Peroxiredoxin I maintains luteal function by regulating unfolded protein response

Unless otherwise stated, all chemicals and reagents used in this study were purchased from Sigma Aldrich (St. Louis, MO, USA).

Female C57/B6J mice (8–10-week-old) of wild type were purchased from Hyochang Bio-Science (Daegu, Korea) and Prdx1-deficient (Prdx1 −/−) female mice (8–10-weeks old) maintained in accordance with the institutional guidelines of the Institutional Animal Care and Use Committee of the Korea Research Institute of Bioscience and Biotechnology (KRIBB, Daejeon, South Korea). Animals were maintained under standard environmental conditions (temperature at 20–22 °C, humidity at 50–60%, and 12-h-dark/light cycles) with free access to food and water.

The short reproductive cycle length observed in rodents is called the estrous cycle. The normal estrous cycle in rodents follows a 4–5-day pattern, the characteristics of which vary with the day [21]. We classified mice on the basis of the predicted time of hormone induced-superovulation as described previously [18]. In all mice, superovulation was achieved by intraperitoneal (i.p) injection of 5 IU pregnant mare’s serum gonadotropin/mouse (n = 5–6 in each group, 8-week-old) (PMSG; Sigma Aldrich), followed 48 h later by injection of 5 IU human chorionic gonadotropin/mouse (hCG; Sigma Aldrich). Ovulation occurred 12 h after the hCG injection, and therefore, 16 h post-hCG injection was considered as the beginning of CL formation. CL tissues were collected from the ovaries of treated mice 16, 24, 48, 72 and 96 h after hCG administration (Fig. 1a). The luteal phase was divided into two groups, namely, the functional stage (16, 24 and 48 h) and the regressing stage (72, 96 h) on the basis of serum progesterone concentration and changes in steroidogenic enzyme expression during progesterone production. Mice were killed, blood was collected, and the ovaries were removed at the indicated time intervals. CL tissues were collected from the ovaries under a dissecting microscope and immediately frozen at − 70 °C until further use. All experiments were performed in triplicate, and each experiment was independently analyzed.

C57BL/6 female mice (8-week-old) were administered tunicamycin (Tm; 0.5 μg/g body weight; Calbiochem, CA, USA) as an ER stress inducer, and tauroursodeoxycholic acid (TUDCA;0.5 μg/g body weight; Calbiochem) or NAC (1.0 μg/g, Sigma) as ER stress inhibitors by i.p. injection at the indicated times (16 h). Saline was injected intraperitoneally in the control animals. After administration, the mice were sacrificed at the indicated times, and blood and CL tissues were collected.

For estimating serum progesterone concentration, blood was collected from female mice after each injection (16, 24, 48, 72 and 96 h after PMSG/hCG treatment). Serum was separated from blood and stored at − 20 °C until it was assayed for progesterone. Progesterone concentration was assessed using a progesterone EIA kit (ALPCO, Windham, NH, USA) according to the manufacturer’s instructions.

Total RNA was isolated from each individual CL tissue using the Trizol reagent (Invitrogen, Inc., CA, USA) according to the manufacturer’s instructions. RNA concentration and purity were measured using a NanoDrop spectrophotometer (ASP-2680; ACTgene, NJ, USA). Next, 1 μg/μl total RNA and the AccuPower® RT-PCR premix (Bioneer Inc., Daejeon, South Korea) were used to synthesize the cDNA. Primers specific for the mouse sequences of interest (Table 1) were designed using the National Center for Biotechnology Information (NCBI) database. PCR was performed using the following condition: 95 °C for 5 min, 30 cycles of 95 °C for 30 s, 55~ 60 °C for 30 s, 72 °C for 30 s, and 72 °C for 5 min. The PCR products were separated by electrophoresis on a 2% agarose gel with a known standard (100 bp ladder, Bioneer Inc.), stained with ethidium bromide, and photographed under UV illumination. Band intensities were quantified using the Image J software (National Institute of Health, Bethesda, MD, USA).

Lysates of CL tissue from each stage were prepared in PRO-PREP protein lysis buffer (iNtRON, Daejeon, Korea). The protein concentration for each sample was estimated using a Bradford dye-binding assay with bovine serum albumin (BSA) as a standard (Bradford, 1976). Aliquots of the proteins (30 μg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 12% gels. After electrophoresis, the separated proteins were transferred onto nitrocellulose (NC) membranes (Pall Corporation, Port Washington, NY, USA). After blocking with 5% non-fat milk in Tris-buffered saline (TBS) with 0.1% Tween 20 at 4 °C under mild agitation, the membranes were incubated with 1:2000 diluted anti-CHOP, anti-p90ATF6, anti-CREB-2, anti-3β-HSD (Santa Cruz, Biotechnology, CA, USA), anti-GRP78/BIP, anti-p-eIF2α, anti-IRE1, anti-eIF2a, anti-JNK, anti-p-SAPK/JNK, anti-caspase-3, anti- β-tubulin (Cell Signaling, Beverly, MA, USA), anti-p50ATF6 (IMG, San Diego, CA, USA), and anti-p-IRE1α (Abcam, Cambridge, MA, USA) antibodies. The membranes were then incubated with secondary horse radish peroxidase (HRP)-conjugated anti-goat/mouse/rabbit IgG (Thermo, Rockford, IL). Antibody binding was detected using an enhanced chemiluminescence (ECL) kit (Advansta, CA, USA). Band intensities were quantified using the Image J software (NIH, MD, USA).

The ovaries were isolated from the body and fixed with 4% formalin (Sigma Aldrich). The sections were stained with hematoxylin and eosin using a standard protocol described previously (Park, et al. 2013b, Park, et al. 2014) [18, 22]. Thin sections were cut with a diamond knife and mounted. The sections were observed under an Olympus BX51 microscope (Olympus, Tokyo, Japan).

The ovaries were fixed in formalin, embedded in paraffin, and cut into 3-μm-thick sections. The sections were deparaffinized and briefly heated. The sections were then treated with a protein blocking solution (Dako, CA, USA) and incubated with anti-phospho eIF2a (Cell Signaling). After washing with 0.1 M TBS containing 0.01% Tween-20 (TBST), the sections were incubated with anti-rabbit polymer (Dako). Peroxidases bound to the antibody complex were visualized after treatment with a 3, 3′-diaminobenzidine (DAB) chromogen substrate solution (Dako). The DAB reaction was monitored under a microscope to determine the optimal incubation time and stopped with several washes of 0.1 M TBS. The immunolabeled sections were dehydrated in a graded ethanol series, defatted in xylene, and mounted. The sections were examined under a bright field Olympus BX51 microscope and images were acquired using an Olympus DP 70 camera (Olympus, Japan).

All experiments were repeated at least three times, and percentage data obtained in this study are presented as the mean ± standard error of the mean (SEM). These data were subjected to one-way analysis of variance (ANOVA), followed by Dunnett’s multiple comparison test. Statistical analysis was performed using paired Student’s t-test. Data obtained from WT and Prdx1 K/O mice CL tissues were compared using the unpaired Student’s t-test. All calculations were performed using the GraphPad Prism 5.0 software package (GraphPad Software, San Diego, CA, USA). P < 0.05 was considered to be statistically significant.

For delineating the timing of the luteal phase during the estrous cycle post-PMSG/hCG injection, we used a methods of hormone injection for synchronization of estrous cycle as previously described by Park et al., [18].

We observed the morphology and determined the number of collected CL tissues microscopically at 16, 24, 48, 72 and 96 h after PMSG/hCG injection (Fig. 1a and b). Next, we determined the serum progesterone level and the expression of 3β-HSD, a classical steroidogenic enzyme involved in progesterone biosynthesis, using the progesterone ELISA kit (ALPCO) and western blot analysis during mice estrous cycle, respectively. Serum progesterone level was significantly increased (P < 0.01) 48 h after the PMSG/hCG injection (Fig. 1c). Protein levels of 3β-HSD showed the same pattern as that of serum progesterone till the CL functional stage, but rapidly decreased after 48 h (Fig. 1d). Based on these results, we concluded that the time point (48 h) corresponding to peak progesterone levels in CL tissues during the entire luteal phase after PMSG/hCG injection should be used for subsequent experiments.

In our previous study, we demonstrated that ER stress affects progesterone production during the functional stage of CL. In addition, ER stress-mediated apoptosis induced CL regression in the mouse luteal phase [18]. Recent studies indicate that ROS production can be induced by ER stress [23]. Therefore, we investigated whether progesterone production is regulated by the ROS scavenger under ER stress in the functional stage of CL. To analyze the relationship between ROS and ER stress during progesterone production in the CL tissue at the peak of progesterone synthesis (48 h after PMSG/hCG injection), we performed the experiment as shown in Fig. 2a. We used Tm as an ER stress inducer, TUDCA as an ER stress inhibitor, or NAC as a ROS scavenger. Mice injected with Tm (0.5 μg/g) showed dynamic deficiency of luteal function with low serum progesterone and reduction in 3β-HSD and P450scc protein levels in the CL. Compared to the Tm-treated group, serum progesterone level recovered significantly (P < 0.05) in the group where NAC (1.0 μg/g) was added after Tm (0.5 μg/g) pre-treatment (Fig. 2b). Addition of NAC as a recovery effect like a TUDCA also significantly increased the protein levels of 3β-HSD and P450scc in CL (Fig. 2c). On the contrary, NAC or TUDCA administration in mice pre-treated with Tm decreased the expression level of CHOP as an ER stress-mediated apoptotic factors during the functional stage of CL (P < 0.001) (Fig. 2c). These results demonstrate that the regulation of ER stress-induced ROS production is required for maintaining progesterone production during the functional stage of CL in mice.

The above results underline the necessity of regulating ROS synthesis for maintaining progesterone production during the functional stage of CL (Fig. 2). We investigated whether PRDXs can act as antioxidants for maintaining progesterone production.

The expression of the PRDXs at 16, 24, 48, 72 and 96 h after PMSG/hCG injection were analyzed (Fig. 3a). Interestingly, PRDX1 was continuously expressed in CL during the entire luteal phase (Fig. 3b), irrespective of the changes in 3β-HSD expression. In contrast, western blotting showed that the other PRDX proteins (2–6) were not expressed continuously during the entire luteal phase (Additional file 1: Figure S1). The PRDX1 level was increased by Tm treatment and recovered by TUDCA at the peak time point of progesterone production (Fig. 3c and d). TUDCA or NAC injection after Tm also decreased PRDX1 level compared to Tm only treatment (Fig. 3c and d). These results indicated that continuously expressed PRDX1 in CL during the entire luteal phase may be involved in ROS regulation for maintaining progesterone production in the functional stage of CL. Therefore, to investigate whether PRDX1 acts as an antioxidant and ROS-regulator during progesterone production in the functional luteal stage after PMSG/hCG injection, we evaluated the changes in progesterone production and expression of ER stress marker genes in the CL tissue of Prdx1 K/O mice and compared them with those of WT mice. First, to determine the CL number, the ovary from Prdx1 K/O and WT mice were subjected to genotyping and H & E staining, respectively (Fig. 4a). The number of formed CL and ovary size in Prdx1 K/O mice dramatically decreased (P < 0.01; fold 2.8, Fig. 4b) 48 h after the PMSG/hCG injection. Similarly, the serum progesterone concentration and protein levels of 3β-HSD and P450ssc were significantly reduced compared to those in the CL of WT mice (Fig. 4c and d). Furthermore, to determine the effect of PRDX1 on progesterone production in CL tissue, we observed the ovary morphology in wild type, and Prdx1 heterozygous and homozygous K/O mice. Although the ovary size in Prdx1 heterozygous mouse was slightly reduced, the levels of serum progesterone and steroidogenic enzymes in the CL tissue of Prdx1 heterozygous mice showed the same pattern as that of WT mice 48 h after PMSG/hCG injection (Additional file 2: Figure S2). These results demonstrate that absence of PRDX1 in the functional luteal stage after PMSG/hCG injection decreased CL number and ovary size compared to those of WT mice, which reduces serum progesterone and protein levels of the steroidogenesis enzymes 3β-HSD and P450ssc.

We investigated whether PRDX1 deficiency is involved in UPR signaling (GRP78/BIP, p-EIF2α, p50ATF6, and p-IRE1) or ER stress-mediated apoptosis (CHOP, p-JNK, and cleaved caspase-3) in the functional luteal stage using western blotting analysis and the CL tissue of WT and Prdx1 K/O mice. As shown in Fig. 5a, the levels of GRP78/BIP, p50ATF6, and p-EIF2α were dramatically increased (P < 0.01) in the CL tissue of Prdx1 K/O mice, but, the levels of total EIF2α and p-IRE1 showed no change compared to those of WT mice. The levels of CHOP, p-JNK, and cleaved caspase-3 were also significantly increased (CHOP; P < 0.05, p-JNK, and cleaved caspase-3; P < 0.01) in the CL of Prdx1 K/O mice (Fig. 5b). In addition, immunohistochemistry (IHC) showed that the levels of p-EIF2α, ATF6, and cleaved caspase-3 were higher in the ovary of Prdx1 K/O mice than in the ovary of WT mice (Fig. 5c). Therefore, these results demonstrated that PRDX1 deficiency during functional luteal stage activates the UPR signaling pathways in response to ER stress-mediated apoptosis.

As shown in Fig. 2, the NAC-mediated decrease in ER stress-induced ROS levels is required for maintaining progesterone production during the luteal functional stage. Thus, we investigated whether NAC injection in Prdx1 K/O mice restores the reduced serum progesterone concentration and protein levels of steroidogenesis enzymes in the ovaries of Prdx1 K/O mice. First, we injected NAC 45 h after the PMSG/hCG injection, and collected CL tissue from the ovary 3 h after the NAC injection. As shown in the Fig. 6a, serum progesterone production, which was rapidly reduced (P < 0.01; compared to WT mice) in the CL tissue of Prdx1 K/O mice, recovered significantly (P < 0.05; compared to Prdx1 K/O mice) after the NAC injection. We also investigated whether NAC treatment of Prdx1 K/O mice recovered steroidogenic enzyme expression and decreased activation of UPR signaling in response to ER stress during the functional luteal stage. We observed that the mRNA levels of the steroidogenic enzymes 3β-HSD, StAR (P < 0.001; compared to Prdx1 K/O mice), and P450scc (P < 0.01; compared to Prdx1 K/O mice) were recovered by NAC treatment (Fig. 6b). In contrast, the mRNA levels of CHOP (P < 0.001 compared to Prdx1 K/O mice), GRP78/BIP, and p50ATF6 (P < 0.05 compared to Prdx1 K/O mice) in the CL tissue of NAC-treated Prdx1 K/O mice were significantly reduced (Fig. 6c). These results suggested that decrease in serum progesterone production and activation of UPR signaling in response to ER stress are restored by injecting NAC, a ROS scavenger, during the luteal stage of PRDX I KO mice.