Introduction
Premature ovarian failure (POF), also known as premature ovarian insufficiency (POI), is a subclass of endocrine and reproductive disorders characterized by amenorrhea, increased gonadotropin level, estrogen deficiency, ovarian atrophy, decreased sexual activity, and decreased fertility in women under the age of 40, which affects approximately 1% of women in the general population [
1,
2]. However, the etiology of POF is unclear. Among the underlying mechanisms, destruction of primordial follicles by toxic agents, activation of proapoptotic signaling pathways, or accelerated follicular recruitment might lead to premature exhaustion of the primordial follicle pool or disturbances of follicle function and accelerated premature follicle depletion, which have been demonstrated to be related to follicular atresia and POF [
3‐
6]. Exploring the association between most of these pathogenic factors and their role in ovarian function is necessary to determine the mechanisms responsible for the heterogeneity of POF.
MicroRNAs (miRNAs) are a class of small (18–22 nucleotides) endogenous noncoding RNAs that can negatively regulate gene expression via post-transcriptional gene silencing [
7,
8]. It has been proposed that aberrant expression and dysfunction of miRNAs might be involved in physiological processes and human diseases, such as the events of mammalian reproduction [
9‐
11]. Accumulating evidence shows that deregulation of miRNAs is closely related to ovarian function, including granulosa cell (GC) apoptosis, oocyte maturation, oocyte apoptosis, recruitment of primordial follicles, and localization of migrating primordial germ cells [
9,
12‐
15]. However, the contribution of miRNAs in the pathogenesis of POF has yet to be determined. Therefore, the aim of the present study was to investigate differentially expressed miRNAs in a mouse POF model and reveal the association between miRNAs and diminished ovarian function, which would help to unveil the possible molecular mechanisms underlying the pathogenesis of POF.
Herein, we established a cisplatin-induced mouse POF model to determine whether cisplatin affects GC apoptosis and primordial follicle activation in the ovary. Next, we analyzed miRNA and messenger RNA (mRNA) profiles and noted that miR-144-3p was significantly downregulated in GCs obtained from cisplatin-induced POF mice. Bioinformatic analysis combined with in vitro and in vivo experiments indicated that miR-144-3p alleviated GC apoptosis and ovarian primordial follicle loss induced by cisplatin by directly targeting the mRNA encoding mitogen-activated protein kinase kinase kinase 9 (MAP3K9). Furthermore, the detailed underlying molecular mechanisms were explored. We demonstrated that miR-144-3p could serve as a potential biomarker for ovarian reserve evaluation.
Materials and methods
Animal experiments
All animal experiments were approved by the Committee on Ethics of Biomedicine Research, Naval Medical University. Female C57BL/6 mice (8 weeks) with normal 4–5-day estrous cycles were used in all experiments described below.
Establishment of the premature ovarian failure mouse model
To establish the chemotherapy-induced POF model, 8-week-old female C57BL/6 mice provided by the laboratory animal center of Naval Medical University were administered with 2 mg/kg cisplatin (Sigma-Aldrich, St. Louis, MO, USA) daily via intraperitoneal injection for 7 days. Mice in the control group were injected intraperitoneally with an equal amount of physiological saline. The estrous cycles were routinely assessed by vaginal smear, and venous blood samples were collected in the diestrus stage after 7 days of treatment.
Enzyme-linked immunosorbent assay (ELISA)
Blood samples of each mouse were collected and centrifuged at 3220×g for 15 min. The serum levels of estradiol (E2) and follicle stimulating hormone (FSH) were measured using ELISA kits (Mlbio, Shanghai, China) according to the manufacturer’s instructions.
Ovarian follicle counting
Seven days after cisplatin treatment, the mice were euthanized, and the ovaries were harvested and fixed using 4% formaldehyde for 24 h, embedded in paraffin, serially sectioned at 5 μm, and mounted on every fifth section. The sections were then stained with hematoxylin and erosion (H&E) for further histological examination. The number of ovarian primordial, primary, secondary, antral and atretic follicles was counted under a light microscope according to the accepted definitions described previously [
16].
Terminal deoxynucleotidyl transferase-mediated dUTP–biotin nick end labeling (TUNEL) assay
An In situ Cell Death Detection Kit (Roche, Germany) was used to detect GCs apoptosis in mouse ovarian tissue sections according to the manufacturer’s instruction. The sections were observed with a fluorescence microscope, and the number of TUNEL-positive GCs was counted.
Isolation of mouse primary ovarian GCs and cell culture
Mouse primary ovarian GCs were isolated from female C57BL/6 mice. The female mice were injected intraperitoneally with pregnant mare serum gonadotropin (PMSG, Solarbio, Beijing, China). After 48 h, bilateral ovaries were isolated mechanically. GCs were obtained under a stereomicroscope and washed with phosphate buffered saline (PBS) three times. The GCs were then collected by brief centrifugation and cultured in Dulbecco’s modified Eagle’s medium (DMEM)/F12 (HyClone, Logan, UT, USA) medium supplemented with 10% fetal bovine serum (Biological Industries, ISV, Kibbutz Beit-Haemek, Israel), 100 U/ml streptomycin and 100 U/ml penicillin (Invitrogen, Waltham, MA, USA). GCs were identified using anti-follicle stimulating hormone receptor (FSHR) antibodies (Abcam, Cambridge, UK). The COV434 cell line (human ovarian GCs) was cultured in DMEM/F12 containing 10% fetal bovine serum and 1% penicillin–streptomycin. All cells were cultured at 37 ℃ in a humidified atmosphere of 5% CO2.
Microarray analysis
GCs from the control and POF groups were used for miRNA and mRNA expression profiling assay. RNA quantity and quality were measured by NanoDrop ND-1000, and RNA integrity was assessed by standard denaturing agarose gel electrophoresis. Sample labeling and array hybridization were performed according to the Agilent miRNA Microarray System with miRNA Complete Labeling and Hyb Kit protocol (Agilent Technology). The differentially expressed genes among the different groups were determined by KEGG and GO analyses. Microarray profiling and data analyses were performed by KangChen Bio-tech, Shanghai, China.
RNA isolation and quantitative real-time reverse transcription PCR (qRT-PCR)
Total RNA from cultured cells was isolated using the TRIzol reagent (Invitrogen). cDNA was synthesized from purified total RNA (1 μg) using a PrimeScript RT reagent kit (Takara, Dalian, China) according to the manufacturer’s instructions. The quantitative real-time PCR (qPCR) step of the qRT-PCR protocol was performed using SYBR Green Real-time PCR Master Mix (Takara) and a Real-time PCR system (Applied Biosystems, Foster City, CA, USA). U6 and
ACTB (encoding β-actin) were used as internal controls for miRNAs and mRNAs, respectively. Fold change was calculated according to the 2
−ΔΔCt method, and data are represented relative to the expression in control cells. The specific primers are listed in Additional file
1: Table S1.
Western blotting analysis
Total proteins from cells were prepared, separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and transferred onto polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). After blocking with 5% milk in Tris-buffered saline Tween 20 (TBST) for 1 h at room temperature, the membranes were incubated with primary antibodies at 4 ℃ overnight. Then, the membranes were incubated with IRdye 800-conjugated goat anti-rabbit IgG and/or IRdye 700-conjugated goat anti-mouse IgG for 1 h at room temperature. An Odyssey infrared scanner (Li-COR Biosciences, Lincoln, NE, USA) was used to visualize the immunoreactive protein bands. The relative intensity of the protein bands was expressed relative to that of the control. The levels of β-tubulin or β-actin were used as internal standards. The following primary antibodies were used: anti-MAP3K9 (Cell Signaling Technology, Danvers, MA, USA; 5029, 1:1000), anti-Caspase-3 (Cell Signaling Technology, 9662, 1:1000), anti-general transcription factor IIH subunit 2 (p44)/proteasome 26S subunit, ATPase 6 (p42) (Cell Signaling Technology, 4695, 1:1000), anti-mitogen-activated protein kinase 14 (MAPK14, or p38) (Cell Signaling Technology, 8690, 1:1000), anti-phosphorylated (p)44/42 (Cell Signaling Technology, 4370, 1:1000), anti-p–p38 (Cell Signaling Technology, 4511, 1:1000), anti-protein kinase B (AKT) (Proteintech, Rosemont, IL, USA;10176-2-AP, 1:1000), anti-p-AKT (Proteintech, 66444-1-Ig, 1:2000), anti-forkhead box O3 (FoxO3A) (Signalway Antibody, Greenbelt, MD, USA; 40937, 1:1000), anti-p-FoxO3A (Signalway Antibody, 12199, 1:1000), anti-β-actin (Proteintech, 66009-1-Ig, 1:5000), and anti-β-tubulin (Proteintech, 10094-1-AP, 1:5000).
Cell transfection
COV434 cells were transfected with miR-144-3p agomir, antagomir, specific small interfering RNAs (siRNAs) for MAP3K9, or negative control (agomir NC, antagomir NC, siNC) (GenePharma, Shanghai, China) using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. After 24 h of transfection, the cells were collected for qRT-PCR and western blotting analysis.
Luciferase reporter assay
A luciferase reporter plasmid containing the wild-type 3′ untranslated region (UTR) of MAP3K9 was constructed and the MAP3K9 3′ UTR containing the mutated has miR-144-3p binding site was cloned into the same reporter plasmid. The above luciferase reporter plasmids were co-transfected with miR-144-3p mimics or negative control (NC) into COV434 cells cultured in 6-well plates using Lipofectamine 3000 (Invitrogen). Subsequently, the luciferase activity of the cell lysates was measured 48 h later using a Dual-Glo Luciferase Assay kit (Promega, Madison, WI, USA) according to the manufacturer’s instructions.
In vivo experiments
The POF mice were created as in '
Establishment of the premature ovarian failure mouse model'. After 7 days of treatment, 20 μl of the miR-144-3p agomir or controls at 1 nmol, with Lipofectamine 3000 (Invitrogen), were injected into the left and right ovarian bursa using insulin syringes, respectively. At 48 h after injection, the ovaries were collected for western blotting assessment and histological examination.
Statistical analysis
All experiments were performed three times. Data are expressed as the mean ± standard deviation (SD) and were analyzed using Student’s t test when the data were normally distributed, using GraphPad Prism 6 software (GraphPad Inc., La Jolla, CA, USA). When the data were not normally distributed, a nonparametric test was applied. The significance of differences among groups was assessed using one-way analysis of variance (ANOVA). Statistical significance was defined as P < 0.05.
Discussion
Anticancer treatment is a well-known risk factor for POF. With the wide application of chemotherapeutic agents in various types of tumors and immune diseases, chemotherapy-induced POF has received increased attention among all the causes of POF. Detailed information focusing on the detrimental effects of chemotherapy exposure on ovarian function has been published [
20,
21]. It is generally believed that chemotherapeutic agents can damage the ovary by inducing prenatal loss of oogonia, direct loss of primordial follicles, accelerated primordial follicle activation, follicular atresia, and damage to the ovarian stroma and the microvascular architecture [
21,
22]. As one of the most commonly used anticancer drugs, cisplatin can cause POF by triggering the activation of primordial follicles and increasing the atresia of ovarian follicles [
21].
GC apoptosis is confirmed to be involved in the development of follicular atresia and follicle loss [
23]. Follicular activation and growth are closely related to the proliferation and anti-apoptosis in granulosa cells. The somatic primordial follicle granulosa cells initiate the activation of primordial follicles and govern the quiescence or awakening of dormant oocytes. Activation of mTORC1 signaling in granulosa cells of primordial follicles accelerates the differentiation of primordial follicle granulosa cells into granulosa cells and cause premature activation of all dormant oocytes and primordial follicles [
24]. Primordial follicle granulosa cells trigger the awakening of dormant oocytes through KIT ligand, and the essential communication network between the somatic cells and germ cells is based on signaling between the mTORC1–KITL cascade in primordial follicle granulosa cells and KIT–PI3K signaling in oocytes [
25]. Many growth factors have been reported to be functional in regulating the activation of primordial follicles in vitro, including leukemia inhibitory factor, basic fibroblast growth factor and platelet-derived growth factor [
17]. Granulosa cell-derived C-type natriuretic factor not only suppresses the final maturation of oocytes to undergo germinal vesicle breakdown before ovulation but also promotes preantral and antral follicle growth. In addition, several oocyte- and granulosa cell-derived factors stimulate preantral follicle growth by acting through wingless, receptor tyrosine kinase, receptor serine kinase, and other signaling pathways [
26]. Oocyte–GC bidirectional communications via signal transduction or direct cell-to-cell contact provide the molecular and structural basis for effective oocyte–GC crosstalk, which is required for adequate follicular growth and maturation [
27]. In this study, we illustrated enhanced GC apoptosis and accelerated primordial follicle activation in a cisplatin-induced mouse POF model. This might accelerate follicular apoptosis and follicle reservoir utilization via multiple molecular reactions. A better understanding of the molecular mechanisms underlying cisplatin-mediated ovarian dysfunction might allow the development of efficient and targeted treatments for patients diagnosed with POF and for infertile women of advanced reproductive age.
Recently, miRNAs have been recognized to play important regulatory roles in ovarian function [
23,
28] and miRNAs are indicated to be involved in POF pathogenesis. Zhang et al. discovered the harmful effects of miR-127-5p on proliferation and DNA repair function of GCs via the
HMGB2 gene (encoding high mobility group box 2) and its predictive value in POF [
29]. Moreover, miR-145 was reported to protect GCs against oxidative stress-induced apoptosis by targeting
KLF4 (encoding Kruppel-like factor 4), thereby preventing abnormal follicular atresia and improving the outcomes of ovarian dysfunction [
30]. It has been found that miR-146b-5p overexpression ameliorates POF in mice by inhibiting the DAB2IP/ASK1/p38-MAPK pathway and γH2A.X phosphorylation [
31]. In the current study, we performed miRNA and mRNA microarray screening on GCs derived from cisplatin-induced POF mice and identified significantly downregulated miR-144-3p in GCs from POF mice. Previous research has reported that the expression of miR-144-3p was greatly reduced in patients with polycystic ovarian syndrome (PCOS) and PCOS rat models. Meanwhile, miR-144-3p overexpression could induce ovarian GCs growth and repress cell apoptosis by targeting
HSP70 (encoding heat shock protein 70), which might function as a novel target for PCOS treatment [
32]. We hypothesized that miR-144-3p might also be involved in the regulation of ovarian function in POF. The findings of this study supported our hypothesis and revealed that miR-144-3p alleviated cisplatin-induced GC damage and primordial follicle activation by repression its downstream target gene,
MAP3K9, which sheds light on the epigenetic mechanism involved in the pathogenicity of POF.
MAP3K9 is an upstream activator of the p38 MAPK signaling pathway. p38 MAPK pathway members function in a variety of cellular processes, including cell growth, proliferation, differentiation, migration, and apoptosis [
33]. Several studies have demonstrated the clear involvement of the p38 MAPK pathway in the response to treatment with chemotherapeutic agents [
34‐
36]. This study explored the correlation between the pathway and GC apoptosis, demonstrating that miR-144-3p overexpression decreased GC cell apoptosis resulting from suppression of the downstream p38 MAPK pathway, while reduced expression of miR-144-3p manifested the opposite results through phosphorylation and activation of p38 MAPK pathway members. Furthermore,
MAP3K9 downregulation abrogated the miR-144-3p inhibition-induced effects partially. These results suggested that miR-144-3p prevents GC cell apoptosis by repressing the p38 MAPK pathway via negatively regulating
MAP3K9.
Activation of dormant primordial follicles is the first step in follicular development and is fundamental to determine the ovarian reserve of females [
37]. Although the exact factors involved in primordial follicle recruitment and growth have yet to be elucidated, the PI3K/phosphatase and tensin homolog (PTEN) signaling pathway has been reported to be the key controller for follicular activation [
38]. In the present study, our results suggested that during the process of primordial follicle initiation in vivo, the addition of miR-144-3p to cisplatin-treated ovaries is accompanied by decreased expression of its target gene,
Map3k9, and suppression of the downstream PI3K/AKT signaling pathway. Therefore, miR-144-3p has a protective effect on cisplatin-induced abnormal activation of primordial follicles by repressing the PI3K/AKT cascade via downregulating
Map3k9.
In conclusion, our data revealed that miR-144-3p significantly downregulated in GCs from POF mice. The in vitro and in vivo experiments revealed that miR-144-3p can protect against cisplatin induced GC cell apoptosis and abnormal activation of primordial follicles via the p38 MAPK and PI3K/AKT pathways, respectively, by directly targeting the MAP3K9 gene. This study clarified the biological effects of miR-144-3p and its detailed molecular mechanisms responsible for POF progression and suggested new strategies for POF management.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.