Background
Retinol (vitamin A) and its derivatives, collectively known as retinoids, play crucial roles in ovarian development and normal physiological function [
1]. Retinol is not biologically active per se, and within cells can be oxidized to retinal and retinoic acid (RA) by dehydrogenases. Most of the cellular actions of retinoids can be accounted for by the transcriptional regulatory activity of RA through their nuclear receptors, known as RA receptors (RARs) and retinoid X receptors (RXRs), which associate with RA response elements (RAREs) within the promoters of retinoid-responsive genes [
1]. RA in ovarian antral follicles enhanced FSH-mediated ovarian follicular cell differentiation and female fertility, and vitamin A deficiency inhibited oocyte development and decreased ovulated oocytes in mice [
2,
3]. RA also plays a crucial role in both nuclear and cytoplasmic maturation of mouse and bovine oocytes [
4,
5] and can also stimulate steroidogenesis, such as testosterone production in human theca cells and estradiol production in mouse granulosa cells [
1,
6]. In addition, ovarian retinoid levels vary with the estrous cycle [
7], and the concentration of retinol is greater in the follicular fluids of the dominant follicles than that of small follicles [
8,
9]. However, the regulatory mechanisms of ovarian retinoid homeostasis have not yet been fully understood.
The data from our laboratory suggest that FSH enhances retinol uptake, accumulation, and metabolism in the mouse ovary (unpublished data), but the mechanisms remain unclear. Retinol-binding protein 4 (RBP4), which acts as the mediator for the systemic and intercellular transport of retinol, plays an important role in cellular retinol influx, efflux, and exchange [
10]; and seems to play an important role in retinol intercellular transport and accumulation in follicular fluids of the dominant follicles. Evidence shows that the RBP4 immunostaining was observed in the layers of theca and granulosa cells of antral follicles with the most intense staining noted in the cells of large and healthy follicles. Furthermore, the levels of RBP4 and retinol in the fluids of large follicles were higher than those in the fluids of medium or small follicles [
8]. High RBP4 levels are also observed in the serum of women with polycystic ovary syndrome (PCOS) and in the fluids from swine follicular cysts [
11,
12]. Based on these data [
8‐
12], the regulation of
Rbp4 expression during follicular development remains an interesting and important point of study and would provide an explanation for the possible mechanisms involved in changing ovarian retinoid levels during follicular development.
The regulatory mechanisms of follicular development and ovarian function are primarily realized through neuroendocrine activities in the hypothalamus–pituitary–ovary (HPO) axial, although early stage occurs independently of the HPO axis. Follicle-stimulating hormone (FSH) or FSH+ luteinizing hormone (LH), which are released by the pituitary gland, principally control follicular development and ovulation by regulating estradiol (E
2) production and the functions of granulosa and theca cells. FSH and LH exert their actions by activating their membrane receptors (namely FSHR and LHR, G-protein coupled receptors) thereby resulting in an increase in intracellular cyclic AMP (cAMP), a second messenger involved in the transduction of hormonal or growth signals that regulate cell functions such as proliferation, differentiation, and metabolism [
13,
14]. Elevated cAMP activates protein kinase A (PKA), which phosphorylates certain transcription factors and regulates downstream gene transcription. However, it is not until puberty (at approximately 4 weeks of age in mouse) that gonadotropin pulsatility is sufficient to stimulate fully-grown mature follicles and induce ovulation.
In the adult female mouse (after 6–8 weeks old), under the influence of reproductive hormones (FSH, LH and E
2), ovarian activities (follicular growth, ovulation, and corpus luteum formation and regression) alternate periodically during the estrous cycle (averages 4–5 days: proestrus for 9–18 h, estrus for about 12 h, metestrus for 24–48 h, and diestrus for 48–72 h) [
15,
16]. During proestrus, some recruited follicles in the ovary start to grow rapidly under the stimulation of FSH [
17]. During the estrus, the dominant follicles mature and ovulate under the influence of FSH, LH, and estradiol [
17‐
19]. After ovulation, the corpus luteum starts to form under the influence of LH and mice enter metestrus [
20]; and during diestrus, the mature corpus luteum releases progesterone which inhibits LH and FSH release by the anterior pituitary and inhibits follicular development [
17]. Metestrus terminates with the regression of the corpus luteum under the influence of the pulses of prostaglandin (PG
F2α) released from the uterus [
20,
21].
The structural protein high mobility group AT-hook 1 (HMGA1) is important in promoter regulation by uncovering chromatin and facilitating the recruitment of multiple transcription factors [
22]. The
cis-regulatory sequences of the mouse
Rbp4 gene contain several copies of an AT-rich motif, known to be the binding site for HMGA1 proteins. It has also been reported that HMGA1 can bind the
Rbp4 promoter and recruit the transcription factors steroidogenic factor 1 (SF-1) and liver receptor homolog 1 (LRH-1) to form an activation complex and induce
Rbp4 transcription in Hepa1 mouse hepatoma cells [
22].
The present study investigated the expression patterns of Rbp4 during different stages of development and the estrous cycle. We also explored the potential regulatory mechanisms governing ovarian Rbp4 expression by estrogen and related gonadotropins in mouse ovaries and examined the involvement of HMGA1, LRH-1, and SF-1. The results of this study are expected to provide mechanistic evidence for the fluctuations in ovarian retinoid levels during follicular development and estrous cycle.
Methods
Animals
Immature (3 weeks old, the same below) and adult (8 weeks old, the same below) BALB/c mice were obtained from the Medical Department of Jilin University (Changchun, China). The mice were raised in an environment with controlled temperature (22–24 °C) and humidity (60–70%) in a 12 h light/dark cycle and provided with food and water ad libitum. Vaginal smear tests were performed on adult female mice (8 weeks old), and ovaries were collected from the mice at various stages of the estrous cycle. Some of the adult female mice (8 weeks old) were mated with male mice, and the ovaries of newborn female mice were collected at different developmental stages (postnatal 1, 2, 3, 4 and 5 weeks). To avoid the effect of endogenous gonadotrophins and E
2, we choose immature female mice (3 weeks old) to investigate hormonal effects on ovarian
Rbp4 expression. They were injected intraperitoneally with a single dose of FSH (10 IU/mouse; Ningbo Second Hormone Factory, Ningbo, China), LH (10 IU/mouse; Ningbo Second Hormone Factory, Ningbo, China), or 17β-estradiol (1 μg/mouse/day in corn oil; Sigma, USA) [
23]. Mice were anesthetized with isoflurane and sacrificed via cervical dislocation at 24 or 48 h after injection, and ovarian samples used for RNA isolation were rapidly placed in EZNATM RNA safe stabilizer reagent (Omega, USA) and then stored at − 80 °C. All animal studies were conducted in accordance with the protocol approved by the Animal Care and Use Committee of Jilin University.
Vaginal smear tests
To retrieve ovarian samples from adult mice (8 weeks old) during each of the four stages of the estrous cycle, we performed the vaginal smear tests to identify the specific estrous cycle stages. Briefly, we placed the tip of a plastic pipette filled with 40 μl of phosphate buffered saline after (PBS; Hyclone, USA) into the mouse vagina, and gently flushed the vagina 3–5 times. Vaginal contents (cells and vaginal fluids) from each mouse were collected into a PCR tube and then smeared onto a glass slide. The evaluation of the vaginal smear images was performed under a microscope IX71 system (Olympus, Japan). Proestrus was characterized by the mostly nucleated and some cornified epithelial cells in the smear; estrus by cornified squamous epithelial cells without visible nuclei; metestrus by the predominance of leucocytes and a few nucleated epithelial and/or cornified squamous epithelial cells; and diestrus was characterized by a predominance of leukocytes [
24,
25]. The reader is directed to further citations for a full description [
24,
25]. Vaginal smear tests were performed at 08:00 and 20:00 each day. When the stages were identified, the mice were sacrificed via cervical dislocation and their ovaries were collected.
Immunohistochemistry
Intact ovaries collected from immature mice (3 weeks old) that were that were untreated controls or treated with FSH for 48 h were fixed in 4% paraformaldehyde (Boster, Wuhan, China) for 12 h at 4 °C. After washed with PBS and dehydrated, these tissues then were embedded in paraffin wax (Thermo, USA); and paraffin-embedded tissue sections (5 μm) were deparaffinized. After washing with PBS, the sections were treated with 10 mM citric acid buffer (pH 6.0) (Boster, Wuhan, China) for antigen activation for 15 min in boiling water. The sections were washed with PBS and blocked by 5% (
v/v) BSA in PBS (Boster, Wuhan, China). Sections were incubated overnight at 4 °C with mouse monoclonal antibody specific to RBP4 at a 1:200 dilution in PBS (Proteintech, USA). In order to validate the specificity of RBP4 antibody, 5% nonimmune goat serum [
26‐
29] or PBS [
30] was used as negative controls instead of RBP4 antibody. After washed by 0.3% (
v/v) Triton X-100 in PBS, sections were incubated with goat-anti-mouse second antibody (Boster, Wuhan, China) for 1 h at room temperature, stained using ABC kits, and counterstained with hematoxylin.
Follicular granulosa cell isolation and culture in vitro
Primary granulosa cells were isolated from immature female mouse (3 weeks old) ovaries, as described previously [
31,
32]. The mice were sacrificed via cervical dislocation after being anesthetized, and the follicles were isolated with no. 5 fine needles. Follicles were then treated with trypsin (Hyclone, USA) for 1 h and filtered using a 100-μm filter (Life Technologies, USA). The isolated granulosa cells were cultured in Dulbecco’s Modified Eagle Medium/F12 1:1 (Hyclone, USA) supplemented with 10% fetal bovine serum (Hyclone, USA), 1% insulin–transferrin–selenium (Sigma, USA), and 1% antibiotics (100 IU/ml penicillin and 100 μg/ml streptomycin; Hyclone, USA) at 37 °C in an atmosphere of 5% CO
2 in air. Twenty-four hours later, non-adherent cells were removed and adherent cells were treated with FSH (100 IU/L), LH (100 IU/L), 8-Br-cAMP (500 μM; Sigma, USA), or H-89 (10 μM; PKA inhibitor; Beyotime, Beijing, China).
RNA interference
siRNA oligonucleotides specific for mouse Hmga1, Sf-1, and Lrh-1 were designed and synthesized by GenePharma (Shanghai, China). The sequences that provided successful knockdown were Hmga1 siRNA, 5′-GAGTCAGAAAGAGCCCAGT-3; Sf-1 siRNA, 5′- GCCTCGATGTGAAATTCCT-3′; and Lrh-1 siRNA, 5′-GCAGAAGAAAGCCCTCATT-3′. The negative control (NC) siRNA was 5′-TTCTCCGAACGTGTCACGT-3′. Cells were transfected at 50–70% confluency with 50 nM of siRNA duplexes using the FuGENE HD transfection reagent (Roche Applied Science, USA) in accordance with the manufacturer’s instructions. FSH was added to the culture medium 24 h after siRNA transfection and cells were treated for an additional 24 h.
Total RNA extraction and real-time quantitative PCR assay
Total RNA from ovarian tissue and granulosa cells cultured in vitro were extracted using an RNAprep pure Micro Kit (Tiangen, Beijing, China) and reverse transcribed into cDNA using a PrimeScript RT reagent kit (Takara, Japan) according to the manufacturer’s instructions. Real-time PCR was performed on a sequence detection system (Agilent Technologies, USA) using a SYBR Premix Ex TaqII kit (Takara, Japan). The gene-specific primers (forward and reverse, respectively) that we used for real-time quantitative PCR amplification were as follows: Rbp4 (NM_011255.3), 5′-AGTCAAGGAGAACTTCGACAAGG-3′, 5′-CAGAAAACTCAGCGATGATGTTG-3′; Hmga1 (NM_016660.2), 5′-GCAGGAAAAGGATGGGACTG-3′, 5′-AGCAGGGCTTCCAGTCCCAG-3′; Sf-1 (NM_139051.3), 5′-CCAGACCTTTATCTCCATTGTCG-3′, 5′-AGTGTCATCTGGTCAGCCACCT-3′); Lrh-1 (NM_030676.3), 5′-TCATGCTGCCCAAAGTGGAGA-3′, 5′-TGGTTTTGGACAGTTCGCTT-3′. The expression level of mouse β-actin (NM_007393.3), 5′-TCTGGCACCACACCTTCTA-3′, 5′-AGGCATACAGGGACAGCAC-3′, was used as an internal reference. The relative gene expression levels were calculated using the 2−ΔΔCt method. All primers were obtained from Sangon Biotech (Shanghai, China), all experiments were repeated at least three times.
Western blot analyses
Protein samples were obtained by homogenizing whole ovaries and by lysing granulosa cells in lysate buffer (Beyotime, Beijing, China) with a 10 μg/ml protease and phosphatase inhibitor cocktail (Thermo, USA). Samples were then centrifuged at 13,000 g at 4 °C. Tissue and cell extracts were normalized to the sample protein concentration, as determined by a protein assay kit (Beyotime, Beijing, China). Proteins (40 μg) were separated via sodium dodecyl sulfate polyacrylamide gel (12%) electrophoresis and transferred to polyvinylidene difluoride membranes. The membranes were then blocked in Tris-buffered saline with Tween 20 (TBST) containing 5% non-fat instant milk and incubated overnight at 4 °C with mouse monoclonal antibodies specific to RBP4 or GAPDH (Proteintech, USA); and rabbit monoclonal antibodies specific to HMGA1, SF-1, or LRH-1 (Abcam, USA) each at 1:5000 dilution. After washing with TBST, the membranes were incubated with horseradish peroxidase-conjugated secondary antibody (Proteintech, USA) for 1 h. Enhanced chemiluminescence (ECL) detection was performed using an ECL system according to the specifications of the manufacturer (Beyotime, Beijing, China). Protein levels were normalized to GAPDH and quantified via densitometry using a Tanon gel imaging system (Tanon, Shanghai, China).
Statistical analyses
The statistical analyses of the data were conducted via one-way ANOVA followed by Tukey’s multiple-range test. Differences were considered significant at P < 0.05. All the statistical analyses were performed using SPSS 22.0 for Windows (StatSoft, USA).
Discussion
Firstly, the present study presented the dynamic expression patterns for
Rbp4 in mouse ovaries during different stages of development and the estrous cycle. The expression of ovarian
Rbp4 remained constant at 1 to 3 weeks postnatally, notably increased at 4 weeks (i.e., peripubertally), and dropped at 5 weeks. Numerous studies have supported the concept that the gonadotropin-releasing hormone (GnRH) neuronal network generates pulse and surge modes of gonadotropin secretion that are critical for puberty [
34], and stimulate the secretion of a mass of local ovarian factors and follicular growth in females. This modulation in
Rbp4 expression may be related to the pituitary gonadotropin wave at the onset of puberty. After puberty, some follicles in ovary start to grow and ovulate irregularly under the influence of serum gonadotropins, and at 5 weeks of age, some mice may be in the follicular-growth phase and some in the post ovulatory period. The different levels of gonadotrophins in these mice may constitute a possible explanation for the drop in
Rbp4 expression at 5 weeks. Furthermore, the expression of
Rbp4 in the ovaries of adult mice (8 weeks old) having normal cycles was found to increase at proestrus and peaked at estrus, the phase with higher serum levels of FSH and LH. In a previous study, higher RBP4 protein concentrations were observed in the follicular fluids of dominant follicles relative to small follicles [
8,
9]. The dynamic expression patterns of
Rbp4 in mouse ovaries during different stages of development and the estrous cycle motivated us to pose an alternative hypothesis: that the expression of ovarian RBP4 is affected by gonadotropins and gonadal steroids and is related to follicular growth.
We then investigated the potential regulatory effects of gonadotropins and estrogen on the dynamic expression of
Rbp4 in the mouse ovary. The expression of
Rbp4 was significantly stimulated by administration of FSH and combined FSH + LH, but unaffected by administration of LH or 17β-estradiol. In addition, intense immunostaining was observed in granulosa cell layers of large antral follicles in the ovaries of mice treated with FSH. To determine whether the induction of
Rbp4 by FSH and FSH + LH occurred due to increased expression of
Rbp4 in granulosa cells, we examined primary granulosa cells cultured in vitro. Our results showed that the expression of
Rbp4 was induced by FSH and combined FSH + LH treatments, and that the induction by combined FSH + LH treatment was stronger than that by FSH alone. The administration of LH alone did not exert an effect. Because the primary granulosa cells were isolated from the immature female mouse (3 weeks old) ovaries without FSH or equine chorionic gonadotropin (eCG) pretreatment and with no or low LHR expression, the cells were mainly immature granulosa cells without LHR expression [
35]. This may be the reason as to why LH alone manifested no obvious effects on granulosa cells and immature ovaries in the present study. However, LHR in granulosa cells can be induced by FSH [
35], and during late maturation of granulosa cells in antral follicles, LH exerts a relatively more notable effect than FSH on cAMP formation [
14,
36,
37], suggesting that LHR density was relatively greater than FSHR density or that LHR was more effectively coupled to cAMP generation. These findings may explain the stronger induction of
Rbp4 expression by FSH + LH than by FSH alone, and the higher levels of RBP4 in the follicular fluids of the large antral follicles [
8].
To determine whether the induction of Rbp4 by FSH was mediated by the cAMP-PKA pathway, we examined the effects of the cAMP analog 8-Br-cAMP and the PKA inhibitor H-89 on the induction of Rbp4 in granulosa cells. The treatment of cells with 500 μM 8-Br-cAMP showed a dramatic increase in Rbp4 expression, whereas 10 μM H-89 significantly prevented the induction of Rbp4 by FSH. Thus, the cAMP-PKA pathway likely participates in the induction of Rbp4 by FSH.
The cis-regulatory sequences of the mouse
Rbp4 gene contain a bipartite promoter, i.e., a proximal region necessary for basal expression and a distal segment that contains several binding sites for the structural HMGA1 proteins that are structural components of chromatin [
22,
38]. HMGA1 binds to the adenine thymine (AT)-rich regions of DNA, changes the structure of DNA and recruit transcription factors to the promoter, and facilitates gene transcription [
39]. There are several AT-rich motifs homologous to the binding sites of HMGA1 in the promoter region of mouse
Rbp4, and HMGA1 proteins can bind upstream sequences of the
Rbp4 promoter. In the present study,
Hmga1 was sensitive to FSH and this response was prevented by the PKA inhibitor H-89; whereas the knockdown of
Hmga1 with siRNA resulted in a dramatic loss of FSH-induced
Rbp4 expression. In addition, HMGA1 can interact with SF-1 and LRH-1 and recruit them to the complex [
22]. In the present study, the depletion of SF-1 and LRH-1 resulted in a dramatic loss of FSH-induced
Rbp4 expression. SF-1 was shown to be expressed at higher levels in theca cells, and also found in granulosa cells; whereas LRH-1 expression was found to be abundant and highly restricted to the granulosa cells of developing follicles [
40,
41]. Consistent with our data, the two transcription factors were shown to be induced in granulosa cells by FSH/cAMP through the activation of PKA [
40,
41], as cAMP signals are mediated largely via PKA [
33] and can be elevated in granulosa cells by FSH [
42,
43] or LH [
44]. Thus, the expression of ovarian
Rbp4 was likely regulated by FSH via the cAMP-PKA pathway, involving HMGA1, SF-1, and LRH-1.