The RNA-binding protein, ZFP36L2, influences ovulation and oocyte maturation

Abstract

ZFP36L2 protein destabilizes AU-rich element-containing transcripts and has been implicated in female fertility. In the C57BL/6NTac mouse, a mutation in Zfp36l2 that results in the decreased expression of a form of ZFP36L2 in which the 29 N-terminal amino acid residues have been deleted, ΔN-ZFP36L2, leads to fertilized eggs that arrest at the two-cell stage. Interestingly, homozygous ΔN-Zfp36l2 females in the C57BL/6NTac strain release 40% fewer eggs than the WT littermates (Ramos et al., 2004), suggesting an additional defect in ovulation and/or oocyte maturation. Curiously, the same ΔN-Zfp36l2 mutation into the SV129 strain resulted in anovulation, prompting us to investigate a potential problem in ovulation and oocyte maturation. Remarkably, only 20% of ΔN-Zfp36l2 oocytes in the 129S6/SvEvTac strain matured ex vivo, suggesting a defect on the oocyte meiotic maturation process. Treatment of ΔN-Zfp36l2 oocytes with a PKA inhibitor partially rescued the meiotic arrested oocytes. Furthermore, cAMP levels were increased in ΔN-Zfp36l2 oocytes, linking the cAMP/PKA pathway and ΔN-Zfp36l2 with meiotic arrest. Since ovulation and oocyte maturation are both triggered by LHR signaling, the downstream pathway was investigated. Adenylyl cyclase activity was increased in ΔN-Zfp36l2 ovaries only upon LH stimulation. Moreover, we discovered that ZFP36L2 interacts with the 3'UTR of LHR mRNA and that decreased expression levels of Zfp36l2 correlates with higher levels of LHR mRNA in synchronized ovaries. Furthermore, overexpression of ZFP36L2 decreases the endogenous expression of LHR mRNA in a cell line. Therefore, we propose that lack of the physiological down regulation of LHR mRNA levels by ZFP36L2 in the ovaries is associated with anovulation and oocyte meiotic arrest.

Figure 1. ΔN-Zfp36l2 oocytes remain arrested at the germinal vesicle stage.

Cumulus oophorus complexes (COCs) were allowed to mature ex vivo. After 20 hours, the oocytes were denuded by pipetting and were morphologically scored as immature (germinal vesicle, GV) or mature (germinal vesicle breakdown, GVBD) using a stereomicroscope. (A) Progression of oocyte maturation. Bars are means ± SD (n = 6). (B) Representative images of live WT and ΔN-Zfp36l2 oocytes. Left panel, intact COCs (40X); center panel, denuded oocytes (80X); right panel, confocal images (40X) of ex vivo matured DAPI stained. (C) Quantification of Zfp36l2 mRNA during oocyte maturation. Total RNA was isolated from cumulus cells, immature (GV) and mature (GVBD) oocytes from heterozygous (Htz) and ΔN-Zfp36l2 mice after ex vivo maturation; Zfp36l2 mRNA expression levels were evaluated by real-time PCR (TaqMan) and normalized to GAPDH. Bars are means ± SD of four independent biological experiments, each in duplicate. (D) Zfp36l2 mRNA levels in immature oocytes were arbitrarily set as 100%, and the mRNA values in mature oocytes were calculated as percentages of the values found in immature oocytes. ‘p’ values were obtained from Student's t-test. ***p<0.001.

Figure 2

(A) PKA inhibition partially rescues ΔN-Zfp36l2 oocyte maturation. WT and ΔN-Zfp36l2 COCs were treated with 0.5 mM of Rp-cAMPS. The oocytes were denuded from the surrounding cumulus cells and classified as immature (GV) or mature (GVBD). Bars are means ± SEM (n = 5). (B) ΔN-Zfp36l2 oocytes exhibit higher levels of basal cAMP. Bars are means ± SEM (n = 5 groups of 20 COC per condition). (C) Adenylyl cyclase (AC) activity in ovarian homogenates of WT and ΔN-Zfp36l2 females. AC activity was determined in the absence (basal) or presence of MnCl₂ (2 mM in place of MgCl₂), LH (bLH 10 mg/mL), forskolin (100 mM), and NaF (10 mM). Values are mean ± SEM (n = 6, in each group). Each reagent (or no reagent, for basal condition) was incubated with the ovarian homogenate for 20 min at 37°C. AC activity is expressed in pmol cAMP formed/min/mg protein. ‘p’ values were calculated using Student's t-test.

Figure 3. Expression of LHR and Zfp36 family mRNAs in synchronized ovaries from WT and ΔN-Zfp36l2 females.

Total RNA was extracted from ovaries harvested on the morning on which the vaginal plug was detected. (A) LHR mRNA levels in ovaries from ΔN-Zfp36l2 mice are significantly higher when compared to WT. (B) Zfp36l2 mRNA was lower in ΔN-Zfp36l ovaries when compared to WT counterparts. (C) LHR and ΔN-Zfp36l2 mRNA levels are inversely correlated, Pearson correlation coefficient  =  -0.9108. The mRNA levels of the other ZFP36 family members, Ttp (D) and Zfp36l1 (E) were similar in WT and ΔN-Zfp36l2 ovaries. ‘p’ values were calculated using Student's t-test.

Figure 4. Interaction of ZFP36L2 protein with LHR mRNA.

RNA electrophoretic mobility shift assays were performed by incubating protein extracts from HEK 293 cells transfected with different constructs with 2×10⁵ cpm of mLHR ARE probes. Four ³²P-labeled mLHR probes were used as shown in panel (C). (A) Overexpressed ZFP36L2-GFP and ΔN-ZFP36L-GFP caused a shift in the electrophoretic mobility of the ARE2197 probe (S, lanes 3 and 4) with respect to the original migration of the probe (lane 1, P). GFP caused no shift (lane 2). Lanes 1, 5, 9: no protein extract; lanes 2, 6, 10 and 13, GFP controls, lanes 3, 4; 7, 8; 11, 12; 14, 15 contain protein extract from cells expressing GFP-tagged WT and ΔN-ZFP36L2, respectively. (B) Overexpressed ZFP36L2-HA and ΔN-ZFP36L2-HA were incubated with the mLHR probe ARE2197 in the absence (lanes 3 and 6, respectively) or presence of HA-antibody. The antibody shifted the migration of both protein-RNA complexes (supershift, SS, lane 4 and 6, respectively). The film was overexposed to clearly show the supershift. (D) Overexpressed hTTP-Flag, hZFP36L1-Flag and ZFP36L2-GFP caused a shift of the electrophoretic mobility of the hTNF-α ARE probe (lanes 3, 4 and 5, respectively) with respect to the probe's original migration (lane 1, P). GFP (lane 2) caused no shift. (E) TTP and ZFP36L1 do not seem to bind to mLHR-ARE2197 probe (lanes 3 and 4, respectively) as strongly as ZFP36L2 (lane 5). The * indicates new bands that appeared only in the presence of TTP and ZFP36L1 (lanes 3 and 4, respectively), and were not seen in the probe alone (lane 1) or in the presence of GFP (lane 2). (F) Western blots showing similar expression levels of proteins used in the gel shift assays mentioned above.

Figure 5. Overexpression of GFP-tagged ZFP36L2 decreases LHR mRNA endogenously expressed in MLTC-1 cells.

Total RNA was extracted from MLTC-1 cells not transfected or transfected with GFP-ZFP36L2 construct. Three different sets of experiments were performed and RNA quantifications were performed in triplicate. MLTC-1 cells do express Zfp36l2 mRNA and protein endogenously, and upon transfection the transcript levels of Zfp36l2 were increased by 8000 fold (A) and an intense band of GFP-ZFP36L2 was detected (B). When GFP-ZFP36L2 was overexpressed; the endogenous LHR mRNA levels were 45% lower than the basal level (C). At the protein level, a decrease of LHR expression was also detected (D). ‘p’ values were calculated using Student's t-test.

Figure 6. Expression of Zfp36l2 and LHR mRNAs in synchronized ovaries from WT animals of both strains, C57BL/6NTac and 129S6/SvEvTac.

Total RNA was extracted from ovaries harvested on the morning that the vaginal plug was detected from each individual female. (A) Zfp36l2 mRNA was similar in WT animals of both strains, C57BL/6NTac (n = 3) and 129S6/SvEvTac (n = 3). (B) Western blot assays also show comparable ZFP36L2 expression in both strains, C57BL/6NTac (n = 2) and 129S6/SvEvTac (n = 2). The (*) denotes an unspecific band. (C) LHR mRNA levels in ovaries from WT 129S6/SvEvTac were significantly lower when compared to WT C57BL/6NTac. ‘p’ values were calculated using Student's t-test.

Figure 7. Proposed model of LH negative feedback involving ZFP36L2 RNA-binding protein.

The binding of LH to LHR activates the cAMP/PKA pathway, which through an unidentified mechanism favors ZFP36L2 binding to LHR mRNA. ZFP36L2 binds specifically to one ARE, located at the 3′UTR of LHR mRNA, promoting degradation of this transcript; thus generating a negative feedback response to LH response.