Abstract

   Mammalian oogenesis features reliance on the mRNAs produced and stored
   during early growth phase. These are essential for producing an oocyte
   competent to undergo meiotic maturation and embryogenesis later when
   oocytes are transcriptionally silent. The fate of maternal mRNAs hence
   ensures the success of oogenesis and the quality of the resulting eggs.
   Nevertheless, how the fate of maternal mRNAs is determined remains
   largely elusive. RNA‐binding proteins (RBPs) are crucial regulators of
   oogenesis, yet the identity of the full complement of RBPs expressed in
   oocytes is unknown. Here, a global view of oocyte‐expressed RBPs is
   presented: mRNA‐interactome capture identifies 1396 RBPs in mouse
   oocytes. An analysis of one of these RBPs, LSM family member 14
   (LSM14B), demonstrates that this RBP is specific to oocytes and
   associated with many networks essential for oogenesis. Deletion of
   Lsm14b results in female‐specific infertility and a phenotype
   characterized by oocytes incompetent to complete meiosis and early
   embryogenesis. LSM14B serves as an interaction hub for proteins and
   mRNAs throughout oocyte development and regulates translation of a
   subset of its bound mRNAs. Therefore, RNP complexes tethered by LSM14B
   are found exclusively in oocytes and are essential for the control of
   maternal mRNA fate and oocyte development.

   Keywords: female infertility, LSM14B, mRNA metabolism, oocytes,
   RNA‐binding proteins (RBPs), ribonucleoprotein (RNP) complexes
     __________________________________________________________________

   Through capturing the mouse oocyte RBPome, LSM14B is identified to be
   an oocyte‐specific RBP that is indispensable for the control of oocyte
   maturation and the fate of maternal mRNAs. The mRNA‐ and
   protein‐interaction partners of LSM14B are revealed, which indicate
   that LSM14B forms dynamic and versatile RNP complexes with proteins and
   mRNAs that are crucial for mammalian oocyte development.

   graphic file with name ADVS-10-2300043-g002.jpg

1. Introduction

   A unique mode of regulation of gene expression exists in oocytes of
   most animal species that ensures the establishment of an
   oocyte‐specific gene expression program to drive oogenesis.^[ [40]^1 ^]
   In mammals, oocyte transcription occurs in the growth phase and
   gradually ceases as the oocyte approaches its final growth phase and
   acquires competence to resume meiosis. This transcriptional silent
   state lasts for the entire process of oocyte meiotic maturation, and
   new transcription event will not take place until after fertilization
   when zygotic genome activation (ZGA) initiates in the early‐stage
   embryos.^[ [41]^2 ^] Maternal mRNAs accumulated in the growing oocytes
   are vital for supporting oocyte maturation, fertilization and
   early‐stage embryo development, and regulation of maternal mRNA
   metabolism is essential for controlling gene expression in oocytes.^[
   [42]^3 ^]

   Oocyte mRNAs are mostly stable before oocyte maturation, a condition
   that enables storage of maternal messages for later use.^[ [43]^1 ^]
   However, the fate of stored maternal mRNAs changes when oocytes reach
   the fully‐grown stage and undergo meiotic maturation. Some mRNAs are
   translated to support oocyte maturation, whereas many others are
   degraded in preparation for establishing the embryo‐specific
   developmental program.^[ [44]^1 , [45]^3 , [46]^4 ^] Specific
   degradation of certain mRNAs also occurs in the growing oocytes, which
   is important for oocyte acquisition of meiotic and developmental
   competence, and for maintaining genome integrity.^[ [47]^5 ^]
   Therefore, the fate of maternal mRNAs ensures the success of oogenesis
   and the quality of the resulting eggs. Nonetheless, how the fate of
   maternal mRNAs is determined remains elusive.

   RNA‐binding proteins (RBPs) are major participants in the regulation of
   oocyte mRNA metabolism,^[ [48]^6 ^] however, the identity and function
   of the full complement of RBPs expressed in oocytes remains deficient.
   Here, after capturing and characterizing the RBPome of mouse oocytes,
   we put forward a global view of the oocyte‐expressed RBPs and
   discovered the role of LSM family member 14B (LSM14B)‐directed
   RNA‐protein interaction networks associated with the regulation of
   maternal mRNA metabolism and oogenesis.

2. Results

2.1. Capture of the Mouse Oocyte RBPome

   To gain a global view of the mRNA interacting RBPs in oocytes, we
   performed mRNA interactome capture (RIC) analysis using 50 000
   fully‐grown oocytes (FGOs) (equivalent to ≈10^8 HeLa cells in the
   amount of total proteins) collected from normal wild type prepubertal
   mice, as previously reported by Hentze Group.^[ [49]^7 ^] This approach
   combines UV cross‐linking and oligo(dT) capture to pull down mRNA‐bound
   RBPs in live oocytes for subsequent quantitative mass spectrometry (MS)
   (Figure [50]S1A,B, Supporting Information). This analysis identified
   1149 and 549 putative RBPs in the first and the second biological
   replicate, respectively, which gave rise to 1396 total putative RBPs in
   two independent experiments (Dataset [51]S1, Supporting Information).
   There were 302 proteins present in both replicates, the intensity
   (iBAQ) of which correlated well between replicates (R = 0.86, p <
   0.0001) (Figure [52]S1C,D, Supporting Information). By comparing the
   1396 putative mouse oocyte RBPs identified here with publicly available
   human and mouse RBP datasets extracted from the EMBL RBPbase
   ([53]https://rbpbase.shiny.embl.de/), we defined 632 overlapping
   proteins as “confirmed RBPs,” while the remaining 764 proteins were
   designated as “unknown RBPs” (Figure [54] 1A; Dataset [55]S1,
   Supporting Information). Gene enrichment analysis suggested that the
   former set was mainly involved in processes related to “mRNA metabolism
   and processing” (Figure [56]1B). In contrast, the latter set was
   associated with diverse processes unrelated to RNA biology
   (Figure [57]S1C, Supporting Information).

Figure 1.

   Figure 1
   [58]Open in a new tab

   Identification and characterization of RBPs expressed in fully‐grown
   mouse oocytes. A) Venn diagram comparing the oocyte RBPs identified
   here with the human and mouse RBP datasets extracted from the EMBL
   RBPbase. B) Bar graph illustrating the pathways associated with the
   “confirmed oocyte RBPs” (defined by the presence in the RBPbase). C)
   The RNA binding properties and RBD compositions of the “confirmed
   oocyte RBPs.” The pie chart demonstrates the number and proportion of
   each type of RBP as categorized by the properties of their binding
   targets. The bar graph shows the number of RBPs with a specified RBD.
   D) Venn diagram comparing the refined list of mouse oocyte RBPs
   (defined by the MS‐peptide number ≥2) with the human and mouse RBP
   datasets. E) Heatmaps illustrating the peptide abundance and functional
   properties of the refined oocyte RBPs that are also present in the
   RBPbase.

   Further manual curation of the 632 “confirmed RBPs” revealed a subset
   of 238 (37.7%) RBPs known to bind RNA, while another subset of 394
   (62.3%) RBPs were not yet annotated for their role in RNA binding
   (Figure [59]1C; Dataset [60]S1, Supporting Information). Among
   well‐annotated RBPs, the vast majority were ribosomal or mRNA‐binding
   proteins (Figure [61]1C; Dataset [62]S1, Supporting Information).
   Moreover, 225 of the “confirmed RBPs” contained at least one defined
   RNA‐binding domain (RBD) in their protein structure, with the most
   prevalent RBDs including RRM, DEAD, Helicase‐C, ZnF‐C2H2, and KH
   (Figure [63]1C; Dataset [64]S1, Supporting Information).

   The list of oocyte RBPs was defined further by including only proteins
   with a detected MS‐peptide number ≥2, resulting in a smaller set of 270
   proteins (188 in the first biological replicate, 255 in the second
   biological replicate, and 143 in both replicates) that comprised the
   oocyte RBPome, 228 of which were found in the EMBL RBPbase
   (Figure [65]1D; Figure [66]S1C and Dataset [67]S1, Supporting
   Information). Two‐thirds of the 228 known RBPs were predicted to
   participate in processes related to “RNA metabolism” and “cell cycle,”
   while the others were involved in “regulation of cytoskeleton
   organization,” “nucleosome assembly/transcription,” “protein
   stability/modification, and glycolysis/gluconeogenesis”
   (Figure [68]1E). Notably, this protein list included both RBPs with
   well‐established, essential functions in oogenesis (e.g., CPEB1,^[
   [69]^8 ^] DAZL,^[ [70]^9 ^] and YBX2 (MSY2)^[ [71]^10 ^]), as well as
   others with currently undefined roles in mammalian oogenesis (e.g.,
   LSM14A/B^[ [72]^11 ^] and RBPMS2^[ [73]^12 ^]). This mouse oocyte
   RBPome provides a rich resource for exploring the fate‐determination
   mechanisms for maternal mRNAs.

2.2. Defining LSM14B as an Oocyte‐Specific RBP Essential for Mouse Oogenesis

   LSM14A and LSM14B are two vertebrate paralogs of the highly conserved
   LSM14 protein family implicated in the regulation of maternal mRNA
   metabolism and oogenesis in various species.^[ [74]^11 , [75]^13 ^]
   They were, therefore, selected to further investigate the potential
   participation of newly identified oocyte RBPs in the regulation of the
   fate of maternal mRNA. Their expression in mouse oocytes was first
   verified by Western blot (WB) and immunofluorescence (IF) analyses
   (Figure [76] 2A–D). LSM14A was detectable in most of the mouse tissues
   and was generally restricted to cumulus and mural granulosa cells of
   the large antral follicles within the ovary (Figure [77]2A,C). LSM14A
   was barely detectable in FGOs although prominent expression was
   observed in oocytes of primordial follicles (Figure [78]2A,C). In
   contrast, LSM14B was exclusively expressed in the oocytes at all
   follicle stages, and not in granulosa cells or any of the other tissues
   (Figure [79]2B,D). These distinct expression patterns between the two
   LSM14 paralogs in mice suggested that LSM14B might function as an
   oocyte‐specific RBP in mammals.

Figure 2.

   Figure 2
   [80]Open in a new tab

   Oocyte‐specific expression and function of LSM14B in mice. A,B) Western
   blot (WB) analysis of the expression of LSM14A and B in various mouse
   tissues, and the cells within the large antral follicle, that is, the
   oocyte, cumulus cell (CC), and mural granulosa cell (MGC) (n = 2). In
   Panel A, 400 fully‐grown oocytes (FGOs) were loaded for comparison with
   various mouse tissues, and 50 FGOs were loaded for comparison with CC
   and MGC. In Panel B, 100 FGOs were loaded for comparison with various
   mouse tissues, and 50 FGOs were loaded for comparison with CC and MGC.
   C,D) Immunofluorescence (IF) staining of LSM14A and B in the ovaries of
   21 day old mice (n = 2). The magnified views of the boxed areas are
   shown in the right panels, with the arrow pointing to the non‐growing
   oocyte in the primordial follicles, * indicating the growing and
   fully‐grown oocytes, and # showing granulosa cells. LSM14A and B are
   stained in red, and DNA is counterstained in blue. Scale bars represent
   100 µm. E) WB and IF staining showing the complete deletion of LSM14B
   protein in Lsm14b‐KO oocytes. LSM14B is stained in red, and DNA is
   counterstained in blue (n = 3). F) Fertility test of the wild type (WT,
   n = 4)‐ and Lsm14b‐KO (KO, n = 6) female mice. G) Oocyte count after
   superovulation of the 24 day old WT‐ and Lsm14b‐KO female mice. Data
   are the mean ± s.e.m (n = 11 and 7 in the WTs and KOs, respectively).
   *p < 0.05, compared with the KOs by student t‐test. H) The rate (left
   bar graph) and representative micrograph (right panel) of pronucleus,
   2‐cell, 4‐cell, morula, and blastocyst embryos formed by ovulated WT‐
   and Lsm14b‐KO oocytes after in vitro fertilization with WT‐ normal
   sperm (n = 2). Scale bars represent 100 µm.

   The possible role of LSM14B in oogenesis was investigated using mice
   carrying a targeted knockout allele of Lsm14b (Lsm14b ^tm1a(KOMP)Mbp)
   induced by reporter‐tagged insertion. In Lsm14b
   ^tm1a(KOMP)Mbp‐homozygous females, LSM14B protein expression in oocytes
   was undetected (Figure [81]2E), verifying that Lsm14b ^tm1a(KOMP)Mbp
   (hereafter referred to as Lsm14b‐KO) is a null allele. The Lsm14b‐KO
   male mice showed normal fertility and were used for colony maintenance.
   However, Lsm14b‐KO female mice were infertile (Figure [82]2F), and
   ovulated slightly fewer eggs than the wild types (WTs) (Figure [83]2G).
   When these mutant eggs were fertilized by WT‐sperm in vitro, neither
   cleavage nor blastocyst formation occurred, instead they gradually
   degenerated in culture (Figure [84]2H). Unlike the ovulated WT‐eggs
   that arrested at metaphase II (MII), the majority (80%) of the
   Lsm14b‐KO ovulated eggs arrested at interphase, with a prominent
   pronucleus (PN)‐like internal structure (Figure [85] 3A). Thus,
   infertility in Lsm14b‐KO females was caused, at least in part, by
   failure of oocytes to progress to metaphase II. Therefore, LSM14B is
   required for oocyte meiotic progression and female fertility in mice.

Figure 3.

   Figure 3
   [86]Open in a new tab

   Compromised progression of meiosis to metaphase II in Lsm14b‐KO
   oocytes. A) The oocytes ovulated by Lsm14b‐KO mice have a visible first
   polar body (PB1) and pronucleus (PN). Data are the mean ± s.e.m (n =
   3). *p < 0.05, compared with the KOs by student t‐test. Representative
   microphotographs of the WT‐ and Lsm14b‐KO‐ ovulated oocytes are shown
   on the right, with IF staining of the oocyte spindle (green),
   chromosome (blue) and F‐actin (red) indicated at the bottom. Arrows
   indicate the PN. Scale bar = 50 µm. B) Kinetics of the emission of PB1
   in WT‐ and Lsm14b‐KO‐oocytes, and the Lsm14b‐KO‐oocytes treated with
   the MPS1 inhibitor (AZ3146) during IVM. C) Defective inactivation of
   SAC in Lsm14b‐KO oocytes. The bar graph shows the quantification of the
   percentage of oocytes with BUBR1 positively stained at the kinetochores
   (active spindle assembly checkpoint) after IVM for 3.5 or 8.5 h. Data
   are the mean ± s.e.m (n = 3). *p < 0.05 by Student's t‐test. The right
   micrographs demonstrate the IF staining of BUBR1 (green), CREST (red),
   and DNA (blue) in the oocytes after IVM for 3.5 or 8.5 h. Scale bar =
   5 µm. D) Changes of the expression of meiotic progression‐related genes
   in Lsm14b‐KO‐oocytes as detected by RNA‐Seq and q‐PCR. The levels of
   the WT groups were normalized to 1 in both the RNA‐Seq and qRT‐PCR
   analyses, which were not shown in the figure. E) Upregulation of WEE1
   and pCDK1‐Y15 (inactive CDK1) protein in the GV‐stage fully‐grown
   oocytes (FGOs) of the Lsm14b‐KO‐mice (WB gel graphs), and rescue of the
   PN‐arrest phenotype in the ovulated Lsm14b‐KO‐oocytes that have a
   visible PN by treatment with WEE kinase inhibitor, PD16685 (the bar
   graph and microphotographs). G) Downregulation of MASTL protein in the
   GV‐stage FGOs of Lsm14b‐KO‐mice (WB gel graphs), and partial rescue of
   the PN‐arrest phenotype in Lsm14b‐KO‐oocytes by microinjection of the
   GV‐stage FGOs with Mastl mRNA followed by in vitro maturation (the bar
   graph and microphotographs). In E,F), spindles, chromosomes, and
   F‐actin are stained in green, blue and red, respectively, Scale bar =
   50 µm. G) Compromised developmental competence in the
   Lsm14b‐KO‐oocytes. The ovulated Lsm14b‐KO‐oocytes that have a visible
   PN were rescued to MII by treatment with WEE kinase inhibitor, PD16685,
   for 6 h, and then fertilized in vitro with normal sperms isolated from
   WT‐mice. The rate and representative micrograph of pronucleus, 2‐cell,
   4‐cell, morula, and blastocyst embryos was shown in the left (bar
   graph) and the right panels, respectively (n = 2). Scale bars represent
   100 µm.

   To characterize the full extent of defects in meiotic progression in
   Lsm14b‐KO oocytes, oocytes were injected with mRNAs encoding green
   fluorescent protein (GFP)‐tagged tubulin and mCherry‐tagged histone. In
   addition to confirming PN formation, in vitro maturation (IVM) and
   live‐cell imaging revealed that MI‐to‐anaphase I (AI) transition and
   the first polar body emission (PBE) were both significantly delayed
   (3–5 h) in Lsm14b‐KO oocytes (Figure [87]3B; Figure [88]S2A,B, and
   Movies [89]S1 and [90]S2, Supporting Information). In accordance with
   the delayed onset of AI, a large proportion (73.03%) of Lsm14b‐KO
   MI‐oocytes were stained positively for BUB1B (BUBR1) expression at
   their kinetochores, indicating prolonged activation of the spindle
   assembly checkpoint (SAC) (Figure [91]3C). Chemical inhibition of the
   upstream TTK (MPS1) kinases by AZ3146 treatment to pass the SAC led to
   complete rescue of the prolonged MI and delayed PBE defects in
   Lsm14b‐KO oocytes (Figure [92]3B). Despite prolonged SAC activation, no
   differences were observed between WT‐ and Lsm14b‐KO oocytes in either
   the timing of meiotic progression to MI, the MI‐spindle assembly and
   chromosome alignment, or microtubule attachment to kinetochores
   (Figure [93]S2A,B,D, and Movies [94]S1 and [95]S2, Supporting
   Information). These results led us to hypothesize that the prolonged MI
   was related to a failure to initiate anaphase, rather than defects
   leading to SAC activation.

   To test this possibility, the activity of the anaphase‐promoting
   complex (APC) was estimated by injecting WT‐ and Lsm14b‐KO oocytes with
   mRNAs encoding Venus‐tagged CCNB1 (Cyclin B1) and mCherry‐tagged PTTG1
   (Securin) and monitored dynamic changes in these proteins in oocytes
   during IVM via live‐cell imaging (Figure [96]S3, Supporting
   Information). While CCNB1‐Venus abruptly declined in WT‐oocytes at
   ≈13 h post IVM, immediately prior to AI onset, the decrease in
   CCNB1‐Venus was considerably less and significantly delayed in
   Lsm14b‐KO oocytes (Figure [97]S3, Supporting Information). Moreover, in
   the absence of Lsm14b, CCNB1‐Venus levels were not partially restored
   after MI‐to‐AI transition as observed in WT oocytes (Figure [98]S3,
   Supporting Information), which could consequently lead to low levels of
   MPF and meiotic entry into interphase. We also observed similar defects
   in PTTG1‐mCherry dynamics in the Lsm14b‐KO oocytes (Figure [99]S3,
   Supporting Information). Collectively, these data indicated that the
   prolonged MI and failed meiotic progression to MII phenotypes of
   Lsm14b‐KO oocyte were most likely caused by delayed activation of the
   APC at the end of MI and an inability to reactivate
   maturation‐promoting factor (MPF) after MI‐to‐AI transition.

2.3. Mis‐Expression of Genes Crucial for Meiotic Progression in Lsm14b‐KO
Oocytes

   To gain more mechanistic insights on how Lsm14b‐KO caused oocyte
   meiotic defects, transcriptome‐ and proteome‐wide changes in gene
   expression in Lsm14b‐KO oocytes were investigated. Transcriptomic
   analysis identified 2285 differentially expressed genes (DEGs; 970 up,
   1315 down) between WT and Lsm14b‐KO immature germinal vesicle
   (GV)‐stage FGOs (Figure [100]S4A, Supporting Information). These were
   enriched in distinct categories of biological pathways
   (Figure [101]S6B,C, and Dataset [102]S2, Supporting Information).
   Upregulation of Bub1, Bub3, Ccnb2, Esco2, Fbxo5 (Emi1), Mapk3 (Erk1),
   Plk1, and Tpx2, and the downregulation of Anapc10, Cdc26 (Anapc12), and
   Cks2 in the Lsm14b‐KO oocytes was found and validated (Figure [103]3D).
   Since these genes are reportedly crucial for the control of oocyte
   meiotic progression,^[ [104]^14 ^] their mis‐expression in oocytes
   could account for the phenotypes of prolonged MI and delayed MI‐to‐AI
   transition, respectively. In addition, Wee1 and Wee2 mRNAs were also
   upregulated in Lsm14b‐KO oocytes, as was WEE1 protein
   (Figure [105]3D,E). As negative regulators of CDK1,^[ [106]^15 ^] WEE
   upregulation could maintain low levels of MPF activity following
   MI‐to‐AI transition, thereby preventing oocyte entry to MII. Indeed,
   pCDK1‐Y15 (i.e., inactive CDK1) levels were elevated in Lsm14b‐KO
   oocytes, and treatment with WEE kinase inhibitor, [107]PD166285,
   effectively induced MII in Lsm14b‐KO oocytes (Figure [108]3E).

   Proteomic analysis identified 326 differentially abundant proteins
   between Lsm14b‐KO and WT oocytes, which were notably enriched in
   “membrane trafficking and protein transport,” “cell cycle,” and
   “‘RNA/amino acid/carbon metabolism” pathways (Dataset [109]S3 and
   Figure [110]S5A–C, Supporting Information). Comparison with the above
   DEGs identified 84 (23 down, 28 up) protein/mRNA pairs common to both
   datasets (Figure [111]S5D–F, Supporting Information). However, no
   change in expression was detected for mRNAs corresponding to a large
   proportion (>84%) of the differential proteins, indicating the
   involvement of posttranscriptional regulation. Among these discrepant
   mRNA/protein pairs, proteomic and WB analysis showed that MASTL protein
   was almost undetectable in Lsm14b‐KO oocytes but abundant in WT
   (Figure [112]3F ; Dataset [113]S3, Supporting Information). Since Mastl
   deletion reportedly induces the same phenotype as Lsm14b‐KO in
   oocytes,^[ [114]^16 ^] whether downregulation of MASTL protein
   contributed to the meiotic defects in Lsm14b‐KO oocytes was
   investigated. Microinjection with Mastl‐mRNA to increase MASTL protein
   levels in Lsm14b‐KO oocytes led to 50% higher numbers of MII‐oocytes
   (Figure [115]3F), indicating that downregulation of MASTL could
   contribute, at least in part, to meiotic failure in Lsm14b‐KO oocytes.

   Interestingly, the Lsm14b‐KO MII‐stage oocytes rescued by the WEE
   kinase inhibitor, [116]PD166285, were able to be fertilized in vitro by
   the WT‐sperm, and formed pronuclei (Figure [117]3G). However, these
   pronucleus‐stage embryos derived from the Lsm14b‐KO oocytes failed to
   cleave and develop further (Figure [118]3G), thus implying that LSM14B
   is also crucial for oocyte acquisition of full developmental
   competence.

2.4. Identification of LSM14B‐Bound mRNAs in Oocytes by LACE‐Seq

   The linear amplification of complementary DNA ends and sequencing
   (LACE‐Seq) method developed by us recently was used to identify
   LSM14B‐bound mRNAs in mouse oocytes at single‐nucleotide resolution.^[
   [119]^17 ^] In total, 78 741 binding peaks were detected genome‐wide,
   mapping to 9253 mRNAs that were enriched in a variety of processes and
   pathways (Dataset [120]S4 and Figure [121]S6A, Supporting Information).
   Fifteen percent of the binding sites were within the 3′UTR of
   LSM14B‐bound mRNAs, with CCUCUC serving as the most enriched hexamer
   (Figure [122] 4A,B). Manual curation of these LSM14B‐bound mRNAs
   further identified a subset of transcripts typified by those encoding
   “star factors,” proteins well‐established to play essential roles in
   oogenesis (Figure [123]4C; Table [124]S2, Supporting Information),
   including 1) RBPs involved in RNA metabolism and translation crucial
   for oocyte nuclear and cytoplasmic maturation, for example, PATL2,^[
   [125]^18 ^] YBX2,^[ [126]^10 ^] and ZAR1;^[ [127]^19 ^] 2) cell
   cycle‐regulatory factors essential for oocyte meiotic progression, for
   example, CKS2,^[ [128]^14g ^] MOS,^[ [129]^20 ^] and WEE2;^[ [130]^21
   ^] 3) key oocyte‐derived paracrine factors (ODPF) and oocyte and
   follicle developmental regulatory factors, for example, GDF9,^[
   [131]^22 ^] GPR3,^[ [132]^23 ^] and PDE3A;^[ [133]^24 ^] 4) Epigenetic
   regulators essential for maintenance of oocyte genome integrity, and
   oocyte‐to‐embryo transition, for example, DPPA3 (Stella),^[ [134]^25 ^]
   NPM2,^[ [135]^26 ^] and TRIM28;^[ [136]^27 ^] 5) components of the
   oocyte‐specific subcortical maternal complex (SCMC) also essential for
   oocyte‐to‐embryo transition, for example, NLRP5 (MATER),^[ [137]^28 ^]
   PADI6,^[ [138]^29 ^] and TLE6;^[ [139]^30 ^] and 6) transcriptional
   factors essential for oocyte development, for example, BRWD1,^[
   [140]^31 ^] FOXO3,^[ [141]^32 ^] and NOBOX.^[ [142]^33 ^] The binding
   of LSM14B to these “star factor” transcripts in oocytes suggests that
   it performs key functions in coordinating RNA‐protein interaction
   networks necessary for normal oocyte and follicle development.

Figure 4.

   Figure 4
   [143]Open in a new tab

   Characterization of LSM14B‐bound RNAs identified in oocytes by CLIP and
   LACE‐Seq. A) Pie chart depicting the genomic distribution of LSM14B
   binding sites. B) Sequence logo representation of the LSM14B‐binding
   consensus calculated from the top‐20 enriched hexamers. C) Network
   illustration of the specific subset of LSM14B‐bound mRNAs typified by
   those encoding known “star factors” of oogenesis. D) qRT‐PCR comparison
   of Mastl mRNA levels in WT‐ and Lsm14b‐KO FGOs. Data are the
   mean ± s.e.m (n = 3). No significant difference between the WTs and KOs
   by Student's t‐test. E) UTR‐reporter assay to evaluate the effect of
   LSM14B binding to 3′UTR on Mastl mRNA translation. “Complete” indicates
   that the 3′UTR sequence of Mastl mRNA is intact, while the “Deleted”
   refers to the binding site for LSM14B is deleted from the 3′UTR
   sequence of Mastl mRNA. Scale bar = 100 µm.

   Additionally, LACE‐Seq revealed that Mastl mRNAs were bound to LSM14B
   through three binding sites, two in the Mastl coding region and one in
   the 3′UTR (Dataset [144]S4, Supporting Information). Since the data
   presented above show that MASTL protein was significantly downregulated
   in Lsm14b‐KO oocytes, we speculated that LSM14B could
   post‐transcriptionally regulate Mastl expression by either binding or
   stabilizing its transcripts or by promoting their translation.
   Supporting this possibility, steady state levels Mastl mRNA did not
   significantly differ between Lsm14b‐KO and WT oocytes (Figure [145]4D),
   indicating that LSM14B did not affect Mastl mRNA stability. To verify
   whether LSM14B could promote MASTL translation, live‐cell imaging of WT
   oocytes was used after injection with complete or LSM14B‐binding
   site‐deleted Mastl‐3′UTRs tethered downstream of an mKate fluorescence
   reporter (Figure [146]S6B, Supporting Information). Co‐injection of
   complete mKate‐fused UTR with EGFP control transcripts in WT oocytes
   resulted in obvious fluorescence expression of both mKate and EGFP
   reporters, whereas injection with the binding site‐deleted Mastl‐3′UTR
   reporter produced nearly undetectable mKate signal, while EGFP was
   fully expressed (Figure [147]4E). Taken together, these data suggested
   that LSM14B can bind to Mastl mRNA and promotes its translation,
   supporting a role for LSM14B in ensuring the translation of
   meiosis‐related transcripts.

2.5. LSM14B Promotes Translation of Oogenesis‐Related mRNAs

   A genome‐wide investigation of the effects of Lsm14b‐KO on mRNA
   translation in oocytes was conducted to obtain a broad perspective of
   LSM14B contributions to regulating oocyte translation. Sequencing the
   polysome‐loaded mRNAs in the GV‐stage FGOs followed by calculating
   translational efficiency (T[E]; i.e., ratio of polysome‐loaded to total
   transcripts) for each mRNA identified 1465 transcripts with
   significantly different T[E] between Lsm14b‐KO and WT oocytes
   (Dataset [148]S5, Supporting Information). A large proportion of
   T[E]‐altered transcripts included mRNAs that were not differentially
   expressed in RNA‐seq analysis (Figure [149] 5A). Comparison with the
   LACE‐seq dataset revealed that 66.3% (984) of the T[E]‐altered mRNAs
   were also bound by LSM14B (Figure [150]5B), 74.5% (733) of which had
   lower T[E] under Lsm14b‐KO, suggesting that LSM14B promotes their
   translation. These mRNAs were mainly involved in “cell cycle,” and
   “protein processing, trafficking, and degradation” processes
   (Figure [151]5C), while the remaining 25.5% (251) had higher T[E] in KO
   oocytes and were enriched in processes related to “kinetochore
   assembly,” “TCA cycle,” and “RNA metabolism,” etc. (Figure [152]5D).
   Further examination indicated that 42 T[E]‐altered mRNAs, 39 of which
   were reduced in Lsm14b‐KO oocytes, were previously reported to be
   crucial for oogenesis (Figure [153]5B).^[ [154]^16 , [155]^24 ,
   [156]^28 , [157]^29 , [158]^33 , [159]^34 ^] These findings indicated
   that LSM14B promotes the translation of some oogenesis‐related mRNAs.

Figure 5.

   Figure 5
   [160]Open in a new tab

   Changes of translation in Lsm14b‐KO‐oocytes. A) Scatterplots
   demonstrating the simultaneous changes of the translational efficiency
   (T[E]) (Y‐axis) and the steady state levels of total mRNA
   (transcriptome) (X‐axis) in the GV‐stage Lsm14b‐KO‐oocytes. B)
   Translational changes of the LSM14B‐bound mRNAs in the GV‐stage
   Lsm14b‐KO‐oocytes. The top Venn diagram shows the overlap of the
   LSM14B‐bound mRNAs (CLIP) with those having a changed T[E], which
   represents the 984 LSM14B‐bound mRNAs with their translation regulated
   by LSM14B. The bottom heatmap shows the changes of the T[E] of the
   specific group of LSM14B‐bound mRNAs that are crucial for oogenesis.
   C,D) Bar graphs illustrating the enriched GO/KEGG terms or canonical
   pathways associated with the mRNAs that are bound by LSM14B and have a
   decreased (C) or an increased (D) translational efficiency (T[E]) in
   Lsm14b‐KO oocytes.

2.6. Interaction of LSM14B with Proteins Essential for Maternal mRNA
Metabolism and Oogenesis

   To determine whether and which proteins might interact with LSM14B to
   control the fate of maternal mRNAs during oogenesis,
   immunoprecipitation (IP) and mass spectrometry (MS) in WT oocytes was
   conducted to screen for potential protein interaction partners. To
   ensure the robustness of IP/MS analyses, ≈8000 GV‐stage FGOs were
   collected in each of two independent experiments. IP/MS identified 305
   putative LSM14B protein interaction partners, enriched in pathways
   related to “RNA metabolism,” “oocyte SCMC formation,” and “gamete
   generation” (Dataset [161]S6 and Figure [162]S7A, Supporting
   Information). More specifically, these proteins included 1) several
   ribosomal proteins and translational initiation factors/regulators
   (e.g., EIF4E1B and EIF4ENIF1); 2) dozens of RBPs, including several of
   the aforementioned “star factors” (e.g., MARF1, PATL2, YBX2, and ZAR1);
   3) components of the oocyte SCMC (e.g., NLRP5, PADI6, and TLE6); 4)
   heat shock proteins; 5) RNA or DNA methylation regulators (e.g.,
   IGF2BP2, YTHDF2, DNMT1, DPPA3, and UHRF1); and 6) other proteins
   necessary for oocyte development (e.g., CAMK2G) (Figure [163] 6A). In
   addition, interactions between LSM14B and some of these putative
   binding partners were verified by Co‐IP and WB analyses
   (Figure [164]6B,C).

Figure 6.

   Figure 6
   [165]Open in a new tab

   Characterization of LSM14B‐bound proteins identified in oocytes by
   IP‐MS. A) Network illustration of LSM14B‐bound proteins that are
   crucial for oogenesis. B) WB validation of binding of LSM14B to the 11
   representative proteins from the IP‐MS dataset. C) WB validation of the
   binding of DDX6, YBX2, MARF1, EIF4ENIF1, and UHRF1 to LSM14B,
   respectively. D) WB analysis of the downregulation of DDX6, PATL2,
   YBX2, EIF4ENIF1, and SPIN1 in Lsm14b‐KO‐oocytes. GAPDH serves as the
   internal control. E,F) Colocalization of LSM14B (in red) with DDX6 (in
   green) in the oocytes of postnatal day‐3 ovaries (E), and the
   non‐growing oocytes isolated from postnatal day‐3 ovaries (F). G) FRAP
   analysis of the phase‐separation property of the LSM14B‐mKate (in red)
   formed P‐body‐like large fluorescent puncta in live early‐stage growing
   oocytes isolated from 6 day‐old mice after microinjection with the
   mRNAs for LSM14B‐mKate. Dynamic changes of the fluorescence intensity
   and morphology of the LSM14B‐mKate puncta before and after photo
   bleaching are shown in the left bar graph and the right micrographs,
   respectively. In E–G panels, DNA is stained in blue; arrows indicate
   the LSM14B and DDX6 co‐formed P‐body‐like large puncta. In G panel,
   insets in the bottom right corner are the magnified views of the boxed
   area before and after photobleaching. Scale bars in Panels E and F
   represent 10 µm. Scale bars in panel G indicate 5 µm.

   Since DDX6 is a central component of processing bodies (P‐bodies)
   responsible for the storage or degradation of translationally repressed
   mRNAs,^[ [166]^35 ^] finding that DDX6 was interacted with LSM14B and
   nearly depleted in Lsm14b‐KO oocytes (Figure [167]6B–D) suggested that
   LSM14B might regulate mRNA metabolism with DDX6 in P‐bodies. Comparing
   LSM14B‐IP/MS dataset with the P‐body protein constituents previously
   described by Hubstenberger et al.^[ [168]^36 ^] revealed that 21
   putative LSM14B‐binding proteins were also present in the pool of
   P‐body‐associated proteins (Figure [169]S7B and Dataset [170]S7,
   Supporting Information), further supporting the likelihood that LSM14B
   was also a component of P‐bodies in oocytes.

   Prominent P‐body structures have been reported in primordial‐follicle
   stage oocytes of neonatal mice.^[ [171]^37 ^] IF co‐staining of LSM14B
   and DDX6 in the ovaries of postnatal day 3 mice to ascertain whether
   LSM14B was indeed present in the oocyte P‐body showed that LSM14 B
   co‐localized with DDX6 in large granular foci within the oocyte
   cytoplasm (Figure [172]6E). The granular pattern of LSM14B‐DDX6
   co‐localization became increasingly prominent when the oocytes were
   isolated from d3‐ovarian follicles and individually stained
   (Figure [173]6F). When mouse endogenous LSM14B and DDX6 were
   ectopically expressed in HEK293 cells, they also co‐localized and
   formed granular structure resembling P‐bodies (Figure [174]S7C,
   Supporting information ).

   P‐bodies are well‐known ribonucleoprotein (RNP) condensates formed
   through phase separation.^[ [175]^38 ^] Fluorescence recovery after
   photobleaching (FRAP) in early‐stage growing oocytes isolated from 6
   day‐old mouse ovaries and HEK293 cells expressing mKate‐tagged mouse
   LSM14B showed that the formation of LSM14B‐containing P‐body‐like
   structures also involves phase separation. Photobleaching specific
   areas of the LSM14B‐mKate aggregate abolished the fluorescent signal,
   but it rapidly returned to detectable levels within 10 s of
   photobleaching in situ, reaching close to its maximum levels by 50 s,
   with the granular structure forming at the original spot
   (Figure [176]6G; Figure [177]S7C and Movie [178]S3,S4, Supporting
   Information). This relatively fast restoration of mKate‐LSM14B
   aggregates after photo bleaching suggested that these LSM14B
   condensates likely formed through liquid‐liquid phase separation.
   Therefore, LSM14B regulates the fate of maternal mRNAs and oogenesis by
   forming dynamic and versatile complexes with proteins that participate
   in mRNA metabolism and oocyte development.

3. Discussion

   Through identification and characterization of the RBPs expressed in
   mouse oocytes, this study expanded the known mouse oocyte RBPome, and
   discovered that LSM14B is an oocyte‐specific RBP that functions in
   mouse oocytes to regulate the fate of maternal mRNAs and oogenesis. The
   mRNA‐ and protein‐ interacting partners of LSM14B were identified, and
   this study showed that LSM14B forms dynamic and versatile complexes
   with proteins and mRNAs that participate in mRNA metabolism and oocyte
   development (see Figure [179] 7 for a summary).

Figure 7.

   Figure 7
   [180]Open in a new tab

   Schematic illustration of the main observations and potential
   mechanisms of LSM14B in the regulation of maternal mRNA metabolism and
   oocyte development in the mouse. Question marks in the top panel
   indicate the other potential mRNA metabolic processes that are
   regulated by the LSM14B‐directed RNP complexes.

   The oocyte‐exclusive expression and function of LSM14B, versus the more
   somatic‐restricted expression of LSM14A found here, not only clarified
   the role of LSM14 in oogenesis but may also help to explain why LSM14
   evolves into LSM14A and LSM14B paralogs in vertebrates. Although it was
   known in oocytes of the lower animal species for >20 years,^[ [181]^39
   ^] the function of LSM14 in oogenesis was unknown until the current
   study. In conjunction with the embryonic lethality phenotype observed
   by others in the Lsm14a–KO mice,^[ [182]^40 ^] the broad expression of
   LSM14A shown here suggests that it functions in the somatic cells. It
   is therefore plausible that as species evolved into the vertebrates,
   LSM14B became oocyte‐specific and replaced the originally defined
   oocyte‐specific functions that were operative in its ancestors. At the
   same time, LSM14A assumed a more diversified expression to support a
   function in somatic cells. The divergence of such an important LSM14
   protein into two paralogs could reflect an evolutionary pressure
   exerted by the complexity of the development of the distinct somatic
   and germ cell lineages in the more advanced species to drive the
   lineage‐specific spatiotemporal regulation of mRNA metabolism.

   Essential oogenic regulators have been identified in previous studies
   and are here referred to as “star factors” of oogenesis. The binding of
   LSM14B to the proteins and/or mRNAs of most of these “star factors”
   (e.g., GPR3, MARF1, MOS, NLRP5, NOBOX, NPM2, PADI6, PDE3A, YBX2, and
   ZAR1) (see Table [183]S2, Supporting Information, for a more extensive
   listing) suggests that it may serve as a facilitator of these “star
   factors” to orchestrate mammalian oogenesis. The central role of LSM14B
   in oogenesis was also supported by its interaction with the mRNAs
   and/or proteins that are involved in several aspects of the control of
   oogenesis. The binding of LSM14B to mRNAs for transcriptional factors,
   epigenetic regulators and factors of RNA metabolism and translation
   suggests that it may participate in oocyte gene expression at various
   levels. While the binding to mRNAs for cell cycle regulators, key ODPFs
   and signaling mediators of oocyte‐granulosa cell communication
   indicates engagement of LSM14B in regulation of oocyte meiotic
   progression and follicular development. Moreover, LSM14B bound to both
   the mRNAs and proteins of most of the components of the oocyte SCMC, a
   functionally conserved multiprotein complex assembled in the oocyte
   subcortical regions that is essential for the oocyte‐to‐embryo
   transition (OET).^[ [184]^41 ^] This suggests that LSM14B is involved
   in regulating the expression and function of the SCMC components.
   Components of the SCMC are encoded by maternal effect genes, deletions,
   or mutations of them compromise OET, and cause female‐specific
   infertility or subfertility and reproductive disorders in mice and
   humans.^[ [185]^41 , [186]^42 ^] Although SCMCs are known to play
   diverse roles during OET, the detailed molecular mechanisms underlying
   the dynamic assembly and function of this maternal proteinaceous
   structure are not known. The observations put forward here therefore
   open new avenues for unraveling these mechanisms.

   LSM14B interacted with many RBPs involving virtually every process of
   RNA metabolism, suggesting its engagement in the formation of versatile
   RNP complexes to regulate maternal mRNA metabolism. Of particular
   importance is its binding to CPEB, DDX6, EDC4, EIF4E1B, EIF4ENIF1
   (4E‐T), MARF1, PABPC1L (ePAB), PATL2, ZAR1, and YBX2, and
   colocalization with DDX6 in the P‐body‐like structure in non‐growing
   primordial follicle‐stage oocytes. YBX2 is a well‐known germ cell
   specific RBP crucial for stabilization and storage of oocyte mRNAs
   before maturation,^[ [187]^10 ^] while the other proteins are involved
   in translational repression and degradation of mRNA. The association
   with these RBPs suggests that LSM14B participates in maternal mRNA
   storage, translational repression and turnover in mouse oocytes, as has
   been proposed previously for LSM14A in the Xenopus oocyte.^[ [188]^11
   ^] However, a prominent increase in the translation of the LSM14B‐bound
   mRNAs in the Lsm14b‐KO oocytes was not observed in the present study,
   rather there was a greater number of the LSM14B‐bound mRNAs with the
   translational efficiency significantly decreased. Moreover, LSM14B
   bound dozens of ribosomal proteins and translation‐initiating factors.
   These results together indicate that the predominant role of LSM14B in
   oocytes is probably not repression of translation, but rather selective
   promotion of the translation of certain class of mRNAs. A similar
   assumption was proposed previously for Xenopus oocytes.^[ [189]^43 ^]
   The association of LSM14B with the recently identified oocyte‐specific
   ribonuclease, MARF1, but not with DCP1 and 2, suggests that LSM14B may
   participate in the regulation of maternal mRNA turnover via an
   oocyte‐specific mechanism that is independent of decapping.^[
   [190]^5b–d ^] The exact role of LSM14B in the regulation of oocyte mRNA
   metabolism is likely context dependent and awaits further
   investigation.

   The only clear phenotypic effect of the absence of LSM14B was inflicted
   on oocyte meiotic maturation and early embryogenesis despite its
   expression throughout oocyte growth and development and association
   with “star factors.” The slight decrease in the number of ovulated eggs
   may reflect a perturbation of processes involved in follicular
   development. More detailed analysis of the dynamics of oocyte and
   follicular development in the mutant ovaries could demonstrate subtle
   effects on these processes in future studies.

   Cheng et al. recently reported that LSM14B, together with DDX6,
   EIF4ENIF1, ZAR1, and YBX2, locates in MARDO, a mitochondria‐associated
   ribonucleoprotein domain existing in FGOs of mouse and several other
   mammalian species and functioning to store maternal mRNAs.^[ [191]^44
   ^] MARDO assembles during oocyte growth and becomes most prominent in
   FGOs, stages when P‐body‐like structure diminishes and disappears.^[
   [192]^37 ^] Our observation that LSM14B colocalized with DDX6 in
   P‐body‐like structure in the non‐growing primordial follicle‐stage
   oocytes thus agrees with its localization in MARDOs during the later
   stages of oocyte development. These results together provide a more
   complete view of the spatiotemporal changes in the localization of the
   LSM14B‐cotaining RNP complexes throughout oocyte development. Whether
   and how depletion of Lsm14b affects the integrity and function of
   oocyte MARDO remains to be explored.

4. Conclusions

   This study shows that the essential oogenic regulators are integrated
   to control the fate of maternal mRNAs, and to enable meiotic maturation
   and probably other processes during oogenesis, as indicated by the
   diverse associations of LSM14B. Given that mutations of genes encoding
   certain RBPs cause oocyte maturation defects or premature ovarian
   insufficiency in human,^[ [193]^6 , [194]^18 , [195]^45 ^] this study
   may also have implications for the diagnosis and treatment of
   infertility resulting from the decline of oocyte quality under
   deleterious conditions such as genetic malfunctions and aging.

5. Experimental Section

Laboratory Mice

   Mice carrying a targeted knockout first, reporter‐tagged insertion with
   conditional potential allele of Lsm14b, Lsm14b^tm1a(KOMP)Mbp, were
   obtained from the Wellcome Trust Sanger Institute and were maintained
   on identical C57BL/6J genetic backgrounds. Lsm14b^tmla/tm1a (hereafter
   referred to as Lsm14b‐KO) mice were produced by mating Lsm14b^tm1a/tm1a
   males with Lsm14b^tm1a/+ females. Mice were genotyped by PCR using
   primers as shown in Table [196]S1, Supporting Information, which
   produces a 652 and 270 bp PCR product for the wild type (WT) and KO
   alleles, respectively. Female fertility test was carried out by mating
   with normal adult C57BL/6J males for at least 8 months. A total of 4
   WT, and 6 Lsm14b‐KO females were fertility tested. All mice were raised
   under the standard conditions at the animal facility of Shandong
   university. All mouse procedures and protocols applied in this study
   were approved by the Institutional Animal Care and Use Committee of
   Shandong university.

Chemicals and Reagents

   Unless otherwise specified, all chemicals and reagents were purchased
   from Sigma‐Aldrich (USA). AZ3146 (Cat No. HY‐14710) and [197]PD166285
   (Cat No. 3785) were purchased from MedicalChemExpress (Shanghai, China)
   and Bio‐Techne/TOCRIS (Minnesota, USA), respectively. Antibodies used
   in this study were shown in Table [198]S3, Supporting Information.

Oocyte Isolation and Culture

   Oocyte isolation and in vitro maturation were carried out in MEM medium
   with Earles’ salts (Thermo Fisher Scientific, USA), supplemented with
   75 µg mL^−1 penicillin G, 50 µg mL^−1 streptomycin sulfate, 0.23 mm
   pyruvate, and 3 mg mL bovine serum albumin using exactly the same
   procedure as described previously.^[ [199]^46 ^] To study the ability
   of oocytes to resume and complete the first meiosis, and the kinetics
   of oocyte meiotic progression, FGOs were matured in MEM medium for
   various hours. Oocyte culture was carried out at 37 °C and 100%
   humidity in an Esco CelCulture CCL‐170T‐8‐IVF incubator (Kringelled,
   Denmark) infused with 5% O[2], and 5% CO[2].

Oocyte In Vitro Fertilization and Embryo Culture

   Mature MII‐stage eggs were collected from oviductal ampulla of the mice
   that were initially primed with eCG for 48 h followed by priming with
   human chorionic gonadotropin (hCG, Ningbo, China) for 14 h. For In
   Vitro Fertilization (IVF), the ovulated oocytes were inseminated with
   normal sperm isolated from B6D2F1 adult males. The formation of
   pronuclear and 2‐cell stage embryos was recorded 8 and 24 h after IVF,
   respectively. The 2‐cell embryos were thereafter transferred into KSOM
   medium for further culture to the blastocyst stage. The formation of
   4‐cell, morula, and blastocyst stage embryos was monitored on day 3,
   day 4, and day 5 after IVF, respectively.

mRNA Interactome Capture

   mRNA interactome capture (RIC) was carried out following the protocol
   as described previously by Castello et al. (2013).^[ [200]^7 ^] This
   approach of RIC combines UV cross‐linking and oligo(dT) capture to pull
   down the RBPs bound to mRNA in live oocytes. The captured RBPs are
   eventually characterized by quantitative mass spectrometry (MS). To do
   so, GV‐stage oocytes were isolated from eCG primed (46 h) 22‐day‐old
   mice and collected in MEM‐alpha containing 5 µm Milrinone to prevent
   resumption of meiosis. The oocytes were washed three times with cold
   PBS‐PVA buffer, and then immediately irradiated on ice by 254 nm UV
   light at 15mJ cm^−2 for 1 min. These UV‐light treated oocytes were
   instantly collected in 1.5 mL centrifuge tubes, frozen with liquid
   nitrogen and stored at −80 °C. A total of 50 000 oocytes were collected
   for two independent experiments, with each replicate using 25 000
   oocytes. The oocytes were lysed in lysis buffer (20 mm Tris‐HCl‐pH7.5,
   500 mm LiCl, 0.5% LiDS, 1 mm EDTA, 5 mm DTT) for 10 min at 4 °C, and
   the resultant lysates were incubated by gentle rotation with 2 mL of
   oligo (dT)[25] magnetic beads (S1419S, New England Biolabs, USA) for
   1 h at 4 °C. The lysates with the beads were then placed on the 50 mL
   magnetic Separation Rack (New England Biolabs) for capturing for
   30 min, with intermittent inversion for several times. After being
   completely captured, the beads were washed six times with wash buffer,
   twice with high‐salt wash buffer (20 mm Tris‐HCl‐pH 7.5, 500 mm LiCl,
   0.1%LiDS, 1 mm EDTA, 5 mm DTT), twice with buffer containing 20 mm
   Tris‐HCl (pH 7.5), 500 mm LiCl, 1 mm EDTA, 5 mm DTT, and twice with
   low‐salt buffer (20 mm Tris‐HCl‐pH 7.5, 200 mm LiCl, 1 mm EDTA, 5 mm
   DTT). The captured proteins and RNAs were eluted with elution buffer
   containing 20 mm Tris‐HCl (pH 7.5) and 1 mm EDTA for 3 min at 55 °C. To
   release proteins, the 500 µL eluate was incubated with 1 µg of RNase A
   (ST576, Beyotime, China) in 50 µL of 10× RNase buffer containing 100 mm
   Tris‐HCl (pH 7.5), 1.5 m NaCl, 0.5%NP‐40, and 5 mm DTT for 1 h at 37
   °C. Proteins were concentrated using an Amicon Ultra‐0.5 Centrifugal
   Filter Unit with Ultracel‐3 membrane (UFC500324, Millipore, USA) and
   sent to the Proteome Core Facility at Nanjing Medical University for
   LC‐MS/MS. Partial protein sample was saved and used for silver
   staining.

LC‐MS/MS Analysis

   Liquid chromatography‐tandem MS (LC‐MS/MS) analysis was performed as
   described previously on EASY nLC 1000 (Thermo Fisher Scientific) and
   LTQ Orbitrap Velos (Thermo Fisher Scientific) according to the
   manufacturer's instruction.^[ [201]^46 ^] The LC gradient was
   formulated with MS buffer A (water with 0.1% FA) and buffer B (ACN with
   0.1% FA). The analytical column (75 µm × 15 cm) was used in the chip LC
   with 2 µm 100 Å Acclaim PepMap100 C18 column. Samples were injected and
   separated with a linear gradient of 3%–100% buffer B over 95 min at a
   flow rate of 0.3 µL min^−1. Parameters of the MS survey scan were set
   with the spray voltage of 1.9 kV, and the quality scan range from 350
   to 1800. Peptide analysis was performed in data‐dependent acquisition
   mode, with the top 20 most intense precursor ions selected for
   subsequent MS/MS. Orbitrap was used for parent ion scan, with the
   resolution set at 60 000. Daughter ion scan (i.e., MS2 sequencing) was
   performed using collision‐induced dissociation (CID) in the ion trap. A
   normalized collision energy of 35% was used. The mass spectrometric raw
   data were searched against the mouse UniProt sequence database using
   MaxQuant version 1.5.2.8 with essentially the default parameters.
   Enzyme specificity was set to trypsin and up to two missed cleavage
   sites were allowed. The identified minimal peptide length was set to
   six amino acids. Cysteine carbamidomethylation (C[2]H[3]NO,
   +57.0215 Da) was set as fixed modification, and methionine oxidation
   (+15.9949 Da) and protein N‐terminal acetylation (C[2]H[2]O, + 42.0106)
   were used as variable modifications. The oocyte RBPome data were
   deposited to the ProteomeXchange Consortium, with the assigned dataset
   accession number PXD034071.

RNA‐Seq Analysis

   RNA‐Seq analysis was carried out on three total RNA or polysome‐bound
   mRNA samples derived from WT and Lsm14b‐KO oocytes as described
   previously.^[ [202]^47 ^] Total RNA was extracted from 20 GV‐stage
   FGOs, while polysome‐bound mRNA was prepared in the polysome profiling
   experiment as described in the following section. The differentially
   expressed transcripts were calculated using default parameter of
   cuffdiff (v2.2.1), and the significantly changed transcripts were
   defined by the criteria of FDR p < 0.05. RNA‐Seq data were deposited in
   the Gene Expression Omnibus (datasets [203]GSE204773). Bioinformatics
   analysis of differentially expressed transcripts was conducted using
   Metascape ([204]http://metascape.org).

Proteomic Analysis

   Microproteomic technology for single cell or low‐input samples was used
   to study the proteomic changes in Lsm14b‐KO oocytes. Triplicate samples
   of WT and Lsm14b‐KO oocytes, with each having 20 GV‐stage FGOs, were
   collected with 1 µL PBS into 0.2 mL low protein binding PCR tubes that
   were pre‐coated with 0.2 µg µL^−1 synthetic peptide (FFWIKVFFIK
   VFFVKIFFVKIFFVKIFFVK). These microsamples were suspended with
   appropriate amount of 50 mm ammonium bicarbonate buffer containing 5 mm
   DTT and sonicated in a water bath at 37 °C for 15 min. The cell lysate
   was then alkylated with 10 mm IAM in dark for 30 min, digested with
   1 µg µL^−1 trypsin at 37 °C for 2 h, and finally subjected to LC–MS/MS.
   Simultaneously, proteins from 1000 GV‐stage oocytes (library sample)
   were extracted with lysis buffer (7 m urea, 2 m thiourea, 0.2% SDS,
   20 mm Tris, pH 8.0–8.5) supplemented with 1 mm PMSF, 2 mm EDTA and
   complete protease inhibitor cocktail for constructing the spectral
   library.

   LC–MS/MS were performed on a Fusion Lumos Tribrid mass spectrometer
   (Thermo Scientific) equipped with an UltiMate 3000 UHPLC (Thermo Fisher
   Scientific). Peptides were loaded on the analytic column (length,
   30 cm; inner diameter, 150 µm) packed in house with 1.7 µm C18
   particles, and eluted at a flow rate of 500 nL min^−1 with a gradient
   of 5%–25% buffer B (0.1% formic acid in 98% acetonitrile) over 90 min,
   25%–35% buffer B over 10 min, and 35%–80% buffer B over a subsequent
   5 min. MS data was acquired in DDA mode, with the scan parameters set
   at a resolution of 2 40 000 (MS1) and 30 000 (MS2), an automatic gain
   control (AGC) target of 4e5 (MS1) and 5e4 (MS2), a maximum injection
   time (IT) of 100 ms (MS1 and (MS2), and a scan range of 350–1500 m z^−1
   (MS1) and 1.6 m z^−1 (MS2), and a first mass of 100 for MS2. For the
   1000‐oocyte library sample, the MS1 scan parameters were set at
   resolution of 1 20 000, AGC target of 1e5, maximum IT of 50 ms, and
   scan range of 350–1500 m z^−1. The MS2 scan parameters were set as
   follows: resolution, 30 000; AGC target, 2e4; maximum IT, 50 ms; Top
   20; and isolation window, 1.6 m z^−1. Unless otherwise specified, all
   the other parameters were set as default.

   Raw data were processed by MaxQuant (version1.5.3.30) for feature
   detection, database searching, and protein quantification with the
   Swiss‐Prot mouse database. Oxidation (M), acetylation (protein N‐term),
   carbamidomethylation (C), deamidation (NQ) and Gln→pyro‐Glu were set as
   variable modifications. Protein identification was achieved by using
   the MaxQuant's integrated Andromeda engine. The identification results
   of the micro‐sample data were filtered at the spectrum level with
   PSM‐level FDR < = 1%, and then further filtered at the protein level
   with Protein‐ Level FDR < = 1% to obtain significant identification
   results. For protein quantification and difference analysis, MaxQuant
   was used to extract peak areas and calculate protein quantitation
   values. Then, according to the set comparison groups, the multiples of
   differences in the proteins in each comparison group were calculated,
   and the significance test was performed using Welch's t‐test.
   Furthermore, screening was performed based on the multiple of
   difference >1.2 and p‐value < 0.05 as the criteria for determining
   significant differences. Proteomic data have been deposited in the
   Proteome Xchange Consortium, with the assigned dataset accession number
   PXD034070.

LACE‐Seq

   GV‐stage FGOs were isolated from the ovaries of eCG primed (46 h) 22‐to
   23‐day‐old mice for LACE‐Seq. Oocytes were washed three times with cold
   PBS‐PVA buffer, and immediately irradiated on ice with UV light at
   400 mJ cm^−2 for two times. The UV light treated oocytes were stored at
   −80 °C until 2000 oocytes required for one experiment were accumulated.
   LACE‐Seq was carried out as described previously,^[ [205]^17 ^] with
   the same experiment repeated independently twice. LACE‐seq data were
   deposited in the Gene Expression Omnibus, with the assigned dataset
   accession number [206]GSE206190.

Polysome Profile Analysis

   Polysome profile analysis was carried out as described previously^[
   [207]^48 ^] with slight modifications. Briefly, 1500 WT and Lsm14b KO
   oocytes were lysed, respectively, in polysome lysis buffer (100 mm KCl,
   0.1%Triton X‐100, 50 mm HEPES, 2 mm MgCl2, 10% glycerol, 100 µg mL^−1
   cycloheximide, 1 mm DTT, 20 unit mL^−1 EDTA‐free RNase inhibitor, and
   EDTA‐free proteinase inhibitor cocktail). Lysates were loaded onto
   15%–55% (w/v) sucrose density gradients prepared with 10 mm Tris‐HCl
   (pH 7.5), 5 mm MgCl2, 100 mm NaCl, and 1 mm DTT, and centrifuged at
   38 000 rpm for 3 h at 4 °C in a Beckman SW41Ti rotor. Gradients were
   fractionated and the absorbance at 254 nm was continuously recorded
   using Gradient Fractionator (BioComp, Canada). The mRNA in the polysome
   fraction (i.e., 11–28^th fractions) were extracted using the RNeasy
   Micro kit (Qiagen, Germany) and sent for RNA‐Seq analysis (See the
   later section). The Polysome‐Seq data were deposited in the Gene
   Expression Omnibus, with the assigned dataset accession number
   [208]GSE206270.

Co‐Immunoprecipitation

   A total of ≈8000 GV‐stage FGOs were lysed using the lysis buffer from
   the Pierce Crosslink Immunoprecipitation Kit (Cat No. 26147, Thermo
   Fisher Scientific). Immunoprecipitation (IP) was then carried out on
   these oocyte lysates using the anti‐LSM14B antibody. The IP products
   were subjected to WB validation followed by Mass Spectrometry (MS)
   analysis. For MS analysis, the IP products from two independent
   experiments were resolved simultaneously on 10% SDS‐PAGE, and the lanes
   corresponding to each IP product were sliced out and sent to the
   Proteomics Core Facility of Nanjing Medical University (Nanjing, China)
   for subsequent analysis. The gel slices were cut into 1 mm^3 particles
   at the Core Facility, destained, reduced, and alkylated, followed by
   the overnight Trypsin in‐gel digestion at 37 °C. LC‐ESI‐MS/MS analysis
   was performed using a Ekspert nano LC 415 system and a TripleTOF 5600
   mass spectrometer (AB Sciex), respectively. The raw data were analyzed
   using MaxQuant (version 1.5.2.8), with the UniProt Mouse database as
   reference. Data were searched using the following parameters: trypsin
   as the enzyme; up to two missed cleavage sites were allowed; 10 ppm
   mass tolerance for MS and 0.05 Da for MS/MS fragment ions;
   propionamidation on cysteine as fixed modification; oxidation on
   methionine as variable modification. The incorporated Target Decoy PSM
   Validator in Proteome Discoverer and the Mascot expectation value was
   used to validate the search results and only the hits with FDR ≤0.01
   and MASCOT expected value ≤0.05 were accepted for discussion. Co‐IP
   data were deposited in the ProteomeXchange Consortium with the assigned
   dataset accession number PXD034185.

qRT‐PCR Analyses

   For qRT‐PCR analysis, total RNA was isolated from oocyte samples using
   the Fast Pure Cell/Tissue Total RNA Isolation Kit (RC101‐01, Vazyme,
   China), and reversed transcribed by the HiscriptIII RT Supermix for
   qPCR (R323‐01, Vazyme). AceQ Universal SYBR Master Mix (Q511‐02,
   Vazyme) was then carried out using primer pairs shown in Table [209]S1,
   Supporting Information. Relative changes in mRNA levels between WT and
   KO oocytes were analyzed by the 2^−ΔΔCt method using Rpl19 as internal
   control, while exogenous mKate mRNA was used as the normalizing control
   for comparison between the same number of FGOs and preimplantation
   embryos.

Preparation of mRNAs for Microinjection

   A live‐imaging and fluorescence reporter‐based 3′UTR assay strategy was
   adopted to assess the effect of LSM14B binding to the 3′UTR of Mastl
   mRNA on its translation.^[ [210]^49 ^] To prepare the construct for
   this assay, the sequence of fluorescent protein mKate was first
   amplified by PCR using the pCMV6‐AN‐mKate plasmid DNA as template, and
   cloned into pCMV6‐AC‐3DDK (Origene, USA) vector. The resulted
   pCMV6‐AC‐mKate‐3DDK construct was then used as the destination vector
   for cloning the Mastl‐3′UTR. Mastl‐3′UTR was amplified by PCR using the
   mouse oocyte cDNA as template, and inserted into the
   pCMV6‐AC‐mKate‐3DDK vector at the site immediately downstream of the
   stop codon of 3DDK. Deletion of the LSM14B binding site
   (TAGTGATGGAGTCTCACTGC) within Mastl‐3′UTR was achieved by using the Mut
   Express MultiS Fast Mutagenesis Kit V2 (Vazyme, China) as tool and the
   pCMV6‐AC‐mKate‐3DDK‐Mastl‐3′UTR plasmid DNA as template. EGFP sequences
   were also cloned into the pCMV6‐AC‐3DDK at the site upstream of 3DDK
   and served as the template for synthesizing control EGFP mRNA. Then,
   sequences of the complete or LSM14B‐binding site‐deleted Mastl‐3′UTRs
   were amplified by PCR using a pair of primers that flank the T7
   promoter and the 3′‐end of Mastl‐3′UTR. The reverse primer had a
   stretch of 20 Ts added at its 5′‐ end in order to make the synthesized
   mRNA having a short oligo A tail. In vitro transcription was finally
   carried out on these PCR products using the Ambion's mMESSAGE mMACHINE
   kit (Thermo Fisher Scientific), and the resulted mRNAs were purified by
   lithium chloride.

   To prepare the construct for synthesizing Mastl mRNA, Mastl CDS was
   amplified by PCR using the mouse oocyte cDNA as template, and inserted
   into the pCMV6‐AC ‐3DDK vector at the site immediately upstream of the
   3DDK. Plasmid DNA of this resulted pCMV6‐AC‐Mastl‐3DDK construct, as
   well as those for GFP‐tagged tubulin, mCherry‐tagged H2B, Venus‐tagged
   CyclinB1, and mCherry‐tagged Securin, were linearized and transcribed
   in vitro using the same Ambion's mMESSAGE mMACHINE kit as mentioned
   above. The transcribed mRNAs were then polyadenylated in vitro using
   the Poly(A) Tailing Kit (Thermo Fisher Scientific), and purified by
   lithium chloride.

Microinjection of Oocytes with mRNA

   The mRNA of purified EGFP and the complete or LSM14B‐binding
   site‐deleted Mastl‐3′UTRs were mixed to make a final concentration of
   100 and 500 ng µL^−1, respectively. The mixed mRNAs were then
   micro‐injected into WT‐GV‐stage FGOs in M2 medium containing 5 µm
   milrinone. ≈10 pL of the mRNAs were injected into the cytoplasm of 1
   FGO. The injected oocytes were first incubated for 12 h in milrinone
   medium to allow the mRNAs to express to levels optimal for fluorescence
   detection, then the expression of mKate and EGFP was recorded under a
   Nikon Ti2‐E‐C2+ confocal microscope (Nikon, Japan).

   For live‐imaging experiments, purified mRNAs for the mCherry‐tagged H2B
   and GFP‐tagged tubulin pair, and Venus‐tagged CyclinB1 and
   mCherry‐tagged Securin pair were mixed at 1:1 ratio to make a final
   concentration of ≈500 and 20 ng µL^−1, respectively for each mRNA, and
   micro‐injected into the GV‐stage FGOs. Live‐imaging was then carried
   out as described previously on an Andor spinning‐disk confocal
   microscope (Andor Technology Ltd, Belfast, Northern Ireland).^[
   [211]^46 ^] Quantification of the fluorescence intensity of the
   recorded images was done using the Image J software
   ([212]http://imagej.nih.gov/ij).

FRAP Analysis

   HEK293T cells were cultured in DMEM (Meilunbio, China) containing 10%
   fetal bovine serum (FBS) at 37 °C with 5% CO2 in a humidified
   incubator. Plasmid DNAs of pCMV6‐AN‐mKate‐Lsm14b and
   pCMV6‐AC‐Ddx6‐EGFP‐3DDK were transfected into these cells using the
   JetPRIME (Polyplus, France) transfection reagent according to the
   manufacturer's protocol. Transfected HEK293T cells were cultured for
   24 h to let the fluorescence tagged protein express to sufficient
   levels for further detection. Potential colocalization of these two
   tagged proteins was then examined under the confocal microscope. To
   express mKate‐tagged LSM14B (LSM14B‐mKate) protein in the early‐stage
   growing oocytes isolated from 6 day old mice, the mRNA for LSM14B‐mKate
   was microinjected into the oocyte cytoplasm, followed by culture of the
   injected oocytes in MEM medium supplemented with 5% serum for 12 h to
   allow the LSM14B‐mKate protein expression. For fluorescence protein
   recovery after photobleaching (FRAP) analysis, rectangular regions of
   interest (ROI) containing the P‐body‐like granular fluorescence puncta
   of LSM14B‐mKate were marked and photobleached using the excitation
   laser line at greater than 50% of the maximum power. Two and 10 photos
   were taken for the HEK293 cells and oocytes, respectively, before
   bleaching, and thereafter, images were captured every 2 s. Fluorescence
   intensities of ROIs over time were quantitated and normalized by
   subtracting the remaining signal after photobleaching and then
   normalized to the mean intensity before photobleaching.

Immunostaining and Confocal Microscopy

   These were performed using the same procedures as described
   previously.^[ [213]^47 ^] Briefly, oocytes were fixed in 4% PFA in PBS
   for 30 min at room temperature, and then permeabilized and blocked for
   1 h in PBS containing 0.1% Triton X‐100 and 10% FBS. They were
   subsequently incubated with primary antibodies (4 °C, overnight) and
   Alexa flour 594/488‐conjugated secondary antibodies (room temperature,
   1 h), respectively. DNA was counterstained with Hoechst 33342 for
   30 min. For assessment of the kinetochore–microtubule attachment in
   oocytes, oocytes that were matured in vitro for 8.5 h were first kept
   on ice for 10 min (cold treatment), and then subjected to IF staining
   of the kinetochore and microtubule using the human anti‐centromere and
   mouse anti‐α‐tubulin antibodies, respectively. IF stained oocyte
   samples were finally mounted in a small drop of antifade on glass
   slides, and subjected to confocal imaging. For IF staining of the
   ovarian sections, freshly isolated ovaries from 3‐ or 21‐day‐old female
   mice were fixed in 4% PFA at 4 °C overnight, and 5 µm thick paraffin
   sections were then prepared. Sections were incubated with the primary
   antibodies, followed by Alexa flour 594‐ or 488‐conjugated secondary
   antibodies, and the DNA was counterstained with Hoechst 33342 for
   10 min. Both the oocyte and ovarian specimens were examined and imaged
   using a laser scanning confocal microscope (Nikon Ti2‐E‐C2+, Nikon).

Western Blot Analysis

   Western Blot analysis was carried out as described previously.^[
   [214]^46 ^] Briefly, oocyte samples were lysed in 2× Laemmli sample
   buffer, and boiled at 100 °C for 5 min to be denatured. The protein
   lysates were resolved on SDS‐PAGE gels and transferred onto
   polyvinylidene difluoride (PVDF) membranes for probing the proteins
   under examination. Beta‐actin (ACTB) or GAPDH was selected to serve as
   internal control of samples.

Pathway Enrichment Analysis

   The differentially expressed genes (proteins or mRNAs) identified by
   proteomics or RNA‐Seq analyses were uploaded to Metascape
   ([215]http://metascape.org), a gene annotation & analysis resource, and
   the significantly associated terms (GO/KEGG terms or canonical
   pathways) were then calculated automatically.

Statistical Analysis

   Statistical analyses were performed using Graphpad Prism software
   (Graphpad software, Inc, USA). Data were presented as mean ± SEM of at
   least three independent experiments. For experiments with only two
   treatments, differences between them were analyzed by Student's t‐test,
   while for experiments involving more than two treatments, differences
   between groups were compared by one‐way ANOVA followed by Tukey's
   Honestly Significant Difference (HSD). p < 0.05 was defined as
   significantly different.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   H.L. and H.Z. contributed equally to this work. Y.‐Q.S., H.L., and Y.X.
   conceived the study; H.L., H.Z., R.S., C.Y., M.L., J.L., and L.S.
   performed the research; H.L., H.Z., Y.X., and Y.‐Q.S. analyzed the
   data; and Y.‐Q.S., H.L., and Y.X. wrote the paper.

Supporting information

   Supporting Information
   [216]Click here for additional data file.^ (2.1MB, pdf)

   Supplemental Movie 1
   [217]Click here for additional data file.^ (4MB, mp4)

   Supplemental Movie 2
   [218]Click here for additional data file.^ (3.9MB, mp4)

   Supplemental Movie 3
   [219]Click here for additional data file.^ (18.1MB, mp4)

   Supplemental Movie 4
   [220]Click here for additional data file.^ (3.2MB, MP4)

   Supplemental Table 1
   [221]Click here for additional data file.^ (977.9KB, xlsx)

   Supplemental Table 2
   [222]Click here for additional data file.^ (124.9KB, xlsx)

   Supplemental Table 3
   [223]Click here for additional data file.^ (28KB, xlsx)

   Supplemental Table 4
   [224]Click here for additional data file.^ (1.5MB, xlsx)

   Supplemental Table 5
   [225]Click here for additional data file.^ (463.4KB, xlsx)

   Supplemental Table 6
   [226]Click here for additional data file.^ (285.6KB, xlsx)

   Supplemental Table 7
   [227]Click here for additional data file.^ (12.2KB, xlsx)

Acknowledgements