Freimer JW et al. (2018), Expression of Alternative Ago2 Isoform Associat...

XB-ART-55955

Curr Biol 2018 Jan 22;282:296-302.e3. doi: 10.1016/j.cub.2017.11.067.

Show Gene links Show Anatomy links

Expression of Alternative Ago2 Isoform Associated with Loss of microRNA-Driven Translational Repression in Mouse Oocytes.

Freimer JW , Krishnakumar R , Cook MS , Blelloch R .

???displayArticle.abstract???
Mouse oocyte maturation, fertilization, and reprogramming occur in the absence of transcription, and thus, changes in mRNA levels and translation rate are regulated through post-transcriptional mechanisms [1]. Surprisingly, microRNA function, which is a major form of post-transcriptional regulation, is absent during this critical period of mammalian development [2, 3]. Here, we investigated the mechanisms underlying the global suppression of microRNA activity. In both mouse and frogs, microRNA function was active in growing oocytes but then absent during oocyte maturation. RNA sequencing (RNA-seq) of mouse oocytes uncovered that the microRNA effector protein AGO2 is predominantly expressed as an alternative isoform that encodes a truncated protein lacking all of the known essential domains. Full-length Ago2 as well as the related Argonautes (Ago1, Ago3, and Ago4) were lowly expressed in maturing mouse oocytes. Reintroduction of full-length AGO2 together with an exogenous microRNA in either mouse or frog oocytes restored translational repression of a target reporter. However, levels of endogenous transcripts remained unchanged. Consistent with a lack of microRNA activity, analysis of transcripts with alternative polyadenylation sites showed increased stability of transcripts with a longer 3' UTR during oocyte maturation. Redundant mechanisms protecting endogenous transcripts and the conserved loss of microRNA activity suggest a strong selection for suppressing microRNA function in vertebrate oocytes.

???displayArticle.pubmedLink??? 29307557
???displayArticle.pmcLink??? PMC5790201
???displayArticle.link??? Curr Biol
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: ago2 ago3 cdkn2c dicer1 myc rpl7 xrn1 zp3 zp3.2
GO keywords: oocyte maturation

???attribute.lit??? ???displayArticles.show???

	Graphical Abstract
	Figure 1. Suppression of miRNA Activity in Maturing Oocytes Is Conserved from Mouse to Xenopus (A) Schematic of constructs used in all subsequent luciferase assays. miR-15a sites were cloned in the 3′ UTR of Renilla luciferase (Luc)—either a 1× site with a perfect match or a 4× site with a bulge match to miR-15a. A control of no sites was also used. Firefly luciferase was used as an injection control. All injected transcripts were transcribed, capped, and polyadenylated in vitro. (B) Luciferase assay in HEK293T cells transfected with or without miR-15a mimic and psiCHECK2, a vector containing Firefly luciferase and Renilla luciferase with either no miR-15a sites, 1× perfect, or 4× bulge sites in the 3′ UTR of Renilla luciferase. N = 2. (C and D) Luciferase assays in Xenopus laevis (C) growing (stage III–V) and (D) GV (stage VI), using 1× perfect and 4× bulge templates, with or without miR-15a mimic. N = 2. (E) Luciferase assay as in (B) but in Xenopus MII eggs, with miR-15a mimic. N = 2. (F) Luciferase assays in oocytes at various stages of postnatal mouse development P15–23 (x axis), with miR-15a mimic. For P15–17, N = 4; P18, N = 3; and GV (P23), N = 4. All luciferase data are a ratio of Renilla signal over Firefly signal. All error bars represent SD. Asterisks (∗) represent significantly different data from the corresponding “no sites” control data, except in (F), where they represent significantly different data from GV (Student’s t test; p < 0.05). See also Figure S1.
	Figure 2. Exogenous hAGO2 and miRNA Can Rescue Translational Suppression of a Reporter (A) RNA-seq coverage of Ago2 isoforms in mouse GV oocytes and embryonic stem cells. Lower tracks represent Gencode Ago2 isoform and Stringtie reconstructed isoforms in oocytes and embryonic stem cells. Arrows denote exons present only in the oocyte isoform. (B) RNA-seq in mouse GV oocytes and mouse ESCs showing expression of the four Ago proteins. (C and D) Luciferase assays in (C) Xenopus GV oocytes or (D) mouse GV oocytes with injection of miR-15a mimic and either wild-type hAGO2 or hAGO2-PAZ10 mutant mRNA. (C) N = 10; (D) AGO2, N = 7; PAZ10, N = 2. (E) Luciferase assays in DICER knockout (KO) mouse GV oocytes, with miR-15a mimic and either a 1× perfect or 4× bulge template, with or without addition of hAGO2. N = 2. Luciferase data are a ratio of Renilla signal over Firefly signal. All error bars represent SD. Asterisks (∗) represent significantly different data from the corresponding “no sites” control data (Student’s t test; p < 0.05). See also Figure S2.
	Figure 3. Both Exogenous hAGO2 and miRNA Are Required to Rescue miRNA Activity but Have Minimal Impact on Endogenous Targets (A) qRT-PCR of combined mouse and human Ago2 mRNA levels in LSL-GFP-myc-hAGO2/+; Zp3-Cre+ and Zp3-Cre− mouse GV oocytes. Primers recognize both mouse and human AGO2. Rpl7 is used as a reference gene. N = 4. (B) Luciferase assay in LSL-GFP-myc-hAGO2/+; Zp3-Cre+ and Zp3-Cre− mouse GV oocytes with no miR-15a mimic and 4x Bulge template. N = 2. (C) Luciferase assay in wild-type mouse GV oocytes with miR-15a mimic and 4× bulge template. N = 4. (D) Luciferase assay in LSL-GFP-myc-hAGO2/+; Zp3-Cre+ and Zp3-Cre− mouse GV oocytes with miR-15a mimic and 4× bulge template. N = 3. (E and F) Illumina Bead Array RNA profiling of (E) LSL-GFP-myc-hAGO2/+; Zp3-Cre+ versus Zp3-Cre− mouse GV oocytes and (F) LSL-GFP-myc-hAGO2/LSL-GFP-myc-hAGO2; Zp3-Cre+ versus Zp3-Cre− mouse GV oocytes injected with miR-15a mimic. Significant differentially expressed transcripts are shown in red (FDR < 0.05). (E) N = 3; (F) N = 4. Data are a ratio of Renilla signal over Firefly signal. All error bars represent SD. Asterisks (∗) represent significantly different data from the corresponding control data (Student’s t test; p < 0.05). See also Figure S3.
	Figure 4. Isoforms with Longer 3′ UTRs Are More Stable in Maturing Oocytes (A) Percentage of distal alternative polyadenylation site usage in GV and MII oocytes. Significant increases in distal alternative polyadenylation usage are colored in orange; significant decreases are colored in blue. Each dot represents one gene. (B) Number of significant changes in alternative polyadenylation ratios resulting in longer 3′ UTRs or shorter 3′ UTRs during the GV to MII transition. (C) Same as (A), except in mouse ESCs and EpiCs. (D) Same as (B), except in mouse ESCs and EpiCs. (E) Boxplots showing change in 3′ UTR length for genes with a significant change in alternative polyadenylation during the GV to MII transition. Left plot shows the ratio of the longer 3′ UTR relative to the shorter 3′ UTR for genes where the distal polyadenylation site is more stable (n = 463 genes). Right plot shows the ratio of the shorter 3′ UTR relative to the longer 3′ UTR for genes where the proximal polyadenylation site is more stable (n = 129 genes). (F) RNA-seq coverage showing stabilization of distal alternative polyadenylation site in GV to MII transition at two genes (Pafah1b1 and Srpk1). Both genes are on the negative strand (transcribed from right to left). Two GV replicates, two MII replicates, Gencode annotated 3′ UTR, and miR-15a TargetScan predicted target site are shown. See also Figure S4 and Table S2.
	Figure S1 - Schematic of reporter UTRs. Related to Figure 1. Schematic of reporter UTR sequences illustrating miR-15a binding to the perfect and bulged sequences in the 3’ UTR of Renilla luciferase.
	Figure S2 – Characterization of Oocyte-specific Ago2 isoform. Related to Figure 2. A) qRT-PCR with primers spanning exons in full length or oocyte specific AGO2. Data is ∆Ct in embryonic stem cells and GV oocytes relative to primers spanning exon 13/14. Embryonic stem cell N=2, oocyte N=3. ND = Not detected. Asterisks (*) represent significantly different data from the corresponding ‘ESC’ control data (Student’s Ttest, p<0.05). Luciferase assays in (B) Xenopus GV oocytes or (C) mouse GV oocytes following injection of miR15a mimic, reporter lacking target sites, plus and minus wild-type AGO2 mRNA. N=10 for B, N=2 for C. D) RTqPCR for Renilla 4x Bulge and Firefly in either control (WT or LSL-GFP-myc-hAgo2/+; Zp3-Cre-) or AGO2 overexpressing (injected with AGO2 mRNA or LSL-GFP-myc-hAGO2/+; Zp3-Cre+) mouse GV oocytes along with miR-15a mimic. Dots represent biological replicates. N=4. All error bars represent standard deviation. N.S. = not significant.
	Figure S3 – Correlations between arrays. Related to Figure 3. A) Log2 quantile normalized intensity values between replicates of LSL-GFP-myc-hAGO2/LSL-GFP-myc-hAGO2; Zp3-Cre+ and Zp3-Cre- mouse GV oocytes injected with miR-15a mimic.
	Figure S4 – Evaluation of additional potential mechanisms that impact miRNA activity. Related to Figure 4. A) Luciferase assay in HEK-293T cells transfected with psiCHECK2, a vector containing Firefly luciferase and Renilla luciferase with either no let-7 sites, mutated sites, or 4X Bulge sites in the 3’ UTR for Renilla luciferase. psiCHECK2 was co-transfected with either GFP, full length hAGO2, or truncated oocyte AGO2. Luciferase data is a ratio of Renilla signal over Firefly signal. Full and truncated AGO2 is not significantly different from GFP control in all cases. All error bars represent standard deviation. N.S. = not significant. B) RNA-Seq expression of Xrn1 in mouse embryonic stem cells and mouse GV oocytes.

References [+] :

ABE, Global gene silencing is caused by the dissociation of RNA polymerase II from DNA in mouse oocytes. 2010, Pubmed