Abstract

   Both genetic and microenvironmental detrimental factors are involved in
   ovarian dysfunction, leading to the increasing rate of involuntary
   childlessness in recent years. Oxidative stress (OS), which is
   characterized by the imbalance of redox system with redundant reactive
   oxygen species (ROS) overwhelming the antioxidant defense, is regarded
   as one of the culprits of ovarian dysfunction. OS causes damage to
   various types of ovarian cells including granulosa cells (GCs),
   jeopardizing the ovarian microenvironment, disturbing follicular
   development and participating in various female reproductive disorders.
   However, the specific molecular pathological mechanisms underlying this
   process have not been fully elucidated. In this study, we found that
   3-nitropropionic acid (3-NP) treatment led to significant IGF2BP1
   downregulation via, at least partially, inducing ROS overproduction.
   IGF2BP1 regulates GCs viability, proliferation, cell cycle and cellular
   senescence by enhancing MDM2 mRNA stability in an m^6A-dependant
   manner. IGF2BP1 overexpression partially rescued 3-NP induced GCs
   damages, while ectopically expressed MDM2 alleviated both 3-NP or
   IGF2BP1-knockdown induced GCs dysfunction. These results reveal an
   epigenetic molecular mechanism underlying OS-related GCs disorders,
   which may help to establish a novel potential clinical marker for
   predicting the GCs status as well as the follicular developmental
   potential.

   Keywords: IGF2BP1, Granulosa cells, Oxidative stress, m^6A
   modification, Cellular senescence, Reactive oxygen species

   Abbreviations: OS, oxidative stress; GCs, granulosa cells; ROS,
   reactive oxygen species; 3-NP, 3-nitropropionic acid; FSH,
   follicle-stimulating hormone; LH, luteinizing hormone; ActD,
   Actinomycin D; m^6A, N6-methyladenosine; UTR, untranslated region

Graphical abstract

   [36]Image 1
   [37]Open in a new tab

Highlights

     * •
       Oxidative stress exposure led to significant downregulation of
       IGF2BP1.
     * •
       IGF2BP1 regulates cell viability, proliferation, cell cycle and
       cellular senescence in GCs.
     * •
       IGF2BP1 regulates GCs functions by enhancing MDM2 mRNA stability in
       an m6A-dependant manner.
     * •
       IGF2BP1 and MDM2 overexpression partially rescued oxidative stress
       induced GCs dysfunction.

1. Introduction

   Postponement of parenthood, unhealthy lifestyle, environmental
   pollution and mental stress are progressively increasing the rate of
   involuntary childlessness [[38]1,[39]2]. In recent years,
   multifactorial female infertility has gradually become one of the most
   arduous problems of reproductive medicine. However, the mechanisms
   underlying female infertility have not been fully elucidated yet.
   Substantial evidences have revealed that both genetic and
   microenvironmental factors are involved in female fertility decline.
   These etiologies collectively contribute to the diminished ovarian
   reserve, ultimately resulting in impaired female fecundity [[40]3].

   Among all the infertility-inducing factors, oxidative stress (OS) is
   regarded as one of the culprits of ovarian dysfunction. As a result of
   the imbalance of redox system, OS causes damages to cells and tissues
   with redundant reactive oxygen species (ROS) overwhelming the
   antioxidant defense. ROS, including peroxides, superoxide, hydroxyl
   radical and singlet oxygen [[41]4], are highly reactive chemicals
   derived from the reduction of molecular oxygen. In normal conditions,
   moderate amount of ROS is generated on the mitochondrial respiratory
   chain and serves as a signaling molecular enabling several
   physiological processes [[42]5]. Nevertheless, when ROS accumulate to
   an extent exceeding the body's scavenging capacity, it will lead to
   varying degrees of OS and, in consequence, multiple kinds of cellular
   damages [[43]6]. OS jeopardizes the ovarian microenvironment, disturbs
   follicular development and participates in various female reproductive
   disorders including endometriosis, polycystic ovary syndrome and
   premature ovarian failure, which make it a key regulator in the
   pathogenesis of female subfertility [[44][6], [45][7], [46][8]].

   Granulosa cells (GCs), a kind of ovarian somatic cell which surrounds
   and closely correlates with the developing oocytes, are considered as a
   multifunctional and pivotal component of the ovarian follicular
   microenvironment. Under the stimulation of follicle-stimulating
   hormone, the single-layered GCs proliferate into multiple layers and
   differentiate into two types, the inner cumulus cells which encircle
   the oocytes and the outer mural GCs which cling to the follicular wall
   [[47]9,[48]10]. GCs participate in oogenesis by supporting the oocyte
   development, producing sex steroids and secreting several growth
   factors [[49]11]. Intact GCs functions are an indispensable
   prerequisite for acquiring oocyte competence and maintaining female
   fertility. However, various detrimental factors, including OS, will
   result in GCs dysfunction and hinder the GCs-oocyte interaction. The
   adverse impact of OS on GCs has been revealed by several studies. Zhou
   et al. found advanced oxidation protein products enhanced intracellular
   ROS generation and caused cell cycle arrest [[50]12]. Another study
   suggested that excessive OS in cumulus GCs lead to cellular senescence
   and subsequent endometriosis-associated infertility [[51]13].
   Additionally, the function of GCs to produce steroid hormones was also
   impaired by OS exposure [[52]14]. All these evidences demonstrate that
   OS will result in GCs dysfunction and impair female fecundity. Despite
   all that, the specific molecular pathological mechanisms underlying
   this process have not been fully elucidated.

   In this study, we found the treatment of 3-nitropropionic acid (3-NP),
   a mitochondrial toxin inducing intracellular OS by targeting complex II
   of electron transport chain, led to significant downregulation of
   Insulin-like growth factor 2 mRNA-binding protein 1 (IGF2BP1) and
   subsequently resulted in MDM2 mRNA instability in GCs. IGF2BP1 is an
   RNA-binding protein which preferentially binds to m^6A modifications
   and helps to stabilize its targeted transcripts [[53][15], [54][16],
   [55][17], [56][18]]. As a potent oncogenic factor, IGF2BP1 regulates
   proliferation, cell cycle progression and stemness of cancer cells
   [[57]17,[58]19,[59]20]. Furthermore, it also modulates cell behaviors
   in pluripotent and mesenchymal stem cells [[60]21,[61]22]. In recent
   years, IGF2BP1 was found to be an essential regulator of intestinal
   functions by facilitating colonic epithelial repair and maintaining the
   intestinal barrier function [[62]23,[63]24]. All these evidences
   revealed that IGF2BP1 is involved in various biological processes,
   especially in cell proliferation. Considering that the massive growth
   of GCs is a crucial supportive condition for follicular development, we
   reasoned that IGF2BP1 may also serve as a key factor regulating GCs
   behavior during folliculogenesis, which has not been, however,
   evidenced by any study up to now. MDM2 is an essential regulator of p53
   in both normal and cancer cells. MDM2 upregulation usually provides
   proliferation advantages to cells, leading to the progression and
   metastasis of cancers. MDM2 and p53 modulate each other via an
   autoregulatory feedback loop, collectively controlling several cellular
   processes including cell proliferation, cell cycle and cell motility
   [[64][25], [65][26], [66][27]]. Additionally, MDM2 also participates in
   several clinically relevant cellular processes in a p53-independent
   manner [[67]28,[68]29]. Previous studies have revealed the role of MDM2
   in regulating GCs apoptosis, proliferation, follicular development and
   oocyte quality [[69][30], [70][31], [71][32]], indicating MDM2 is a key
   factor in maintaining normal GCs functions and female fertility. The
   specific mechanisms by which MDM2 is regulated in GCs, however, still
   remain unclear.

   The present study first revealed the involvement of IGF2BP1 in
   meditating OS-induced GCs dysfunction. IGF2BP1 downregulation in GCs
   destabilized MDM2 mRNA, resulting in impaired cell viability, inhibited
   cell proliferation, arrested cell cycle and exacerbated cellular
   senescence. These results decipher a novel OS-related mechanism
   meditating damages and dysfunctions in GCs, which may facilitate to
   develop new therapeutic strategies to improve the follicular
   developmental potential and subsequent reproductive outcomes.

2. Materials and methods

2.1. Animal handling and ethics approvals

   Female C57BL/6J mice (8-week-old, SPF-class) were obtained from Charles
   River, Beijing, China. All mice were adapted for 3 days after their
   purchase and were maintained under a controlled temperature (26 ± 2 °C)
   with 12 h (h) light/dark conditions. Fifty mice were randomly dived
   into two groups (the control group and the 3-NP group, n = 25 per
   group). Mice in the 3-NP group received intraperitoneal injections of
   20 mg/kg 3-NP (diluted with PBS) administration every morning for 2
   weeks, while mice in the control group received an equal volume of PBS.
   The body weight of mice was recorded before and after 3-NP
   administration. Vaginal smears were obtained before sacrifice to
   monitor the estrous cycles of mice [[72]33] and subsequent assays were
   performed using mice at the same phase of estrous cycle. Ovaries were
   removed and weighed immediately. The ovary coefficient was defined as
   the ratio of the ovary weight to the body weight of the mouse. For
   preparation of paraffin sections, the ovaries were immediately fixed in
   4% paraformaldehyde, embedded in paraffin, and sectioned into
   5-μm-thick slices. For frozen sections, the ovaries were directly
   immersed in liquid nitrogen, embedded in a freezing compound (OTC) and
   sectioned into 10-μm-thick slices. The paraffin sections were stained
   with hematoxylin and eosin (H&E) and imaged using a virtual slide
   scanner VS120 (Olympus, Japan). All stages of follicles (primordial,
   primary, small secondary, large secondary and antral follicles) were
   detected and classified. The number of follicles from each ovary was
   calculated and compared between each group.

   The experiments were approved by the Institutional Animal Care and Use
   Committee, Huazhong University of Science and Technology in accordance
   with the National Research Council's “Guideline for the Care and Use of
   Laboratory Animals”.

2.2. Cell culture and treatment

   The human granulosa tumor-derived cell line COV434 was purchased from
   Wuhan Warner Bio Technology Co., Ltd. The human granulosa-like tumor
   cell line KGN and human embryonic kidney cell line HEK293T were
   obtained from Procell Life Science&Technology Co., Ltd. All the cells
   were maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco,
   USA) supplemented with 10% fetal bovine serum (Gibco, USA) and 1%
   penicillin/streptomycin (Gibco, USA) at 37 °C in a humidified incubator
   containing 5% CO[2], and stored in FBS with 10% DMSO at −80 °C. To
   stimulate ROS production or induce energy depletion, COV434 cells (70%
   confluence) were treated with 10 mM 3-NP (Sigma-Aldrich, USA), 150 μM
   hydrogen peroxide (H[2]O[2], Sinopharm, China),2.5 μM antimycin A
   (Maokang biotechnology, China) or 1.5 μM Carbonyl cyanide
   4-(trifluoromethoxy)phenylhydrazone (FCCP, MedChemExpress, China) for
   24 h. Pretreatment of N-acetylcysteine (NAC, MedChemExpress) was
   performed at concentration of 2.5 mM for 2h. Transfection of cells with
   DNA or siRNAs was performed using Lipofectamine 3000 (Thermo Fisher
   Scientific, USA) according to the manufacturer's instructions. For
   siRNA-meditated IGF2BP1 or METTL3 knockdown, cells were transfected
   with 50 nM IGF2BP1- or METTL3-directed siRNA as well as its negative
   controls (RiboBio, China). For IGF2BP1 or MDM2 overexpression, cells
   were transfected with pcDNA3.1-IGF2BP1 (Genomeditech Co., Ltd, China)
   and pCMV-MDM2 (Vigene Biosciences, China) as well as their respective
   vectors as negative controls. The siRNA sequences used are provided as
   follows:

   si-IGF2BP1: 5′- GGCTCAGTATGGTACAGTA -3'.

   si-METTL3: 5′- GCAAGAATTCTGTGACTAT -3'.

2.3. ROS and superoxide anion quantification

   A reactive oxygen species assay kit (Beyotime) and superoxide anion
   indicator (Beyotime) were used following the manufacturers’
   instructions. The cells were rinsed and incubated with the fluorescent
   probe DCFH-DA and Dihydroethidium in the dark at 37 °C for 25 min.
   After 3 washes with DMEM, cells were incubated with Hoechst 33,342 for
   nuclear staining. Fluorescent images were captured with a Zeiss LSM 900
   Confocal Laser Scanning Microscope (Carl Zeiss, Germany).
   Quantification of ROS or superoxide anion in COV434 cells or ovarian
   frozen sections was performed by analyzing the fluorescent intensity.

2.4. Cell viability and proliferation assay

   Cell viability was determined using the Cell Counting Kit-8 (CCK-8,
   Yeasen, China). After cell treatment, 10μL/well CCK-8 solution was
   added to each well of the 96-well plate and incubated at 37 °C for 2 h.
   The spectrophotometric absorbance increase at 450 nm was measured with
   a Synergy HTX Multi-Mode Reader (BioTek, USA). Cell proliferation
   capacity was analyzed with Cell-Light EdU Apollo567 In Vitro Kit
   (RiboBio, China). Briefly, cells were labeled with EdU by incubating
   with 50 μM EdU medium for 2 h and were fixed with 4% paraformaldehyde
   in phosphate buffer saline (PBS). After washed with 2 mg/mL glycine
   solution for 5 min, cells were permeabilized with 0.5% Triton X-100 in
   PBS for 10 min and then incubated with 1x Apollo solution for 30min
   while protected from light. For subsequent DNA staining, cells were
   incubated with 1x Hoechst 33,342 for 30 min. All the solution used was
   prepared following the manufacturers’ instructions.

2.5. Flow cytometry

   The cell cycle phase distribution was determined using the Cell Cycle
   Analysis Kit (Beyotime, China). Briefly, cells were digested with
   trypsin and subsequently fixed with 70% ethanol at 4 °C for 12 h. The
   fixed cells were then stained with PI solution (prepared according to
   the manufacturers’ instructions) at 37 °C for 30 min, and data was
   acquired using NovoCyte Flow Cytometer Systems (Agilent, USA).

2.6. SA-β-Gal staining

   SA-β-Gal staining was performed using a Senescence β-Galactosidase
   Staining Kit (Beyotime). Cells and frozen sections of mouse ovaries
   were fixed in fixation solution (provided in the kit) for 15min. Next,
   cells and frozen sections were incubated with staining solution
   (prepared according to the manufacturers’ instructions) at 37 °C.
   SA-β-Gal activity was examined after incubation for 48 h.

2.7. Hormone assays

   The levels of hormones in the serum were measured using commercialized
   enzyme-linked immunosorbent assay (ELISA) kits in accordance with the
   manufacturer's instructions. The ELISA kits for estradiol, progesterone
   and follicle-stimulating hormone were purchased from Meimian (China).
   The ELISA kit for luteinizing hormone was from CUSABIO (USA).

2.8. Quantitative real-time PCR (qRT-PCR)

   Total RNA was extracted from COV434 cells using RNA-easy Isolation
   Reagent (Vazyme, China) and reverse-transcribed using HiScript II 1st
   Strand cDNA Synthesis Kit (Vazyme). For measurement of mRNA targets,
   qRT-PCR was performed on a Quantagene q225 real-time PCR system (Kubo,
   China) using Taq Pro Universal SYBR qPCR Master Mix (Vazyme) with
   custom-designed primers. The RT-qPCR conditions were set as follows:
   95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s and 60 °C for
   30 s. The relative expression of candidate mRNAs was normalized to
   β-actin (ACTB) and fold changes were calculated using the 2^−ΔΔCt
   method. All experiments were performed in biological and technical
   triplicates and presented as the means ± SD. The primer sequences are
   listed in [73]Supplementary Table 1.

2.9. Western blot

   Total protein from the COV434 cells and mouse ovaries was extracted
   using RIPA reagent (Solarbio, China) supplemented with protease
   inhibitor cocktails (Abmole, China). After being denatured and
   separated on a 10% polyacrylamide SDS-PAGE gel, proteins were
   transferred to a Polyvinylidene fluoride (PVDF) membrane (Millipore,
   USA). Protein expression was analyzed with indicated primary antibodies
   by using goat anti-rabbit or anti-mouse HRP-conjugated secondary
   antibodies. The protein bands were visualized using a
   electrochemiluminescence (ECL) reagent kit (EpiZyme, China) and imaged
   with Bio-Rad Gel Imaging Systems (Bio-Rad, USA). Antibodies used are
   indicated in [74]Supplementary Table 2.

2.10. Immunofluorescence (IF) staining

   For cultured cells, COV434 cells were fixed with 4% paraformaldehyde
   and permeated with 0.5% Triton X-100 in PBS. For mouse ovarian tissue,
   the paraffin sections were treated with antigen retrieval buffer in
   boiling after dewaxing and rehydration. The cells and sections were
   then blocked with goat serum and incubated with indicated primary
   antibodies and secondary antibodies. For nuclear staining, the cells
   and sections were incubated with fluorescent dye DAPI staining reagent
   (Servicebio, China). Fluorescence imaging was performed using a Zeiss
   Axio Observer 5 fluorescence microscope (Carl Zeiss). Antibodies used
   are indicated in [75]Supplementary Table 2.

2.11. Measurement of ATP levels

   The intracellular ATP levels were measured using an ATP Assay Kit
   (Beyotime, China). Briefly, COV434 cells were lysed with ATP lysis
   buffer then the supernatant was collected after centrifugation
   (12,000 g, 4 °C, 5 min). The working solution was added to the opaque
   96-well plate first to deplete background ATP, after which the
   reference standard and samples were added to the detection well, and
   the relative light unit (RLU) value was measured with a Synergy HTX
   Multi-Mode Reader (BioTek). All the solution used was prepared
   following the manufacturers’ instructions.

2.12. RNA-sequencing

   Total RNA was isolated from control and 3-NP treated COV434 cells by
   using Trizol reagent (Invitrogen, USA) as described by the
   manufacturer. The quality and quantity of RNA samples were assessed
   with NanoDrop 2000 Spectrophotometer (Thermo Fisher, USA). Library
   construction and RNA sequencing were performed by Annaroad Gene
   Technology (Beijing) Co., Ltd. Sequencing reads were aligned to the
   USCS hg38.0 Homo sapiens reference genome using HISAT2 software with
   default settings (859). Fragments per kilobase million (FPKM) was used
   to quantify transcript expression levels. Analysis of differentially
   expressed genes (DEGs) was performed using DESeq2. Significant DEGs
   were identified with a P value < 0.05 and a |log[2](fold change)| >1.
   KEGG enrichment analysis and gene set enrichment analysis (GSEA) were
   performed with clusterProfiler R package (860). The data reported in
   the present study have been deposited in the Gene Expression Omnibus
   (GEO) database under GEO accession number [76]GSE211679.

2.13. RNA co-immunoprecipitation (RIP)

   For RNA co-immunoprecipitations (RIP) cells (1 ×10^7 per condition)
   were cross-linked with 150 mJ/cm^2 using a UV crosslinker (Analytik
   Jena, Germany) and then lysed on ice using RIP lysis buffer (50 mM Tris
   pH 8, 150 mM NaCl, 0.5% sodium deoxycholate, 1% Triton X-100, 5 mM
   MgCl2, 1 mM DTT, 1x proteinase inhibitor cocktail and 40U/mL RNase
   inhibitor). Anti-IGF2BP1-antibody or IgG were incubated with ProteinA/G
   magnetic beads (Beyotime) at 4 °C for 1 h. Cleared lysates were
   incubated with antibody-beads complexes at 4 °C overnight, followed by
   three washing steps with high salt buffer (50 mM Tris pH 8, 1 M NaCl,
   1 mM EDTA, 1% NP-40, 0,5% sodium deoxycholate) and two washing steps
   with wash buffer (20 mM Tris pH 8, 150 mM NaCl, 0.5% NP-40), Protein
   enrichment was analyzed by western blotting. Co-purified RNAs were
   extracted using TRIZOL and analyzed by RT-qPCR.

2.14. Luciferase reporter assay

   The 3′UTR of MDM2 contains putative IGF2BP1 binding sites. Mutated
   3′UTR of MDM2 were synthesized using a QuikChange II XL Site-Directed
   Mutagenesis Kit (Stratagene), and the fragment was amplified by PCR.
   COV434 cells were co-transfected with the 3′UTR luciferase vector,
   pRL-TK and pcDNA3.1-IGF2BP1 using Lipofectamine 3000 (Invitrogen, USA).
   Renilla luciferase was used as internal reference. After 48 h of
   transfection, luciferase activities were measured using a
   Dual-Luciferase Reporter Assay System (Promega) following the
   manufacturer's instructions.

2.15. Statistical analysis

   Statistical analyses were performed using GraphPad Prism 8.0 (version
   8.0c; GraphPad Software, Inc., USA). All of the data were displayed as
   the mean ± SD and statistical significance of equally distributed data
   was tested by an unpaired parametric Student's t-test. Otherwise, a
   non-parametric Mann–Whitney-test was performed. Differences with P
   values < 0.05 were considered statistically significant.

3. Results

3.1. 3-NP caused GCs dysfunction in vitro

   In this study, 3-NP, a mitochondrial toxin inducing OS by irreversibly
   inhibiting the succinate dehydrogenase enzyme in the complex II of
   electron transport chain, was used to simulate OS status of GCs both in
   vitro and in vivo. In agreement with the results acquired from other
   kinds of tissues such as the brain, liver and heart [[77][34],
   [78][35], [79][36]], 3-NP (10 mM) exposure resulted in massive ROS
   accumulation in granulosa cell line COV434 ([80]Fig. 1A). Specifically,
   the level of superoxide anion, one of the major species of ROS, was
   also determined by dihydroethidium fluorescent probe and the result
   showed that the amount of superoxide anion also dramatically increased
   in 3-NP treated cells, suggesting the redox homeostasis was impaired in
   GCs after 3-NP treatment ([81]Fig. 1A). We next tried to investigate
   the further impact of 3-NP on GCs. First, we performed CCK-8 and EdU
   assay to assess cell viability and proliferative ability respectively,
   finding that both cell viability and proliferation were obviously
   inhibited after 3-NP treatment ([82]Fig. 1B and C). Given that previous
   studies have proved that the ROS level varies during the cell cycle
   progression [[83]37] and can lead to cell cycle dysregulation
   [[84]38,[85]39], we next determined the cell cycle phase distribution
   with flow cytometry. As seen in [86]Fig. 1D, 3-NP treatment caused an
   obvious increase in the percentage of G1- or G2/M-phase cells, which
   reflected cell cycle arrest at the G1 and G2/M phase during the cell
   cycle progression. Since ROS are one of the stressors that may lead to
   senescence-associated phenotypes [[87]40], we next evaluated the level
   of cellular senescence in GCs with or without 3-NP stimulation. The
   proportion of senescent cells, as determined by SA-β-Gal staining, was
   evidently increased in the 3-NP group, indicating a aggravated cellular
   senescent level ([88]Fig. 1E). This result was further validated by
   Western blot analysis of several senescence-related markers, p16, p21,
   p53 and Cyclin B1 ([89]Fig. 1F). Collectively, these data demonstrated
   that 3-NP treatment inhibited cell proliferation and exacerbated
   cellular senescence in GCs in vitro.

Fig. 1.

   [90]Fig. 1
   [91]Open in a new tab

   3-NP caused GCs dysfunction in vitro. (A) COV434 cells were treated
   with 10 mM 3-NP for 24h. Superoxide anion and ROS contents in control
   (PBS) or 3-NP treated (3-NP) cells were measured and quantified by
   measuring fluorescent intensity. (B) Cell viability changes after
   treatment with different concentrations of 3-NP were analyzed with
   CCK-8 assay. (C) Cell proliferation was determined by EdU assay and
   quantified by the percentage of positive cells. Representative images
   are shown in the left panel. (D) Cell cycle distribution analyses of
   control or 3-NP treated cells were analyzed with PI-labeled flow
   cytometry. (E) Cellular senescence levels were measured with SA-β-Gal
   staining and quantified by the percentage of positive cells. (F)
   Western blot analysis of indicated proteins in COV434 cells with or
   without 3-NP treatment. β-actin served as loading control. Bar: 50 μm
   *, P < 0.05; **, P < 0.01; ***, P < 0.001.

3.2. 3-NP caused GCs dysfunction and impaired follicular development in
C57BL/6J mice

   To further characterize the impact of 3-NP treatment on GCs, changes of
   mouse ovaries have been investigated in C57BL/6J mice. As shown in
   [92]Fig. 2A and B, intraperitoneal administration of 3-NP did not
   result in significant changes on mouse body weights, but obviously
   decreased ovary coefficients (defined as ovary/body weight ratio),
   which reflected a global detrimental effect of 3-NP on mouse ovaries.
   To determine whether 3-NP administration resulted in a disturbed redox
   balance in vivo, DCFH-DA staining was performed to detect ROS levels of
   mouse ovaries. The intensive fluorescent signals manifested massive ROS
   accumulation in mouse ovaries, especially in GCs, of the 3-NP group
   ([93]Fig. 2C), suggesting 3-NP stimulates an OS status in GCs in vivo
   as well. To further evaluate the 3-NP induced detrimental effect on
   mouse ovaries, we next analyzed the changes of pituitary hormone and
   sex hormone levels as the hormonal regulations within the HPG axis is
   crucial for maintaining normal folliculogenesis. The levels of
   estradiol, progesterone, follicle-stimulating hormone (FSH) and
   luteinizing hormone (LH) in mouse serum samples were measured by ELISA.
   The estradiol level in the 3-NP group was decreased compared to the
   control group whereas the level of progesterone was not obviously
   altered ([94]Figs. S1A and B). Since the synthesis of estradiol is
   largely relied on intact GCs, decreased estradiol levels suggest the
   hormone synthesis function of GCs was impaired by 3-NP administration.
   FSH is a pituitary hormone that can stimulate estradiol secretion and
   is simultaneously modulated by estradiol levels through a feedback
   regulatory mechanism. The level of FSH elevated after 3-NP
   administration ([95]Fig. S1C), which is consistent with the decreased
   level of estradiol. LH, another pituitary hormone promoting follicular
   development, was also modestly increased in mouse serum after 3-NP
   treatment, although this change was not significant ([96]Fig. S1D). We
   next evaluated proliferative ability of GCs with Ki-67 IF staining.
   Comparing to the control group, the proportion of Ki-67 positive cells
   dramatically decreased in 3-NP treated mouse ovaries, indicating the
   proliferative capacity of GCs was impaired ([97]Fig. 2D). GCs
   proliferation plays a pivotal role in follicular development, we then
   counted follicles at different stages in mice. As shown in [98]Fig. 2E,
   the numbers of primordial, primary and small secondary follicles were
   not obviously affected, while the numbers of large secondary and antral
   follicles were significantly decreased in mice received 3-NP treatment,
   which further validated the declined proliferative ability of GCs.
   SA-β-Gal assay was then performed to determine the cellular senescent
   levels in mouse ovaries. As expected, strong SA-β-Gal staining was
   observed in the 3-NP group rather than the control group, illustrating
   an accumulation of senescent cells in ovarian follicles was induced by
   3-NP treatment ([99]Fig. 2F). Protein levels of p16, p21, p53 and
   Cyclin B1 also exhibited consistent changing trends ([100]Fig. 2G),
   further indicating that 3-NP treatment caused aggravated cellular
   senescence in vivo as well. These results suggest that 3-NP exposure
   can also induce GCs dysfunction in vivo.

Fig. 2.

   [101]Fig. 2
   [102]Open in a new tab

   3-NP caused GCs dysfunction and impaired follicular development in
   C57BL/6J mice. (A, B) Body weights and ovary coefficients of C57BL/6J
   mice were recorded after 14 days of 3-NP/PBS intraperitoneal
   administration. (C) ROS levels of ovaries were measured. (D)
   Proliferative capacity of GCs was evaluated by detecting Ki-67, a
   marker of cell proliferation, with immunofluorescent staining. (E)
   Representative images of H&E staining showing ovarian histology and
   follicle morphology at different developmental stages. Follicle counts
   in ovaries of mice from the control or 3-NP group were analyzed. (F)
   Representative images of SA-β-Gal staining indicating senescent cells
   in ovaries. (G) Western blot analysis of indicated proteins in ovarian
   tissue. β-actin served as loading control. Bar: 100 μm ns, not
   statistically significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

3.3. Impact of 3-NP on global transcriptome profiling by RNA-seq

   Aiming to determine associated pathways and regulators involved in the
   process of 3-NP induced GCs dysfunction, we next performed RNA-seq in
   3-NP treated COV434 cells. Comparative transcriptome analysis revealed
   the impact of 3-NP on global transcriptome, with the heatmap to
   visualize gene expression differences between 3-NP and control group
   ([103]Fig. 3A). A total of 28,943 genes have been identified and
   analyzed, and 3538 genes were considered as differentially expressed
   genes (|log[2](fold change)| > 1 and adjusted p-value < 0.05 were set
   as the cut-off values), among which 1964 genes were upregulated and
   1574 genes were downregulated, as shown in the volcano plot ([104]Fig.
   3B). The result of KEGG pathway enrichment analysis was illustrated by
   dot plot ([105]Fig. 3C). Several enriched pathways, such as ‘Cell
   cycle’ and ‘p53 signaling pathway’, validated our previous findings
   that 3-NP treatment caused cell cycle disorder in GCs. In addition,
   GSEA was performed to further characterize transcriptome changes caused
   by 3-NP treatment. The ridge plot visualized expression distributions
   of core enriched genes in the enriched categories ([106]Fig. 3D). Genes
   in several cell cycle associated gene sets, including ‘HALLMARK MITOTIC
   SPINDLE’, ‘HALLMARK G2M CHECKPOINT’ and ‘KEGG CELL CYCLE’, were mainly
   negatively regulated in 3-NP treated cells, whereas genes in the
   category of ‘HALLMARK HYPOXIA’ showed an upregulated trend upon 3-NP
   treatment ([107]Fig. 3E), providing an evidence for the 3-NP induced
   disturbance of redox homeostasis.

Fig. 3.

   [108]Fig. 3
   [109]Open in a new tab

   Impact of 3-NP on global transcriptome profiling by RNA-seq. (A)
   Expression heatmap showing differential RNA-seq feature counts for
   COV434 cells with or without 3-NP treatment. This heatmap was built on
   normalized gene read counts. C = CONTROL; N = 3-NP. (B) Volcano plot
   obtained from DESeq2 analysis showing DEGs (|log2 fold change| ≥ 1 and
   P < 0.01) after 3-NP treatment. (C) KEGG pathway analysis of DEGs.
   GeneRatio is defined as (count of core enrichment genes)/(count of
   pathway genes). (D) The ridge plot visualizing the expression
   distributions of core enriched genes for GSEA enriched Hallmark gene
   sets. (E) GSEA results of the Hallmark pathway ‘MITOTIC SPINDLE’,
   ‘HYPOXIA’, ‘G2M CHECKPOINT’ and KEGG pathway ‘Cell Cycle’ are shown.

3.4. 3-NP treatment downregulates IGF2BP1 in GCs

   We next try to identify the key regulator which is responsible for 3-NP
   induced GCs dysfunction. Having demonstrated that cell cycle
   dysregulation is one of the pivotal mechanisms underlying GCs damages,
   we reasoned that a cell cycle regulating factor might be involved in
   this process. Previous studies have revealed that IGF2BPs, an
   RNA-binding protein family facilitating to stabilize its target RNAs
   [[110][15], [111][16], [112][17], [113][18]], play a key role in
   controlling cell cycle and regulating cancer cell proliferation
   [[114]20,[115][41], [116][42], [117][43]]. IGF2BPs knockdown or
   depletion impaired cell cycle progression and limited tumor growth
   [[118]20,[119]41]. Intriguingly, IGF2BP1, rather than IGF2BP2 and
   IGF2BP3 was observed to be dramatically downregulated in 3-NP treated
   COV434 cells according to our RNA-seq data ([120]Fig. 4A). In addition,
   the result of GSEA demonstrated the gene set ‘HALLMARK E2F TARGETS’ was
   enriched and mainly negatively regulated in 3-NP treated cells
   ([121]Fig. 3D). It has been revealed that IGF2BP1 is a
   post-transcriptional enhancer of E2F-driven gene expression and IGF2BP1
   depletion led to significant downregulation of E2F-driven genes
   [[122]20]. These results are consistent with our data and suggest
   IGF2BP1 may serve as part of the regulatory mechanism underlying
   3-NP-induced GCs dysfunction.

Fig. 4.

   [123]Fig. 4
   [124]Open in a new tab

   3-NP treatment downregulates IGF2BP1 in GCs. (A) Bar plots showing the
   expression levels of IGF2BPs in COV434 cells with or without 3-NP
   treatment. Data were derived from the RNA-seq and quantified by
   normalized FPKM. (B) RT-qPCR verification of RNA-seq results. (C)
   Western blot analysis of IGF2BP1 protein levels in COV434 cells (upper
   panel) and mouse ovaries (bottom panel). (D) Representative images of
   IGF2BP1 IF staining in COV434 cells. Relative quantification was
   performed using fluorescent intensity. (E) Representative images of
   IGF2BP1/Ki-67 double staining in ovarian slides. Bar: 50 μm **,
   P < 0.01; ***, P < 0.001.

   To validate this result, we performed RT-qPCR and Western blot to
   evaluate the mRNA and protein levels of IGF2BP1. It was found that both
   mRNA and protein levels were obviously decreased after 3-NP treatment
   ([125]Fig. 4B and C). In addition, IF staining of IGF2BP1 was performed
   in COV434 cells. The fluorescence signal distribution revealed the
   cytoplasmic localization of IGF2BP1, and the fluorescent intensity was
   obviously decreased in the 3-NP group, further confirming that 3-NP
   treatment inhibited IGF2BP1 expression in GCs ([126]Fig. 4D). Since
   3-NP induced mitochondrial dysfunctions not only lead to increased ROS
   generation, but also to the impairment of ATP production, we next
   investigated whether the IGF2BP1 downregulation in GCs is due to
   excessive ROS or energy depletion. Aside from 3-NP, we used antimycin A
   (an inhibitor of mitochondrial electron transport chain complex III),
   FCCP (a mitochondrial oxidative phosphorylation uncoupler) and H[2]O[2]
   to treat COV434 cells and measured the changes of ROS levels,
   intracellular ATP contents and IGF2BP1 expression levels upon these
   different treatments. As shown in Figure S2A, S2B and S2C, exposure to
   3-NP and H[2]O[2] significantly elevated ROS levels but only moderately
   decreased intracellular ATP production. FCCP treatment resulted in a
   drastic decrease in ATP contents but ROS elevation was relatively mild
   in FCCP-treated cells, whereas antimycin A caused pronounced changes in
   both ROS and ATP levels. IGF2BP1 expression has been inhibited to
   varying degrees by all these four treatments. The differences between
   H[2]O[2] group and 3-NP/Antimycin A group suggest that IGF2BP1
   expression is under regulation of both ROS and ATP levels, whereas the
   difference between the 3-NP and FCCP group indicates that GCs may be
   more sensitive to ROS elevation than to ATP reduction in terms of
   IGF2BP1 expression ([127]Fig. S2D). This hypothesis was further
   validated by the result of pretreatment of NAC, an ROS scavenger,
   partially reversing 3-NP induced IGF2BP1 downregulation ([128]Fig. S2D,
   lane8 vs 7). However, how and to what extent IGF2BP1 expression is
   modulated by those ROS-independent cellular disorders need further
   exploration. In addition, we measured the changes of IGF2BP1 levels
   upon 3-NP exposure in KGN and HEK293T cell lines. Both
   immunofluorescent staining and Western blot demonstrated that IGF2BP1
   expression was decreased after 3-NP treatment in KGN cells ([129]Figs.
   S2E and F) as well as in HEK293T cells, but in a much slighter degree
   ([130]Figs. S2G and H), suggesting that 3-NP induced IGF2BP1
   downregulation is not a GCs-specific response.

   Furthermore, IGF2BP1 downregulation upon 3-NP treatment was also
   observed in vivo. Western blot analysis demonstrated dramatically
   decreased IGF2BP1 protein levels in ovaries of 3-NP administered mice
   ([131]Fig. 4C). IF staining validated this result. As shown in
   [132]Fig. 4E, intensive IGF2BP1 fluorescent signals were detected in
   both GCs and oocytes of the control group. While in 3-NP treated mice,
   the fluorescent intensity in ovaries was generally decreased, and the
   decline in GCs was much more obvious than that in oocytes, suggesting
   3-NP treatment can lead to IGF2BP1 downregulation in GCs both in vitro
   and in vivo.

3.5. IGF2BP1 regulates cell proliferation, cell cycle and cellular senescence
in GCs

   We next investigate the effects of IGF2BP1 in GCs through loss- and
   gain-of-function experiments. IGF2BP1 was silenced or overexpressed by
   using IGF2BP1-directed siRNA or pcDNA3.1-IGF2BP1 plasmids respectively.
   The silencing and overexpression efficiency were validated by RT-qPCR
   and western blotting ([133]Figs. S3A and B). As expected, CCK-8 assay
   indicated that cell viability was significantly impaired by IGF2BP1
   knockdown at 72 h post-transfection ([134]Fig. 5A). Cell proliferative
   capacity was evaluated by EdU assay. The decreased proportion of
   EdU-positive cells validated the inhibited cell proliferation upon
   IGF2BP1 silencing ([135]Fig. 5C). In contrast, ectopically expressed
   IGF2BP1 facilitated GCs proliferation and enhanced GCs viability as
   early as 24 h post-transfection ([136]Fig. 5B, D). We next performed
   flow cytometry to analyze the impact of IGF2BP1 silencing or
   overexpression on cell cycle distribution. As shown in [137]Fig. 5E,
   IGF2BP1 knockdown impaired G1/S transition and increased the proportion
   of cells in G2/M phase, whereas IGF2BP1 overexpression facilitated G1/S
   transition, but did not significantly affected the proportion of G2/M
   cells ([138]Fig. 5F), indicating IGF2BP1 is involved in maintaining
   normal cell cycle progression in GCs. Given that cell cycle arrest,
   which has been observed in IGF2BP1-depleted GCs, is one of the
   hallmarks of senescent phenotypes [[139]44], we next investigated
   whether IGF2BP1 expression affects cellular senescence in GCs. As shown
   in [140]Fig. 5G, IGF2BP1 knockdown obviously increased the percentage
   of SA-β-Gal-positive cells, whereas IGF2BP1 overexpression resulted in
   relatively light staining, but did not significantly affect the
   positive rate ([141]Fig. 5H). We further strengthen this conclusion by
   determining the protein levels of senescence-associated markers. As
   expected, expression of p16, p21 and p53 was enhanced after IGF2BP1
   knockdown and inhibited after IGF2BP1 overexpression while the level of
   Cyclin B1 exhibited an opposite trend in response to IGF2BP1 silencing
   or overexpression ([142]Fig. 5I). These findings suggest that IGF2BP1
   is involved in regulating GCs proliferation, cell cycle and cellular
   senescence. Furthermore, IGF2BP1 knockdown inflicted a 3-NP-simialr
   effect on GCs, indicating IGF2BP1 downregulation may be partially
   responsible for 3-NP induced GCs dysfunction.

Fig. 5.

   [143]Fig. 5
   [144]Open in a new tab

   IGF2BP1 regulates cell proliferation, cell cycle and cellular
   senescence in GCs. (A, B) COV434 cells were transfected with
   control/IGF2BP1-directed siRNA (siCTRL/siIGF2BP1) or
   pcDNA3.1-vector/pcDNA3.1-IGF2BP1 (vector/IGF2BP1). Cell viability
   changes were measured 24/48/72h after transfection. (C, D) Cell
   proliferative capacity was determined by EdU assay and quantified by
   the percentage of positive cells. (E, F) Cell cycle distribution
   changes upon IGF2BP1 knockdown or overexpression were analyzed with
   PI-labeled flow cytometry. (G, H) Cellular senescence levels were
   measured with SA-β-Gal staining and quantified by the percentage of
   positive cells. (I) Representative Western blot analysis of indicated
   proteins in COV434 cells after IGF2BP1 knockdown of overexpression.
   β-actin was used as loading control. a, lane1 vs 2; b, lane3 vs 4. Bar:
   50 μm ns, not statistically significant; *, P < 0.05; **, P < 0.01;
   ***, P < 0.001.

3.6. IGF2BP1 rescues 3-NP induced GCs dysfunction

   To further investigate if the adverse effects of 3-NP on GCs are at
   least partially mediated by IGF2BP1 downregulation, we next examined if
   ectopically expressed IGF2BP1 could restore part of the GCs functions
   hindered by 3-NP treatment. GCs were treated with 3-NP in the presence
   of ectopically expressed IGF2BP1 or vehicle control, after which the
   protein level of IGF2BP1 was validated by Western blot ([145]Fig. 6E).
   As illustrated in [146]Fig. 5A, the result of CCK-8 and EdU assay
   demonstrated that the impaired GCs viability and proliferative capacity
   were partially reversed by IGF2BP1 overexpression ([147]Fig. 6A and B).
   We next examined whether 3-NP induced cell cycle disorders can also be
   corrected by ectopically expressed IGF2BP1. As shown in [148]Fig. 5E,
   the elevated proportion of G1- and G2/M-phase cells was obviously
   deceased in the IGF2BP1 overexpressed cells, indicating IGF2BP1
   restoration facilitated cell cycle progression ([149]Fig. 6C). In
   accordance with these results, the number of senescent cells, which was
   dramatically increased after 3-NP exposure, exhibited an obvious
   downward trend in IGF2BP1-overexpression group ([150]Fig. 6D).
   Furthermore, protein levels of senescent markers validated the result
   of SA-β-GAL assay, with p16, p21, p53 inhibited and Cyclin B1 rebounded
   after IGF2BP1 overexpression ([151]Fig. 6E). These findings suggest
   that IGF2BP1 overexpression partially alleviates 3-NP induced GCs
   dysfunction and IGF2BP1 downregulation meditates part of the 3-NP
   induced GCs damages.

Fig. 6.

   [152]Fig. 6
   [153]Open in a new tab

   IGF2BP1 rescues 3-NP induced GCs dysfunction. (A) COV434 cells were
   transfected with pcDNA3.1-vector or pcDNA3.1-IGF2BP1, 3-NP was added to
   the medium 48h after transfection. Cell viability was measured 24h
   after 3-NP treatment. (B) Cell proliferation was analyzed by EdU assay
   and quantified with the percentage of positive cells. (C) Cell cycle
   distribution changes were analyzed with PI-labeled flow cytometry. (D)
   Senescent cells were indicated with SA-β-Gal staining. (E)
   Representative Western blot analysis of indicated proteins in
   COV434 cells. β-actin served as loading control. a, lane1 vs 2; b,
   lane2 vs 3. Bar: 50 μm *, P < 0.05; **, P < 0.01; ***, P < 0.001.

3.7. IGF2BP1 recognizes and regulates MDM2 in an m^6A-dependent manner

   Having demonstrated that IGF2BP1 is one of the potential effectors
   during 3-NP induced GCs dysfunction, we next aimed to reveal which
   factor is responsible for the detrimental impact meditated by IGF2BP1
   downregulation. Given that IGF2BP1's main and conserved role in
   cancer-derived cells relies on the 3′UTR- and m^6A-dependent regulation
   of mRNA stability [[154]20,[155]45], we reasoned that IGF2BP1's
   involvement in 3-NP induced GCs dysfunction might also be associated
   with its property as an RNA binding protein. We combined differentially
   expressed genes identified by RNA-seq (performed in COV434 cells with
   or without 3-NP treatment) and IGF2BP1-binding transcripts profiled by
   RNA immunoprecipitation sequencing (RIP-seq, [156]GSE90639), 601 genes
   were found in the intersection of these two gene lists ([157]Fig. 7A).
   KEGG pathway enrichment analysis was performed to further screen the
   possible targeted mRNAs. As a result, several pathways such as ‘Cell
   cycle’ and ‘Cellular senescence’, which are highly consistent with GCs
   changes after 3-NP treatment or IGF2BP1 knockdown, were enriched
   ([158]Fig. 7B). Since both the p53 and FoxO signaling pathways are also
   important in regulating cell cycle and cellular senescence, we obtained
   the overlapping elements among these four pathways, eventually MDM2,
   together with two other genes encoding the Cyclin proteins, was
   shortlisted for the possible IGF2BP1 targets ([159]Fig. 7C).

Fig. 7.

   [160]Fig. 7
   [161]Open in a new tab

   IGF2BP1 recognizes and regulates MDM2 mRNA in an m^6A-dependent manner.
   (A) Venn diagram showing the intersection of DEGs from RNA-seq and
   enriched transcripts from IGF2BP1 RIP-seq. (B) KEGG pathway enrichment
   analysis of genes obtained from the intersection. (C) Venn diagram
   showing the intersection of KEGG pathway ‘Cell cycle’, ‘p53 signaling
   pathway’, ‘FoxO signaling pathway’ and ‘Cellular senescence’. (D, E)
   RT-qPCR and Western blot analysis of mRNA and protein levels of MDM2 in
   COV434 cells after 3-NP treatment and IGF2BP1 knockdown/overexpression.
   a, lane1 vs 2; b, lane3 vs 4; (F) RT-qPCR analysis of MDM2 mRNA levels
   0/3/6h after ActD treatment in IGF2BP1 depleted/overexpressed
   COV434 cells. (G) Map of SRAMP-predicted m^6A sites and IGF2BP1-RIP
   peaks on MDM2 mRNA. (H) Co-purification of MDM2 mRNAs with IGF2BP1 in
   wildtype, control- (siCTRL) or METTL3-depleted(siMETTL3) COV434 cells
   was analyzed by RIP using anti-IGF2BP1 antibodies and RT-qPCR analysis.
   (I) RT-qPCR and Western blot analysis of mRNA and protein levels of
   MDM2 in COV434 cells after IGF2BP1 overexpression and/or METTL3
   knockdown. a, lane2 vs 1, P < 0.001; b, lane3 vs 1, P < 0.01. (J)
   Schematic representation showing the reporters harboring wildtype or
   m^6A-enriched-region deleted fragments of SOX2-3′UTR (left panel). Dual
   luciferase reporter analysis of MDM2-3′UTR-driven firefly luciferase
   reporter constructs in COV434 cells with ectopically expressed IGF2BP1
   (right panel). ns, not statistically significant; *, P < 0.05; **,
   P < 0.01; ***, P < 0.001.

   We next examined the mRNA and protein levels of MDM2 after 3-NP
   treatment and IGF2BP1 knockdown/overexpression. As expected, MDM2 was
   downregulated in 3-NP treated and IGF2BP1-knockdowned cells, and
   upregulated in IGF2BP1 overexpression cells ([162]Fig. 7D and E).
   Furthermore, the effect of IGF2BP1 expression alteration on MDM2 mRNA
   stability was investigated by the Actinomycin D (ActD) chase
   experiment. Forty-eight hours after the transfection, 5 μg/mL ActD was
   added to the medium and total RNA was collected 0, 3 and 6h after ActD
   administration. As shown in [163]Fig. 7F, IGF2BP1 knockdown and ectopic
   expression resulted in the impaired and enhanced stability of MDM2 mRNA
   respectively, further validating the role of IGF2BP1 in mRNA
   metabolism. In view of IGF2BP1 had been identified as an m^6A reader
   recognizing and binding transcripts with m^6A modifications, we
   explored whether m^6A modifications are essential for IGF2BP1-MDM2
   interaction in GCs. The m^6A sites of MDM2 mRNA were predicted by SRAMP
   database [[164]46] and further validated by MeRIP-seq ([165]GSE178095),
   while the IGF2BP1-binding sites were determined by RIP-seq
   ([166]GSE90639). It was found that m^6A sites and IGF2BP1-binding sites
   highly overlapped and were both enriched around the stop codon and
   3′UTR of MDM2 mRNA, suggesting IGF2BP1 might recognize and regulate
   MDM2 mRNA in an m^6A-dependent manner ([167]Fig. 7G). We next performed
   IGF2BP1 RIP assay in WT and METTL3-knockdown cells to validate this
   hypothesis. Given that METTL3 serves as a key component of m^6A
   methyltransferase complex, METTL3 knockdown ([168]Fig. S3C) will
   significantly decreases m^6A levels of the whole transcriptome, thus
   hampering the interaction between m^6A readers and their target
   transcripts. As it turned out, the interaction between IGF2BP1 and MDM2
   mRNA, which was evidenced by RIP assay performed in WT cells, was
   obviously impeded by METTL3 knockdown ([169]Fig. 7H). This result was
   further strengthened by changes of the MDM2 protein level in response
   to METTL3 knockdown. As depicted in [170]Fig. 7I, MDM2 expression was
   respectively enhanced and suppressed after IGF2BP1 overexpression
   (lane2 vs 1) and METTL3 knockdown (lane3 vs 1), and the IGF2BP1
   meditated MDM2 upregulation was largely attenuated by METTL3 inhibition
   (lane4 vs 3). We next expected to test whether the presence of the
   m^6A-harboring fragment is required for IGF2BP1 binding to MDM2‐3′UTR.
   Luciferase reporter constructs harboring either wildtype or mutant
   (containing deletion of the m^6A enriched region defined by MeRIP-seq)
   fragment of MDM2‐3′UTR were generated, as shown in [171]Fig. 7J (left
   panel). Overexpression of IGF2BP1 induced a significant increase in
   luciferase activity in cells transfected with WT reporter harboring
   wildtype 3′UTR of MDM2 mRNA, but such elevation was impaired by
   deletion of the m^6A enriched region of the reporter ([172]Fig. 7J,
   right panel). Taken together, these results indicated that the m^6A
   modifications on MDM2‐3′UTR are indispensable for IGF2BP1 recognizing
   and binding to 3′UTR region, thus regulating MDM2 mRNA stability in
   damaged GCs.

3.8. MDM2 rescues GCs dysfunction induced by 3-NP or IGF2BP1 knockdown

   To explore whether MDM2 inhibition partially mediates the GCs
   dysfunction induced by 3-NP treatment and IGF2BP1 downregulation, we
   observed the effect of MDM2 overexpression on GCs functions in damaged
   GCs. The overexpression efficiency was verified by RT-qPCR and Western
   blot ([173]Fig. S3D). The COV434 cells were transfected with pCMV-MDM2
   plasmids either in the presence of IGF2BP1-directed siRNA or followed
   by 3-NP treatment, and the evaluation of cell functions was performed
   as described previously. Both cell viability and cell proliferative
   capacity, which were impaired by 3-NP treatment and IGF2BP1 knockdown,
   have been efficiently rescued after MDM2 overexpression ([174]Fig. 8A,
   B, C, D). As expected, since MDM2 has already been identified as a
   pivotal factor in cell cycle regulation, its overexpression also
   facilitated the cell cycle progression in both 3-NP treated or IGF2BP1
   knockdown cells ([175]Fig. 8E, F, [176]Figs. S4A and B). We next
   examined whether ectopically expressed MDM2 alleviated cellular
   senescence via performing SA-β-Gal assay and measuring protein levels
   of senescence-associated markers, finding that MDM2 supplement
   significantly reduced the number of senescent cells ([177]Fig. 8G, H,
   [178]Figs. S4C and D)and resulted in partial recovery of protein levels
   of senescence-related hallmarks ([179]Fig. 8I). These results further
   validated the hypothesis that IGF2BP1 plays a role in regulating GCs
   function by maintaining the stability of MDM2 mRNA, which uncovers a
   novel molecular regulatory mechanism underlying the OS-related GCs
   dysfunction and subsequent ovarian follicular developmental disorder.

Fig. 8.

   [180]Fig. 8
   [181]Open in a new tab

   MDM2 rescues 3-NP or IGF2BP1-knockdown induced GCs dysfunction. (A, B)
   3-NP treated or IGF2BP1-depleted COV434 cells were transfected with
   pCMV-vector or pCMV-MDM2. Cell viability was measured with CCK-8 assay.
   (C, D) Cell proliferation was analyzed by EdU assay and quantified with
   the percentage of EdU-positive cells. Representative images were shown
   in the left panel. (E, F) Cell cycle phase distribution was determined
   by PI-labeled flow cytometry. (G, H) Cellular senescence was quantified
   by the percentage of SA-β-Gal-positive cells. (I) Representative
   Western blot analysis of indicated proteins in COV434 cells. β-actin
   served as loading control. a, lane1 vs 2; b, lane2 vs 3, c, lane4 vs 5;
   d, lane5 vs 6. Bar: 50 μm *, P < 0.05; **, P < 0.01; ***, P < 0.001.

4. Discussion

   Serving as the supporting somatic cells in follicles, proliferating GCs
   control the progression of folliculogenesis and provide the
   microenvironment required for the acquisition of a meiotically
   competent oocyte [[182]47,[183]48]. GCs dysfunction has been proved to
   be involved in various female reproductive diseases [[184][49],
   [185][50], [186][51]], thus making detrimental impact on female
   fertility. OS is one the culprits of GCs dysfunction. The imbalance
   between oxidants and antioxidants brings about broad-ranging and
   significant effects on cells, leading to impaired proliferation,
   increased apoptosis and abnormal cell cycle progression
   [[187]52,[188]53]. Several studies have revealed the negative
   implications of OS on GCs [[189]13,[190]54,[191]55], the molecular
   mechanisms underlying which, however, still remain unclear.

   Here, we observed the impact of 3-NP, a mitochondrial toxin targeting
   complex II of electron transport chain and inducing various
   mitochondria-related disorders, on GCs both in vitro and in vivo.
   Similar to previous studies, GCs exhibited limited proliferative
   capacity and exacerbated cellular senescence following 3-NP exposure.
   In consequence, 3-NP induced GCs dysfunction resulted in follicular
   development disorder and may be responsible, as least partially, for
   the ovary coefficient reduction in mice. Considering OS is such a
   common phenomenon that participates in kinds of female reproductive
   diseases, whether and how GCs dysfunction regulates the progression of
   these diseases and subsequent fertility loss worth further exploration.

   IGF2BP1, an RNA binding protein enhancing the stability of its target
   transcripts, has been found to be significantly downregulated after
   3-NP treatment and may serve as a pivotal regulator meditating OS
   associated GCs dysfunction. The conserved proliferation-stimulating
   role of IGF2BP1 has been evidenced by massive studies in various kinds
   of cancers. This cancer-promoting effect is largely relied on its
   m^6A-dependent stabilization of several oncogene mRNAs
   [[192]20,[193]56,[194]57]. However, the evidences revealing the role of
   IGF2BP1 in non-cancer cells are much more limited. Manieri et al. found
   that IGF2BP1 functions as a key regulator of colonic injury and repair
   by facilitating the induction of Ptgs2 mRNA in colonic mesenchymal stem
   cells [[195]58]. While in immune thrombocytopenia, IGF2BP1 is involved
   in regulating apoptosis of mesenchymal stem cells [[196]22]. Comparing
   to its cancer-related effect, the role of IGF2BP1 in normal cells has
   been rarely investigated and to be more specific, the participation of
   IGF2BP1 in OS or female reproductive diseases has not been unveiled
   yet. Our studies indicate 3-NP treatment induced dramatic IGF2BP1
   downregulation partially by stimulating ROS overproduction. The
   depletion of IGF2BP1 in GCs impairs cell viability, restrains cell
   proliferation, hinders cell cycle progression and aggravates cellular
   senescence. In contrast, IGF2BP1 overexpression results in the opposite
   effects in GCs and partially rescues 3-NP induced GCs dysfunction.
   These results further validated the pro-proliferative effect of
   IGF2BP1, which revealed another key regulator of OS-associated cell
   damage and broadened the range of cell types where IGF2BP1 plays a role
   in regulating cell cycle. Given that GCs proliferation is required for
   follicular development and oocyte maturation, whether and how much the
   level of IGF2BP1 in GCs varies in female reproductive diseases need
   more investigation. In addition, further evidences are needed to
   understand the specific impact of IGF2BP1 on the whole transcriptome
   scale and other aspects of cell behaviors in GCs.

   Previous studies have demonstrated that IGF2BP1 is an RBP which
   functions by enhancing the stability of its targeted transcripts, so we
   next explored the possible downstream mRNAs that is regulated by
   IGF2BP1 and responsible for the IGF2BP1-mediated GCs dysfunction. We
   combined results of RNA-seq performed in COV434 cells (with or without
   3-NP treatment) and IGF2BP1 RIP-seq, eventually MDM2 was identified as
   the IGF2BP1 target. In accordance with previous studies, IGF2BP1 acts
   as a post-transcriptional enhancer of MDM2 expression, increasing its
   mRNA stability and subsequent protein expression. One of the crucial
   observations of our study is that IGF2BP1-controlled MDM2 expression is
   apparently m^6A-dependent. IGF2BP1 binding sites and m^6A modification
   sites are highly overlapping and peaking around 3′UTR of MDM2 mRNA,
   which is consistent with previously reported m^6A-enriched regions
   [[197]59,[198]60]. Furthermore, the knockdown of METTL3, an essential
   catalytic subunit required for methyltransferase activity,
   significantly repressed IGF2BP1-MDM2 interaction in GCs. This
   emphasizes and further validates the recently reported m^6A reader role
   of IGF2BP1.

   MDM2 is an E3 ligase for p53 and for itself. In cancer cells, it
   inhibits p53 activity in several pathways, thus promoting cell cycle
   progression and uncontrolled cell proliferation [[199]61]. Our results
   indicate that MDM2 is under the m^6A-dependent and post-transcriptional
   control of IGF2BP1 in GCs. MDM2 partially meditates IGF2BP1-directed
   proliferating-stimulating effect. Overexpression of MDM2 facilitates
   cell cycle phase transition and alleviates cellular senescence induced
   by 3-NP treatment and IGF2BP1 knockdown. These findings expand the
   pro-proliferating role of MDM2 and strongly suggest that
   METTL3-catalyzed m^6A modification is a conserved and indispensable
   mechanism for cell cycle progression in both caner and non-cancer
   cells.

   GCs dysfunction is a pathological condition that can be found in a
   variety of female reproductive diseases. Multi-source detrimental
   factors from these diseases inflict damages on GCs, contributing to
   steroidogenesis disorder, folliculogenesis disarrangement and oocyte
   incompetence, ultimately resulting in infertility [[200]31,[201]62].
   Therefore, unveiling the molecular regulatory network meditating GCs
   dysfunction can be helpful in interpreting and further analyzing those
   GCs-related pathological phenomena and clinical manifestations. This
   study has focused on clarifying the role of IGF2BP1 in 3-NP induced GCs
   damages. However, the effect of IGF2BP1 knockdown or overexpression on
   cell viability, proliferation, cell cycle and cellular senescent levels
   was also explored in normal GCs, which partially demonstrated the
   regulatory effect of IGF2BP1 on GCs behaviors and functions, including
   but not limited to the OS condition. Given that fully functional GCs
   are indispensable for normal follicular development, these findings
   provide a novel potential clinical marker for predicting the GCs status
   as well as the follicular developmental potential, which, however,
   requires further researches to reveal the in vivo effect of IGF2BP1 on
   GCs.

5. Conclusion

   In summary, this study elucidated the critical role of IGF2BP1 in
   regulating 3-NP induced GCs damages. We revealed IGF2BP1 is essential
   for maintaining normal cell cycle progression and proliferation in GCs
   by enhancing MDM2 mRNA stability in an m^6A-dependant manner. Our
   present study provides insights into the biological function of IGF2BP1
   in modulating GCs growth and follicular development, which provides
   evidences for taking IGF2BP1 as a novel therapeutic target in tackling
   OS induced ovarian dysfunction.

Funding

   This work was supported by National Natural Science Foundation of China
   (92068120) and The Key R&D Program of Hubei Province (2022BCB024).

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

Footnotes

   ^Appendix A

   Supplementary data to this article can be found online at
   [202]https://doi.org/10.1016/j.redox.2022.102492.

Contributor Information

   Ling Zhang, Email: Zhling312@163.com.

   Huaibiao Li, Email: lihuaibiao@hust.edu.cn.

   Wenpei Xiang, Email: wpxiang2010@hust.edu.cn.

Appendix A. Supplementary data

   The following is the Supplementary data to this article:
   Multimedia component 1
   [203]mmc1.docx^ (3.5MB, docx)

Data availability

   The data reported in the present study have been deposited in the Gene
   Expression Omnibus (GEO) database.

References