochondria was normal (Fig. 5Gc,Gf). These data demonstrate that depletion of eef1a1l2 alleviates ER stress and UPR in the nanog mutant, indicating that transcriptional activation of eef1a1l2 in nanog mutant oocytes induces ER stress and UPR, thus leading to defects of oogenesis.
Depletion of eef1a1l2 rescues early embryonic development defects in the nanog mutant
Given that depletion of eef1a1l2 rescues the oogenesis defect and alleviates ER stress and UPR in the nanog mutant, we wondered whether depletion of eef1a1l2 would have a rescue effect on the early embryonic development defects. We examined the early embryonic development phenotype of the double mutant. The morphological phenotypes of WT, Mnanog (nanog−/− female cross with WT male), Mnanog,Meef1a1l2 (nanog−/−,eef1a1l2−/− female cross with WT male), Mnanog,MZeef1a1l2 (nanog−/−,eef1a1l2−/− female cross with eef1a1l2−/− male), and Mnanog,Zeef1a1l2 (nanog−/−,eef1a1l2+/− female cross with eef1a1l2−/− male; nanog+/−,eef1a1l2−/− embryos were genotyped) were recorded at 0.2, 8, 12 and 24 hpf. As described in Fig. 1A, blastomere cells stacked at the animal pole and were unable to complete the gastrulation in Mnanog embryos. Until 24 hpf, Mnanog embryos still showed abnormal shapes and died gradually (Fig. 6A). Surprisingly, either getting rid of maternal increased eef1a1l2 in Mnanog,Meef1a1l2 embryos, or eliminating both maternal and zygotic increased eef1a1l2 in Mnanog,MZeef1a1l2 embryos, could effectively rescue the developmental defects of Mnanog (Fig. 6A). Mnanog,Meef1a1l2 and Mnanog,MZeef1a1l2 embryos both exhibited ameliorated epiboly movement at gastrulation stage and improved axial formation except for the telencephalon defect at prim stage. However, Mnanog,Zeef1a1l2 embryos only lacking the zygotic eef1a1l2 exhibited a similar phenotype as Mnanog (Fig. 6A), indicating that the maternally provided eef1a1l2 mRNA in Mnanog,Zeef1a1l2 still led to developmental failure. In summary, the Mnanog,Meef1a1l2 or Mnanog,MZeef1a1l2 embryos displayed rescued early embryonic development only if eef1a1l2 was completely disrupted in the nanog mutant oocytes. These data indicate that the maternal activation of eef1a1l2 in nanog mutant oocytes not only leads to oocyte maturation defects, but also results in early developmental defects of Mnanog embryos.
Furthermore, we investigated the rescue effects using a set of molecular markers representing different functions of Nanog in early development. Several studies have proved that maternal Nanog directly activates mxtx2 to regulate endoderm and extra-embryonic formation through the Nanog-mxtx2-Nodal pathway (Xu et al., 2012; Gagnon et al., 2018; Veil et al., 2018).We confirmed that expression of mxtx2 was absent in Mnanog, but the disappearance of mxtx2 transcripts could be restored in Mnanog,Meef1a1l2 (Fig. 6B, Fig. S6A). During zebrafish ZGA, together with Pou5f3 and SoxB1, maternal Nanog initiates the transcription of the first major wave of zygotic genes and directly activates microRNA miR-430; maternal mRNAs are then cleared by miR-430 post-ZGA (Giraldez et al., 2006; Lee et al., 2013). Expression of the zygotic genes miR-430 and blf failed to be activated, and the maternal mRNA sod1, which is targeted by miR-430, also failed to be removed post-ZGA in Mnanog (Fig. 6B, Fig. S5B-D). In contrast to the rescue of mxtx2 expression, the defects of ZGA and maternal mRNA clearance could not be rescued in Mnanog,Meef1a1l2 embryos (Fig. 6B, Fig. S5B-D). These results illustrate that, during oogenesis, maternal Nanog safeguards oocyte development by suppression of activation of eef1a1l2 as a transcriptional repressor, and during early embryogenesis the transcriptional suppression of eef1a1l2 by Nanog is mainly required for transcription initiation of mxtx2 and yolk syncytial layer formation.
All these results helped us to decipher a picture of molecular regulatory mechanisms during oogenesis and early embryonic development (Fig. 6C). During WT oogenesis, Nanog acts as a transcriptional repressor, with certain co-repressors, and directly inhibits the transcription of the eukaryotic translation elongation factor eef1a1l2, contributing to translational control in oocytes. After fertilization, Nanog acts as a transcriptional activator, together with Pou5f3 and SoxB1, to initiate ZGA. In oocytes produced by nanog−/− females, the transcriptional inhibition of eef1a1l2 is absent, and ectopic maternal proteins are translated and accumulated, thus inducing ER stress and excessive UPR, leading to oocyte developmental defects. Fertilized embryos derived from the nanog−/− defective eggs fail to undergo normal gastrulation. In oocytes produced by nanog−/−,eef1a1l2−/− females, however, the translation level of maternal mRNAs is mitigated owing to the lack of functional eEF1A1l2, and ER stress and UPR are alleviated, therefore leading to normal oogenesis. Therefore, it is likely that Nanog shifts from acting as a transcriptional repressor to a transcription activator during the oocyte-to-zygote transition, with both of these functions being essential for early embryogenesis.
DISCUSSION
The function of Nanog at early embryonic developmental stages has been well characterized in previous studies (Veil et al., 2018, 2019; He et al., 2020; Pálfy et al., 2020). However, as a maternally expressed gene, its role in oocyte development and maturation is still unknown. In this study, we demonstrate that zebrafish Nanog is essential for oocyte development and maturation, and has a lasting effect on early embryogenesis. Loss of maternal Nanog causes impaired oocyte maturation, deficient egg activation and early embryo developmental failure. Mechanistically, Nanog transcriptionally represses the expression of the translation elongation factor eef1a1l2 to maintain a relative translational control state during oocyte development. In contrast, in nanog mutant oocytes, ectopic transcriptional activation of eef1a1l2 elevates global translation activity and causes ER stress and UPR. Depletion of eef1a1l2 rescues the oogenesis defects and embryonic development defects of the nanog mutant. Taken together, our results delineate the mechanisms underlying a general role of Nanog as a translational repressor during oogenesis.
Maternal mRNAs synthesized in the oocyte initiate the development of future generations. Some maternal mRNAs are either somatic or germline determinants and must be translationally repressed until embryogenesis (Richter and Lasko, 2011; Flora et al., 2018). A long-recognized mechanism of translational regulation during oocyte development acts by controlling mRNA poly(A)-tail length (Richter, 2007; Weill et al., 2012). The observation of short poly(A) tails in oocytes led to the proposal that short poly(A) tails help mask maternal mRNAs and promote translational repression. Certain mRNAs, such as c-mos and several cyclin genes, are then targeted for cytoplasmic polyadenylation, and the lengthened poly(A) tails in turn cause translational upregulation of these mRNAs during oocyte maturation in Xenopus (Mcgrew et al., 1989; Sheets et al., 1994; Barkoff et al., 1998). In our study, however, cyclin B1 and zp3b mRNAs were still translationally activated in nanog mutant early oocytes, although they were translationally silent in WT early oocytes, suggesting that their poly(A) tail should be relatively short. The increased translational level of Cyclin B1 and Zp3b in nanog mutant oocytes was suppressed in nanog and eef1a1l2 double mutant oocyte (Fig. S5A-D), assuming that the changes in the translation level of maternal mRNA in nanog-null oocytes are mainly dependent on the transcriptional activation of eef1a1l2 rather than on the size of the poly(A) tail of mRNAs. Another widely studied mechanism of translational regulation acts through sequence-specific regulators, mostly RNA-binding proteins, which post-transcriptionally maintain translational repression of mRNAs containing targeted cis-regulatory elements in gametogenesis and early embryogenesis (Tadros et al., 2007; Sha et al., 2017), as has been shown for some germline mRNAs, such as oskar, pgc and nanos, during oocyte maturation in Drosophila (Smibert et al., 1996; Kugler and Lasko, 2009; Flora et al., 2018). However, different from these mechanisms, in this study we demonstrated a translational control mechanism mediated by Nanog, which transcriptionally inhibits the expression of a translational elongation factor, eEF1A1l2, and controls the maternal mRNA translational activity during oogenesis. The translational control mediated by Nanog is relatively a global one, and does not depend on the specificity of mRNA sequence and on the size of the poly(A) tail of mRNAs. Thus, this study reveals a mechanism of translational control regulated by Nanog to promote oogenesis and early embryonic development.
Genes that are not transcribed in oocytes and mature eggs are considered as non-maternal genes or zygotic genes. In theory, the transcription of non-maternal genes should be suppressed in oocytes to safeguard oocyte development and maturation and early embryonic patterning. Abnormal activation and expression of non-maternal genes will change the cell fate of the oocyte, impair oocyte development, and even lead to oocyte apoptosis. Based on the expression pattern of eef1a1l2 during oogenesis and early embryonic development in WT (Fig. 3B,C, Fig. S3), we found that eef1a1l2 has no maternal expression in oocytes, indicating that eef1a1l2 is a non-maternal gene and is not employed during oogenesis. However, eef1a1l2 is precociously transcribed in nanog mutant oocytes, leading to overactivation of global translational activity, which in turn impairs oocyte development and maturation. This finding implies that Nanog may protect oogenesis and early embryogenesis by suppressing the transcriptional activation of non-maternal genes.
Studies in human and mouse embryonic stem cells have shown that Nanog acts as both a transcriptional activator and a transcriptional repressor (Boyer et al., 2005; Liang et al., 2008). As a transcriptional activator, Nanog, in cooperation with Pou5f3 and Sox2, transcriptionally activates the expression of genes responsible for stem cell self-renewal and maintenance of pluripotency, whereas as a transcriptional repressor Nanog is associated with repression complexes and transcriptionally represses the expression of genes related to differentiation and development. In this study, we conclude that Nanog acts as a transcriptional repressor to suppress the transcription of eef1a1l2, and speculate that Nanog safeguards oogenesis by suppressing eef1a1l2 during oocyte development in zebrafish. However, Nanog is known to act as a transcription activator in ZGA and shapes the embryo during zebrafish gastrulation. For instance, Nanog initiates ZGA together with Pou5f3 and SoxB1, and miR-430 is directly activated by Nanog and is responsible for clearance of maternal mRNA during MZT (Lee et al., 2013). Further studies have shown that Nanog binds to the high nucleosome affinity regions center and synergistically opens chromatin along with Pou5f3 and Sox19b, priming genes for activity during ZGA in zebrafish (Veil et al., 2019; Pálfy et al., 2020). Nanog directly activates mxtx2 and regulates the formation of extra-embryonic tissue and embryonic architecture (Xu et al., 2012; Gagnon et al., 2018; Veil et al., 2018). Quantitative imaging also shows that Nanog cooperates with Pou5f3 to promote ventral fate (Perez-Camps et al., 2016). These studies illustrate that Nanog switches from transcriptional repressor to transcriptional activator during the oocyte-to-zygote transition. As a homeodomain protein, Nanog binds to target genes at the homeobox domain, but the question of which repressive partner interacts with Nanog to exert the gene silencing function in oocytes needs further investigation
MATERIALS AND METHODS
Zebrafish maintenance
All the zebrafish used in this study were maintained and raised as previously described (Westerfield, 1995) at the China Zebrafish Resource Center of the National Aquatic Biological Resource Center (CZRC-NABRC, Wuhan, China; http://zfish.cn). WT embryos were collected by natural spawning from the AB strain. Oocyte developmental stages were classified according to previous studies (Selman et al., 1993; Lubzens et al., 2010). Developmental stages of mutant embryos were indirectly determined by observation of WT embryos born at the same time and incubated under identical conditions. Experiments involving zebrafish were performed under the approval of the Institutional Animal Care and Use Committee of the Institute of Hydrobiology, Chinese Academy of Sciences, under protocol number IHB2014-006.
Generation of nanog and eef1a1l2 double mutants
Mnanog was generated by crossing nanog−/− females with WT males as previously described (He et al., 2015, 2020). The eef1a1l2 mutant was generated in a nanog mutant background using CRISPR/Cas9. The gRNA target and PAM sequence (underlined) of eef1a1l2 was 5′-GGCCACCTCATTTACAGTGTGG-3′; pT3TS-zCas9 was used for Cas9 mRNA transcription; capped Cas9 mRNA was generated using the T3 mMessage Machine kit (AM1344, Ambion). gRNA was generated by in vitro transcription using T7 RNA polymerase (Promega). Cas9 mRNA and gRNA were co-injected into embryos created by crossing nanog+/− females and nanog−/− males at the one-cell stage. Mnanog/Meef1a1l2 was obtained by crossing nanog−/−,eef1a1l2−/− females with WT males. The primers used for mutant screening are listed in Table S2.
Morphological analysis of ovaries and oocytes
After anesthesia by immersion in 0.16 mg/ml tricaine methanesulfonate (MS-222), we dissected the intact gonadal tissues from WT, nanog−/− and nanog−/−,eef1a1l2−/− adult zebrafish (4 months post-fertilization) and calculated the GSI (gonad weight/body weight×100%). For embryos, chorion elevation distance and oocyte diameter were measured at 15 mpf using ImageJ. Oocyte diameter was measured as the longest distance in the vertical direction of the animal-vegetal axis. Chorion diameter was considered to be the longest length of chorion when it was fully inflated.
Follicle isolation, in vitro culture and GVBD assay
Ovaries were dissected from adult females and transferred into oocyte sorting medium, made from 90% Leibovitz's L-15 medium (Gibco) and 10% fetal bovine serum (Boehringer Ingelheim) with 100 μg/ml Penicillin-Streptomycin (Gibco). Follicles (follicle-enclosed oocytes) were manually separated and divided into five groups based on oocyte size and vitellogenic state: primary growth stage (stage I), previtellogenic stage (stage II), vitellogenic stage (stage III), full-growth stage (stage IV) and mature oocytes (follicles after GVBD in vitro; stage V). Ovulated maturation oocytes were defined as eggs. Different stages of follicles were gently separated using two tweezers in a dish covered with 1% agarose.
Dissociated stage IV follicles were transferred into oocyte culture medium (OCM) by gentle pipetting. OCM was made from 90% Leibovitz's L-15 medium and 10% fetal bovine serum with 1 μg/ml DHP (Cayman Chemical). Sorted oocytes were cultured at 28°C for 2 h according to a previous study (Nair et al., 2013). GVBD rates were determined in a unified standard by ImageJ. The concentrations of different inhibitors added to OCM were: 4EGI-1 (25 ng/μl, Santa Cruz Biotechnology), ISRIB (5 μM, Selleck), GSK2606414 (50 nM, Selleck).
CG staining
Ovulated eggs at 10 mpa in water were collected and fixed with 4% paraformaldehyde (PFA) overnight prior to further steps. CGs were visualized by staining embryos with 50 μg/ml FITC-conjugated Maclura pomifera agglutinin (Vector Laboratories, FL-1341) as previously described (Mei et al., 2009).
SDS-PAGE and Coomassie staining
SDS-PAGE and Coomassie staining were performed following the established protocol (Schägger, 2006). To obtain yolk protein, ten follicles at indicated stage were lysed in 500 μl TNE buffer, made from 10 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA and 1% Triton X-100; 10 μl lysate was loaded for SDS-PAGE and Coomassie staining as previously described (Sun et al., 2018). Intensity measurement was carried out using ImageJ.
Western blot analysis
GFP protein was purchased form DIA-AN Biotechnology and 20 pg of GFP protein per embryo was co-injected at the one-cell stage. Injected embryos or dissected ovaries were homogenized using RIPA (P0013B, Beyotime). Western blot was carried out as previously described (Ye et al., 2019). Primary antibodies and dilutions for western blot were: GAPDH (2058, DIA-AN, 1:3000), mCherry (BE2026, Easybio, 1:2000), GFP (2057, DIA-AN, 1:2000), Hspa5 (11587-1-AP, Proteintech, 1:2000), Ddit3 (AC532, Beyotime, 1:2000), S6 (2217S, CST, 1:1000), pS6 (2215S, CST, 1:1000), Cyclin B1 (A2056, ABclonal, 1:1500), Zp3b (A13156, ABclonal, 1:1500).
RNA-seq and analysis
Total RNA of ovulated eggs of WT and nanog homozygous were extracted using TRIzol Reagent (Invitrogen) and mRNA was enriched using oligo-dT magnetic beads. First-strand cDNAs (from purified mRNA) were synthesized using random hexamers. The PCR-amplified cDNA was purified using AMPure XP beads, then 1 μl cDNA was validated using an Agilent 2100 Bioanalyzer. Sequencing libraries were generated using the Illumina TruSeq RNA sample preparation kit v2 according to the manufacturer's recommendations. Clustered library preparations were sequenced on an Illumina HiSeq 2000 machine and 100 bp single-end reads were generated. Clean reads, with low quality reads removed from the raw data, were mapped to the zebrafish GRCz10 reference genome using TopHat2 (Kim et al., 2013). HTSeq v0.6.1 was used to count the read numbers mapped to each gene. Then, the RPKM of each gene was calculated to determine gene expression levels (Trapnell et al., 2010). Differential expression analysis was performed using DESeq (Anders and Huber, 2010). Genes with an adjusted P<0.05 as calculated by DESeq were considered differentially expressed.
Proteomics
Ovulated eggs of WT and nanog homozygous were pooled and homogenized for quantitative proteomic analysis. The iTRAQ analysis was performed as previously described (Miao et al., 2017). The UniPort proteome sequence for Danio rerio was used for database searching.
ChIP-PCR
ChIP assays were performed with a ChIP assay kit (Cell Signaling Technology) as described (Wei et al., 2014). Briefly, two ovaries of Tg(CMV:nanog-myc) at 6 mpf were dissected and lysed for ChIP assay. Immunoprecipitation was carried out using an anti-Myc antibody (Cell Signaling Technology). Immunoprecipitation of genomic eef1a1l2 in immunoprecipitated chromatin was detected by PCR. Primers specific for the eef1a1l2 promoter region were used and the sequences are listed in Table S2. The exon of the ribosomal protein rpl5b served as a negative control, with primers 5′-GGGGATGAGTTCAATGTGGAG-3′ (forward) and 5′-CGAACACCTTATTGCCAGTAG-3′ (reverse), as described (Belting et al., 2011).
TEM analysis
Isolated stage I/II follicles from different genotypic ovaries were collected into a test tube and fixed with 100 μl 2.5% glutaraldehyde at 4°C overnight. Sample preparation for TEM was carried out according to a previously described protocol (Zhang et al., 2022) and observed under a Hitachi HT7700 transmission electron microscope.
In situ hybridization
PCR-amplified sequences of genes of interest were used as templates for the synthesis of an antisense RNA probe, labeled with digoxigenin-linked nucleotides. Whole-mount in situ hybridization (WISH) on embryos was performed as described previously (Thisse and Thisse, 2008). For in situ hybridization on frozen section, adult ovaries were stripped and embedded in Optimal Cutting Temperature compound (O.C.T., Sakura Finetek) and sectioned at 10 μm. The procedures of hybridization followed a previous study (Zhang et al., 2020).
Immunofluorescence
For whole-mount immunofluorescence, embryos were collected and fixed in 4% PFA overnight at 4°C. For immunofluorescence on cryosections, sections were prepared as for in situ hybridization and fixed in 4% PFA for 20 min at room temperature. Embryos and slides were immunostained as described in previous studies (He et al., 2020; Zhang et al., 2022). Primary antibodies and dilutions were: cleaved Caspase 3 (#9661, Cell Signaling Technology, 1:1000), Myc (#2276, Cell Signaling Technology, 1:1000).
RT-qPCR
Total RNA was extracted from samples using TRIzol (Invitrogen). RNA was reverse-transcribed with the PrimeScript RT reagent kit (Thermo Fisher Scientific) and the relative abundance of target mRNAs was examined with gene-specific primers. gapdh was used as a normalization control. Sequences of PCR primers are listed in Table S2. RT-qPCR was performed using the SYBR Green Supermix from Bio-Rad on a Bio-Rad CFX96 detection system.
Stem-loop RT-PCR of miR-430a
Stem-loop RT-PCR was performed to quantify the expression of miR-430a as previously described (Chen et al., 2005). Total RNAs were reversely transcribed using miR-430a-specific primers and U6 was used as internal control. The PCR primers of miR-430a and U6 used in this study have been described in a previous study (He et al., 2020).
TUNEL assay
Ovaries were dissected from WT and nanog−/− adult fish at 4 months post-fertilization and cryosections were prepared as for in situ hybridization. The samples were sectioned at 10 μm thickness. The TUNEL cell death assay was performed using the In Situ Cell Death Detection Kit (Roche) according to the manufacturer's instructions. Images were obtained using a laser scanning confocal microscope (Leica SP8).
Statistical analysis
GraphPad Prism 8.3.0 software was used for statistical analyses and statistical graphs. Significance of differences between means was analyzed using unpaired, one-tailed Student's t-tests. Sample sizes are indicated in the figure legends. Data are shown as mean±s.d. and P-values indicated as follows: *P<0.05, **P<0.01, ***P<0.001 0.001. ‘NS’ indicates no significant difference.
Acknowledgements
We thank other members of the Sun laboratory for method advice and manuscript discussions, and Linglu Li from the China Zebrafish Resource Center for assistance with raising fish.
Footnotes
Author contributions
Conceptualization: M.H., S.J., Y.S.; Methodology: M.H., S.J.; Formal analysis: M.H., S.J.; Investigation: M.H.; Resources: H.W.; Data curation: R.Z.; Writing - original draft: M.H.; Writing - review & editing: D.Y., Y.S.; Supervision: Y.S.; Project administration: M.H., Y.S.; Funding acquisition: M.H., Y.S.
Funding
This work was supported by National Natural Science Foundation of China (32025037, 32273134 and 31702323), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDA24010108), the Ministry of Science and Technology of the People's Republic of China (2018YFA0801000 and NK2022010207) and the State Key Laboratory of Freshwater Ecology and Biotechnology (2019FBZ05).
Data availability
RNA-seq data in this study have been deposited in Science Data Bank (doi:10.57760/sciencedb.04575) and in BioProject with accession number PRJNA633216. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the iProX partner repository with the dataset identifier PXD036250.
Peer review history
The peer review history is available online at https://journals.biologists.com/dev/lookup/doi/10.1242/dev.201213.reviewer-comments.pdf
References
Competing interests
The authors declare no competing or financial interests.