ZAR1 and ZAR2 orchestrate the dynamics of maternal mRNA polyadenylation during mouse oocyte development

Wu, Yu-Ke; Su, Ruibao; Jiang, Zhi-Yan; Wu, Yun-Wen; Rong, Yan; Ji, Shu-Yan; Liu, Jingwen; Niu, Zhuoyue; Li, Zhiyi; Xue, Yuanchao; Lu, Falong; Fan, Heng-Yu

doi:10.1186/s13059-025-03593-8

Research
Open access
Published: 08 May 2025

ZAR1 and ZAR2 orchestrate the dynamics of maternal mRNA polyadenylation during mouse oocyte development

Yu-Ke Wu¹,
Ruibao Su²,
Zhi-Yan Jiang¹,
Yun-Wen Wu¹,
Yan Rong^1,3,
Shu-Yan Ji⁴,
Jingwen Liu^5,6,
Zhuoyue Niu^5,6,
Zhiyi Li⁷,
Yuanchao Xue²,
Falong Lu^5,6 &
…
Heng-Yu Fan^1,3,8

Genome Biology volume 26, Article number: 120 (2025) Cite this article

550 Accesses
Metrics details

Abstract

Background

During meiosis, the oocyte genome keeps dormant for a long time until zygotic genome activation. The dynamics and homeostasis of the maternal transcriptome are essential for maternal-to-zygotic transition. Zygotic arrest 1 (ZAR1) and its homolog, ZAR2, are RNA-binding proteins that are important for the regulation of maternal mRNA stability.

Results

Smart-seq2 analysis reveals drastically downregulated maternal transcripts. However, the detection of transcript levels by Smart-seq2 may be biased by the polyadenylated tail length of the mRNAs. Similarly, differential expression of maternal transcripts in oocytes with or without Zar1/2 differs when analyzed using total RNA-seq and Smart-seq2, suggesting an influence of polyadenylation. Combined analyses using total RNA-seq, LACE-seq, PAIso-seq2, and immunoprecipitation-mass spectrometry reveals that ZAR1 may target the 3’-untranslated regions of maternal transcripts, regulates their stability in germinal vesicle stage oocytes, and interacts with other proteins to control the polyadenylation of mRNAs.

Conclusions

The jointly analyzed multi-omics data highlight the limitations of Smart-seq2 in oocytes, clarify the dynamics of the maternal transcriptome, and uncover new roles of ZAR1 in regulating the maternal transcriptome.

Peer Review reports

Background

During meiosis, oocyte transcription remains silent since germinal vesicle breakdown (GVBD) until the beginning of zygotic genome activation (ZGA) [1,2,3,4,5]. Accordingly, the dynamics of the maternal transcriptome are of great concern. Previous studies have reported the acute degradation of maternal mRNA during the transition from GV to metaphase II (MII) stage oocytes [6]. This process has led to the identification of numerous meiosis-coupled decay genes (or oocyte decay genes, simplified as O-decay genes). The degradation of O-decay genes is critical in the progression of meiosis [7]. However, a growing body of evidence suggests that the dynamics of maternal RNA commonly detected by Smart-seq2, a low-input technology for constructing next-generation sequencing library [8], may not be accurate enough. Even with strict thresholds, there are genes found to be upregulated during GV to MII stage oocytes [9], which contradicts the known transcriptionally dormant state. Experimental evidence is required to explain this bias, and the actual dynamics of the maternal transcriptome remain unclear.

Both Smart-seq2 and total RNA-seq were designed for high-throughput transcriptome analysis through next-generation sequencing (Additional file 1: Fig. S1 A). Smart-seq2 has been more commonly used for oocytes because of the lower input required [8]. The oligo-d(T) primer used in Smart-seq2 can anneal to the polyadenylated tail (poly(A) tail) of mRNA through complementary pairing between thymine and adenine bases [8]. Many recent studies have reported that the poly(A) tail of maternal mRNA undergoes drastic changes during meiosis [10,11,12]. It remains unclear whether these alterations affect the capture efficiency of oligo-d(T) primers. The poly(A)-inclusive full-length RNA isoform sequencing method version 2 (PAIso-seq2) [10, 11] is a recently developed technology that enables the detection of complete poly(A) information in RNA molecules without relying on oligo-d(T) primers. As both PAIso-seq2 and a recent edition of total RNA-seq are available for relatively low-input samples, they are suitable for detecting poly(A) dynamics and the abundance of maternal transcripts.

As the transcriptome will not be renewed for a long time, how oocytes maintain homeostasis of the transcriptome deserves further elucidation. Zygote arrest-1 (Zar1) is an oocyte-specific gene. Oocytes lacking ZAR1 undergo normal meiosis but are arrested at the one- or two-cell stages after fertilization [13, 14]. Zar1-like (Zar1 l or Zar2) is an ortholog of Zar1 [15, 16], which has been reported to repress translation in immature Xenopus oocytes [17] and regulate epidermalization in early embryos [18]. However, Zar2 has only a minor effect on mouse fertility [13]. Using Smart-seq2, we previously observed abnormal accumulation of maternal mRNAs in Zar1/2^–/– MII-stage oocytes. Based on this observation, we hypothesized that ZAR1 promotes the degradation of maternal mRNAs during meiosis [13]. However, more recent research revealed that ZAR1 participates in the assembly of the mitochondria-associated ribonucleoprotein domain (MARDO), a non-membrane structure responsible for storing translationally repressed mRNAs and stabilizing maternal transcripts prior to GVBD [19]. This reported function of ZAR1 appears to contradict our previous findings. Given that total RNA-seq was employed in the MARDO study, it would be worthwhile to investigate if the discrepancies in the results are influenced by biases associated with sequencing methodologies.

MARDO has been reported to store maternal mRNAs in dormant oocytes. However, its RNA content remains largely uncharacterized [19]. ZAR1 has been identified as an RNA-binding protein (RBP) containing a CxxC zinc-finger domain at its C-terminus, which supports its ability to interact with RNA [13, 14, 16, 20, 21]. As the main component of MARDO, ZAR1-target genes are hypothesized to be closely linked with MARDO-stored mRNAs. However, the profile of ZAR1-target RNAs has not been reported due to the high requirements of previously available technologies. Recently, linear amplification of complementary DNA ends and sequencing (LACE-seq) [22] has enabled the accurate identification of RBP-binding sites using low-input samples. This technology is well-suited for detecting ZAR1-target RNA in oocytes.

In this study, we integrated Smart-seq2, total RNA-seq, and PAIso-seq2 data to improve the understanding of maternal RNA dynamics during meiosis, redefine O-decay genes, correct previously inaccurate hypotheses regarding the function of ZAR1, and uncover its novel role in orchestrating maternal RNA polyadenylation. Moreover, by integrating LACE-seq and immunoprecipitation-mass spectrometry (IP-MS) data, we provided the first report on ZAR1-interacting mRNAs and proteins. The data provides novel insights that may contribute to a better understanding of the mechanisms underlying ZAR1 function.

Results

The dynamics of mRNA detected by Smart-seq2 can be biased by polyadenylation

During meiosis after GVBD, transcription of the maternal genome keeps silent, and the gradual diminishing of the maternal transcriptome has been observed for years (Fig. 1A). Genes with transcripts degraded during this period are referred to as O-decay genes. Smart-seq2, also known as poly(A)-tailed RNA sequencing technology (Additional file 1: Fig. S1 A), is commonly used to explore maternal RNA dynamics and define O-decay genes during meiosis [8]. According to the sequencing results, many genes were downregulated in MII versus GV stage oocytes (n = 3091, Fig. 1B). The relative RNA copy number calculated by ERCC was also significantly downregulated in MII stage oocytes (~ 60% in Fig. 1C, Additional file 2: Table S1). However, among the differential expressing genes (DEGs) defined by Smart-seq2, there were also many upregulated genes (Fig. 1B), which is contrary to the dormant state of transcription after GVBD. The same tendency was evident in the results detected by Smart-seq3 (Additional file 1: Fig. S1B) [23, 24].

According to the Smart-seq2 and Smart-seq3 processes, oligo-d(T) primers were used to capture RNA [25] (Additional file 1: Fig. S1 A); therefore, the DEGs might have been influenced by these primers. To validate this, we primed reverse transcription with oligo-d(T) primers to mimic Smart-seq2, using GV and MII stage oocytes as samples, and compared the capture efficiency with cDNA reverse transcription primed by random primers. The RT-qPCR results indicated that these two different RT primers could lead to widely divergent results (Fig. 1D and Additional file 1: Fig. S1 C). Thus, the O-decay genes defined by Smart-seq2 might also be biased.

The oligo-d(T) primer was primarily designed to capture the poly(A) sequence by denaturing to it. Therefore, the biased results might be attributed to dynamic polyadenylation. To investigate this, PAIso-seq2 analysis [10, 11] (Additional file 1: Fig. S2) was employed to detect the dynamics of polyadenylation status during meiosis [10, 11]. Interestingly, a drastic deadenylation in oocytes during the transition from GV to MII stage was observed (Fig. 1E–G, Additional file 3: Table S2).Over 90% of the downregulated genes identified by Smart-seq2 during this transition were found to be deadenylated (Fig. 1H, I). The mean length of the poly(A) tail was reduced to 8 nt (Fig. 1E), which is significantly shorter than the length of oligo-d(T) primers. This observation suggests a potential bias in capture efficiency of oligo-d(T) primers.

Total RNA-seq defines convincing dynamics of maternal transcriptome

To avoid the bias of Smart-seq2, we employed total RNA-seq, a newer library construction technology capable of capturing transcripts independently of the poly(A) tail by using random primers during reverse transcription (Additional file 1: Fig. S1 A), to identify the DEGs in MII versus GV stage oocytes (Additional file 1: Fig. S3 A). As expected, the global transcripts diminished, with only a few upregulated genes (Fig. 2A, B, Additional file 4: Table S3). A similar trend was observed in previously published data (Additional file 1: Fig. S3B) [26], which more accurately reflect the physiological transcriptional silencing during meiosis compared those in Smart-seq2. However, the degree of transcript degradation observed with total RNA-seq was relatively lower than that detected with Smart-seq2 (~ 30% in total RNA-seq; Fig. 2B), and the number of O-decay genes identified by total RNA-seq was much smaller than that defined by Smart-seq2 (Fig. 2A). Given that total RNA-seq displayed better sensitivity (Additional file 1: Fig. S3 C and S3D), it provides a more reliable depiction of the dynamics of maternal RNA during meiosis.

We further analyzed the O-decay genes detected by total RNA-seq. Firstly, based on the poly(A) length, translational efficiency (TE) [27], and protein abundance (Additional file 1: Fig. S4) patterns of GV, MI, and MII stage oocytes, we grouped the genes into clusters (Additional file 1: Fig. S5, Additional file 5: Table S4). We then examined these three features of the O-decay genes and observed that, as compared to the global pattern (represented by all genes), the downregulated genes displayed a higher proportion of the cluster with the longest poly(A) tails in GV stage oocytes, but immediately shortened in MI stage oocytes (Fig. 2D). These results were consistent with those from the other two groups, showing that the proportion of O-decay genes with the highest TE and protein intensity in GV stage oocytes was significantly higher than the background (Fig. 2E, F). These results suggest that O-decay genes are likely polyadenylated and translated primarily in GV stage oocytes, and subsequently degraded after fulfilling their functional roles.

Furthermore, by comparing the DEGs detected by total RNA-seq and Smart-seq2, we observed that the downregulated genes were better correlated between the two technologies, whereas most of the upregulated and downregulated genes detected by Smart-seq2 were relatively stable in the total RNA-seq results (Fig. 2C). Since Smart-seq2 may be biased due to its dependence on oligo-d(T) primers, which target poly(A) tails, we hypothesize that maternal RNA undergoes more extensive polyadenylation or deadenylation than degradation during meiosis. Given the differences in how total RNA-seq and Smart-seq2 are processed in oocytes, previous studies using Smart-seq2 to examine the dynamics of maternal RNA may also be biased.

Maternal mRNAs are unstable in Zar1/2 ^–/–GV stage oocytes

The absence of maternal ZAR1 in mice leads to female infertility by resulting in developmental arrest at the one-to-two-cell stage [13, 14, 16] (Additional file 1: Fig. S6). In our previous studies using Smart-seq2, we observed abnormally accumulation of many genes in Zar1/2^–/– MII versus GV stage oocytes. These DEGs were found to overlap with O-decay genes defined by Smart-seq2, suggesting that ZAR1 plays a role in promoting the degradation of maternal transcripts [13]. However, a more recent study identified ZAR1 as a component of MARDO, a mitochondria-associated ribonucleoprotein domain, which is believed to accumulate and protect maternal transcripts, and the deletion of ZAR1 leads to abnormal degradation of maternal transcripts [19] (Additional file 1: Fig. S6).

To illustrate this contradiction, we performed total RNA-seq to detect DEGs in Zar1/2^–/– oocytes (Additional file 1: Fig. S3 A). According to the results, the trends of DEGs in Zar1/2^–/– versus WT GV stage oocytes and Zar1/2^–/– versus WT MII stage oocytes detected by total RNA-seq are both mainly downregulated (Fig. 3A, B and Additional file 1: Fig. S7 A). Moreover, the DEGs in Zar1/2^–/– MII stage oocytes were largely inherited from Zar1/2^–/– GV stage oocytes (Fig. 3 C, 50.12% among DEGs in Zar1/2^–/– MII stage oocytes detected by total RNA-seq), which aligns with the recent finding that ZAR1 stabilizes maternal RNA in GV stage oocytes [19].

Moreover, we found that the downregulated genes in Zar1/2^–/– GV stage oocytes were mostly correlated with the genes that ought to be downregulated or stable during the GV to MII stage oocytes (Additional file 1: Fig. S7B; 61.76% among DEGs in Zar1/2^–/– GV stage oocytes), and the O-decay genes were also significantly downregulated in GV stage oocytes (Fig. 3H and Additional file 1: Fig. S7 C). However, the proportion of downregulated genes shared between Zar1/2^–/– versus WT MII stage oocytes and WT MII versus GV stage oocytes was much smaller than that shared between Zar1/2^–/– versus WT GV stage oocytes and WT MII versus GV stage oocytes (Additional file 1: Fig. S7B and S7D; 8.20% and 26.33%, respectively). These findings indicate that the O-decay genes were degraded normally during meiosis in Zar1/2^–/– oocytes, consistent with the trend shown in Additional file 1: Fig S7 C. Overall, these results indicate that ZAR1 may prevent the O-decay genes from premature degradation in GV stage oocytes to ensure these genes function properly. In addition, the upregulated genes detected by total RNA-seq in Zar1/2^–/– MII stage oocytes were negatively correlated with the downregulated genes in GV to MII stage oocytes (Additional file 1: Fig. S7D, 18.86% of upregulated genes among all DEGs in Zar1/2^–/– MII stage oocytes detected by total RNA-seq), suggesting that degradation during meiosis was also disrupted in Zar1/2^–/– oocytes.

Disruption of polyadenylation during meiosis in Zar1/2 ^–/– GV stage oocytes

We reanalyzed the Smart-seq2 data and found that the abnormal accumulation of many RNA transcripts still appeared in Zar1/2^–/– MII stage oocytes (Fig. 3D, E), which differed significantly from the results detected by total RNA-seq. Then we used oligo-d(T) and random primers for RT-qPCR to detect the transcript levels in Zar1/2^–/– versus WT MII stage oocytes, and the results revealed a bias (Fig. 3F, G). Furthermore, most of the DEGs in Zar1/2^–/– MII stage oocytes detected by Smart-seq2 remained unchanged when assessed by total RNA-seq (Fig. 3I; 62.25%). Specifically, those abnormally upregulated genes detected by Smart-seq2 in Zar1/2^–/– versus WT MII stage oocytes were deadenylated during meiosis (Fig. 3J).

The findings indicated that polyadenylation in Zar1/2^–/– oocytes may also be disrupted. To assess this possibility, we processed PAIso-seq2 to detect the differentially polyadenylated genes (DPGs) in Zar1/2^–/– oocytes (Fig. 4A–D). As expected, the poly(A) tails of DPGs in Zar1/2^–/– MII stage oocytes were abnormally elongated (Fig. 4C, D). The poly(A) length of abnormally upregulated genes in Zar1/2^–/– versus WT MII stage oocytes was elongated (Fig. 4E), and the dynamics of poly(A) length were correlated with the DEGs detected by Smart-seq2 in Zar1/2^–/– MII stage oocytes (Fig. 4F). Thus, the abnormally upregulated genes detected by Smart-seq2 in Zar1/2^–/– MII stage oocytes were likely due to abnormal poly(A) length.

Next, we explored the temporal dynamics of DPGs in Zar1/2^–/– MII stage oocytes by comparing them with DPGs in Zar1/2^–/– versus WT GV stage oocytes and in WT GV versus MII stage oocytes. These comparison revealed that the DPGs in Zar1/2^–/– MII stage oocytes were negatively correlated with those during meiosis (Fig. 4G), but were not inherited from Zar1/2^–/– GV stage oocytes (Fig. 4H), suggesting abnormal polyadenylation both in Zar1/2^–/– GV stage oocytes and during meiosis.

The collective findings indicated that the abnormal polyadenylation of genes in Zar1/2^–/– MII stage oocytes interfered with Smart-seq2 detection. Polyadenylation events were disrupted both in Zar1/2^–/– GV stage oocytes and during meiosis.

ZAR1 stabilizes the maternal transcript by binding to the 3′ UTR region

ZAR1 is an RBP that binds to maternal transcripts via a CxxC zinc finger domain in its C-terminus [13, 14, 16]. (Additional file 1: Fig. S6). Accordingly, it is possible that ZAR1 functions as an RBP that selectively binds to maternal transcripts and regulates their stability and polyadenylation in oocytes.

To confirm this, LACE-seq was employed to accurately detect the ZAR1-target transcripts in GV stage oocytes with an antibody that specifically recognizes ZAR1 (Additional file 1: Fig. S8 A and S8B, Additional file 6. Table S5) [13]. To avoid non-specific signals, we combined the groups using Zar1/2^–/– oocytes as samples with those using IgG as the negative control (NC) group, while all the WT samples were merged into the test group (ZAR1) (Additional file 1: Fig. S8 C and S8D). Many confirmed ZAR1-target genes could also be detected by LACE-seq (Additional file 1: Fig. S8E) [13]. Screening of ZAR1-target genes with a fold change of at least 5 in the ZAR1 versus NC group identified over 8000 candidate genes (Additional file 1: Fig. S8 F), which were subsequently analyzed.

First, we assessed the differential expression of ZAR1-target genes in GV stage oocytes and found that ZAR1-target DEGs were predominantly downregulated in Zar1/2^–/– GV stage oocytes. More than half of the abnormally downregulated genes were targeted by ZAR1, and their proportion was much higher than that of the abnormally upregulated genes (Fig. 5A; 54.8% downregulated genes and 29.6% upregulated genes were targeted by ZAR1). To clarify the relationship between ZAR1 targeting and mRNA stability, we defined the binding strength based on read counts detected by LACE-seq. Genes with the highest (top 500) and lowest (bottom 500) ZAR1 binding strength were then categorized into separate groups. The fold changes of the top 500 and bottom 500 genes in Zar1/2^–/– versus WT GV stage oocytes were compared. As expected, the top 500 genes showed significantly lower fold changes in Zar1/2^–/– GV stage oocytes compared to the bottom 500 genes (Fig. 5B). This provided further evidence that ZAR1 binding stabilizes maternal RNA.

The data indicated that ZAR1 primarily targets the 3′ UTR and CDS of transcripts with 40.3% and 57.6% of the peaks annotated to these two features, respectively (Fig. 5C). The genes were divided into three groups based on the features to which they were bound: genes with peaks annotated mainly to the 3′ UTR (n = 1715), mainly to CDS (n = 2553), and annotated to both the 3′ UTR and CDS (n = 1122) (Fig. 5D). The fold changes of these genes in Zar1/2^–/– versus WT GV stage oocytes were also compared across the gene groups. Genes with ZAR1 binding mainly to the 3′ UTR were expressed at significantly lower levels compared to genes in the other two groups (Fig. 5E). These results suggested that ZAR1 may bind to the 3′ UTR to further stabilize mRNAs.

Furthermore, we were curious about whether the targeted genes could be characterized by their dynamics during meiosis, so we analyzed the poly(A) and TE features of the ZAR1-target genes, dividing them according to their binding strengths (Fig. 5G, H, Additional file 1: Fig. S5). Interestingly, the enrichment of ZAR1-target genes with both poly(A) tail lengths and TE highest in GV stage oocytes decreased as binding strength increased, suggesting that ZAR1 may not bind to the previously defined O-decay genes. In addition, as binding strength enhanced, the ZAR1-target genes with poly(A) length sustained in GV-MI stage oocytes, as well as those highest in MI stage oocytes, both increased. This pattern was similar to the genes with the highest TE in MI stage oocytes or those sustained in MI-MII stage oocytes. These findings suggest that ZAR1 may play a significant role in regulating the dynamics of gene translation after GVBD (Fig. 5F–H).

ZAR1 indirectly regulates the polyadenylation of maternal transcript

As polyadenylation of maternal transcripts was significantly disrupted in Zar1/2^–/– oocytes (Fig. 4A–D), we also compared the correlation between differential polyadenylation and ZAR1 binding. Approximately 40% of the DPGs were ZAR1-target genes in both Zar1/2^–/– GV and MII stage oocytes, with no significant preference (Additional file 1: Fig. S9 A and S9B). The fold change in poly(A) length in Zar1/2^–/– versus WT GV and MII stage oocytes between the ZAR1-target top 500 genes and bottom 500 genes showed no significant difference (Additional file 1: Fig. S9 C and S9D), nor did the genes with specific binding features (Additional file 1: Fig. S9E and S9 F).

We then wondered whether the poly(A) length of genes could characterize ZAR1 binding. First, we divided the genes expressed in GV stage oocytes into two groups: a short-poly(A)-carrying group (Short) and a long-poly(A)-carrying group (Long) using a 30-nt boundary as defined in a recent study (Additional file 1: Fig. S10 A and S10B) [28]. Interestingly, we found that the proportion of genes with no poly(A) significantly decreased in ZAR1-target group. This decrease may explain the increased proportion of genes with other poly(A) lengths (Additional file 1: Fig. S10 A). However, although the percentage increase of genes with 0 < pA ≤ 30 was slightly higher than that of genes with 30 < pA ≤ 105, the fold changes in both groups were quite low (less than 1.3) in the ZAR1-target group compared to the global group (Additional file 1: Fig. S10B). Thus, we proposed that ZAR1 targeting is independent of poly(A) length.

Then we discussed whether ZAR1 has specific regulatory effects on genes with different poly(A) lengths. Interestingly, we found that the genes in the short group are more likely to be abnormally polyadenylated in Zar1/2^–/– GV stage oocytes but deadenylated in Zar1/2^–/– MII stage oocytes (Additional file 1: Fig. S10 C). In contrast, for the long group genes, the number of deadenylated genes in Zar1/2^–/– versus WT GV stage oocytes was significantly higher than the number of polyadenylated genes, whereas in Zar1/2^–/– MII stage oocytes, the number of deadenylated genes was lower than that of polyadenylated genes (Additional file 1: Fig. S10 C). These results indicate that ZAR1-target long-tailed genes are more susceptible to disruption when ZAR1 is deleted in GV stage oocytes, and this effect persists into MII stage oocytes. Additionally, we assessed the stability of ZAR1-tareget short- or long-tailed genes and found that genes in both groups tended to be downregulated in Zar1/2^–/– GV stage oocytes (Additional file 1: Fig. S10D and S10E). This impact persists into the MII stage, as the trends of DEGs remain the same as in Zar1/2^–/– GV stage oocytes (Additional file 1: Fig. S10 F and S10G). These results are similar to the global trends (Fig. 3A, B), indicating that the stability of targeted maternal RNA in Zar1/2^–/– oocytes is independent of their poly(A) length.

The collective findings indicate that ZAR1 may indirectly regulate the maternal polyadenylation, suggesting that it interacts with other proteins to regulate polyadenylation in Zar1/2^–/– oocytes.

ZAR1-targeted maternal mRNAs exhibit abnormal polyadenylation and stability upon ZAR1/2 deletion

As the maternal transcriptome was globally disrupted in Zar1/2^–/– oocytes (Fig. 3 A, B and 4 A–D), we further examined whether ZAR1 regulates the stability or polyadenylation of key maternal mRNAs to ensure their proper expression. We divided the ZAR1-target genes into 4 groups based on both their polyadenylation and stability in Zar1/2^–/– versus WT GV or MII stage oocytes: genes with normal polyadenylation and stable (Normal Stable), the genes with normal polyadenylation but unstable (Normal Unstable), the genes with abnormal polyadenylation but stable (Abnormal Stable) and the genes with abnormal polyadenylation and unstable (Abnormal Unstable) (Additional file 1: Fig S11 A and S11B, Additional file 7: Table S6). As expected, only half of the genes were abnormally polyadenylated in Zar1/2^–/– oocytes, whereas over 70% of the genes were unstable (Additional file 1: Fig. S11 A). This corresponds to our findings that ZAR1 has no preference for polyadenylation status (Additional file 1: Fig. S9), but for the stability (Fig. 5).

We further examined whether key maternal genes are influenced by ZAR1. Interestingly, many known maternal genes were disrupted in Zar1/2^–/– oocytes (Additional file 1: Fig. S11 A and S11B). For example, Cnot7, Ccnb1, and Wee2, of which the protein intensity had been reported to be disrupted in Zar1/2^–/– oocytes, were targeted by ZAR1 and were unstable (Additional file 1: Fig. S11 A, B). Besides, many other genes that are indispensable for meiosis like Cnot6 l [29], Pabpc1 [28], and Polr2j were also abnormal in either polyadenylation or stability (Additional file 1: Fig. S11 A and S11B). However, due to our strict threshold, many confirmed ZAR1-target genes were excluded from the ZAR1-target list [13]. To further explore the effect of ZAR1 on the maternal transcriptome, we included these genes in Additional file 1: Fig. S11 C. As expected, all of these genes were abnormal in Zar1/2^–/– oocytes (Additional file 1: Fig. S11 C). For example, the RNA degradation factor Btg4 [30], which belongs to the Abnormal Unstable groups, has been reported to fail in expression in Zar1/2^–/– oocytes. Other meiosis-related genes, such as Cpeb1 [31], Lsm14b [32], Igf2bp2 [33], Zp2 [34], and Ybx2 [35], were also disrupted in Zar1/2^–/– oocytes, suggesting a potential role for ZAR1 in ensuring the expression of maternal RNAs to maintain meiosis.

ZAR1 interact with other proteins to regulate the stability of maternal RNA

IP-MS was performed to identify candidate proteins interacting with ZAR1. The analysis identified 715 ZAR1-interacting proteins (Additional file 1: Fig. S12 A, Additional file 7: Table S6). Since the IP involved whole cell lysates from the ovaries, nuclear proteins that may also be enriched by ZAR1 antibodies were excluded from the analysis. Finally, 420 candidate ZAR1-interacting proteins were identified (Additional file 1: Fig. S12B). Some of these candidate proteins have been previously reported as mitochondrial proteins, which correlated to MARDO and may interact with ZAR1, further supporting the reliability of the results (Additional file 1: Fig. S12 C and S12D) [13]. However, when ovary lysates were used to identify ZAR1-interacting proteins in oocytes via IP-MS, the presence of a significant number of cumulus cells in the sample led to the detection of homologs of functional oocyte proteins, such as LSM14 s. LSM14B is highly expressed in oocytes and interacts with ZAR1, whereas LSM14 A, a homolog more highly expressed in cumulus cells than LSM14B, was detected by IP-MS [36] (Additional file 1: Fig. S12 C).

Gene ontology analysis was then performed to explore the ZAR1-interacting proteins. As expected, biological processes such as regulation of RNA metabolic processes, RNA stability, mRNA processing, and cytoplasmic translation were significantly enriched (Fig. 6A–D). Among the putative proteins detected by IP-MS, some have been reported to be essential for regulating stability and polyadenylation in oocytes, such as poly(A) binding protein (PABP) [37]. PABPN1 promotes cytoplasmic polyadenylation during meiosis in oocytes [38] and its deletion severely impedes oocyte maturation [39]. PABPC1 and PABPC4 have also been identified as putative ZAR1-interacting proteins [28, 40], while the homolog PABPC1L is reportedly vital. Additionally, Y-box-binding protein 2 (YBX2) stabilizes maternal mRNA in mouse oocytes [35].

YTH N6-methyladenosine RNA-binding protein F (YTHDF) [41] and insulin-like growth factor 2 mRNA-binding protein (IGF2BP) [42] are m⁶A readers that bind to m6 A-containing RNA and regulate its stability. Among the putative ZAR1-interacting proteins, YTHDF1 and IGF2BP1 were detected. The homologs YTHDF2 [43] and IGF2BP2 [33] are known to stabilize m6 A-containing mRNA in oocytes. As expected, IGF2BP2 can be co-immunoprecipitated by ZAR1 (Additional file 1: Fig. S12E). Furthermore, the location of m⁶A-containing mRNA in GV stage oocytes is similar to that of MARDO [44], suggesting that m⁶A-containing mRNA may also be stored in MARDO in oocytes by indirectly interacting with ZAR1.

Many other proteins have been reported to regulate the stability or polyadenylation of RNA in various cell lines or in vitro. For example, trinucleotide repeat containing adaptor 6B (TNRC6B) scaffolds mRNA and deadenylases (such as PAN3 and NOT1), promoting deadenylation and degradation through the miRNA pathway [45]. Deletion of TNRC6B suppresses the silencing caused by the miRNA-induced silencing complex. ELAV-like RNA-binding protein 1 (ELAVL1) binds to mRNA with a short poly(A) tail, preventing its degradation and promoting translation [46, 47]. According to a published study, ELAVL2, a homolog of ELAVL1, is highly expressed in oocytes and displays a similar location to ZAR1 in GV stage oocytes [48], suggesting a potential interaction with ZAR1.

The collective findings suggest that ZAR1 may interact with other proteins to regulate the homeostasis of the maternal transcriptome. The candidate proteins, which lack further research in oocytes, imply the existence of unclear pathways involved in maintaining maternal transcriptome.

Chromatin compaction cannot be maintained in Zar1/2 ^–/–MII stage oocytes

The arrest of Zar1/2^♀–;♂+ embryos at the zygote stage might be attributed to a failure in maintaining MII stage oocytes. To test this, MII stage oocytes were cultured in vitro starting 14 h after hCG injection. As expected, Zar1/2^–/– MII stage oocytes fragmented or died 40 h after hCG injection, which occurred more rapidly than that in WT MII stage oocytes (Fig. 7A, B). Unexpectedly, some Zar1/2^–/– MII stage oocytes were unable to maintain condensed chromatin and instead formed pronucleus-like (PN-like) structures 40 h post-hCG injection (Fig. 7A).

To determine whether these abnormalities were present at earlier stages, immunofluorescence was performed to examine spindle alignment and chromatin states. Zar1/2^–/– MII stage oocytes exhibited aberrant spindle alignment as early as 21 h after hCG injection, with only ~ 10% showing diffused chromatin at that time. Both abnormalities worsened 32 h after hCG infection (Fig. 7C–E). Previous studies have reported that LSM14B, a MARDO co-localized protein, is essential for maintaining oocytes at the MII stage, as its deletion results in misaligned spindles and the formation of PN-like structures due to decondensed chromatin [19, 36, 49]. We hypothesized that similar phenotypes would be observed in Zar1/2^–/– MII stage oocytes. Surprisingly, Lsm14b mRNA levels remained unaltered despite the disruption of polyadenylation in Zar1/2^–/– oocytes (Fig. 7F and Additional file 1: Fig. S13 A), while Mastl, another gene known to cause PN-like structure when deleted in oocytes and reported to be downregulated in Lsm14b-deleted oocytes, remained mostly stable at the transcript level (Additional file 1: Fig. S13B).

Given the observation of spindle misalignment 21 h after hCG injection, RT-qPCR was performed on Zar1/2^–/– MII stage oocytes (Fig. 7G). As expected, many genes were downregulated in Zar1/2^–/– oocytes, suggesting that ZAR1 plays a role in stabilizing these genes, thereby contributing to the maintenance of oocytes at the MII stage.

Discussion

Owing to the limited number of oocytes, Smart-seq2 has been widely used to detect the dynamics of the maternal transcriptome during MZT. Whether the limitation of the technology misled the detected trends remains unclear. In this study, to elucidate the dynamics of maternal mRNA, we analyzed datasets obtained using Smart-seq2, total RNA-seq, and PAIso-seq2. The degradation of the maternal transcriptome during meiosis, as detected by total RNA-seq, was less severe compared to that detected by Smart-seq2 (Fig. 2B). Moreover, according to PAIso-seq2, the maternal transcriptome was dramatically deadenylated to an average of 8-nt in MII stage oocytes (Fig. 1E), which likely interferes with the detection of the oligo-d(T) primer used in Smart-seq2 [8], as it consists of 30 thymidine residues. Therefore, we suggests that Smart-seq2 may not suitable for samples where the poly(A) tail is too short, such as in MII stage oocytes, as it may lead to biased detection. In contrast, total RNA-seq, which does not rely on poly(A) tails to capture RNA, is more appropriate.

In addition, we redefined the O-decay genes using total RNA-seq and found that these genes were primarily translated in GV stage oocytes, with high protein levels observed in both GV and MI stage oocytes (Fig. 2D–F). It is conceivable that these genes must be translated before GVBD, after which their transcripts needed to be cleaned (Fig. 8A). Further evidence is required to support this.

The maternal transcriptome remains relatively stable during meiosis, suggesting that polyadenylation plays a more critical role in regulating the meiotic process than previously recognized. Zar1 is a key gene involved in oocyte maturation and early embryonic development [13, 14, 16]. Recently, Zar1 was reported to stabilize maternal transcripts in GV stage oocytes by participating in the assembly of MARDO [19]. This function contrasts with its previously reported role in promoting the degradation of maternal transcripts during meiosis. Since previous studies have utilized Smart-seq2 to detect DEGs, we suspect that disruptions in polyadenylation in Zar1/2^–/– oocytes may lead to misleading results.

As expected, the differential expression trends of DEGs detected by total RNA-seq in Zar1/2^–/– oocytes differed from those detected by Smart-seq2, revealing that more genes had downregulated transcripts than upregulated ones in Zar1/2^–/– oocytes (Fig. 3A–D). These findings suggest that ZAR1 plays a role in preventing maternal transcript degradation. Additionally, PAIso-seq2 results showed that polyadenylation was disrupted during meiosis. Some genes with abnormally polyadenylated transcripts in Zar1/2^–/– MII stage oocytes were misidentified by Smart-seq2, leading to incorrect interpretation (Fig. 4E).

In addition, genes with abnormally polyadenylated transcripts in MII stage oocytes were negatively correlated with the trend of polyadenylation during meiosis, but weakly correlated with those abnormally adenylated in GV stage oocytes (Fig. 4G, H). Thus, in addition to maintaining the stability of the maternal transcriptome, ZAR1 regulates polyadenylation both in GV stage oocytes and during meiosis.

ZAR1 acts as an RBP with a zinc finger domain at the C-terminus [13, 16, 21]. However, its binding profile remains unclear. In this study, we used LACE-seq to screen for ZAR1-target genes. Combined with total RNA-seq, we obtained evidence suggesting that ZAR1 may regulate the stability of maternal RNA by binding to its 3′ UTR (Fig. 5C–E). However, the regulation of polyadenylation by ZAR1 is indirect, as there is no preference for ZAR1 binding to DPGs in Zar1/2^–/– oocytes (Additional file 1: Fig. S9 and S10). Furthermore, ZAR1-target genes are distinct from O-decay genes and are primarily translated in the later stages of meiosis (Fig. 5F, H). Therefore, we hypothesized that ZAR1 may bind to the transcripts that are required for later meiosis, inhibiting their translation in GV stage oocytes to ensure their translational activation during later stages of meiosis.

According to Additional file 1: Fig. S11, we found that many important maternal RNAs targeted by ZAR1, detected by LACE-seq, were disrupted in terms of polyadenylation and stability in Zar1/2^–/– oocytes. Notably, Btg4 [30], a trigger for maternal mRNA degradation via CCR4-NOT complex, was reported to fail translation during meiosis in Zar1/2^–/– oocytes [13], and showed abnormal deadenylation and degradation in both Zar1/2^–/– GV and MII stage oocytes (Additional file 1: Fig. S11 C). Additionally, Ccnb1 transcripts, which belong to the Abnormal Unstable group, may lead to the failure of Cyclin B1 accumulation in Zar1/2^–/– MII stage oocytes [13]. Other maternal mRNAs, including Lsm14b, Cpeb1, were also disrupted to varying extents (Additional file 1: Fig. S11), suggesting that ZAR1 regulates the expression of specific maternal transcripts to ensure normal meiosis. However, further evidence is needed to explore the underlying mechanisms behind these abnormalities.

However, we did not find a conserved motif in the LACE-seq results (data not shown). Since ZAR1 contains only an RNA-binding domain but lacks a specific RNA-recognition domain, we hypothesize that ZAR1 may recognize transcripts through other mechanisms, such as interacting with other RBPs. Among ZAR1-interacting proteins, PABPN1 [39] and YBX2 [35] are known to regulate the polyadenylation and stability of the maternal transcriptome. Additionally, several putative proteins with homologs that regulate the oocytes transcriptome, including YTHDFs [43] and IGF2BPs [33, 42], which control mRNA homeostasis through binding to m6 A modifications, and LSM14s [36], which act as hubs for RBP-mRNA interactions, were identified. Moreover, proteins like ELAVL2 [48], which are known to regulate mRNA stability in vitro, were also detected. These findings suggest that ZAR1 may interact with other RBPs to regulate the stability and polyadenylation of maternal transcripts in oocytes. Therefore, ZAR1 likely interact with other RNA-regulatory proteins to determine target genes and maintain proper polyadenylation (Fig. 8B, C).

In addition, we observed that in Zar1/2^–/– MII stage oocytes, the maintenance of meiotic arrest was disrupted, accompanied by abnormal tubulin distribution, oocyte fragmentation, and partial chromatin agglutination failure (Fig. 7). Previous studies have shown that the deletion of other proteins, such as Lsm14b [36] and Mastl [50], also led to the formation of PN-like nuclei at the time when the oocytes should reach the MII stage. These defects were primarily due to failed activation and phosphorylation of CDK1. However, although the poly(A) tail of Lsm14B was significantly shorter in Zar1/2^–/– versus WT oocytes (Additional file 1: Fig. S11 and S12 A), Mastl appeared to be unaffected in Zar1/2^–/– oocytes (Additional file 1: Fig. S11 and S12B). Moreover, only a relatively small percentage of Zar1/2^–/– oocytes exhibited a phenotype similar to those lacking these two genes (Fig. 7). Since Cyclin B1 levels have been reported to be significantly lower in Zar1/2^–/– oocytes [13], we suspect that the failure to maintain pCDK1-T161 may be responsible for the inability of the oocytes to remain arrested at the MII stage. Additionally, we observed that the transcripts of genes crucial for meiotic maturation were degraded at 32 h post-hCG treatment during in vitro culture (Fig. 7G), indicating that the stability of maternal mRNAs are also disrupted after the MII stage. Taken together, these findings suggest that ZAR1 and ZAR2 may play a role in maintaining MII stage oocytes through mechanisms distinct from Lsm14B or Mastl.

Conclusion

We clarified the limitations of Smart-seq2 in comparison to total RNA-seq, providing new insights into the dynamics of the maternal transcriptome in oocytes and highlighting the critical role of polyadenylation during meiosis. We also offered suggestions for the appropriate application of Smart-seq2 (Fig. 8A). Additionally, through multi-omics analysis, we addressed the misinterpretations caused by Smart-seq2, conducted a more comprehensive exploration of ZAR1’s regulation of the maternal transcriptome (Fig. 8B, C), and presented an interaction protein profile that serves as a resource for future studies on the regulatory network and functions of MARDO-related proteins driven by ZAR1.

Methods

Collection and culture of mouse oocytes

All mice used in this study were maintained in a specific pathogen-free environment under suitable life conditions at the Laboratory Animal Center of Zhejiang University. Zar1/2^–/– mice were generated using the CRISPR/Cas9 system as described previously [13]. As all Zar1/2^–/– mice had a C57BL6/ICR hybrid strain background, we used ICR mice as the wild type in this study.

Oocytes were collected from mice aged 4–6 weeks, which were injected with 5 IU of pregnant mare serum gonadotropin (PMSG; Ninbo Sansheng Pharmaceutical, China) and euthanized 44 to 48 h later. GV stage oocytes were collected using M2 medium (Sigma-Aldrich, USA; Cat. No. M7167) supplemented with milrinone (5 μM) to maintain the arrest state at GV stage. For collecting MI stage oocytes, GV stage oocytes were collected in M2 medium without milrinone, cultured in mini-drops of M16 medium (Sigma-Aldrich; Cat. No. M7292), and covered with mineral oil (Sigma-Aldrich; Cat. No. M5310) at 37℃ in a 5% CO₂ atmosphere for 6 h. Oocytes without GVs were collected as MI stage oocytes. For the collection of MII stage oocytes, 5 IU of human chorionic hormone per mice (hCG; Ningbo Sansheng Pharmaceutical, China) was injected 44 h after injection of PMSG. The mice were euthanized 14–16 h after hCG administration, and the MII stage oocytes were collected from the oviducts using an injector. The connection between granule cells and MII stage oocytes was digested with hyaluronidase, and clean MII stage oocytes were harvested for further use.

PAIso-seq2

One hundred GV and MII stage oocytes harvested from wild type (WT) and Zar1/2^–/– mice, and 100 MI stage oocytes harvested from WT mice were used for PAIso-seq2 library construction, as previously described [10, 11], with minor modification. Briefly, total RNA was extracted using TRIzol Reagent (Invitrogen, cat. no. 15596026). Then rRNA was removed using StarPure rRNA depletion Pools (BEIJING FOREVERSTAR BIOTECH, Cat.FS-R1051), and the sample was further cleaned up with RNA Clean & Concentrator-5 kit (Zymo Research; Cat. No. R1015). The barcode-containing adapter was ligated to the 3′-end of rRNA-removed RNA. After purification using an RNA Clean & Concentrator-5 kit (Zymo Research; Cat. No. R1015), the RNA was reverse transcribed to cDNA with template switching to add the second end PCR handler using a unique molecular identifier (UMI)-containing template-switching oligo (TSO). Full-length cDNA was amplified using primers matching the 3’ adaptor and TSO. The full-length cDNAs were concatenated into long molecules which was used for SMRTbell library preparation and sequenced on the PacBio Revio platform (GrandOmics).

PAIso-seq2 data analysis

The analysis pipeline was the same as previously described [10, 11]. Briefly, the obtained PacBio HiFi reads were segmented into single PAIso-seq2 cDNA reads according to the concatenating primer (Additional file 8: Table S7). Reads with successive barcodes were used for subsequent analysis. The segmented reads were split into samples according to their barcodes (Additional file 8: Table S7), and the UMI sequences for each read were extracted simultaneously. Duplicated reads were removed according to the UMI sequence, and the clean reads were aligned to the mm10 genome (gencode, vM25) using minimap2 (v2.24) [51]. poly(A) tails of the mapped clean reads were extracted and annotated following the previous published pipeline [12]. The poly(A) length of each gene was calculated as the geometric mean of poly(A) tail lengths of all its transcripts.

Total RNA-seq library construction

Thirty GV or MII stage oocytes were harvested from WT and Zar1/2^–/– female mice as described for each group, and three groups per sample were collected for replication. Harvested oocytes were washed three times with phosphate-buffered saline (PBS) containing 0.2% (wt/vol) bovine serum albumin (BSA) and transferred into 1.5-ml RNase/DNase-free PCR tubes. The TRIzol-chloroform-isopropanol method was used to obtain the total RNA. Briefly, 750 μl TRIzol reagent and 1 μl 1:1000 diluted External RNA Controls Consortium (ERCC), an RNA spike-in mixture for reproducible gene expression measurements, were added to each sample and mixed thoroughly. The cells were lysed at room temperature (RT) for 5 min before adding 500 μl chloroform and mixing thoroughly by vortexing. The preparation was centrifuged at 12,000 g at 4℃ for 15 min. The top phase was collected, with a small portion left behind to avoid contamination, ensuring the volume of the top phase remained consistent. An equal volume of isopropanol and 1 μl of GeneElute LPA (Sigma-Aldrich; Cat. No. 56575) was added, mixed, and frozen at − 20℃ overnight. Next, the mixture was centrifuged at 12,000 g at 4℃ for 30 min. The supernatant was removed, and the pellet was washed once with fresh 75% ethanol at 7500 g at 4℃ for 5 min then re-suspended in 12 μl deionized water.

The extracted RNA was used for total RNA-seq (Additional file 1: Fig. S1 A) with rRNA Depletion Module (H/M/R) (ABclonal, USA; Cat. No. RK20348) and library construction using Fast RNA-seq Lib Prep Kit V2 (ABclonal; Cat. No. RK20306). Briefly, rRNA was hybridized with probes, digested with RNase H, and then cleaned with DNase I. rRNA-depleted RNA was purified using beads (AFTMag NGS DNA Clean Beads, ABclonal, Cat. no. RK20257) and fragmented into 200–350-nt fragments with random primers hybridized. The fragmented and hybridized RNA was then reverse-transcribed to cDNA using random primers in a first–second-strand synthesis strategy. After ligating with the truncated adapter, the cDNA was purified using AFTMag NGS DNA Clean Beads (ABclonal; Cat. No. RK20257). Then approximately 11 amplification cycles were performed to amplificate and ligate the i5 and i7 indexes to the cDNA. The cDNA library was pooled for next-generation sequencing (NGS) on the Illumina Novaseq 6000 platform in the PE150 mode.

Total RNA-seq and Smart-seq2 data analyses

Smart-seq2 data from our previous study [13] were used for this study. Both Smart-seq2 data and total RNA-seq data were analyzed with a renewed analysis pipeline and genome mapping, as described below. Briefly, raw reads were trimmed using Trim-Galore (v.0.6.10) and mapped to the mm10 genome (genecode vM25) using STAR (v.2.7.10a) [52]. Uniquely mapped reads were quantified using FeatureCounts (v.2.0.2) [53] and further calibrated using ERCC spike-in as described previously [13]. Differentially expressed genes (DEGs, adjusted P < 0.01, average |log2 fold change|> 1) were analyzed using the DESeq2 [54] R package. The ERCC-calibrated count lists are provided in Additional file 2: Table S2 and Additional file 3: Table S3.

Reverse transcription quantitative real-time RCR (RT-qPCR)

Ten oocytes per sample were collected in an RNA-free PCR tube and lysed in 2 μl cell lysis buffer containing 0.2% Triton X-100 and 2 IU/μl recombination RNase inhibitor, followed by reverse transcription with random primer and the SuperScript II 1 st strand cDNA synthesis kit (TaKaRa Bio, Japan; Cat. No. 6210 A). The obtained cDNA was used for RT-qPCR using Power SYBR® Green PCR Master Mix (Thermo Fisher Scientific, USA; Cat. No. 4367659) and an Applied Biosystems 7500 Real-Time PCR System (Thermo Fisher Scientific). For relative mRNA level normalization, endogenous Gapdh was used for WT MII versus GV stage oocytes, and Actb was used for WT versus Zar1/2^–/– MII stage oocytes. All the primers used are listed in Additional file 10: Table S8.

LACE-seq library construction

The LACE-seq method was previously described [22] (Additional file 1: Fig S8 A). Briefly, 2 μg IgG per sample ZAR1 antibody was coupled to the activated protein A/G magnetic beads. Thirty GV stage oocytes per sample were harvested from WT and Zar1/2^–/– female mice as previously described and irradiated twice on ice with ultraviolet-C light at 400 mJ for RBP-RNA crosslinking. Next, the cells were lysed on ice, then antibody-coupled beads were added to the lysate, and the ZAR1-target RNA was immunoprecipitated using beads and fragmented by micrococcal nuclease (MNase). RNA dephosphorylation and the first-strand synthesis of reverse transcription were both performed on beads; the 3’ linker and biotin-conveyed T7-RT primer were then added to the RNA at this step. The first-strand cDNA was released and captured by streptavidin beads. Subsequent poly(A) tailing, second strand proceeding, and pre-amplification were all performed on the beads, and the products were purified for further in vitro transcription amplification. The obtained RNA was purified for further reverse transcription and PCR barcoding. Barcoded libraries were sequenced on an Illumina HiSeq 2500 platform (Novogene).

LACE-seq data analysis

The analysis pipeline of the LACE-seq data was the same as previously described [22], with slight adjustments. Briefly, Cutadapt (v.1.18) [55] was used to remove the adapter sequences. The adapters, poly(A) tail in the 3′-untranslated repeat (3′ UTR) and the two bases inserted between adapter and the reads were removed with the following parameters: -f fastq -q 30,0 -a ATCTCGTATGCCGTCTTCTGCTT -m 18 –max-n 0.25 –trim-n; -f fastq -a A{15} -m 18 -n 2; and -f fastq -u 5. The trimmed reads were mapped to the mouse pre-rRNA genome using Bowtie2 (v.2.3.5.1) [56]. The remaining reads were mapped to the mm10 genome using Bowtie2 with basic parameters. The output SAM files were transferred to BAM files using SAMtools (v.1.0.0) [57], and duplicating reads were removed using Sambamba (v.0.7.1) [58]. The remaining clean reads were used for visualization, peak calling, and calculation of the counts of certain genes.

For visualization, deeptools (v.3.5.2) software [59] was used to transfer BAM files into BigWig format with the following parameters: bamCoverage -bs 10, and IGV software(v.2.14.0) was used to visualize the reads. For peak calling and annotation, reads that overlapped with the repeat genome (https://www.repeatmasker.org/species/mm.html) were removed using Bedtools (v.2.26.0) [60] because approximately 37% of the reads could be annotated into the repeat sequence (data not shown). Then, macs2 (v.2.2.8) software [61] was used with the following parameters: callpeak -f BED –keep-dup all –fe-cutoff 5 -p 0.001 –extsize 151 –nomodel-g mm. ChIPseeker (v.1.5.1) [62] was used to annotate the peaks.

To calculate the count, BAM files were used to quantify the reads using FeatureCounts (v2.0.2) [53]. The ratio between the counts of WT and negative control (NC) was considered as the binding strength of ZAR1.

Western blot

Gel loading buffer containing 2% sodium dodecyl sulfate (SDS) and β-mercaptoethanol was used to lyse oocytes or immunoprecipitation (IP) samples. Fifty GV stage oocytes were collected as described for antibody tests. Proteins were separated by SDS–polyacrylamide gel electrophoresis (PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, USA). After transfer, non-specific proteins on the membrane were blocked with 5% non-fat milk in Tris-buffered saline with Tween 20 (TBST) at RT for 30 min. The membranes were then incubated with the primary antibodies at 4℃ overnight. After washing with TBST, horseradish peroxidase (HRP)-linked secondary antibodies (Jackson ImmunoResearch Laboratories, USA) were added. Finally, the protein bands in the membranes were detected by lab-made enhanced chemiluminescence in a dark room. The primary antibodies used are listed in Additional file 10: Table S8.

Cell culture and plasmid transfection

Human embryonic kidney 293 T (HEK293 T, RRID: CVCL_0063) cell line was obtained from atcc, and mouse ovary surface epithelium (MOSE) was originally established by Dr. Qinghua Shi (Molecular&Cell Genetics Laboratory, University of Science and Technology of China) and kindly provided to our laboratory. The MOSE cell line is not registered in Cellosaurus and therefore does not have an RRID. Both cell lines were not authenticated and were not tested for mycoplasma contamination.

The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen, USA) containing 10% fetal bovine serum (FBS; Hyclone, USA) and 1% penicillin–streptomycin solution (Gibco, USA) at 37℃ with 5% CO₂. For transient plasmid transfection, Lipofectamine 2000 (Invitrogen) was used as a carrier. Briefly, the plasmid and Lipofectamine 2000 were mixed thoroughly by vortexing in Opti-Medium Essential Medium (Opti-MEM, Gibco). After incubation at RT for 30 min, the mixture was added to HEK293 T cells with the medium transferred into FBS-free DMEM. After 6 h of transfection, the medium was replaced with FBS-containing DMEM, and cells containing plasmids were cultured for 48 h for further use.

Immunoprecipitation-mass spectrometry (IP-MS)

pDEST-Flag-Zar1 and pDEST-Flag were transfected into HEK293 T cells for 48 h. The cells were then lysed with lysis buffer (50 mM Tris–HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, and 1% Triton) with protease inhibitor added. Twenty ovaries from 10 mice without PMSG injection were harvested with lysis buffer. After centrifugation, half of the ovarian lysate was mixed with FLAG-ZAR1 expressing cell lysate, and the other half was mixed with the FLAG expressing cell lysate. Anti-FLAG affinity gels (Sigma, Cat. No. A2220) were added to the lysates for IP. After rocking at 4℃ for over 2 h, the gels were collected by low-speed centrifugation (2000 rpm, 3 min) and washed four times with lysis buffer. Proteins on the gel were resolved by SDS-PAGE. The gel was stained by Coomassie brilliant blue and destained, and the resolved proteins were used for low-input MS. The samples were analyzed using an Orbitrap mass spectrometer (Thermo Fisher Scientific; Cat. No. HFX).

Low-input MS

For sample collecting, 50 oocytes at the GV, MI, and MII stages per sample were harvested as previously described. After removing the zona pellucida using hydrochloric acid containing M2 medium (HCl:M2 = 1:150), the oocytes were washed with DPBS (containing 0.3 M sucrose) to remove foreign proteins. Clean oocytes were collected for further processing or stored at − 80℃.

Sample preparation for low-input MS was conducted following the in-solution protocol. Briefly, the oocytes were lysed in a lysis buffer containing Tris (pH 8.0), NaCl, 80 Ω H₂O, and protease inhibitor and nucleic acids were fragmented by sonication. Proteins were precipitated by acetone and denatured under 4℃ with urea added. After reducing the disulfide bond of the protein using tris (2-carboxyethyl) phosphine hydrochloride and alkylating with iodoacetamide, the protein was diluted and enzymatically digested with trypsin. Finally, after removing salt ions from the fragmented peptides, the samples were analyzed using an Orbitrap mass spectrometer (Thermo Fisher Scientific; Cat. No. HFX).

MS data were analyzed using MaxQuant (v1.6.10.43), and the Mus musculus reference genome was obtained from the Swiss-Prot data source. For the protein intensity analysis, the average number was calculated for each sample.

Co-immunoprecipitation (co-IP)

pDEST-HA and pDEST-HA-Igf2bp2 plasmids were respectively co-transfected with pDEST-Flag-Zar1 into MOSE cell line for 48 h. After washing twice with PBS, the cells were lysed by home-made cell lysis buffer (CLB) (20 mM Tris–HCl (pH 7.5), 80 mM NaCl, 5 mM MgCl₂, 0.1% Triton-X100, 10% Glycerinum, and 20 mM NaF containing) with protease inhibitor added. Then the lysates were rocked at 4℃ for 15 min and centrifuged at 16,000 g, 4℃ for 10 min. Part of (~ 20 μl) the up-flow of the lysate was collected as input, and the remaining was co-incubated with EZview™ Red Anti-HA Affinity Gel (Millipore, Cat. No. E6779) by rocking at 4℃ for 2 to 4 h. Then after washed 5 times by CLB, the proteins were eluted by Sodium dodecyl sulfate (SDS) sample buffer for WB, the uncropped images can be found in Additional file 9: Image S1 and S2.

Immunofluorescence

MII stage oocytes were harvested as described previously and cultured in M16 medium until required. The oocytes were fixed in PBS containing 4% paraformaldehyde for 30 min, permeabilized in PBS containing 0.3% Triton X-100 for 20 min, and blocked with PBS containing 1% BSA for 30 min. The blocked samples were incubated with fluorescein isothiocyanate (FITC) isomer conjugated α-tubulin and 4′, 6-diamidino-2-phenylindole (DAPI) for 30 min. Samples were imaged by confocal microscopy using an LSM710 instrument (Zeiss, Germany). Antibodies used are listed in Additional file 10: Table S8.

Data availability

Raw Total RNA-seq (GSE276332) [63], PAIso-seq2 (GSE276408) [64] and LACE-seq (GSE276330) [65] data have been deposited in the NCBI Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov). IP-MS (PXD055584) [66] and low-input MS (PXD055762) [67] data have been deposited in the Proteome Xchange Consortium (http://www.proteomexchange.org). The previously published Smart-seq2 and Smart-seq3 dataused in this work was from NCBI GEO accession number GSE135787 [68] and GSE185732 [69], respectively. The previously published total RNA-seq data used in Additional file: Fig. S3B was from GSA accession number CRA003985 [70]. The previously published Ribo-lite data used in Additional file: Fig. S5B was from NCBI GEO accession number GSE165782 [71]. The immunofluorescence images used for quantification in Fig. 7 have been deposited in Figshare [72] with a reserved DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.6084/m9.figshare.28781060. All the data analyzing pipelines have been described in Methods. No other scripts and software were used other than those mentioned in the Methods section.

References

Abe K, Inoue A, Suzuki MG, et al. Global gene silencing is caused by the dissociation of RNA polymerase II from DNA in mouse oocytes. J Reprod Dev. 2010;56(5):502–7.
Article CAS PubMed Google Scholar
Wassarman PM, Letourneau GE. RNA synthesis in fully-grown mouse oocytes. Nature. 1976;261(5555):73–4.
Article CAS PubMed Google Scholar
Moore GP, Lintern-Moore S, Peters H, et al. RNA synthesis in the mouse oocyte. J Cell Biol. 1974;60(2):416–22.
Article CAS PubMed PubMed Central Google Scholar
Jukam D, Shariati SAM, Skotheim JM. Zygotic Genome Activation in Vertebrates. Dev Cell. 2017;42(4):316–32.
Article CAS PubMed PubMed Central Google Scholar
Asami M, Lam B, Ma MK, et al. Human embryonic genome activation initiates at the one-cell stage. Cell Stem Cell. 2022;29(2):209–16.
Article CAS PubMed Central Google Scholar
Jiang Z, Fan H. Five questions toward mRNA degradation in oocytes and preimplantation embryos: when, who, to whom, how, and why? Biol Reprod. 2022;107(1):62–75.
Article Google Scholar
Sha Q, Zheng W, Wu Y, et al. Dynamics and clinical relevance of maternal mRNA clearance during the oocyte-to-embryo transition in humans. Nat Commun. 2020;11(1):4917.
Article CAS PubMed PubMed Central Google Scholar
Picelli S, Björklund ÅK, Faridani OR, et al. Smart-seq2 for sensitive full-length transcriptome profiling in single cells. Nat Methods. 2013;10(11):1096–8.
Article CAS Google Scholar
Zou Z, Zhang C, Wang Q, et al. Translatome and transcriptome co-profiling reveals a role of TPRXs in human zygotic genome activation. Science, 2022,378(6615).
Liu Y, Nie H, Zhang Y, et al. Comprehensive analysis of mRNA poly(A) tails by PAIso-seq2. Sci China Life Sci. 2023;66(1):187–90.
Article CAS Google Scholar
Liu Y, Zhao H, Shao F, et al. Remodeling of maternal mRNA through poly(A) tail orchestrates human oocyte-to-embryo transition. Nat Struct Mol Biol. 2023;30(2):200–15.
Article CAS PubMed Central Google Scholar
Liu Y, Zhang Y, Wang J, et al. Transcriptome-wide measurement of poly(A) tail length and composition at subnanogram total RNA sensitivity by PAIso-seq. Nat Protoc. 2022;17(9):1980–2007.
Article CAS Google Scholar
Rong Y, Ji S, Zhu Y, et al. ZAR1 and ZAR2 are required for oocyte meiotic maturation by regulating the maternal transcriptome and mRNA translational activation. Nucleic Acids Res. 2019;47(21):11387–402.
Article CAS Google Scholar
Wu X, Viveiros MM, Eppig JJ, et al. Zygote arrest 1 (Zar1) is a novel maternal-effect gene critical for the oocyte-to-embryo transition. Nat Genet. 2003;33(2):187–91.
Article CAS PubMed Google Scholar
Yamamoto TM, Cook JM, Kotter CV, et al. Zar1 represses translation in Xenopus oocytes and binds to the TCS in maternal mRNAs with different characteristics than Zar2. Biochimica Biophysica Acta. 1829;2013(10):1034–46.
Google Scholar
Wu Y, Fan H. Revisiting ZAR proteins: the understudied regulator of female fertility and beyond. Cell Mol Life Sci. 2022;79(2):92.
Article CAS PubMed Central Google Scholar
Heim A, Niedermeier M L, Stengel F, et al. The translation regulator Zar1l controls timing of meiosis in Xenopus oocytes. Development, 2022,149(21).
Nakajima Y, Okamoto H, Kubo T. Expression cloning of Xenopus zygote arrest 2 (Xzar2) as a novel epidermalization-promoting factor in early embryos of Xenopus laevis. Genes Cells. 2009;14(5):583–95.
Article CAS PubMed Google Scholar
Cheng S, Altmeppen G, So C, et al. Mammalian oocytes store mRNAs in a mitochondria-associated membraneless compartment. Science, 2022,378(6617).
Charlesworth A, Yamamoto TM, Cook JM, et al. Xenopus laevis zygote arrest 2 (zar2) encodes a zinc finger RNA-binding protein that binds to the translational control sequence in the maternal Wee1 mRNA and regulates translation. Dev Biol. 2012;369(2):177–90.
Article CAS PubMed PubMed Central Google Scholar
Wu X, Wang P, Brown CA, et al. Zygote arrest 1 (Zar1) is an evolutionarily conserved gene expressed in vertebrate ovaries. Biol Reprod. 2003;69(3):861–7.
Article CAS PubMed Google Scholar
Su R, Fan L, Cao C, et al. Global profiling of RNA-binding protein target sites by LACE-seq. Nat Cell Biol. 2021;23(6):664–75.
Article CAS PubMed Google Scholar
Hagemann-Jensen M, Ziegenhain C, Chen P, et al. Single-cell RNA counting at allele and isoform resolution using Smart-seq3. Nat Biotechnol. 2020;38(6):708–14.
Article CAS PubMed Google Scholar
Ozadam H, Tonn T, Han CM, et al. Single-cell quantification of ribosome occupancy in early mouse development. Nature. 2023;618(7967):1057–64.
Article CAS Google Scholar
Liao J, Gong Z. Sequencing of 3’ cDNA clones using anchored oligo(dT) primers. Biotechniques. 1997;23(3):368–70.
Article CAS PubMed Google Scholar
Wu Y, Xu X, Qi M, et al. N(6)-methyladenosine regulates maternal RNA maintenance in oocytes and timely RNA decay during mouse maternal-to-zygotic transition. Nat Cell Biol. 2022;24(6):917–27.
Article CAS PubMed Google Scholar
Xiong Z, Xu K, Lin Z, et al. Ultrasensitive Ribo-seq reveals translational landscapes during mammalian oocyte-to-embryo transition and pre-implantation development. Nat Cell Biol. 2022;24(6):968–80.
Article CAS Google Scholar
Cheng S, Schuh M. Two mechanisms repress cyclin B1 translation to maintain prophase arrest in mouse oocytes. Nat Commun. 2024;15(1):10044.
Article CAS Google Scholar
Sha Q Q, Yu J L, Guo J X, et al. CNOT6L couples the selective degradation of maternal transcripts to meiotic cell cycle progression in mouse oocyte. EMBO J, 2018,37(24).
Yu C, Ji SY, Sha QQ, et al. BTG4 is a meiotic cell cycle-coupled maternal-zygotic-transition licensing factor in oocytes. Nat Struct Mol Biol. 2016;23(5):387–94.
Article CAS PubMed Google Scholar
Mendez R, Barnard D, Richter JD. Differential mRNA translation and meiotic progression require Cdc2-mediated CPEB destruction. EMBO J. 2002;21(7):1833–44.
Article CAS PubMed Central Google Scholar
Zhang T, Li Y, Li H, et al. RNA-associated protein LSM family member 14 controls oocyte meiotic maturation through regulating mRNA pools. J Reprod Dev. 2017;63(4):383–8.
Article CAS PubMed PubMed Central Google Scholar
Liu HB, Muhammad T, Guo Y, et al. RNA-Binding Protein IGF2BP2/IMP2 is a Critical Maternal Activator in Early Zygotic Genome Activation. Adv Sci (Weinh). 2019;6(15):1900295.
Article Google Scholar
Nishio S, Emori C, Wiseman B, et al. ZP2 cleavage blocks polyspermy by modulating the architecture of the egg coat. Cell. 2024;187(6):1440–59.
Article CAS PubMed PubMed Central Google Scholar
Zhang Z, Liu R, Zhou H, et al. YBX2-dependent stabilization of oocyte mRNA through a reversible sponge-like cortical partition. Cell Res. 2023;33(8):640–3.
Article CAS PubMed Central Google Scholar
Li H, Zhao H, Yang C, et al. LSM14B is an Oocyte-Specific RNA-Binding Protein Indispensable for Maternal mRNA Metabolism and Oocyte Development in Mice. Adv Sci (Weinh). 2023;10(18):e2300043.
Article PubMed Google Scholar
Reyes JM, Ross PJ. Cytoplasmic polyadenylation in mammalian oocyte maturation. Wiley Interdiscip Rev RNA. 2016;7(1):71–89.
Article CAS PubMed Google Scholar
Zhao LW, Zhu YZ, Wu YW, et al. Nuclear poly(A) binding protein 1 (PABPN1) mediates zygotic genome activation-dependent maternal mRNA clearance during mouse early embryonic development. Nucleic Acids Res. 2022;50(1):458–72.
Article CAS PubMed Google Scholar
Dai X X, Pi S B, Zhao L W, et al. PABPN1 functions as a hub in the assembly of nuclear poly(A) domains that are essential for mouse oocyte development. Sci Adv, 2022,8(43):eabn9016.
Seli E, Lalioti MD, Flaherty SM, et al. An embryonic poly(A)-binding protein (ePAB) is expressed in mouse oocytes and early preimplantation embryos. Proc Natl Acad Sci U S A. 2005;102(2):367–72.
Article CAS PubMed Google Scholar
Fu Y, Zhuang X. m(6)A-binding YTHDF proteins promote stress granule formation. Nat Chem Biol. 2020;16(9):955–63.
Article CAS PubMed Google Scholar
Huang H, Weng H, Sun W, et al. Recognition of RNA N(6)-methyladenosine by IGF2BP proteins enhances mRNA stability and translation. Nat Cell Biol. 2018;20(3):285–95.
Article CAS PubMed Central Google Scholar
Ivanova I, Much C, Di Giacomo M, et al. The RNA m(6)A Reader YTHDF2 Is Essential for the Post-transcriptional Regulation of the Maternal Transcriptome and Oocyte Competence. Mol Cell. 2017;67(6):1059–67.
Article CAS PubMed PubMed Central Google Scholar
Sui X, Hu Y, Ren C, et al. METTL3-mediated m(6)A is required for murine oocyte maturation and maternal-to-zygotic transition. Cell Cycle. 2020;19(4):391–404.
Article CAS PubMed Central Google Scholar
Braun JE, Huntzinger E, Fauser M, et al. GW182 proteins directly recruit cytoplasmic deadenylase complexes to miRNA targets. Mol Cell. 2011;44(1):120–33.
Article CAS PubMed Google Scholar
Calder MD, Watson PH, Watson AJ. Culture medium, gas atmosphere and MAPK inhibition affect regulation of RNA-binding protein targets during mouse preimplantation development. Reproduction. 2011;142(5):689–98.
Article CAS PubMed Google Scholar
Fan XC, Steitz JA. Overexpression of HuR, a nuclear-cytoplasmic shuttling protein, increases the in vivo stability of ARE-containing mRNAs. EMBO J. 1998;17(12):3448–60.
Article CAS PubMed Central Google Scholar
Chalupnikova K, Solc P, Sulimenko V, et al. An oocyte-specific ELAVL2 isoform is a translational repressor ablated from meiotically competent antral oocytes. Cell Cycle. 2014;13(7):1187–200.
Article CAS PubMed PubMed Central Google Scholar
Wan Y, Yang S, Li T, et al. LSM14B is essential for oocyte meiotic maturation by regulating maternal mRNA storage and clearance. Nucleic Acids Res. 2023;51(21):11652–67.
Article CAS PubMed Google Scholar
Adhikari D, Diril MK, Busayavalasa K, et al. Mastl is required for timely activation of APC/C in meiosis I and Cdk1 reactivation in meiosis II. J Cell Biol. 2014;206(7):843–53.
Article CAS PubMed Google Scholar
Li H. Minimap2: pairwise alignment for nucleotide sequences. Bioinformatics. 2018;34(18):3094–100.
Article CAS PubMed PubMed Central Google Scholar
Dobin A, Davis CA, Schlesinger F, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21.
Article CAS PubMed Google Scholar
Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics Bioinformatics. 2014;30(7):923–30.
Article CAS PubMed Google Scholar
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550.
Article PubMed PubMed Central Google Scholar
Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnetjournal. 2011;17:1.
Google Scholar
Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9.
Article CAS PubMed PubMed Central Google Scholar
Li H, Handsaker B, Wysoker A, et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009;25(16):2078–9.
Article PubMed PubMed Central Google Scholar
Tarasov A, Vilella AJ, Cuppen E, et al. Sambamba: fast processing of NGS alignment formats. Bioinformatics. 2015;31(12):2032-4.
Ramírez F, Dündar F, Diehl S, et al. deepTools: a flexible platform for exploring deep-sequencing data. Nucleic Acids Res. 2014;42(Web Server issue):W187–91.
Article PubMed Central Google Scholar
Quinlan AR, Hall IM. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics. 2010;26(6):841–2.
Article CAS PubMed PubMed Central Google Scholar
Feng J, Liu T, Qin B, et al. Identifying ChIP-seq enrichment using MACS. Nat Protoc. 2012;7(9):1728–40.
Article CAS Google Scholar
Yu G, Wang L, He Q. ChIPseeker: an R/Bioconductor package for ChIP peak annotation, comparison and visualization. Bioinformatics. 2015;31(14):2382–3.
Article CAS PubMed Google Scholar
Yu-Ke Wu, Ruibao Su, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong, Shu-Yan Ji, Jingwen Liu, Zhuoyue Niu, Zhiyi Li, Yuanchao Xue, Falong Lu, Heng-Yu Fan. The Dominant Role of Polyadenylation in Regulating Maternal mRNA Dynamics in Dormant Oocytes and the Function of ZAR1 in this Process_total RNA-seq. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE276332 (2025).
Yu-Ke Wu, Ruibao Su, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong, Shu-Yan Ji, Jingwen Liu, Zhuoyue Niu, Zhiyi Li, Yuanchao Xue, Falong Lu, Heng-Yu Fan. The Dominant Role of Polyadenylation in Regulating Maternal mRNA Dynamics in Dormant Oocytes and the Function of ZAR1 in this Process_PAIso-seq2. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE276408 (2025).
Yu-Ke Wu, Ruibao Su, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong, Shu-Yan Ji, Jingwen Liu, Zhuoyue Niu, Zhiyi Li, Yuanchao Xue, Falong Lu, Heng-Yu Fan. The Dominant Role of Polyadenylation in Regulating Maternal mRNA Dynamics in Dormant Oocytes and the Function of ZAR1 in this Process_LACE-seq. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE276330 (2025).
Yu-Ke Wu, Ruibao Su, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong, Shu-Yan Ji, Jingwen Liu, Zhuoyue Niu, Zhiyi Li, Yuanchao Xue, Falong Lu, Heng-Yu Fan. The Dominant Role of Polyadenylation in Regulating Maternal mRNA Dynamics in Dormant Oocytes and the Function of ZAR1 in this Process, IP-MS. Datasets. Proteome Xchange Consortium. https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD055584 (2025).
Yu-Ke Wu, Ruibao Su, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong, Shu-Yan Ji, Jingwen Liu, Zhuoyue Niu, Zhiyi Li, Yuanchao Xue, Falong Lu, Heng-Yu Fan. The Dominant Role of Polyadenylation in Regulating Maternal mRNA Dynamics in Dormant Oocytes and the Function of ZAR1 in this Process, MS data for oocytes. Datasets. Proteome Xchange Consortium. https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD055762 (2025).
Yan Rong1, Shu-Yan Ji, Ye-Zhang Zhu, Yun-Wen Wu, Li Shen and Heng-Yu Fan. ZAR1 and ZAR2 are Required for Oocyte Meiotic Maturation by Regulating the Maternal Transcriptome and mRNA Translational Activation. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE135787 (2019).
Hakan Ozadam, Tori Tonn, Crystal M Han, Alia Segura, Ian Hoskins, Shilpa Rao, Vighnesh Ghatpande, Duc Tran, David Catoe, Marc Salit, Can Cenik. Single cell quantification of ribosome occupancy in early mouse development. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE185732 (2023).
You Wu, Xiaocui Xu, Meijie Qi, Chuan Chen, Mengying Li, Rushuang Yan, Xiaochen Kou, Yanhong Zhao, Wenqiang Liu, Yanhe Li, Xuelian Liu, Meiling Zhang, Chengqi Yi, Hongbin Liu, Junhong Xiang, Hong Wang, Bin Shen, Yawei Gao, Shaorong Gao. Distinct roles of N6-methyladenosine in maternal-zygotic-transition. Datasets. Genome Sequence Archieve. https://ngdc.cncb.ac.cn/gsa/browse/CRA003985 (2022).
Zhuqing Xiong, Kai Xu, Zili Lin, Feng Kong, Qiujun Wang, Yujun Quan, Qian-qian Sha, Fajin Li, Zhuoning Zou, Ling Liu, Shuyan Ji, Yuling Chen, Hongmei Zhang, Jianhuo Fang, Guang Yu, Bofeng Liu, Lijuan Wang, Huili Wang, Haiteng Deng, Xuerui Yang, Heng-yu Fan, Lei Li, Wei Xie. Ultrasensitive Ribo-seq reveals translational landscapes during mammalian oocyte-to-embryo transition and pre-implantation development. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE165782 (2022).
Yu-Ke Wu, Ruibao Su, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong, Shu-Yan Ji, Jingwen Liu, Zhuoyue Niu, Zhiyi Li, Yuanchao Xue, Falong Lu, Heng-Yu Fan. Immunofluorescence (IF) analysis of chromatin states in WT and Zar1/2^–/– MII stage oocytes 21 or 32 h following hCG injection. Datasets. Gene Expression Omnibus. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE276330 (2025).

Download references

Acknowledgements

We thank Dr. C.C. Cao for sharing the detailed LACE-seq analysis process. Mr. Y.W. Zhang for sharing detailed PAIso-seq2 analysis process. The first author sincerely appreciates Ms. X. Zhang, M.Sc., for her unwavering support, encouragement and warm companionship.

Peer review information

Wenjing She was the primary editor of this article and managed its editorial process and peer review in collaboration with the rest of the editorial team. The peer review history is available in the online version of this article.

Funding

This work was supported by the National Key Research and Development Program of China (2024YFA1803000), the Natural Science Foundation of Zhejiang Province (LD22 C060001), and the National Natural Science Foundation of China (31930031).

Author information

Authors and Affiliations

MOE Key Laboratory for Biosystems Homeostasis and Protection and Innovation Center for Cell Signaling Network, Life Sciences Institute, Zhejiang University, Hangzhou, 310058, China
Yu-Ke Wu, Zhi-Yan Jiang, Yun-Wen Wu, Yan Rong & Heng-Yu Fan
Key Laboratory of RNA Biology, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
Ruibao Su & Yuanchao Xue
Zhejiang Key Laboratory of Precise Protection and Promotion of Fertility, Department of Obstetrics and Gynecology, School of Medicine, Sir Run Run Shaw Hospital, Zhejiang University, Hangzhou, 310016, China
Yan Rong & Heng-Yu Fan
Institute of Medical Genetics and Development, Key Laboratory of Reproductive Genetics (Ministry of Education) and Women’s Hospital, Zhejiang University School of Medicine, Zhejiang, 310006, China
Shu-Yan Ji
State Key Laboratory of Molecular Developmental Biology, Institute of Genetics and Developmental Biology, Beijing, China
Jingwen Liu, Zhuoyue Niu & Falong Lu
University of Chinese Academy of Sciences, Beijing, 100049, China
Jingwen Liu, Zhuoyue Niu & Falong Lu
Department of Computer Science and Engineering, College of Engineering, The Ohio State University-Columbus, Columbus, Ohio, USA
Zhiyi Li
Center for Biomedical Research, Shaoxing Institute, Zhejiang University, Shaoxing, 312000, China
Heng-Yu Fan

Authors

Yu-Ke Wu
View author publications
You can also search for this author inPubMed Google Scholar
Ruibao Su
View author publications
You can also search for this author inPubMed Google Scholar
Zhi-Yan Jiang
View author publications
You can also search for this author inPubMed Google Scholar
Yun-Wen Wu
View author publications
You can also search for this author inPubMed Google Scholar
Yan Rong
View author publications
You can also search for this author inPubMed Google Scholar
Shu-Yan Ji
View author publications
You can also search for this author inPubMed Google Scholar
Jingwen Liu
View author publications
You can also search for this author inPubMed Google Scholar
Zhuoyue Niu
View author publications
You can also search for this author inPubMed Google Scholar
Zhiyi Li
View author publications
You can also search for this author inPubMed Google Scholar
Yuanchao Xue
View author publications
You can also search for this author inPubMed Google Scholar
Falong Lu
View author publications
You can also search for this author inPubMed Google Scholar
Heng-Yu Fan
View author publications
You can also search for this author inPubMed Google Scholar

Contributions

Y.K. and H.Y. designed the experiment and wrote the manuscript. Y.K. performed most of the experiments and bioinformatics analysis of all the sequencing data. R.B. performed the library construction of LACE-seq. Z.Y. J. performed the MS, IP-MS, and the raw data processing. Y.W. and Z.Y. L. helped with the bioinformatics analysis. J.W. and Z.Y. N. performed the library construction of PAIso-seq2. Y.C. and F.L. revised the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Heng-Yu Fan.

Ethics declarations

Ethics approval and consent to participate

Animal experiments were performed in strict compliance with the guidelines and regulations of Zhejiang University, and the study protocol (ZJU20220002) received approval from the University’s Institutional Animal Care and Use Committee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

13059_2025_3593_MOESM1_ESM.pdf

Additional file 1. Fig. S1 Smart-seq2 detection can be biased by oligo-d(T). Pipeline for total RNA-seq and Smart-seq2 and supplementary information for Fig. 1. Fig. S2 Quality control of PAIso-seq2 data. The clean ccs number and corrrelation of PAIso-seq2 samples. Fig. S3 Quality control of total RNA-seq data. Correlation of total RNA-seq samples, cited total RNA-seq data and comparison between efficiency between total RNA-seq and Smart-seq2. Fig. S4 Quality control of MS data. Simple pipeline for MS sample preparation, correlation and detected proteins of MS samples. Fig. S5 Cluster for PAIso-seq2, TE and MS data. Genes clusters grouped according to the poly(A), TE and protein intensity during meiosis. Fig. S6 Summary for reported dynamics and functions of ZAR1 and ZAR2 in oogenesis and maternal-to-zygotic transition. The location, dynamics and reported information of ZAR1 and ZAR2. Fig. S7 Analysis for the dynamics of O-decay genes in Zar1/2^–/– oocytes. The RNA intensity and degradation extend of O-decay genes in Zar1/2^–/– oocytes detected by total RNA-seq. Fig. S8 Quality control of LACE-seq data. Simple pipeline for LACE-seq library construction, ZAR1 antibody efficiency and data quality control of LACE-seq data. Fig. S9 DPGs of ZAR1-target genes. The polyadenylation status of ZAR1-target genes in Zar1/2^–/– oocytes. Fig. S10 Differential expression of ZAR1-target genes related to poly(A) tail length. The stability status of grouped ZAR1-target genes according the poly(A) length in Zar1/2^–/– oocytes. Fig. S11 The dynamics of ZAR1-targeted maternal genes. Clustered genes according to the stability and polyadenylation status and the dynamics of maternal genes in Zar1/2^–/– oocytes. Fig. S12 Quality control of IP-MS data. The screening standard and the quality control of IP-MS data. Fig. S13 RNA level of Lsm14b and Mastl in Zar1/2^–/– and WT oocytes. Supplementary information for Fig. 7.

13059_2025_3593_MOESM2_ESM.xlsx

Additional file 2. Table S1. Count list of Smart-seq2.xlsx. This file contains two sheets: the ‘Count list’ sheet presents the read counts of each gene detected by Smart-seq2, and the ‘ERCC%’ sheet shows the percentages of reads mapped to the ERCC reference genome

Additional file 3. Table S2. Poly(A) length of genes.xlsx. The poly(A) length of each gene detected by PAIso-seq2

13059_2025_3593_MOESM4_ESM.xlsx

Additional file 4. Table S3. Count list of total RNA-seq.xlsx. This file contains two sheets: the ‘Count list’ sheet presents the read counts of each gene detected by total RNA-seq, and the ‘ERCC%’ sheet shows the percentages of reads mapped to the ERCC reference genome

13059_2025_3593_MOESM5_ESM.xlsx

Additional file 5. Table S4. Cluster information for PAIso-seq2, TE and protein.xlsx. This file contains the gene clusters corresponding to Additional file 1: Fig. S5. It includes three sheets: the ‘pA_cluster’ sheet presents genes clustered based on polyadenylation status dynamics during meiosis as detected by PAIso-seq2; the ‘TE_all_9_cluster’ sheet shows the genes clustered based on translational efficiency dynamics during meiosis as detected by Ribo-lite; and the ‘Protein_cluster’ sheet contains genes clustered based on protein abundance dynamics during meiosis as detected by low-input MS

13059_2025_3593_MOESM6_ESM.xlsx

Additional file 6. Table S5. ZAR1-target gene list.xlsx. This file contains two sheets: the ‘ZAR1-target FC > 5’ sheet lists the screened ZAR1-target genes identified from LACE-seq data, and the ‘Peaks_to_features’ sheet shows genes with peaks located in the 3’ UTR or CDS regions, while the ‘3’ UTR_CDS_both’ column indicates genes with overlapping peaks in both regions, while the ‘3’ UTR_only’ or ‘CDS_only’ columns contains genes with peaks exclusively in the 3’ UTR or CDS, respectively

Additional file 7. Table S6. ZAR1-interact protein list.xlsx. Screened ZAR1-interact proteins by using low-input MS

13059_2025_3593_MOESM8_ESM.xlsx

Additional file 8. Table S7. Primer information for PAIso-seq2.xlsx. The concatenating primer sequences to separating concatenated reads and the barcode sequences used for separating samples from PAIso-seq2 data

13059_2025_3593_MOESM9_ESM.pdf

Additional file 9. Image 1 and 2: Uncropped western blot images corresponding to Additional file 1: Fig. S12E, showing the co-IP results for FLAG-ZAR1 and HA-IGF2BP2/HA using anti-HA affinity gels for IP

13059_2025_3593_MOESM10_ESM.xlsx

Additional file 10. Table S8. RT primer and antibody information.xlsx. This file contains two sheets: the ‘PCR primers’ sheet presents the names, target genes, applications and sequences of RT primers, and the ‘Antibody’ sheet shows the protein name, manufacture, applications and the working dilution of antibodies used in this research

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Wu, YK., Su, R., Jiang, ZY. et al. ZAR1 and ZAR2 orchestrate the dynamics of maternal mRNA polyadenylation during mouse oocyte development. Genome Biol 26, 120 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13059-025-03593-8

Download citation

Received: 19 October 2024
Accepted: 27 April 2025
Published: 08 May 2025
DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13059-025-03593-8

ZAR1 and ZAR2 orchestrate the dynamics of maternal mRNA polyadenylation during mouse oocyte development

Abstract

Background

Results

Conclusions

Background

Results

The dynamics of mRNA detected by Smart-seq2 can be biased by polyadenylation

Total RNA-seq defines convincing dynamics of maternal transcriptome

Maternal mRNAs are unstable in Zar1/2 –/– GV stage oocytes

Disruption of polyadenylation during meiosis in Zar1/2 –/– GV stage oocytes

ZAR1 stabilizes the maternal transcript by binding to the 3′ UTR region

ZAR1 indirectly regulates the polyadenylation of maternal transcript

ZAR1-targeted maternal mRNAs exhibit abnormal polyadenylation and stability upon ZAR1/2 deletion

ZAR1 interact with other proteins to regulate the stability of maternal RNA

Chromatin compaction cannot be maintained in Zar1/2 –/– MII stage oocytes

Discussion

Conclusion

Methods

Collection and culture of mouse oocytes

PAIso-seq2

PAIso-seq2 data analysis

Total RNA-seq library construction

Total RNA-seq and Smart-seq2 data analyses

Reverse transcription quantitative real-time RCR (RT-qPCR)

LACE-seq library construction

LACE-seq data analysis

Western blot

Cell culture and plasmid transfection

Immunoprecipitation-mass spectrometry (IP-MS)

Low-input MS

Co-immunoprecipitation (co-IP)

Immunofluorescence

Data availability

References

Acknowledgements

Peer review information

Funding

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Additional information

Publisher’s Note

Supplementary Information

Rights and permissions

About this article

Cite this article

Share this article

Keywords

Genome Biology

Contact us

Maternal mRNAs are unstable in Zar1/2 ^–/–GV stage oocytes

Disruption of polyadenylation during meiosis in Zar1/2 ^–/– GV stage oocytes

Chromatin compaction cannot be maintained in Zar1/2 ^–/–MII stage oocytes