Alleviation of endoplasmic reticulum stress protects against cisplatin-induced ovarian damage

Six-week-old wild-type female C57BL/6 J mice were from the Southern Medical University Animal Center (Guangzhou, China). The mice were housed in a temperature- and humidity-controlled animal facility and maintained on a 12-h light/dark cycle. They were acclimated for 5 days before the experiment, with free access to a commercial rodent diet and tap water. All animal experiments were approved by the Southern Medical University Committee on the Use and Care of Animals and were performed in accordance with the Committee’s guidelines and regulations.

Twenty mice were randomly and evenly divided into two groups. The experimental group received intraperitoneal injections of CDDP (2.5 mg/kg/d in saline) (Sigma-Aldrich, Shanghai, China) for 7 consecutive days, and the control group were injected with saline. This dosage was according to a previous study [19] and our preliminary results showing that this is the minimum dose causing significant morphological changes within 7 days when compared with the control group. After anesthesia induced with 10% chloral hydrate, the mice were killed and their ovaries rapidly dissected. Ovaries of three mice randomly selected from each group were subjected to a proteomic analysis with iTRAQ (Fitgene Biotech, Guangzhou, China). The remaining ovaries were used for real-time quantitative PCR (qPCR) and a western blotting analysis. Ovaries for histological examination were fixed in 4% formaldehyde.

The frozen ovaries were grinded by liquid nitrogen and then lysed in 500 μl of LC3 SDS lysis buffer (Add 1× PMSF before use) containing 7 M Urea, 2 M Thiourea, 20 mM Tris base and 0.2% SDS. Then the samples were sonicated and centrifuged to collect the supernatant. Every 250 μL of supernatant was precipitated overnight at − 20 °C with 1 mL acetone. After discarding the acetone and air drying, the resulting pellet was dissolved in lysate L3 (no SDS). Then the sample was sonicated on ice and centrifuged at 12,000×g for 20 min at 4 °C, the supernatant was collected. Protein concentration was estimated by the Bradford method (Fitgene Biotech, Guangzhou, China).

iTRAQ labeling was performed according to the manufacturer’s protocol (Applied Biosystems, Sciex). Briefly, 200 μg of each protein sample was reduced with TCEP Reducing Reagent at 60 °C for 1 h,and alkylated with MMTS Cysteine-Blocking Reagent at room temperature for 30 min. Then, proteins were digested with trypsin (Promega, USA) at 37 °C at a ratio of 1:50 (enzyme-to-substrate) overnight. Each sample was labeled separately with two of the eight available tags (control: 114 and 116 tags; cisplatin: 117 and 119 tags). All labeled peptides were pooled together.

The Ultimate 3000 HPLC system (Dionex, USA) equipped with a 2.00-mm-inner diameter *100-mm-long Gemini-NX 3u C18110Acolumns (Phenomenex, USA) was used for High-pH fractionation. Peptides were loaded onto the column and washed isocratically at 95% eluent A (20 mM HCOONH4, 2 M NaOH) (pH 10). The iTRAQ-tagged peptides fractionation was performed using a linear binary gradient from 15 to 50% B (20 mM HCOONH 4, 2 M NaOH, 80% ACN) (pH 10) at 0.2 ml/min for more than 45 min. Finally, the column was washed at 90% B for 10 min and returned to 95% A for 10 min. Set the UV detector was at 214/280 nm, and fractions were collected every 1 min. 10 fractions were pooled and dried by vacuum centrifuge for subsequent nano-reversed phase liquid chromatography (nano-LC) fractionation.

The fraction was resuspended in loading buffer (0.1% FA,2% ACN) and separated with an Ultimate 3000 nano-LC system equipped with a C18 reverse phase column (100-μm inner diameter, 10-cm long, 3-μm resin from MichromBioresources, Auburn, CA). Separate the peptides with the following parameters: 1) mobile phase A:0.1% FA, 5% ACN, dissolved in water; 2) mobile phase B: 0.1% FA, 95% ACN; 3) flow rate: 300 nl/min; 4) gradient: B-phase increased from 5 to 40%,70 min. Then, the LC eluent was subject to Q Exactive (Thermo Fisher) in an information dependent acquisition mode. MS spectra were acquired across the mass range of 400–1250 m/z in high resolution mode (> 30,000) using 250 ms accumulation time per spectrum. A maximum of 20 precursors per cycle were chosen for fragmentation from each MS spectrum with100 ms minimum accumulation time for each precursor and dynamic exclusion for 20 s. Tandem mass spectra were recorded in high sensitivity mode (resolution > 15,000) with rolling collision energy on and iTRAQ reagent collision energy adjustment on.

The ovaries were fixed in 4% formaldehyde for 24 h, embedded in paraffin wax, sectioned at 4 μm thickness, and mounted on glass slides for hematoxylin and eosin (H&E) staining. At least five sections (taken 100 μM apart) from an ovary and 5 ovaries from each group were photographed for follicle assessment. A follicle was deemed present if the oocyte contained a germinal vesicle, and was counted and classified according to its health and developmental stage, as a healthy (including primordial, primary, secondary, and antral follicle) or atresia follicle, according to previously described criteria [20].

For the immunofluorescence analysis, the sections were incubated with the antibodies summarized in Additional file 1: Table S1, followed by incubation with Alexa-Fluor-488-labeled secondary antibodies (Molecular Probes, MA, USA) and 4′,6-diamidino-2-phenylindole (Invitrogen, Carlsbad, CA, USA) to visualize the nuclei. The immunofluorescent images were obtained using FluoView FV1000 confocal microscopy (Olympus, Tokyo, Japan).

Cell apoptosis in the follicles was evaluated in sections using a terminal-deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) assay for the situ visualization of DNA fragmentation with the commercial DeadEnd™ Fluorometric TUNEL System (Promega, Madison, WI, USA). Images were obtained with a FluoView FV1000 confocal microscope. Every 25th section in 5 ovaries from each group were analyzed, and the level of apoptosis was expressed as the total number of apoptotic cells on five sections from an ovary of each mouse.

The mice were anesthetized to collect their blood via cardiocentesis. The blood samples were centrifuged at 3000×g for 10 min and the serum collected. The plasma levels of FSH and estradiol (E2) hormone were measured with ELISA kits (Elabscience Biotechnology, Wuhan, China).

Total ovarian RNA was purified with TriGene Reagent, and then processed to cDNA with the RETROscript Reverse Transcription Kit, and amplified and quantified with the RealStar Power SYBR Kit (all from GenStar BioSolutions, Beijing, China) with a StepOnePlus Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). The gene-specific primers (Sangong Biotech, Shanghai, China) for qPCR were: For Hspa5, forward: 5′-ACTTGGGGACCACCTATTCCT-3′, reverse: 5′-GTTGCCCTGATCGTTGGCTA-3′; for Hsp90ab1, forward: 5′- GTCCGCCGTGTGTTCATCAT-3′, reverse: 5′-GCACTTCTTGACGATGTTCTTGC-3′. The expression of Hspa5 and Hsp90ab1 were normalized to that of glyceraldehyde 3-phosphate dehydrogenase (Gapdh).

Ovaries were homogenised in lysis buffer and boiled in sodium dodecyl sulfate (SDS) loading buffer. The protein extracts were then subjected to 6–12% SDS-PAGE and electrotransferred to nitrocellulose membranes (GE Healthcare Life Sciences, Beijing, China). The membranes were blocked with 5% nonfat dry milk for 1 h at room temperature, washed, and incubated with the indicated primary antibody at 4 °C overnight. The membranes were further washed, incubated with Peroxidase-AffiniPure Goat Anti-Rabbit or -Mouse IgG (H + L) (Jackson ImmunoResearch, West Grove, PA, USA) for 1 h at room temperature, washed again, and finally visualized with an enhanced chemiluminescence kit (PerkinElmer, Waltham, MA, USA). The primary antibodies used for the immunoblotting analysis are summarized in Additional file 1: Table S1.

Seventy-two mice were evenly and randomly divided into four groups, which received intraperitoneal injections of 0.9% saline as control group, and experimental groups were intraperitoneal injections of CDDP (2.5 mg/kg/d), CDDP + 4-PBA (2.5 + 150 mg/kg/d, respectively) [21], and CDDP + 3-methyladenine (3-MA; 2.5 + 10 mg/kg/d), respectively [22] (4-PBA and 3-MA were from Sigma-Aldrich), for 1, 3 and 7 days. The mice were euthanized at the indicated time points, and their ovaries were isolated for histological analysis and a western blotting assay.

The peptide data were analyzed using Protein Pilot Software 4.0 (AB SCIEX, CA), and the parameters were set as follows: Cys alkylation: MMTS; ID focus: biological modifications; Digestion: typsin; Database: Uniprot_MOUSE; Search effort: thorough ID. All experiments were performed at least in duplicate. Data are presented as means ± SD, and differences between groups were analyzed with an independent-samples t-test and among groups with one-way ANOVA followed by the SNK test (SPSS 25.0, SPSS Inc., Chicago, IL, USA). If the data were not normally distributed, the nonparametric Mann-Whitney test was used. A χ2 test was used for rate comparisons. p < 0.05 was considered statistically significant.

To investigate the molecular changes underlying CDDP-induced ovarian damage, we generated a mouse model of POI by injecting mice intraperitoneally with CDDP for 7 days, and then subjecting their ovaries to proteomic screening with an iTRAQ analysis. The expression levels of 214 proteins were significantly upregulated (fold change ≥1.5) and the levels of 180 proteins were significantly downregulated (fold change ≥ − 1.5) (Additional file 1: Table S2). Of these 394 differentially expressed proteins, HSPA5 and HSP90AB1, two well-established markers for ERS and UPR [14], were of our particular interest, because they were downregulated 2.5-fold and 2.3-fold, respectively, in the CDDP-treated ovaries (Table 1). Furthermore, a STRING analysis showed that both of them were at the core of signaling network (Additional file 1:Figure S1). To confirm this proteomic result, we used qPCR to show that the relative mRNA expression of Hspa5 and Hsp90ab1 was significantly lower in the CDDP group than in the control group (Fig. 1a). Western blotting further confirmed the reduction in their protein levels in the CDDP group (Fig. 1b, c). Although the proteomic analysis indicated a more dramatic decline in the protein levels of FADS2 and HSD11B2, which, however, were not confirmed by western blotting (Additional file 1:Figure S2). These results indicated that ERS was strongly associated with CDDP-induced ovarian damage.

We next characterized the expression profiles of HSPA5 and HSP90AB1 in the CDDP-treated ovaries with immunofluorescence. Both HSPA5 and HSP90AB1 were predominantly localized to the cytoplasm around the nucleic of granulosa cells from secondary and antral follicles, modestly expressed in those from primary follicles, while hardly present in those of from primordial follicles (Fig. 2a), indicating that ERS in granulosa cells may play a role in CDDP-induced follicle loss. Therefore, two human granulosa tumor cell lines, KGN and COV434, were used for an in vitro study. CDDP at 50 μM significantly decreased the proteins expression of HSPA5 and HSP90AB1 in KGN cells at 24 h,thus we treated cells with the dose in the following cell experiments (Fig. 2b). CDDP (50 μM) induced time-dependent changes in HSPA5 and HSP90AB1 protein levels in the KGN cells, with an increase at 4 h, which were then decreased at 24 h (Fig. 2c), indicating that ERS was induced upon exposure to CDDP. Autophagy related 12 (ATG12) is positively involved in the assembly of the autophagosome. P62 binds to polyubiquitinated proteins and LC3 on the autophagosomal membrane to target aggregates to the autophagosomes for degradation, and its expression negatively correlates with autophagosome formation [28]. CDDP treatment induced a time-dependent increase in the expression of ATG12 and a reduction in P62, and also an increase in cleaved PARP protein, a marker of cell apoptosis (Fig. 2d). Interestingly, when either the HSPA5 or HSP90AB1 gene was knocked down with RNA interference (RNAi), the levels of both ATG12 and cleaved PARP proteins were greatly reduced (Fig. 2d). This finding suggested that ERS induced cell autophagy and apoptosis in response to excessive CDDP treatment. This was verified by introducing 4-PBA, an ERS inhibitor, into CDDP-treated cells, where it significantly relieved CDDP-induced ERS, cell autophagy, and cell apoptosis (Fig. 2e). A flow-cytometric analysis of annexin V/PI staining further confirmed that CDDP-induced cell apoptosis was largely prevented by 4-PBA (P < 0.05, Fig. 2f, g). These results together demonstrated that the alleviation of ERS attenuated CDDP-induced cell autophagy and apoptosis.

We next examined whether inhibiting autophagy protected cells from CDDP-induced apoptosis. DALgreen staining showed that although both 3-MA and 4-PBA efficiently reduced the production of CDDP-induced autophagosomes (P < 0.05) (Fig. 3a, b). Western blotting also showed that 3-MA alleviated CDDP-induced autophagy level by reducing the expression of ATG12 and by increasing that of P62, whereas the protein levels of cleaved PARP, HSPA5, and HSP90AB1 remained stable (Fig. 3c). The finding that 3-MA negligibly reduced CDDP-induced cell apoptosis was further supported by the results of flow cytometry (p > 0.05, Fig. 3d, e). These findings collectively suggested that lowering the autophagy level with 3-MA neither reduced cell apoptosis nor alleviated ERS during CDDP treatment.

We then compared the protective effects of 4-PBA and 3-MA against CDDP-induced ovarian damage in vivo. Female mice were treated with CDDP combined with 4-PBA or 3-MA for 1, 3, or 7 days. A histological examination showed that the morphologies of follicles at different stages of development were parallel across all the groups on day 1, except that the healthy follicles in CDDP group decreased, compared with the saline group ((P < 0.05, Fig. 4a-c). By day 3, the CDDP-treated group had significantly fewer healthy follicles and more atresia follicles than the saline-treated group (P < 0.01), which was predominantly attributable to the loss of primordial and antral follicles (Fig. 4b, c). By day 7, the longer exposure to CDDP seriously damaged the healthy follicles and produced excessive atresia follicles than the saline-treated group (both P < 0.01). Besides primordial and antral follicles, more secondary follicles were damaged, whereas the primary follicles appeared not susceptible. We speculate that it is associated with follicular recruitment and CDDP-induced granulosa cell apoptosis in growing follicles. The more granulosa cells in the growing population of follicles, the more susceptible they were to CDDP. These data support a previous report that chemotherapeutic agents over activate the primordial follicles and damage the growing population of follicles, leading to the premature depletion of follicle reserves [29]. As expected, 4-PBA, but not 3-MA, markedly ameliorated CDDP-induced follicle loss and preserved more healthy follicles by protecting follicles from atresia (Fig. 4b, c). Consistent with this, a TUNEL assay showed that CDDP-induced granulosa cell apoptosis peaked on day 3 (P < 0.01), and was markedly ameliorated by 4-PBA (P < 0.05) but not by 3-MA (P > 0.05, Fig. 5a, b). By day 7, the intensity of granulosa cell apoptosis had decreased in the CDDP group compared to the 3rd day, and 4-PBA slightly reduced the number of apoptotic cells (Fig. 5a, b). Consistent with the histological results, hormone measurements with ELISAs showed that exposure to CDDP for 7 days clearly reduced the plasma E₂ content, and especially, increased the FSH level (both P < 0.01, Fig. 5c), which is considered as the most important biochemical indicator of POI in clinical. The 4-PBA treatment clearly-inhibited the CDDP-induced increase in FSH, but not the reduction in E₂, whereas and the 3-MA treatment had negligible effects on these changes (Fig. 5c). We propose that the secretion of FSH is affected by the negative feedback of E₂ and FSH receptors, the production of E₂ decreased due to excessive granulosa cell apoptosis induced by CDDP, which in turn stimulates the secretion of FSH from pituitary. 4-PBA reduced granulosa cell apoptosis and may preserve FSH receptors, therefore, the secretion of FSH was not increased significantly. Furthermore, an immunoblotting analysis of whole-mount ovarian proteins showed that in vivo treatment with 4-PBA blocked the increases in the protein levels of ATG12 and cleaved PARP, but not of HSPA5 and HSP90AB1during CDDP treatment (Fig. 5d). This may indicate different responses to 4-PBA between in vivo and in vitro. The CDDP-induced increase in cleaved PARP was resistant to 3-MA treatment, although 3-MA reduced the protein level of ATG12 and increased P62 (Fig. 5d). These results together provided in vivo evidence supporting the notion that the alleviation of ERS attenuated CDDP-induced granulosa cell death and ovarian damage.