Background
Follicle-stimulating hormone (FSH) is a pleiotropic hormone produced by the pituitary and exerts its action on the ovary by inducing proliferation, differentiation and steroidogenesis in the granulosa cells of growing pre-ovulatory follicles. The initial development of primordial follicles (PF) is believed to be FSH independent
[
1]. However, several reports have shown that appreciable amount of FSH is present in serum of perinatal mice
[
2], rats
[
3], hamsters
[
4,
5] and humans
[
6]. Elevated FSH levels are found to be associated with altered ovarian function including diminished ovarian reserve, pre-mature ovarian failure, infertility, ovarian hyper-stimulation syndrome, poor quality of eggs, menopause
[
7,
8] and also a huge body of literature exists suggesting an association with ovarian cancers
[
9,
10]. At the same time, FSH acts as a survival factor during PF transition to primary/secondary follicles in serum-free ovarian cortical tissue culture
[
11]. Studies in hamster using anti-FSH antibodies have shown that FSH action is also required for formation of primordial follicles
[
12]. They showed that a single dose of FSH specific polyclonal antibody during gestational period could significantly affect the number of primordial follicles after birth. Demeestere et al.
[
13] recently concluded that FSH possibly coordinates development of both germline and somatic compartments of the mouse follicle. FSH is also known to exert similar dual effect on both the germline and somatic cells in the testis. Rao et al.
[
14] reported that by blocking the action of FSH using phage-expressed extracellular domain of FSHR results in azoospermia after 100 days in non-human primates although FSHR are localized only on the surface of Sertoli cells. It remains rather unclear as to how FSH can exert action on both somatic and germline compartment of the ovarian follicles and in the testis, although FSHR are expressed only on the somatic cells namely granulosa cells of developing follicles and on Sertoli cells in the testis. Although indirect action on germ cells through paracrine-autocrine interactions is suggested, mechanism underlying FSH action on the mammalian gonads needs further investigation.
Several published reports using ovarian tissue or immortalized ovarian epithelial cell lines show that besides granulosa cells, FSH receptors (FSHR) are also localized on the normal ovary surface epithelium
[
15‐
17] and in ovarian tumor surface epithelium
[
9,
18‐
26]. Oocytes have also been reported to express FSHR. FSHR mRNA was detected in the oocytes and cleavage stage mouse embryos
[
27]. FSHR is reported to be 20 fold higher in the human oocytes compared to granulosa cells
[
28]. Various techniques like
in situ hybridization, RT-PCR, Q- PCR and immunolocalization studies have been used to demonstrate FSHR in the OSE of mice, rats, cows and humans
[
9,
15,
17,
19]. Few studies have also shown that OSE undergoes rapid proliferation in response to PMSG
[
29‐
31], based on BrdU uptake or PCNA staining. All these studies suggest that FSH acting through FSHR localized in OSE may have an important role in ovarian biology. It is imperative to mention here that these published reports assume greater significance, since recently it has been reported by several groups
[
32‐
38] that the OSE also houses ovarian stem cells which may contribute to postnatal oogenesis. Hence, PMSG induced FSH action in adult mouse can serve as a good model to understand the probable mechanism of FSH action on stem cells as well as in new follicular development. Our group earlier reported that OSE in mice, rabbits, sheep, monkeys and women harbors two distinct stem cell populations that include the pluripotent very small embryonic-like stem cells (VSELs) and their immediate descendants ‘progenitors’ ovarian germ stem cells (OGSCs)
[
32,
39]. We have further implicated these stem cells in postnatal oogenesis and proposed a working model for the same
[
39].
The concept of postnatal oogenesis in mammalian ovary is a highly debated area of research and several investigators have contributed both in favor
[
36,
39‐
42] and against
[
43,
44] in the recent past. Tilly’s and co-workers
[
36] were the initial group to generate evidence against the concept of fixed number of eggs at the time of birth and their recent paper
[
45] provides unequivocal evidence to support the presence of mitotically active germ cells that can generate oocytes both
in vitro and
in vivo. The existing controversy regarding existence of stem cells in adult mammalian ovary appears to be more of technical nature
[
46] or markers
[
47] used to study the stem cells rather than absence of stem cells itself. Recently, Gamwell and co-workers
[
48] recently reported a side population in mouse OSE that excludes Hoechst dye and express Sca-1 along with Nanog and Kit.
The present study was undertaken to study the effect of PMSG on mouse ovary with a focus on stem cells located in the OSE. Besides normal estrus cycle, ovaries were studied 2 and 7 days after PMSG treatment. Normally 2 days are required to stimulate maturation of follicles in a standard super-ovulation protocol using PMSG and hence 2 days was chosen to study the standard effect. A longer time period of 7 days was chosen to understand delayed effect and to get insights into the effect of FSH on ovarian stem cells. What is the effect of FSH when the ovary is super-ovulated during assisted reproduction, how does FSH treatment affect the OSE, does it have any effect on the stem cells and whether postnatal oogenesis and PF assembly is affected by FSH treatment are open questions that remain to be answered.
Materials & methods
The study was approved by Institute Animal Ethics Committee and was carried out using in-house bred Swiss mice in the experimental animal facility at our Institute. Mice were housed in a temperature and humidity controlled room on a 12hr light/12hr darkness cycle with free access to food and water.
Different stages of estrus cycle were monitored by daily examination of vaginal smears of one month old female mice. Briefly, the vaginal swabs were prepared using sterile PBS dipped cotton swabs, fixed in 100% methanol, stained with Giemsa and observed under a 90i microscope (Nikon, Tokyo, Japan). Ovaries were collected during different stages of estrus cycle namely estrus, proestrus and diestrus (DE). Ovaries were also collected from mice treated with subcutaneous injection of 5IU PMSG (National Hormone & Peptide Program, Harbor-UCLA Medical Center, California, USA) after 2 days (2D) and 7 days (7D) of treatment. All PMSG injections were done in normal cycling mice at DE stage and mice at DE stage were used as controls for the experiments. The experimental design and number of animals used in the present study are shown in Table
1. The collected ovaries were fixed in 4% formalin for histology and immunolocalization studies and also in TRIZOL for RNA extraction to analyze various gene transcripts of interest by RT-PCR and Q-PCR. The ovaries were carefully dissected from all the adhering extraneous tissue under a stereomicroscope with the help of a micro scissor prior to freezing for RNA.
Table 1
Experimental design of the study
Estrous | 2 | 4 | 4 | - |
Pro-estrus | 1 | 2 | 2 | - |
Di-estrus | 10*
| 20 | 8 | 12 |
2D PMSG | 12*
| 24 | 14 | 10 |
7D PMSG | 12*
| 24 | 14 | 10 |
Histological studies
Formalin fixed ovaries were processed and embedded in paraffin using standard protocols. 5μm thick sections of the embedded ovaries were prepared and stained with Hematoxylin and Eosin for studying the histo- architecture during different phases of estrus cycle and after treatment with PMSG. The representative areas were photographed using 90i microscope and data recorded.
Immunohistochemistry (IHC)
Immunolocalization studies were carried out for various antigens listed in Table
2. Briefly, the paraffin sections were deparaffinized, hydrated and antigen retrieval was performed with boiling SSC buffer of pH 6 in microwave for 5 min (cytoplasmic antigens) or 15 min (nuclear antigens). After cooling, the slides were washed with Tris buffer saline (TBS) for 5 min and processed further for blocking with 10% NGS and 1% BSA in TBS. After removing excess blocking reagent, the slides were incubated with primary antibody diluted with 1% BSA in TBS for two hrs at room temperature. For negative control, the antibody was omitted and sections were incubated with 1% BSA in TBS. In case of FSHR detection alone, negative control included the pre-immune sera from rabbit in which the primary antibody was raised. The detection was done using anti mouse Vecta ABC kit (Vector Laboratories, USA) according to manufacturer’s instructions. Final color development was done using DAB (Biogenex, USA). After counterstaining with Haematoxylin, the slides were observed under 90i microscope. Representative areas were photographed.
Table 2
Details of antibodies used for immunolocalization studies
FSHR Polyclonal | FSHR: membrane bound receptor of FSH | Gift | 1:100 | To study cells responsive to PMSG | |
PCNA Monoclonal | PCNA: nuclear protein that functions during DNA replication | Sigma | 1:5000 | To study proliferating cells | |
OCT-4 Polyclonal | OCT-4 isoforms: OCT-4A is a nuclear transcription factor and OCT-4B is cytoplasmic | Abcam | 1:100 | To study pluripotent and progenitor stem cells | |
SCP3 Polyclonal | SCP3: a nuclear protein required for homologous recombination during meiosis I | Abcam | 1:500 | To study presence of cells undergoing meiosis | |
VASA Polyclonal | MVH: germ cell specific cytoplasmic protein expressed from primordial germ cell stage | R&D System | 1:500 | To study early germ cells and oocytes | |
RNA isolation and cDNA synthesis
Total RNA was extracted by standard protocol using Trizol (Invitrogen, Carlsbad- CA, USA) and treated with DNase I (Amersham Biosciences, Piscataway, NJ) to remove any genomic DNA present. First-strand cDNA was synthesized using the iScript cDNA synthesis kit (Bio-Rad, USA) according to the manufacturer’s instructions. Briefly, 1μg of total RNA was incubated with 5x reaction mix and reverse transcriptase mix. The reaction was carried out in G-STORM thermocycler (Gene Technologies, Braintree, UK). The reaction mix was first incubated at 25°C for 5 min, then at 42°C for 30 min and finally at 85°C for 5 min.
Both RT-PCR and Q-PCR were employed to study gene transcripts expression for the following markers:
pluripotent stem cell specific markers (
Oct-4A: isoform A of octamer binding transcription factor 4
[
53,
54] and
Nanog: homeobox transcription factor also expressed in primordial germ cells
[
55]);
early primordial germ cell markers (
Oct-4, represented here as total Oct-4 as it includes all isoforms of Oct-4 including isoform B expressed in cytoplasm of OGSCs
[
32]),
Stella
[
56],
Fragilis
[
57],
Mvh mouse vasa homolog, plays a role in oogenesis, germ cell and follicle development);
oocyte specific markers (
Nobox newborn ovary homeobox-encoding gene is essential for folliculogenesis
[
58],
HoxA10 homoeobox gene restricted to primordial and early primary follicular oocytes
[
58]);
pre-meiotic marker (
Stra8 stimulated by retinoic acid gene 8 essential for meiotic initiation
[
59]);
meiotic marker (
Scp3 synaptonemal complex protein 3 and
Dmc1 required for, meiosis specific homologous recombination) and PMSG action mediator (
Fshr). Each sample was obtained by pooling ovaries from two different animals.
Reverse –transcription PCR
RT-PCR was carried out to study the expression of meiotic markers (Stra 8, Scp3 and Dmc1) using Dream Taq DNA polymerase (Fermentas Life Sciences; Vilnius, Lithuania). Briefly, the cDNA mix (2 μl) was amplified using 0.2mM of each primer (Table
3), 1.25 unit of DreamTaq DNA polymerase (Fermentas) in 1x Dream Taq buffer (Fermentas) and 0.2mM dNTPs in a G-STORM thermocycler. Amplification was carried out for 35 cycles, with each cycle consisting of denaturation at 94°C for 30 sec, annealing at the specified temperature (Table
3) for 20 sec, and extension at 72°C for 30 sec. The products were analyzed on 2% agarose gel stained with 0.5 μg/ml ethidium bromide (Bangalore Genei, Bangalore, India). The product size was approximated using a 100-bp DNA ladder (Bangalore Genei). The negative control did not include cDNA in the reaction mixture. Normal adult mouse testis was used as positive control.
Table 3
Primers used for RT- PCR and Q-PCR
Fshr
| F: TGGAGGCGGCAAACCTCTGAAC | 65 |
R: TCTGGCTTTGGCGAGCAGGTC |
Oct-4A
| F: CCATGTCCGCCCGCATACGA | 61 |
R: GGGCTTTCATGTCCTGGGACTCCT |
Total | F: CCTGGGCGTTCTCTTTGGAAAGGTG | 68 |
Oct-4 | R: GCCTGCACCAGGGTCTCCGA |
| F: CAGGAGTTTGAGGGTAGCTC | 61 |
R: CGGTTCATCATGGTACAGTC |
Stella | F: ACGCTTTGGATGATACAGACGTCC | 67 |
R: GCGCTTTGAACTTCCCTCCGGA |
Fragilis | F: GGGGTGACTGAGCTGGGGGAA | 65 |
R: TGTCCCTAGACTTCACAGAGTAGGC |
Nobox | F: AGGGTGCTGAGAGGGTGGCAG | 66 |
R: GGCGATACTAGTGCCCCAGGAC |
HoxA10 | F:CACTTGTCCGGCACCCCTTCG | 65 |
R: TTCGCCTTTGGAACTGCCTTGAC |
Mvh | F: AGTGGAAGTGGTCGAGGTGGT | 61 |
R: TGCCGGTGGTGCATCATGTCC |
Scp3 | F: TGTTGCAGCAGTGGGAACTGGAT | 68 |
R: CCATCTCTTGCTGCTGAGTTTCCA |
Dmc-1 | F: TGGGGAATTGGCTGAACGCC | 68 |
R: CAGGCATCTCGGGGCTGTCATAA |
Stra8 | F: GTTTCCTGCGTGTTCCACAAG | 61 |
R: CACCCGAGGCTCAAGCTTC |
18s | F:GGAGAGGGAGCCTGAGAAAC | 61 |
R: CCTCCAATGGATCCTCGTTA |
| F: GTCCCGTAGACAAAATGGTGA | 58 |
| R: TGCATTGCTGACAATCTTGAG | |
Quantitative PCR
Q-PCR was carried out to study the expression of various gene transcripts mentioned in Table
3. The expression levels of these gene transcripts in relation to housekeeping gene transcript 18s were estimated by CFX96 Real-time PCR system (Bio-Rad Laboratories, Hercules, CA) using SYBR Green chemistry (Bio-Rad).
The primers and their respective annealing temperatures used in the study are mentioned in Table
3. All the primers had efficiency close to 100% and all primers except Gapdh and Nanog are exon-exon spanning primers. The amplification conditions included initial denaturation at 94°C for 3 min followed by 40 cycles comprising of denaturation at 94°C for 30 secs, primer annealing for 20 secs and extension at 72°C for 30 secs. The final step included incubation at 94°C for 20 secs to remove any secondary structures followed by melt curve analysis. The fluorescence emitted at each cycle was collected during the extension step of each cycle. The homogeneity of the PCR amplicons was verified by running the products on 2% agarose gels and also by studying the melt curve. All PCR amplifications were carried out in duplicate. Mean Ct values generated in each experiment using the CFX Manager software (Bio-Rad) were used to calculate the mRNA expression levels. Since ΔCt is inversely proportional to relative mRNA expression levels, the levels were calculated manually by the ΔCt method.
Discussion
Present study demonstrates that pluripotent stem cells VSELs and progenitors OGSCs reported earlier by our group in adult mice, rabbit, sheep, monkey and human ovary
[
32,
39] are possibly regulated by FSH and are responsible for postnatal oogenesis and follicular assembly during adult reproductive life. Normal adult mice ovaries were studied during different stages of estrus cycle and after 2 and 7 days of PMSG treatment by histology, immunolocalization (FSHR, PCNA, OCT-4, SCP3 and MVH), and RT-PCR/ Q-PCR for various gene transcripts including pluripotent (Oct-4A, Nanog); early germ cell (total Oct-4, Stella, Fragilis, Mvh), oocyte-specific (Nobox & HoxA10); pre-meiosis (Stra8) and meiosis (Scp3, Dmc1) specific markers. The process of ovarian stem cell differentiation and meiosis giving rise to new oocytes, which get enclosed by the granulosa cells to assemble as PFs in the OSE appears to be a normal feature of adult ovary (Figure
1
2). The granulosa cells arise by epithelial-mesenchymal transition of OSE as suggested earlier by us based on
in vitro evidence
[
32]. Present study provides
in vivo evidence in favor of our earlier observations that OSE cells may be involved in the formation of granulosa cells (Figure
6B,C, E & F). PMSG treatment further augments the process of postnatal oogenesis and PF assembly resulting in increased PFs by D7 (Figure
3,
4,
5,
6).
PMSG is known to exert its actions on mammalian ovary through FSH-FSHR interaction. In the present study, besides granulosa cells of bigger follicles, FSHR was detected in both OSE and oocytes of PF in agreement with earlier reports
[
15‐
17,
19,
27,
28] and an up-regulation was observed after PMSG treatment at both protein by immunohistochemistry (Figure
7) and transcript (Figure
10) level. Presence of FSHR in locations like oocytes and OSE suggests that the various diverse functions of FSH are probably not exerted only through the classical FSHR present on the granulosa cells of growing follicles. Sairam’s group has earlier reported that FSH may exert diverse actions through alternatively spliced FSHR isoforms based on their studies from sheep and mice ovaries
[
62‐
64].
In the current study, upregulation of FSHR after PMSG treatment correlated well to the proliferation observed in OSE. OSE proliferation and multilayer appearance which were normally observed in control ovaries at few focal areas were increased after PMSG treatment (Figure
3 &
4). Ovarian sections on 2D of PMSG treatment showed the presence hyperplastic epithelial cells and small protuberances bulging from the ovarian surface (Figure
3) which subside by 7D (Figure
4,
5,
6). PF cohorts normally observed with hyper-proliferative OSE regions in control ovaries, increased significantly after 7D of PMSG treatment in the ovarian cortex (Figure
4 &
5). This correlated well with increased mRNA expression for HoxA10- a primordial follicular oocyte marker
[
58] and Nobox that is expressed abundantly in primordial and primary follicular oocytes
[
58] post PMSG treatment. Occurrence of meiosis was supported by expression of pre-meiotic (Stra8) and meiotic (Scp3 and Dmc 1) markers. All the three markers were detected in control as well as all PMSG treated ovarian samples. A distinct up regulation of Scp3 was noted after PMSG treatment, both at the protein (Figure
8 H-J) and mRNA transcript level (Figure
9 &
10). As reported earlier
[
65], SCP3 staining was observed throughout the nuclei suggesting that the oocytes are in prophase I of meiosis and was gradually lost in more developed oocytes. The PF at birth are generally arrested in diplotene stage of prophase 1 of meiosis. SCP3 expression is restricted to earlier stages i.e. zygotene and pachytene stage. Thus expression of Scp3 and pre-meiotic marker Stra8 in adult ovary are considered direct evidence in support of postnatal oogenesis and have been reported earlier by other groups also
[
36,
66].
It is intriguing to point out that OSE which expresses FSHR also lodges VSELs as observed in rabbit, sheep, monkey, human
[
32] and also mouse ovaries
[
39]. The VSELs with nuclear OCT-4 differentiate into ‘progenitors’ ovarian germ stem cells (OGSCs) with cytoplasmic OCT-4 and undergo spontaneous differentiation into oocyte-like structures
in vitro
[
32,
37]. Possibly, in the present study, the pluripotent VSELs were activated by PMSG treatment (increased expression of Oct-4A, Nanog Figure
8 &
10), undergo proliferation (increased PCNA staining and Oct-4A expression, Figure
8) and differentiation (increased expression of stella, fragilis, total Oct-4, Vasa and MVH Figure
8 &
10) resulting in meiosis (Stra-8, Scp3 and Dmc1 Figure
8,
9,
10) and formation of oocyte nests with prominent intercellular cytoplasmic bridges that assemble to form PF in the OSE (Figure
2 &
6). These nests later shift to the cortex and are visualized as small cohorts on 7D after PMSG treatment. This data demonstrates the mechanism of oogenesis occurring
in situ in adult ovary for the first time in OSE similar to well-studied spermatogenesis in the testicular seminiferous tubules where the VSELs are lodged at the basement membrane
[
67,
68]. The process is regulated by FSH and is in contradiction to the existing paradigm that initial PF growth is gonadotropin independent
[
69,
70]. The results of the present study support the model for postnatal oogenesis proposed by us earlier
[
39].
Messinis et al.
[
70] have reviewed a huge body of literature and summarized the current understanding of the role played by FSH during folliculogenesis. According to existing understanding, the ‘intercycle peak’ of FSH is crucial and responsible for ‘cyclic recruitment’ of follicles. Ovary experiences waves of folliculogenesis and a cohort of follicles start growing in each cycle, however, only one becomes dominant and others undergo atresia
[
69]. On treating with gonadotropins, FSH window gets widened and multiple follicles get selected and start to grow. Whether more follicles get recruited due to the treatment or FSH prevents atresia in the already recruited cohort of PF is still controversial. Results of the present study add another dimension to this by demonstrating that FSH also exerts action on the OSE and induces neo-oogenesis and PF assembly from the stem cells residing in the OSE.
The prevailing notion of extrusion of atretic oocytes from the ovary surface
[
71‐
73] or that the PF may get pushed to the ovarian surface because of the growing follicles in response to the PMSG treatment (and thus increased number of PF are visualized after treatment in the present study) may not be valid since this reasoning cannot account for distinct increased pattern of pluripotent and meiotic markers in the developing follicles (Figure
7,
8,
9) observed in the present study. Also the earlier published reports of BrdU incorporation in OSE
[
34,
36] may not be due to mtDNA synthesis – rather it may be due to stem cell activity that gets augmented by PMSG.
The extensive germ cell nest formation, initiation of meiosis, PF assembly etc. are well documented and studied during fetal ovarian development in the existing literature
[
60]. It is interesting to point out that the stage of fetal development when PF assembly occurs is actually associated with several folds higher levels of FSH in females
[
2‐
6,
74]. Thus FSH is directly related to PF assembly and regular intercycle peak of FSH may be stimulating the VSELs to result in differentiation and PF assembly. Lei et al.
[
74] reported that the high levels of circulatory FSH in perinatal period and increased FSHR mRNA in mouse ovaries facilitate germline nest breakdown and PF assembly.
Conclusions
We propose that PMSG treatment through induction of FSH-FSHR action possibly stimulates the pluripotent VSELs residing in the OSE leading to proliferation and differentiation of OGSCs to oocytes and primordial follicle assembly. These results are in contradiction to the existing paradigm that ovaries have fixed number of follicles in adulthood and that early PF growth is independent of FSH.
The results of the present study are in direct agreement with postnatal oogenesis proposed by Tilly’s group in 2004
[
36]. We propose that intercycle FSH peak is probably associated with cyclic recruitment of VSELs in the OSE throughout reproductive life and results in neo-oogenesis and PF assembly. Several fold elevated FSH levels in female fetus during peri-natal period
[
2‐
6] are possibly responsible for increased PF assembly well documented during fetal development. Moreover, since VSELs exist in various body organs, it would be interesting to study whether only ovarian VSELs are selectively stimulated by FSH or even VSELs in testis
[
67,
68] or bone marrow
[
75] etc. are also regulated by FSH. It is indeed intriguing that despite exploiting the ovaries for so many years as a source of eggs, present study has provided newer information towards the mechanism of action of FSH on the ovaries. Elevated FSHR3 (an isoform of FSHR) in OSE has been implicated in ovarian cancers
[
25]. It is highly likely that the VSELs residing in the OSE may also be implicated in the ovarian cancers as has been proposed earlier
[
76,
77].
Competing interests
Authors declare no competing interests.
Authors’ contributions
DB was responsible for conceptualizing the study, data interpretation and manuscript preparation. KS was responsible for RT PCR and Q PCR analysis, data compilation and manuscript preparation. PG was responsible for immunolocalization studies and data compilation. HM was responsible for histology studies. All read and approve the final manuscript.