Effect of resistin on estradiol and progesterone secretion from human luteinized granulosa cells in culture

ABSTRACT

Information on the role of resistin on steroidogenesis is limited to animal studies. The aim of this study was to investigate the effect of various doses of resistin on estradiol and progesterone secretion from human luteinized granulosa cells in culture. Granulosa cells were obtained from follicular fluid aspirated from 50 women undergoing in vitro fertilization (IVF) treatment. The cells were cultured for 48 h after a 24 h pre-incubation period. The effect of resistin at dosages 1, 10 and 100 ng/ml alone or in combinations with FSH (10 and 100 ng/ml) on steroidogenesis was investigated. Estradiol and progesterone were measured by radioimmunoassays in culture supernatants at 24 h and 48 h. FSH treatment increased both estradiol and progesterone secretion. Resistin suppressed basal estradiol (at 1 ng/ml) and progesterone secretion (at all concentrations tested). When resistin (all concentrations) was combined with FSH (100 ng/ml), it eliminated the stimulatory effect of FSH on the secretion of estradiol and progesterone. This study indicates an inhibitory effect of resistin on the secretion of estradiol and progesterone by human luteinized granulosa cells in vitro. It is likely that this adipokine locally affects ovarian function in women.

Abbreviations: 3β-HSD: 3β-hydroxysteroid dehydrogenase; CAP1: cyclase-associated protein 1; DCN: decorin; FIZZ: Found in Inflammatory Zones; hCG: human chorionic gonadotropin; IGF1: insulin-like growth factor type 1; IVF: in vitro fertilization; PCOS: polycystic ovary syndrome; RIA: radioimmunoassay; ROR1: receptor tyrosine kinase-like orphan receptor-1; TLR4: Toll–like receptor 4

KEYWORDS:

• Resistin
• estradiol
• progesterone
• granulosa cells
• ovary

Introduction

Resistin, a 12.5 kD protein, also known as “Found in Inflammatory Zones (FIZZ)”, was discovered in 2001 (Steppan et al. 2001). This protein, mainly expressed in adipose tissue, is also expressed in many other tissues at variable levels (Jamaluddin et al. 2012). A receptor for resistin has not been identified yet. However, it has been shown that resistin can bind Toll–like receptor 4 (TLR4) on the surface of human leucocytes and rat hypothalamus (Tarkowski et al. 2010; Benomar et al. 2013). This receptor has been identified in many other cell types in humans and various animals (Nishimura et al. 2005). Another possible action of resistin is via receptor tyrosine kinase-like orphan receptor-1 (ROR1) in 3T3-L1 preadipocytes (Sanchez-Solana et al. 2012) or adenylyl cyclase-associated protein 1 (CAP1) in human monocytes (Lee et al. 2014). Moreover, an isoform of decorin (DCN) serves as a functional receptor in adipocyte progenitors for resistin and this is possibly involved in white adipose tissue expansion (Daquinag et al. 2011).

Resistin affects the sensitivity of target tissues to insulin. Evidence has been provided that this adipokine plays a role in reproductive function. Studies in animals have demonstrated that resistin in a species-dependent manner is expressed in various parts of the ovarian follicles including granulosa and theca cells as well as in the oocyte (Maillard et al. 2011; Rak-Mardyla et al. 2013; Singh et al. 2015). In porcine follicles, it was found that the expression of resistin is regulated by FSH, LH and locally produced substances, such as insulin-like growth factor type 1 (IGF1) (Rak et al. 2015). Regarding human ovary, expression of resistin has been also found in the oocyte, the granulosa and the theca cells (Niles et al. 2012; Reverchon et al. 2013).

Studies in animals have provided evidence that resistin may affect ovarian steroidogenesis, but the results have been equivocal. In particular, it has been shown that in rats resistin increased the secretion of progesterone without affecting estradiol, while in cows it decreased both estradiol and progesterone secretion from granulosa cells in vitro (Maillard et al. 2011). Additionally, large follicles obtained from prepubertal pigs expressed higher amounts of resistin than small ones, while recombinant resistin, through the up-regulation of specific enzymes, stimulated the secretion of progesterone and androgen from such follicles without affecting the secretion of estradiol (Rak-Mardyła et al. 2013).

Information regarding the role of resistin in the regulation of ovarian steroidogenesis in humans is limited. There are only two studies; in the first, theca interna cells were obtained from the ovaries of two regularly cycling premenopausal women during hysterectomy (Munir et al. 2005). In agreement with data in pigs (Rak- Mardyla et al. 2013), it was shown that in a 3-day culture of the theca cells resistin enhanced the expression of 17α-hydroxylase activity only in the presence of forskolin or forskolin combined with insulin (Munir et al. 2005). In the second study, granulosa cells were obtained from women undergoing IVF/ICSI treatment (Reverchon et al. 2013). It was shown that resistin did not affect basal estradiol and progesterone secretion in the presence or in the absence of FSH, but it eliminated the IGF1-stimulated secretion of these steroids. In that study, only a single dose of 10 ng/ml resistin was used and therefore it is not known what higher or lower dosages would do. It is evident that there is no consistency between the animal and the human data, while information in humans regarding the effect of resistin on ovarian steroidogenesis is scant.

The aim of the present study was to investigate the effect of various doses of resistin on the secretion of estradiol and progesterone by human luteinized granulosa cells in culture.

Results and discussion

Effect of resistin on basal steroids secretion

The present study has investigated the effect of resistin on ovarian steroidogenesis in human luteinized granulosa cells in vitro. In particular, the effects of increasing concentrations of resistin (1, 10, 100 ng/ml) on estradiol and progesterone secretion were studied after 24 or 48 h in culture and in the presence or absence of different concentrations of FSH (10, 100 ng/ml).

The amount of estradiol secreted by the granulosa cells after 48 h in culture was significantly higher than that secreted after 24 h (control values: 4688 ± 670 and 2040 ± 238 pg/ml, respectively, P = 0.01; Figure 1A), while that of progesterone was similar (control values: 1096 ± 214 vs. 1001 ± 103 ng/ml; Figure 2A). These results were comparable to previous studies from our group, indicating that though granulosa cells estradiol secretion increases in a time-dependent manner that of progesterone remains stable between 24 and 48 h of culture (Prapa et al. 2015).

Figure 1. Effect of resistin on estradiol secretion from human luteinized granulosa cell cultures after 24 h and 48 h of culture. Effect of (A) resistin at various dosages; (B) combination of 1 ng/ml resistin with FSH (10 and 100 ng/ml), (C) combination of 10 ng/ml resistin with FSH (10 and 100 ng/ml), and (D) combination of 100 ng/ml resistin with FSH (10 and 100 ng/ml). The asterisk ‘*’ indicates statistically significant difference from control; the sharp ‘#’ indicates statistically significant difference from 10 ng/ml FSH; the section mark ‘§’ indicates statistically significant difference from 100 ng/ml FSH. The number of symbols’ repetition represents: 1, p < 0.05; 3, p < 0.001.

Figure 2. Effect of resistin on progesterone secretion from human luteinized granulosa cell cultures after 24 h and 48 h of culture. Effect of (A) resistin at various dosages, (B) combination of 1 ng/ml resistin with FSH (10 and 100 ng/ml), (C) combination of 10 ng/ml resistin with FSH (10 and 100 ng/ml), and (D) combination of 100 ng/ml resistin with FSH (10 and 100 ng/ml). The asterisk ‘*’ indicates statistically significant difference from control; the sharp ‘#’ indicates statistically significant difference from 10 ng/ml FSH; the section mark ‘§’ indicates statistically significant difference from 100 ng/ml FSH. The number of symbols’ repetition represents: 1, p < 0.05; 2, p < 0.01; 3, p < 0.001.

Regarding resistin, an inhibitory effect was observed on both estradiol and progesterone basal secretion. In particular, it was found that though this peptide hormone after 24 h in culture and at every dose tested did not affect basal estradiol secretion, after 48 h in culture and at the dose of 1 ng/ml it significantly reduced the secretion of estradiol as compared to control (P < 0.05, Figure 1A). In addition, resistin significantly reduced basal progesterone secretion at all three dosages tested, as compared to control both after 24 and 48 h in culture (P < 0.01, Figure 2A).

In the literature, there are only two studies in humans regarding the effect of resistin on steroidogenesis, and only one concerns the effect of this peptide on granulosa cells. In the first study, theca interna cells were obtained from two normally menstruating premenopausal women and cultured with or without 10 ng/ml resistin (Munir et al. 2005). It was shown that resistin alone did not affect 17α-hydroxylase activity, but it was enhanced in the presence of forskolin or in combination of forscolin and insulin. This indicates a possible stimulatory effect of resistin on androgen secretion from ovarian theca cells. No granulosa cells were included in this study (Munir et al. 2005). In the second study, luteinized granulosa cells were obtained from women undergoing IVF treatment and were cultured in vitro (Reverchon et al. 2013). It was found that oocytes and granulosa cells at all stages of development as well as theca cells from large follicles express resistin at mRNA and protein level. It was also shown that resistin at the ‘physiological’ concentration of 10 ng/ml did not affect basal estradiol and progesterone secretion (Reverchon et al. 2013). Our study also showed no effect of 10 ng/ml resistin on estradiol secretion, but in general, the effect of resistin on the secretion of both steroids was inhibitory.

The difference between the present and the Reverchon et al. (2013) study could be partly related to the culture conditions, but it also demonstrates the need for further research. It should be noted that in contrast to our study, only one dose of resistin (10 ng/ml) was used in the study of Reverchon et al. (2013). Although authors stated that similar results were obtained with 100ng/ml resistin, these results were not shown.

Furthermore, TLR4 and ROR1 receptors as well as CAP1 and DCN have all been documented to serve as resistin targets in different cell types (Tarkowski et al. 2010; Daquinag et al. 2011; Sánchez-Solana et al. 2012; Benomar et al. 2013; Lee et al. 2014) and at least two of them (TLR4 and DCN)  appear  to be expressed on granulosa cells (Herath et al. 2007; Peng et al. 2016). Thus, as far as estradiol secretion is concerned, the variations observed between different resistin concentrations tested in our study can be the result of a concentration-dependent interaction of resistin to its various targets present on granulosa cells.

Effect of resistin on FSH-induced steroids secretion

Regarding FSH, not unexpectedly, the high dose of this hormone stimulated steroidogenesis in the present study. In particular, though FSH at the dose of 10 ng/ml, reduced the secretion of progesterone from the granulosa cells (48 h; P < 0.01, Figure 2B), at the dose of 100 ng/ml, it significantly increased the secretion of this steroid (24 h) as compared to control (P < 0.05, Figure 2B) and that of estradiol (48 h) as compared to both control and 10 ng/ml FSH-treated cells (P < 0.001, Figure 1B).

Furthermore, according to our results, the stimulatory effect of FSH could be modified by resistin. In particular, combination of 100 ng/ml FSH with 1 ng/ml resistin eliminated the stimulatory effect of FSH (P < 0.001), so that estradiol values did not significantly differ from those of control (Figure 1B). At the same dose resistin when combined with 10 ng/ml FSH did not significantly affect the secretion of estradiol, as compared to control or to FSH alone (10 ng/ml). However, combination of 10 ng/ml resistin with 10 ng/ml FSH significantly reduced the secretion of estradiol as compared to control (P < 0.05, Figure 1C). Similar to 1 ng/ml, 10 ng/ml resistin in combination with 100 ng/ml FSH eliminated the stimulatory effect of FSH on estradiol secretion after 48 h in culture (P < 0.001, Figure 1C). Regarding the dose of 100 ng/ml resistin, its combination with FSH eliminated the stimulatory effect of 100 ng/ml FSH on estradiol secretion after 48 h in culture and the values did not differ significantly from that of control (P < 0.001, Figure 1D). However, when 100 ng/ml resistin was combined with 10 ng/ml FSH, it reduced the secretion of estradiol both at 24 h and 48 h of culture as compared to control (P < 0.05), FSH 10 ng/ml (P < 0.05) and FSH 100 ng/ml (P < 0.05 and P < 0.001, respectively, Figure 1D).

In terms of FSH-induced progesterone secretion from the granulosa cells, a significant reduction was observed when 1 ng/ml resistin was combined with 10 ng/ml FSH, as compared to control both after 24 and 48 h in culture (P < 0.001, Figure 2B). The suppression was also significant as compared to 10 ng/ml FSH only at 24 h (P < 0.001) as well as to 100 ng/ml FSH at both time points (P < 0.001). Similarly, the combination of 1 ng/ml resistin with 100 ng/ml FSH significantly suppressed progesterone secretion as compared to control and 100 ng/ml FSH at both time points, and to 10 ng/ml FSH at 24 h (Figure 2B). The combination of 10 ng/ml resistin with 10 ng/ml or 100 ng/ml FSH and that of 100 ng/ml resistin with 100 ng/ml FSH also significantly suppressed the secretion of progesterone as compared to control and to 100 ng/ml FSH at both time points or to 10 ng/ml FSH at 24 h (Figure 2C,D). It is evident that resistin at all dosages was able to eliminate the stimulatory effect of the high dose FSH on estradiol and progesterone secretion.

These findings are in contrast with those of the Reverchon et al. (2013) study in which, the addition of resistin to the culture system did not affect the stimulatory action of 10^–8 M FSH on estradiol and progesterone secretion, but it was able to eliminate the stimulatory effect of IGF-1 on the secretion of these two steroids. The discrepancy between the Reverchon et al. (2013) and the present study, in terms of resistin-FSH interaction, is difficult to explain. One reason can be the differences in FSH concentrations tested; the FSH concentration used in the study of Reverchon et al. (2013) is more than 3 times less than the lower dose tested here. The culture conditions and the single dose of resistin used by Reverchon et al. (2013) might also be implicated; however, in our study, even the lower dosage of resistin (1 ng/ml) was able to prevent the stimulatory action of FSH on estradiol and particularly progesterone secretion.

Based on our results, there was no specific dose-dependent pattern of resistin effects, at least not within the concentrations tested; however, the effect of resistin on basal and FSH-induced ovarian steroidogenesis was mainly inhibitory. More studies are needed to examine further this issue. The present study demonstrates for the first time an inhibitory action of various dosages of resistin on human ovarian steroidogenesis particularly on basal progesterone secretion and on FSH-stimulated estradiol and progesterone secretion by human luteinized granulosa cells. It is suggested that resistin may be a local regulator of ovarian function in women.

Experiments in animals have also provided conflicting evidence. Specifically, in bovine granulosa cells, resistin decreased basal, but not IGF1-stimulated progesterone secretion, while in rat granulosa cells it increased both basal and IGF1-induced progesterone secretion (Maillard et al. 2011). In both species, estradiol secretion was not affected by resistin (Maillard et al. 2011). An increase in progesterone secretion by resistin via the up-regulation of various enzymes of steroidogenesis was also reported in ovarian follicles from prepubertal pigs (Rak-Mardyła et al. 2013). This species’ difference has been also confirmed in cattle, in which resistin enhanced FSH and IGF1-induced secretion of estradiol, but not progesterone, in granulosa cells obtained from large follicles, while it attenuated the IGF1-induced secretion of estradiol and progesterone in granulosa cells obtained from small follicles (Spicer et al. 2011). There are, therefore, many differences in the results obtained from different animal species as well as between animals and humans.

An explanation for these disparities is difficult, as the mechanism via which resistin can affect the production of estradiol and/or progesterone in the granulosa cells is not clear. A resistin-specific receptor has not been identified yet; as mentioned above, TLR4 and ROR1 receptors as well as CAP1 and DCN have all been documented to serve as resistin targets in different cell types (Tarkowski et al. 2010; Daquinag et al. 2011; Sánchez-Solana et al. 2012; Benomar et al. 2013; Lee et al. 2014) and at least two them (TLR4 and DCN) are known to be expressed on granulosa cells (Herath et al. 2007; Peng et al. 2016). Resistin binding to TLR4, ROR1, CAP1, and DCN can activate various signaling pathways, including Akt, MAPK, Stat-3, PPARγ and NF-kB (Hsieh et al. 2014; Rak et al. 2017). However, the role of these intracellular signaling pathways in resistin-mediated actions in the granulosa cells is not known. Thus far, it has only been suggested that PPARγ is the key factor controlling resistin expression and steroidogenic function in porcine ovarian follicles (Rak-Mardyła and Drwal 2016). Furthermore, data from experiments in human granulosa cells suggest an association between the inhibition of IGF1-stimulated secretion of estradiol and progesterone by resistin and the reduction in the expression of aromatase and cholesterol side-chain cleavage cytochrome P450, but not the expression of 3β-hydroxysteroid dehydrogenase (3β-HSD) or steroidogenic acute regulatory protein (Reverchon et al. 2013). Nevertheless, an inhibitory action of resistin on the expression of aromatase, 3β-HSD, and 17β-HSD has been shown in porcine ovarian follicles (Rak et al. 2015). Resistin has been also shown to reduce IGF1-induced IGF1 receptor as well as MAPK in human granulosa cells (Reverchon et al. 2013). It is evident that there is not a unique pathway via which resistin affects steroidogenesis in granulosa cells. Despite species-specific ovarian expression of resistin, investigation of other intracellular signaling molecules might provide further insights into the mechanism of action of this protein.

The clinical importance of the present findings demonstrating an inhibitory action of resistin on human ovarian steroidogenesis is unclear at present. It is also unknown if the actions of resistin are part of the physiological mechanisms regulating normal steroidogenesis. Resistin is known to increase insulin resistance and thus may be a link between obesity and polycystic ovary syndrome (PCOS; Spritzer et al. 2015). As serum resistin levels may be increased in women with PCOS (Munir et al. 2005), this protein, through affecting steroidogenesis, is likely involved in the reproductive axis functional abnormalities observed under these pathologic conditions. More studies are needed to examine this hypothesis.

Materials and methods

Human granulosa cell cultures

Human luteinized granulosa cells were obtained from 50 women, aged 25–40 years old, who underwent IVF treatment due to tubal pathology or male-related infertility and gave written informed consent. The study was approved by the Institutional Ethical Committee: Faculty of Medicine, University of Thessaly, Greece (1789/15/04/2011). Recombinant FSH was used for ovarian stimulation in the context of a GnRH long agonist protocol. For induction of final follicle and oocyte maturation, human chorionic gonadotropin (hCG) was injected, while oocyte recovery was attempted 34-36 h later. Granulosa cells’ isolation from the follicular fluid was performed as previously described using four circles of PBS buffer [phosphate buffer saline (Biochrom AG, Berlin, Germany) containing 0.1% BSA (bovine serum albumin; Sigma, Bornem, Belgium)] washes and centrifugations (Karamouti et al. 2008).

Following isolation, cells were seeded in 24-well plates (5x10⁴ viable cells/1ml medium/well) and cultured in FBS enriched culture medium containing 10% FBS, 2% L-glutamine and 1% Penicillin/Streptomycin for 24 h in order to attach to the dishes, form a monolayer and recover from gonadotropin desensitization due to hCG in vivo stimulation imposed as part of the IVF patient’s treatment (Foldesi et al. 1998; Lambert et al. 2000). After this 24h-preincubation period, cells were incubated for 24 or 48 h in serum-free medium containing 10^–6 M androstendione (NIDDK’s National Hormone and Peptide Program, Harbor-UCLA Medical Centre, Los Angeles, CA) as a substrate and in the presence/absence of recombinant human resistin [1, 10 and 100 ng/ml; Bio-Rad, USA (Cat No. PHP200)] or/and FSH [10 and 100 ng/ml; Recombinant Human FSH alpha/beta Protein, R&D Systems, Inc., USA (Cat No. 5925-FS)]. All cultures were performed at 37°C in 5% CO₂ atmosphere while at the end of the 24 and 48 h incubation periods supernatants were collected and stored at 20°C until estradiol and progesterone level assessment.

Cell viability was evaluated using trypan blue staining after 24 and 48 h of resistin in culture. Resistin didn’t affect cell viability as compared to control (absence of hormone) at both time points studied (data not shown).

Hormone assays

Estradiol and progesterone levels in culture supernatants were measured using commercially available human radioimunoassay (RIA) diagnostic kits (KIP0629 and KIP1458, respectively; DIASource Europe SA, Louvain-la-Neuve, Belgium). The lower limits of detection and the intra- and inter-assay coefficients of variation for estradiol were 2 pg/ml, 4.9–5.9% and 6.2–8.1% while that of progesterone were 0.05 ng/ml, 3.3–4.1% and 6.5–8.6%, respectively. Results were expressed in pg/ml for estradiol and ng/ml for progesterone.

Statistical analysis

Data are expressed as means±SEM, while values represent the mean of 17b-estradiol or progesterone released from 5–10 different cell cultures. For the statistical analysis of data, GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego California USA) was used. Two-way ANOVA was performed in order to locate the source (time and/or treatment) of estradiol and progesterone release variations between groups, while one-way ANOVA followed by the Bonferroni post-test was used in order to assess the significance of estradiol and progesterone differences between treatments; p-values < 0.05 were considered statistically significant.

Author contributions

Conceptualization, IEM; Data curation, AV; Format analysis, CIM; Methodology, EK; Project administration, AV, EK, GA; Resources, PG, IEM; Assay PG; Software, AV; Supervision, IEM; Validation, AV; Visualization, KD; Writing – original draft, CIM; Writing review and editing, AG, AD, AV, IEM.

Disclosure statement

No potential conflict of interest was reported by the authors.