Genome-wide analysis of translation reveals a critical role for deleted in azoospermia-like (Dazl) at the oocyte-to-zygote transition

Abstract

Oocyte maturation, fertilization, and early embryonic development occur in the absence of gene transcription. Therefore, it is critical to understand at a global level the post-transcriptional events that are driving these transitions. Here we used a systems approach by combining polysome mRNA profiling and bioinformatics to identify RNA-binding motifs in mRNAs that either enter or exit the polysome pool during mouse oocyte maturation. Association of mRNA with the polysomes correlates with active translation. Using this strategy, we identified highly specific patterns of mRNA recruitment to the polysomes that are synchronized with the cell cycle. A large number of the mRNAs recovered with translating ribosomes contain motifs for the RNA-binding proteins DAZL (deleted in azoospermia-like) and CPEB (cytoplasmic polyadenylation element-binding protein). Although a Dazl role in early germ cell development is well established, no function has been described during oocyte-to-embryo transition. We demonstrate that CPEB1 regulates Dazl post-transcriptionally, and that DAZL is essential for meiotic maturation and embryonic cleavage. In the absence of DAZL synthesis, the meiotic spindle fails to form due to disorganization of meiotic microtubules. Therefore, Cpeb1 and Dazl function in a progressive, self-reinforcing pathway to promote oocyte maturation and early embryonic development.

Figure 1.

Genome-wide analysis of transcripts recovered from polyribosomes during oocyte maturation. (A) Schematic representation of the stages of oocyte maturation and early embryo development. Fully grown mouse oocytes are arrested in prophase I with prominent nucleus/nucleolus, termed GV. Within 3 h after LH/hCG stimulation, oocytes re-enter the cell cycle with breakdown of the nuclear membrane (GVB [GV breakdown]) and progress to MI in 4–6 h. Oocytes extrude the first polar body (PB) at 9–10 h and arrest at MII between 10 h and 12 h after LH/hCG stimulation. (Adapted with permission from Development [from Oh et al. 2000], http://dev.biologists.org.) (B) Comparison of transcripts associated with polysomes in MI or MII with those recovered from GV oocytes. Transcripts showing less than twofold changes between (constitutively translated, class I) are in yellow, those significantly decreased more than twofold (class II, repressed; P < 0.05) are in blue, and transcripts significantly increased more than twofold (class III, activated) are in red. FDR < 5% and P < 0.05 were used in this analysis. (C) Pie chart depicting the relative distribution of transcripts in the three classes using the GV and MII data. The number of the transcripts included in each class is reported. (see the Materials and Methods for inclusion criteria). (D) Stability of class II and class III transcripts during oocyte maturation. The two major classes of regulated transcripts (class II/repressed and class III/activated) were compared with available databases of total oocyte transcripts (Su et al. 2007). The criteria for classification of stable and unstable transcripts are reported in the Materials and Methods.

Figure 2.

Motifs enriched in the 3′UTR of class III transcripts differentially regulated during oocyte maturation. (A) Sequence logo representation of motifs enriched in class III transcripts. These representative motifs were identified using an unbiased search for sequences enriched in class III transcripts. See the Materials and Methods for details. (B) Relationship between number of CPEs and transcript movement to the polysome fraction. The 3′UTRs of 4645 transcripts were scanned for canonical CPEs (UUUUAU or UUUUAAU) and then subdivided into classes according to the number of putative CPEs detected. The data were compared with a pool of 1954 transcripts without a recognizable CPE and plotted as the log₂ of MII/GV fold changes in abundance in the polysome fraction. (C) Relationship between the presence of DAZL elements and transcript recruitment to the polysomes during oocyte maturation. Transcripts were scanned for putative DAZL elements, and transcripts with no DAZL elements (2402 transcripts) or one or two or more Dazl were plotted against the log₂-fold GV/MII change in transcript levels in the polysome fraction.

Figure 3.

Dazl mRNA translation and protein accumulation during maturation. (A) qPCR analysis of Dazl mRNA in polysome and subpolysome/RNP fractions and whole-cell lysates during oocyte maturation. Each point is the mean ± SEM of three to five biologically different samples. The microarray data are included for comparison. (B) DAZL protein accumulation at the oocyte-to-zygote transition. A representative experiment of the four experiments performed is reported. Accumulation of CPEB1 and α-tubulin was used as a control. (C) Quantification of the intensity of the DAZL immunoreactive band from different experiments (mean ± SEM; N = 4).

Figure 4.

Regulation of the Dazl mRNA translation during oocyte maturation. (A) CPEB1 protein coimmunoprecipitates with Dazl transcripts. GV stage oocytes were lysed and extracts were immunoprecipitated with nonimmune IgG or CPEB1 antibodies. mRNAs recovered in the immunoprecipitation pellets were quantitated by qPCR. The data are corrected for the IgG background and reported as the mean ± SEM of three separate experiments. (B) Polyadenylation of Ccnb1 and Dazl during oocyte maturation in vivo. After stimulation with hCG, GV, MI, and MII stage oocytes were isolated and used for PAT assays. A representative experiment of the three experiments performed is reported. (C) Polyadenylation of Dazl mRNA in MI is disrupted in Cpeb1 MO-injected oocytes. After injection, oocytes were preincubated in 2 μM Milrinone overnight, then cultured in inhibitor-free medium to induce maturation. Samples were collected at indicated stages and used for PAT assays. (D) DAZL protein accumulation depends on CPEB1. GV stage oocytes and oocytes with polar bodies were collected to assess CPEB1 and DAZL protein levels, respectively. The bar graph on the right reports the densitometric quantification of the DAZL signal (mean ± SEM, N = 3). (E) Dazl mRNA autoregulation. RNA from a construct with renilla luciferase (RL) coding region fused to the 3′UTR of Dazl was coinjected in GV oocytes with either control MOs or Dazl MOs. Oocytes were collected at different times of maturation and extracts were assayed for luciferase activity (mean ± SEM; N = 4).

Figure 5.

DAZL is involved in the translation of Tex19.1. (A) Transcripts enriched in DAZL immunoprecipitates of oocyte extracts. After stimulation with hCG, MI stage oocytes were harvested as described. Oocyte lysates were immunoprecipitated with DAZL-specific antibody and selected transcripts were detected by qPCR. (B,C) Tex19.1 is recruited to the polysomes during oocyte maturation. Microarray analysis (B) and qPCR (C) of Tex19.1 mRNA in the polysome, subpolysome/RNP fractions, and total transcripts are shown. (D) Diagram of the luciferase constructs injected in oocytes. (E) Mutation of DAZL-binding elements impaired the expression of reporter during oocyte maturation. After overnight incubation, oocytes were cultured in inhibitor-free medium to allow maturation. Samples were collected at indicated culture times for assays of luciferase activity. (F) Decreased expression of Tex19.1-RL in DAZL knockdown oocytes. Luciferase activity assay was performed in extracts from oocytes coinjected with Tex19.1-RL and Dazl MOs or control MOs.

Figure 6.

Dazl is required for oocyte maturation and early embryo development. (A) Morpholino down-regulation of DAZL protein. Oocytes injected with control or Dazl MOs were preincubated overnight with 2 μM Milrinone and then cultured in inhibitor-free medium for maturation. Oocytes were collected at indicated times and used for Western analysis to detect DAZL protein levels. A representative experiment of the five experiments performed is reported. (B) Oocytes injected with Dazl MOs show a slow GVB time course. (C) Decreased polar body extrusion in Dazl-deficient oocytes. A group of oocytes was also injected with human DAZL protein. The number above the bar indicates the total number of oocytes injected in each group and the number of oocytes extruding a polar body. (D) Impaired fertilization in Dazl-deficient oocytes. Cumulus cell-enclosed oocytes were injected with control or Dazl MOs. After 15–16 h of incubation, they were used for in vitro fertilization. The fertilization rate was assessed by scoring the number of two-cell embryos after an additional overnight culture.

Figure 7.

Spindle localization and defective spindle formation in Dazl-deficient oocytes. (A) Representative patterns of DNA, tubulin, and DAZL localization in oocytes injected with control MOs, Dazl MOs, or Dazl MOs plus hDAZL protein. (B) Quantification of spindle phenotypes in oocytes injected with Dazl MOs in the presence or absence of hDAZL protein. (C) Details of spindle localization of DAZL. (D) qPCR of Tpx2 mRNA in the polysome and subpolysome/RNP fractions and total transcripts are shown. (E) TPX2 expression during meiosis is dependent on Dazl. Oocytes were injected with control MOs or Dazl MOs and incubated in meiotic arresting medium for 20 h. Oocytes were then cultured in medium that allows maturation overnight and collected for Western blot analysis using antibodies against TPX2 and α-tubulin. A representative experiment of the three experiments performed is reported. (F) Dazl is required for translation of a RL reporter fused to the 3′UTR of Tpx2. Luciferase activity was measured in extracts from oocytes coinjected with Tpx2-RL and Dazl MOs or control MOs. The data are mean ± SEM of three independent experiments.