BTG4 is a key regulator for maternal mRNA clearance during mouse early embryogenesis.

Dear Editor, The maternal to zygotic transition (MZT) is a crucial process in the early development of almost all animals, during which maternal mRNAs are degraded and the zygotic genome is activated (Li et al., 2013). How maternal mRNAs are degraded is one of the long-standing questions in the field of developmental and reproductive biology. Recently, high-throughput sequencing and genetic studies have determined that the elimination of maternal mRNAs is accomplished by two modes: the first mode is dependent on maternally encoded transcripts, while the second mode relies on zygotic transcription (Yartseva and Giraldez, 2015). The first mode is well characterized in Drosophila, in which the maternally encoded RNA-binding proteins SMAUG and PUMILIO are important mediators of maternal mRNA clearance (Semotok et al., 2005; Gerber et al., 2006). The second mode is exemplified by zygotically transcribed miR-430 in zebrafish that directly targets and triggers the deadenylation and subsequent clearance of maternal mRNAs (Giraldez et al., 2006). Despite these encouraging findings in model animals of lower species, the mechanisms governing the selective maternal mRNA clearance are still largely unclear in mammals. Deadenylation is the first and often ratelimiting step accounting for mRNA turnover (Garneau et al., 2007). The CCR4–NOT complex plays predominant roles in mRNA deadenylation, whereby it is involved in the SMAUGand PUMILIO-mediated maternal mRNA clearance in Drosophila (Semotok et al., 2005; Gerber et al., 2006). Previous studies have shown that the anti-proliferative Tob/BTG protein family members are important players in CCR4– NOT-mediated mRNA deadenylation and subsequent degradation (Winkler, 2010). To investigate the potential roles of Tob/ BTG proteins during mouse MZT, we first systematically analyzed the expression patterns of all Tob/BTG members (Tob1, Tob2, and Btg1–4) during this process by using RNA-seq data sets that are publicly available. Strikingly, among all the six members, Btg4, the main functional domain of which is highly conserved among vertebrates, showed an absolutely dominant and specific expression pattern during the MZT (Supplementary Figures S1 and S2A). Next, we examined the spatio-temporal expression pattern of Btg4 by quantitative real-time PCR (qRT-PCR) and western blot. Btg4 was exclusively present in ovaries and testes (Supplementary Figure S2B). In oocytes and preimplantation embryos, Btg4 mRNA was highly expressed in germinal vesicle stage (GV) oocytes and gradually perished during mouse preimplantation development, decreasing by ~90% by the 2cell stage (Figure 1A). At the protein level, however, the translation of Btg4 was largely initiated in metaphase II (MII) oocytes and was peaked at the 1-cell (1C) stage (Figure 1B, Supplementary Figure S2C), depending on the cytoplasmic polyadenylation elements and polyadenylation hexanucleotide AAUAAA sequence in the 3′ UTR (Supplementary Figure S3). To investigate the physiological role of Btg4, we depleted Btg4 using the CRISPR/Cas9 system that targets the first exon, and a targeted frame-shift mutant with a 116-bp deletion (Btg4) was successfully obtained (Supplementary Figure S4A and B). qRT-PCR and western blot showed that Btg4 was efficiently disrupted at both RNA and protein levels (Supplementary Figure S4C and D). Disruption of Btg4 had no effect on mouse viability and the fertility of male mice, but led to the infertility of female mice (Figure 1C). Histological analysis of paraffin sections from wild-type (WT) and Btg4 ovaries and superovulation experiments demonstrated that the oogenesis and fertilization of Btg4 oocytes were grossly normal (Supplementary Figure S5), but the development of embryos from Btg4 female mice was arrested at 1–2-cell stage (Figure 1D), indicating essential roles of Btg4 during the MZT of mouse early embryogenesis. We further performed transcriptome analysis on WT and Btg4 GV oocytes, MII oocytes, and 1C embryos (Supplementary Tables S1–S3). Although the transcriptome profile of Btg4 GV oocytes was comparable to that of WT, >46% (6401/13770) and 20% (2898/ 14058) mRNAs were upregulated in Btg4 MII oocytes and the resulting 1C embryos, respectively (Figure 1E). These findings were consistent with the temporal protein expression pattern and thus the functions of Btg4 during mouse preimplantation development. Moreover, all the 12 genes, generated by overlapping the top 50 significantly upregulated genes in MII oocytes and 1C embryos, could be well validated by qRT-PCR, except for one gene with a very low expression level (Figure 1F, Supplementary Figure S6). To test whether the aberrant upregulation of maternal transcripts is caused by the inefficient deadenylation of maternal mRNAs, poly(A) tail length (PAT) assay (Supplementary Figure S7A) was performed with several representative transcripts that were abundant in GV oocytes but significantly reduced in MII oocytes (Ma et al., 2015). The results showed that these transcripts in Btg4 MII oocytes and the resulting 1C embryos failed to be deadenylated and were thus resistant to degradation (Figure 1G and Supplementary Figure S7B). As the CCR4–NOT complex