New insights into the ovulatory process in the human ovary

Abstract

BACKGROUND

Successful ovulation is essential for natural conception and fertility. Defects in the ovulatory process are associated with various conditions of infertility or subfertility in women. However, our understanding of the intra-ovarian biochemical mechanisms underlying this process in women has lagged compared to our understanding of animal models. This has been largely due to the limited availability of human ovarian samples that can be used to examine changes across the ovulatory period and delineate the underlying cellular/molecular mechanisms in women. Despite this challenge, steady progress has been made to improve our knowledge of the ovulatory process in women by: (i) collecting granulosa cells across the IVF interval, (ii) creating a novel approach to collecting follicular cells and tissues across the periovulatory period from normally cycling women, and (iii) developing unique in vitro models to examine the LH surge or hCG administration-induced ovulatory changes in gene expression, the regulatory mechanisms underlying the ovulatory changes, and the specific functions of the ovulatory factors.

OBJECTIVE AND RATIONALE

The objective of this review is to summarize findings generated using in vivo and in vitro models of human ovulation, with the goal of providing new insights into the mechanisms underlying the ovulatory process in women.

SEARCH METHODS

This review is based on the authors’ own studies and a search of the relevant literature on human ovulation to date using PubMed search terms such as ‘human ovulation EGF-signaling’, ‘human ovulation steroidogenesis’, ‘human ovulation transcription factor’, ‘human ovulation prostaglandin’, ‘human ovulation proteinase’, ‘human ovulation angiogenesis’ ‘human ovulation chemokine’, ‘human ovulatory disorder’, ‘human granulosa cell culture’. Our approach includes comparing the data from the authors’ studies with the existing microarray or RNA-seq datasets generated using ovarian cells obtained throughout the ovulatory period from humans, monkeys, and mice.

OUTCOMES

Current findings from studies using in vivo and in vitro models demonstrate that the LH surge or hCG administration increases the expression of ovulatory mediators, including EGF-like factors, steroids, transcription factors, prostaglandins, proteolytic systems, and other autocrine and paracrine factors, similar to those observed in other animal models such as rodents, ruminants, and monkeys. However, the specific ovulatory factors induced, their expression pattern, and their regulatory mechanisms vary among different species. These species-specific differences stress the necessity of utilizing human samples to delineate the mechanisms underlying the ovulatory process in women.

WIDER IMPLICATIONS

The data from human ovulation in vivo and in vitro models have begun to fill the gaps in our understanding of the ovulatory process in women. Further efforts are needed to discover novel ovulatory factors. One approach to address these gaps is to improve existing in vitro models to more closely mimic in vivo ovulatory conditions in humans. This is critically important as the knowledge obtained from these human studies can be translated directly to aid in the diagnosis of ovulation-associated pathological conditions, for the development of more effective treatment to help women with anovulatory infertility or, conversely, to better manage ovulation for contraceptive purposes.

REGISTRATION NUMBER

N/A.

Graphical Abstract

Investigations of the ovulatory process in humans have identified specific signaling pathways and factors expressed in follicular cells and leukocytes, and involved in the complex process of ovulation and luteinization. EGF, epidermal growth factors; PGs, prostaglandins, LUFS, luteinized unruptured follicle syndrome; PCOS, polycystic ovarian syndrome. Image created with BioRender.com

Introduction

The mid-cycle surge of LH sets in motion dramatic physiological and morphological changes in the dominant follicle of the ovary, culminating in ovulation and subsequent transformation of the ruptured follicle into a corpus luteum. Ovulation encompasses multilateral changes in follicle morphology, including the rupture of the follicle wall, extrusion of a mature oocyte enclosed in expanded cumulus cells, and collapse of the remaining follicular structure. Coinciding with these morphological changes, follicular cells of preovulatory follicles undergo physiological changes, including a shift in steroidogenesis from estradiol to progesterone production and a swift transition in the cell cycle from proliferation to terminal differentiation. The surge level of LH, or the administration of ovulatory hCG, upon binding to their cognate receptors, LHCGR, activates various yet interconnected signaling pathways in preovulatory follicular cells. The activation of these signaling pathways leads to the induction and activation of a diverse array of cellular and extra-cellular factors that exert coordinated actions to bring about these morphological and physiological changes necessary for ovulation and luteinization.

Over the past few decades, studies using animal models such as rodents, ruminants, and non-human primates have provided foundational information on LH/hCG-induced mediators involved in the periovulatory process (reviewed in Richards (2007), Stouffer et al. (2007), Duffy et al. (2019)). Among these LH/hCG-induced mediators are various signaling molecules, transcription factors, steroid hormones, secretory proteins, and proteinases along with their inhibitors. Although these animal studies have revealed several key ovulatory LH/hCG-induced mediators and pathways required for successful ovulation and luteinization, there have been large gaps in our understanding of the ovulatory process in the human ovary. This is mainly due to extremely limited access to tissue of preovulatory follicles obtained before and at defined hours after the LH surge throughout the ovulatory period from normally cycling women. Moreover, there is a lack of well-established in vitro models that can recapitulate the complex cellular events induced by the LH surge/ovulatory hCG stimulation or mimic the follicle rupture observed in vivo in ovulatory follicles in humans.

Our laboratories have been uniquely qualified to overcome some of these challenges because we have collected human pre- and peri-ovulatory follicles across the entire ovulatory period. This was accomplished by collecting pre- and peri-ovulatory follicles from normally cycling women at defined hours after an ovulatory hCG administration that was used to mimic the endogenous LH surge. Moreover, we have established a primary human granulosa cell culture model in which hCG-induced increases in the expression of key ovulatory genes in granulosa cells of periovulatory follicles could be mimicked. Using these in vivo and in vitro models, we have begun to characterize the expression profile of key ovulatory genes, investigate the regulatory mechanisms by which the LH surge/ovulatory hCG stimulation induces the expression of these genes, and finally explore their specific function in the ovulatory follicle.

The current review contains a synopsis of the ovulatory process, focusing on the aspects common to most species studied and highlighting key ovulatory genes that are essential for successful ovulation in animal models. We then describe in vivo and in vitro model systems used in our previous studies and summarize up-to-date findings of the mechanisms underlying the ovulatory process in the human ovary based on recent publications from our laboratory and others over the last several years.

General overview of the ovulatory process in mammals

There are several essential steps for successful ovulation that are common across vertebrate animal models. The foremost is the initial step that triggers the ovulatory process: the activation of the luteinizing hormone/choriogonadotropin receptor (LHCGR) in granulosa and theca cells of a dominant follicle(s) by the LH surge or ovulatory hCG administration. The activation of LHCGR, a G-protein-coupled receptor, triggers the release of various secondary messengers that transmit their signals inside follicular cells through their respective signaling pathways. These signals rapidly spread through gap junctions between cells inside the preovulatory follicle. It is well established that ovulatory LH/hCG activates adenylate cyclase and increases intracellular cAMP levels as a primary intracellular signaling molecule (Mason and Marsh, 1975; Richards et al., 1979; Davis et al., 1986; Strauss et al., 1988; Chin et al., 2004). The increased cAMP leads to the activation of cAMP-dependent protein kinase A (PKA), which transduces the signals by phosphorylating its target proteins in the cytoplasm and cAMP-response element-binding protein (CREB) in the nucleus, ultimately exerting transcriptional regulation (Salvador et al., 2002; Panigone et al., 2008). This PKA secondary messenger system is widely regarded as the primary signaling pathway mediating ovulatory LH/hCG action in the preovulatory follicle. Other protein kinase pathways such as PKC, PI3K, and p38MAPK have also been shown to be the downstream signaling pathways activated by ovulatory LH/hCG (Davis et al., 1986; Maizels et al., 2001; Salvador et al., 2002; Fan et al., 2008; Breen et al., 2013). However, the experimental evidence of these specific downstream signaling pathways varies depending on the species and is often limited to single species.

In addition to these pathways, studies using various species showed that the ovulatory LH/hCG-activated cAMP-PKA signaling pathway led to rapid activation of the EGF signaling network through the induction of EGF-like factors (e.g. AREG, EREG, BTC, or NRG1) and phosphorylation of EGF receptor (EGFR) family members (EGFR, ERBB1, ERBB2, ERBB3) in the preovulatory follicle (Park et al., 2004; Panigone et al., 2008; Noma et al., 2011). The activation of EGFR by its ligands stimulates its intrinsic tyrosine kinase activity, transducing the signal to downstream kinases, notably the RAS-mitogen-activated protein kinase kinase-MAPK1/3 (ERK1/2) pathway in granulosa and cumulus cells of preovulatory follicles (Fan et al., 2008; Panigone et al., 2008).

These ovulatory LH/hCG-activated signaling pathways ultimately lead to the induction or activation of various transcription factors. These transcription factors play central roles in the ovulatory process by regulating the transcription of genes that exert specific intracellular and extra-cellular actions involved in critical aspects of ovulatory processes such as steroidogenesis, tissue remodeling, angiogenesis, and inflammatory responses. Among these induced transcription factors are the progesterone receptor (PGR), C/EBPA/B, FOS/activator protein 1 (AP-1), and RUNX1/2. Studies using transgenic knockout mouse models have further demonstrated that the ovulatory induction of these transcription factors in granulosa cells of preovulatory follicles is essential for successful ovulation and luteinization by regulating the transcription of a diverse array of genes (Robker et al., 2000; Fan et al., 2011; Xie et al., 2015; Lee-Thacker et al., 2020; Park et al., 2020). Downstream genes of these transcription regulators include factors involved in cholesterol and progesterone synthesis, tissue contraction/remodeling, prostaglandin (PG) production and secretion, and luteinization (reviewed in Kim et al. (2009a), Robker et al. (2018), Duffy et al. (2019)). The coordinated and temporally regulated expression of effector genes and their action on ovulatory follicles are necessary for the precisely controlled process of follicular rupture, maturation of the oocyte, extrusion of cumulus–oocyte complex, and subsequent transformation of the ovulated follicle into the corpus luteum. Indeed, transgenic mouse studies of these effector genes, such as Ptgs2, Adamts1, and Edn2, confirmed the essential role of these factors in the ovulatory process (Lim et al., 1997; Shozu et al., 2005; Cacioppo et al., 2017). Therefore, this review focuses on those genes known to be up-regulated and to play an important role in the ovulatory process in animal models described above.

In understanding the ovulatory process in humans, it is essential to establish an in vivo model where all the necessary changes in dominant follicles during the ovulatory period in normally cycling women can be assessed. Secondly, for the further dissection of the mechanism(s) underlying the ovulatory process in humans, it is critical to establish an in vitro system that can closely mimic in vivo ovulatory changes. Below, we describe in vivo models, followed by the detailed procedure paradigm of an in vitro model used to study the ovulatory process in our laboratory and others.

Experimental models used for the investigation of the human ovulatory process

In vivo models

The biggest challenge in studying the ovulatory process in humans has been the extremely limited access to dominant follicles obtained at defined time points before and after the LH surge from normally cycling women. Historically, researchers have used human ovarian tissues removed from patients undergoing scheduled surgery because of reproductive tract-associated diseases. However, the use of these tissues has limited value as it was difficult to accurately define the exact time of the menstrual cycle, and very few tissue samples have been collected in the time interval from the preovulatory stage just before the initiation of the LH surge to until rupture of the follicle has occurred. This time interval, spanning >36 h, would be the most important time span to obtain tissue samples. Follicular aspirates of IVF patients have also been used as a source of granulosa/lutein cells since these samples are relatively easy to obtain, and their collection timing and hormone profiles can be specified and documented during the FSH priming period. The caveat of this sampling method is that collection is at a single time point, lacking comparison groups, and that the cells are obtained at a very late ovulatory stage. Moreover, the IVF procedure uses an ovarian stimulation protocol, and this stimulation may not necessarily mimic a physiological situation. To overcome this caveat, Wissing et al. (2014) employed the paradigm in which granulosa cells were collected twice from the same patients; first, before administration of recombinant hCG (rhCG) or GnRH agonist to trigger ovulation, and second, at 36 h after rhCG administration during oocyte retrieval for IVF (Fig. 1). However, the collection timing of granulosa cells before ovulation triggering varied among patients. Nonetheless, the microarray analysis of granulosa cells from these nine paired samples generated a list of genes whose expression was differentially regulated after ovulation induction in human granulosa cells of preovulatory follicles (Wissing et al., 2014). Subsequent bioinformatics analysis of these data predicted that a diverse array of pathways, such as inflammation, angiogenesis, extra-cellular matrix remodeling, and cell cycles, are involved in the ovulatory process in humans. More recently, Poulsen et al. (2020a) expanded this approach by collecting human granulosa cells twice from each patient, first, prior to the ovulation induction or 12-, 17- or 32-h post-rhCG administration and second, at the time of oocyte retrieval (36-h post-ovulatory induction). Microarray analysis of these granulosa cells showed massive changes in the transcriptome (13 345 differentially expressed genes out of 135 750 transcripts detected, fold change >2 or <-2, false discovery rate <0.05 compared before stimulation) across the five time points. This microarray study showed two transcriptional up-regulation points (first at 12 h and second at 36 h after ovulation induction) (Poulsen et al., 2020a). The authors stated that up-regulated genes at the first peak indicate initiation of an inflammatory response, while the list of genes at the second peak represents effector functions of inflammation such as vasodilation, angiogenesis, coagulation, chemotaxis, and tissue remodeling (Poulsen et al., 2020a). This transcriptomic analysis was a critical first step in enhancing our understanding of the ovulatory process in humans. Yet, granulosa cell transcriptional changes may not portray all aspects of the ovulatory process. This is because ovulation is a complex event accomplished by coordinated communications and actions between/among multiple cell types such as the oocyte, cumulus cells, granulosa cells, theca cells, endothelial cells, and immune cells.

Figure 1.

In vivo sample collection of granulosa cells from women undergoing IVF/ICSI treatment. Individually dosed rFSH or hMG was given, starting on Days 2–3 of the menstrual cycle. From stimulation Days 5–6, GnRH antagonist was administrated daily. When at least three follicles reached 17 mm in diameter, rhCG or GnRH agonist was given to trigger the ovulatory process. Each woman donated the content of two follicles: one before (0 h, T0) or at 12 h (T12), 17 h (T17), or 32 h (T32), and another one at the time of retrieval at 36 h (T36). The T0 follicle aspiration was taken an average of 11.9 h prior to the ovulation trigger. Granulosa cells and follicular fluid were collected and processed for gene expression analysis and hormone measurement. Image created with BioRender.com.

In an effort to address these issues, Brannstrom and colleagues established a unique in vivo experimental model where ovulatory changes can be examined in dominant follicle tissues and follicular cells collected at multiple and defined time points throughout the periovulatory period of women undergoing laparoscopic surgery (Fig. 2). As shown in the paradigm, dominant follicles were collected from women (aged 30–38 years) with regular menstrual cycles and who had not taken hormonal contraceptives for at least three months prior to the study. The women were monitored by transvaginal ultrasound for two to three menstrual cycles before surgery to ascertain the occurrence of regular cycles and to monitor the growth of a dominant follicle during the follicular phase. For patients undergoing tubal ligation for sterilization, the entire dominant follicle was collected by laparoscopic excision at the time points outlined below (Fig. 2). The patients were divided into four groups: pre-, early, late, and post-ovulatory phases. In the preovulatory group, surgery was performed when the follicle reached >14 and <17.5 mm in diameter, prior to the endogenous LH surge. These patients were not administered hCG. The remaining women were given rhCG (Ovitrelle^®, 250 µg) and divided into three groups: early ovulatory (surgery between 12- and <18-h post-rhCG), late ovulatory (surgery between 18- and 34-h post-rhCG), and post-ovulatory (surgery between 40- and 70-h post-rhCG) phases. To confirm that these patients had normal hormonal patterns before the LH surge and after hCG administration, blood samples were taken at surgery and analyzed for serum progesterone and estradiol. The whole intact follicle was removed using laparoscopic scissors and processed for either immunohistochemical analyses or the isolation of granulosa and theca cells (Figs 2 and 3). For cell isolation, follicles were first bisected by a scalpel, and granulosa cells were collected by gently scrubbing the follicle using a glass needle. The theca layer was peeled off from the follicle wall. These samples were used to isolate total RNA for gene expression analysis (Figs 2 and 3).

Figure 2.

In vivo collection of dominant follicles throughout the periovulatory period from regularly cycling women. Menstrual cyclicity and the size of the dominant follicle were monitored by transvaginal ultrasound. When the dominant follicle reached 14–17 mm, rhCG was given to trigger the ovulatory process. Surgery was performed at one of four predetermined phases: (i) preovulatory phase (Pre), prior to the spontaneous LH surge when the follicle was between 14 and 17 mm; (ii) early ovulatory phase (EO), between 12 and <18 h after hCG; (iii) late ovulatory phase (LO), between 18 and 34 h or less after hCG; or (iv) post-ovulatory phase (PO), between >44–70 h after hCG administration. The dominant follicle was dissected from the ovary and used either for immunohistochemistry or to harvest granulosa cells and thecal tissue for gene expression analysis. Image created with BioRender.com.

Figure 3.

An image of a dissection of a dominant human follicle of the early ovulatory (EO) stage using laparoscopic scissors and forceps. (A) The dissected whole intact follicle of ∼2 cm in size. (B) Blood vessels can be seen on the surface of the follicle. OC: ovarian cortex; OS: ovarian stroma; FW: follicle wall.

In vitro models

Dominant follicle tissues and granulosa cells collected at defined hours throughout the periovulatory period from women have been extremely valuable in identifying factors expressed in the ovulatory follicle. The equally important next step is to delineate how these factors and specific ovulatory LH/hCG-induced pathways control and participate in the ovulatory process in humans. In animal models, the routine approach has been to use granulosa cells isolated from preovulatory follicles or to utilize preovulatory follicles dissected from ovaries obtained prior to the LH surge. The granulosa cells or preovulatory follicles isolated from the ovary were typically treated with LH or hCG to mimic the endogenous LH surge and then treated with agents that can activate or inhibit the expression or actions of specific factors or ovulatory pathways. However, in humans, this approach is not practical considering the extremely limited availability of granulosa cells isolated from preovulatory follicles obtained right before the LH surge from normally cycling women. Alternatively, several research groups have developed human granulosa cell lines using ovarian tumor samples (e.g. COV434, KGN, and HTOG) or via gene transfection (e.g. SVOG, HGL5, GC1a, HGrC1, HO23, and HGP53) (Havelock et al., 2004; Ota et al., 2006; Bayasula et al., 2012), or have utilized granulosa cells isolated from follicular aspirates from IVF patients. However, the cell lines often have underlying problems for studying the mechanisms involved in the ovulatory changes as they lack key characteristics of preovulatory granulosa cells, such as the responsiveness to LH or the induction of known ovulatory genes. On the other hand, granulosa cells from IVF patients also have their own limitations as these cells have already experienced the ovulatory dose of hCG ∼36 h prior to collection and should rather be referred to as granulosa/lutein cells. Subsequently, these cells are not responsive to LH/hCG treatment during the first few days in culture due to the desensitization of LHCGR (Jalkanen et al., 1986; Emi et al., 1991). However, Breckwoldt et al. (1996) showed that these IVF-derived human granulosa/lutein cells could become responsive to FSH and hCG when precultured for 7 days in the hormone-free medium. For instance, FSH and hCG treatments increased cyclic AMP levels, and progesterone and estradiol production in this 7-day acclimated human granulosa/lutein cell model (Breckwoldt et al., 1996). Building on these reports, we have carefully characterized this primary cell culture system with a few modifications (Fig. 4). Briefly, luteinizing human granulosa cells were isolated from follicular aspirates of IVF patients using a Percoll-gradient centrifugation method to remove red blood cells. The isolated cells from each patient were plated in OptiMEM media containing 10% fetal calf serum containing antibiotic-antimycotic and acclimated for 6 or 7 days in culture with media changed every 24 h. These cells produced progesterone, and their levels in conditioned media were highest during 48–96 h after the initiation of cultures, but then declined to the level of the first 24 h culture period by Day 6 (Al-Alem et al., 2015). During this 6/7-day acclimation period, the expression of key ovulatory genes such as PGR, PTGS2, and AREG was rapidly and markedly down-regulated (Al-Alem et al., 2015). In addition, the expression of CD45, a leukocyte marker, was also completely reduced, indicating the elimination of leukocytes during this acclimation period. More importantly, when human granulosa/lutein cells (hGLCs) were treated in vitro with hCG (1 IU/ml) on Day 6 or 7, we found a rapid and transient increase in the expression of PGR, PTGS2, and AREG as well as the production of progesterone and prostaglandins (PGE₂ and PGF_2a) (Al-Alem et al., 2015; Choi et al., 2017b). This hCG-induced transient increase in PGR, PTGS2, and AREG expression mimicked the in vivo expression profile of these genes in human preovulatory follicles collected throughout the periovulatory period (Choi et al., 2017b). These findings are critical as they demonstrate that this in vitro system can recapitulate in vivo changes in three key ovulatory pathways: progesterone/PGR, prostaglandins, and epidermal growth factor (EGF) signaling pathways. Furthermore, the hGLC model can be utilized to delineate the mechanisms by which these three key mediators coordinate the ovulatory process in humans.

Figure 4.

Experimental paradigm using in vitro cultures of human granulosa/lutein cells. (A) The in vitro experimental paradigm using human granulosa/lutein cells as described (Al-Alem et al., 2015). Follicles were aspirated at 36 h after ovulation induction with rhCG and cumulus–oocyte complexes were removed from follicular aspirates for fertility treatment. The remaining follicular aspirates were subjected to Percoll-gradient centrifugation to remove blood cells. The isolated luteinizing granulosa cells were plated and cultured for 6 days, with a change of media every 24 h. At the end of acclimation, the cells were washed with the fresh media and treated ± hCG ± selected agents. The cells and culture media were collected after defined hours and used for downstream analyses. (B) Representative microscopic images of human granulosa/lutein cells at the time of plating (Day 0), 6 days after preincubation (Day 6), and 36 h after hCG treatment (1 IU/ml). Magnification of all images, 10×.

In the following sections, we focus on key ovulatory genes whose expression was induced and regulated by ovulatory hCG stimulation based on the data obtained from the in vivo and in vitro models described above. These genes were grouped according to their known functions in the ovulatory process. The data were also compared with the existing microarray data generated using granulosa cells collected throughout the preovulatory period from IVF patients (Fig. 5, Tables 1 and 2) (Wissing et al., 2014; Poulsen et al., 2020a). For comparison, the information on specific genes generated using animal models, such as rodents and monkeys, is also included when appropriate.

Figure 5.

Temporal changes in gene expression profile in granulosa cells of follicle aspirates collected throughout the ovulatory period. The expression levels of each gene were extracted from the list of differentially regulated genes in the microarray dataset from Poulsen et al. (2020a). Log2-fold change to 0 h levels was calculated using the mean(log2) value of the gene at each time point.

Table 1.

Differential regulation in gene expression in human, monkey, and mouse ovaries during the ovulatory period.

Up-regulated in all three species	AREG, EREG, STAR, PGR, NR3C1, RUNX1, RUNX2, FOS, JUNB, FKBP5, PTGS2, PLA2GA4, SLCO2A1, MMP19, ADAMTS1, ADAMTS4, ADAMTS9, TIMP1, TFPI2, THBS1, NTS, SCG2, ADAM9
Down-regulated in all three species	CYP17A1, HAD17B1, CYP19A1, HSD11B2, NR3C2
Different among the three species	Gene symbol	Human	Monkey	Mouse
	BTC	up	no	up
	NRG1	no	up
	CYP11A1	up	down	up
	HSD3B1	up	no	down
	HSD3B2	up	up	n/a
	HSD11B1	up		no
	CEBPA	no	up
	CEBPB	n/a	up
	CEBPD	n/a	up	up
	CEBPZ	up	down	up
	JUN	up	n/a	up
	JUND	no	up
	FKBP4	up	down	up
	PTGES	up		down
	AKR1C1	up		down
	ABCC4	up		no
	PTGER1	no	up
	PTGER2	up		no
	PTGER3	down	up	no
	PTGER4	no	n/a	up
	PTGFR	no	up
	MMP1	up		n/a
	MMP10	up		n/a
	ADAMTS3	down	up	up
	ADAMTS7	up	no
	ADAMTS15	up	no	down
	ADAM10	up	no	down
	ADAM12	up	down
	ADAM23	down	n/a	up
	TIMP2	up	no	down
	TIMP3	down	no	up
	TIMP4	no	n/a	up
	VEGFA	down	up
	ACE2	up	no
	CXCL8	up	no
	CXCL16	up	up	no
	CCL20	up	no

Human data are based on microarray datasets (Wissing et al., 2014; Poulsen et al., 2020b).

Monkey data are based on a microarray dataset (Xu et al., 2011).

Mouse data are based on RNA-seq datasets (Shirafuta et al., 2021; Dinh et al., 2023).

Inside the column, ‘up’ and ‘down’ are denoted when the gene expression was up-regulated and down-regulated, respectively, at any time point after ovulation induction (e.g. hCG or GnRH agonist administration) compared to that before ovulation induction (>2-fold change); ‘no’ is denoted when no significant changes are found throughout the ovulatory period. If the data are unavailable, the column is denoted with ‘n/a’.

Table 2.

Datasets for ovulation studies from humans, monkeys, and mice.

Species	Methodology	Material	Time points	Reference	Accession no.
Human	Microarray	GC and CC	36 h after rhCG from IVF	Kõks et  al. (2010)	GSE18559
Human	Microarray	GC and CC	36 h after rhCG from IVF	Grondahl et al. (2012)	E-MEXP-3641
Human	Microarray	GC	before or 36 h after rhCG from IVF	Wissing et al. (2014)	E-MTAB-2203
Human	RNA-seq	CC	36 h after rhCG from IVF or from IVM	Yerushalmi et al. (2014)	GSE50174
Human	RNA-seq	GC	34–36 h after rhCG from IVF or natural cycle	Lu et al. (2019)	GSE124177
Human	Microarray	GC	Before, 12, 17, 32, or 36 h after rhCG or after GnRH agonists from IVF	Poulsen et al. (2020b)	GSE133868
Monkey	Microarray	GC	0, 12, 24, and 36 h after rhCG	Xu et al. (2011)	GSE22776
Mouse	Microarray	CC	0, 8, and 16 h after hCG	Hernandez-Gonzalez et al. (2006)	GSE4260
Mouse	RNA-seq	GC	0, 4, and 12 h after hCG	Shirafuta et al. (2021)	GSE167940

GC, granulosa cells; CC, cumulus cells.

Key ovulatory factors

EGF-like factors

It was first reported in the mouse ovary that ovulatory hCG stimulation induces transient increases in the expression of three members of EGF-like factors [amphiregulin (AREG), epiregulin (EREG), and betacellulin (BTC)] in preovulatory follicles (Park et al., 2004). Later, another member of EGF-like growth factor, neuregulin (Nrg1), was also found to be rapidly and transiently up-regulated in granulosa cells of ovulatory follicles in mice (Noma et al., 2011). These factors function in cells by activating the EGF family of receptor tyrosine kinases, EGFR (ErbB1), ErbB2, ErbB3, and/or ErbB4 (Wieduwilt and Moasser, 2008). Accompanying the induction of EGF-like factors, the presence and activation of their receptors were also reported in granulosa and cumulus cells of ovulatory ovaries in mice (Noma et al., 2011). In vitro functional studies suggested that these EGF-like growth factors play a critical role in cumulus expansion and oocyte meiotic maturation (Park et al., 2004; Noma et al., 2011). Subsequent transgenic studies using Areg^−/−, Ereg^−/−, Areg^−/−; Egfr^wa/wa, Nrg1^flox/flox; Cyp19^cre mice further proved that ovulatory LH-induced increases in EGF-like factors and their receptor activation are necessary for oocyte maturation, cumulus cell expansion, and follicular rupture (Hsieh et al., 2007; Kawashima et al., 2014). In women, the expression of AREG is rapidly and transiently up-regulated in both granulosa and theca cells of ovulatory follicles (Choi et al., 2017b). Poulsen et al. (2020a) also reported a transient increase in AREG and EREG mRNA, peaking at 12 h after ovulatory induction from women undergoing fertility treatment (Fig. 5). Consistent with these findings, AREG protein levels were also increased in human follicular fluid, with the highest level detected at 17-h post-hCG (Poulsen et al., 2019). However, different from mice, NRG1 mRNA was not regulated, while BTC mRNA levels showed a relatively minor increase (<0.2-fold) in granulosa cells of human follicles during the early ovulatory period (Poulsen et al., 2020a). Similar to humans, in rhesus monkey follicles, AREG and EREG mRNA levels, but not BTC and Nrg1, were increased at 12 h, then declined at 24 h after hCG stimulation, and then increased again at 36 h in the ruptured, but not unruptured, follicles (Xu et al., 2011). Consistent with the in vivo findings, hCG treatment increased the level of mRNA for AREG and EREG, not BTC in hGLCs (Park et al., 2004; Ashkenazi et al., 2005; Al-Alem et al., 2015; Choi et al., 2017b). Using hGLCs, we also demonstrated that treatment with AG1478, an EGFR tyrosine kinase inhibitor, suppressed hCG-induced AREG and EREG expression, suggesting the auto-regulation of EGF-like factor expression (Choi et al., 2017b). In addition, treatment with RU486, an antagonist for PGR and NR3C1 (glucocorticoid receptor), also reduced hCG-induced increases in AREG and EREG expression, suggesting the possible role of progesterone/PGR or glucocorticoid/NR3C1 in regulating the expression of EGF-like factors (Choi et al., 2017b). Zamah et al. (2010) demonstrated that AREG is the most abundant EGF-like growth factor in human follicular fluid samples from hCG-stimulated follicles. In that study, the treatment with human follicular fluid obtained at oocyte retrieval in IVF caused cumulus cell expansion and oocyte maturation in cultured mouse follicles. In contrast, the depletion of AREG using anti-AREG antibody completely abolished the ability of the human follicular fluid to induce cumulus expansion and oocyte maturation (Zamah et al., 2010). Together, these data indicate that in humans, the LH surge/ovulatory hCG administration induces marked increases in AREG and EREG expression, but not BTC and NRG1, in preovulatory follicles after hCG administration, and these EGF-like growth factors stimulate cumulus cell expansion and oocyte maturation.

Steroidogenic factors

It is well known that the ovulatory LH surge/hCG stimulation triggers the rapid shift in the steroidogenesis pathway from predominantly estradiol/androstenedione synthesis to progesterone production in preovulatory follicles in many species including humans (Goff and Henderson, 1979; Chaffin et al., 1999a; Komar et al., 2001; Poulsen et al., 2020a; Johannsen et al., 2024). This shift is necessary for luteal transition and has been shown to be primarily mediated by changes in the expression profile of genes involved in steroidogenesis. In humans, microarray data showed a significant increase in levels of mRNA for STAR, a protein involved in cholesterol transport to the mitochondrial inner membrane, after ovulatory stimulation (Fig. 5) (Wissing et al., 2014; Poulsen et al., 2020b). Meanwhile, the expression of CYP11A1, an enzyme converting cholesterol to pregnenolone, was increased and highest at 17 h after ovulation induction but decreased thereafter (Fig. 5). The levels of mRNA for HSD3B1/2, enzymes that convert pregnenolone to progesterone, were also transiently increased after ovulation induction, with the levels increasing at 12 and 17 h and then slightly declining at 36 h (Fig. 5) (Poulsen et al., 2020a). Corresponding to these changes, the concentration of progesterone in the follicular fluid of the same patient samples was dramatically increased within 12 h after ovulation induction (Poulsen et al., 2020b; Johannsen et al., 2024). In contrast, levels of androgens, estrone, and estradiol were decreased in the follicular fluid, consistent with the down-regulated expression of genes involved in steroid synthesis, including CYP17A1, HSD17B1, and CYP19A1 (Fig. 5) (Poulsen et al., 2020a,b; Johannsen et al., 2024). Similarly, in our hGLC experiments, ovulatory hCG treatment induced rapid increases in the levels of mRNA and protein for STAR at 12 h, which continued to 36 h, whereas the increase in CYP11A1 and HSD3B2 expression was modest and only detected at 36 h (Jeon et al., 2023). These data together suggest the rapid shift in follicular levels of estradiol and progesterone is caused primarily by the up-regulation of STAR, CYP11A1, and HSD3B1/2 and down-regulation of CYP17A1, HSD17B1, and CYP19A1 in human preovulatory follicles. Accompanying this shift in sex steroid hormone profiles, the expression of their respective receptors is also up-regulated during the ovulatory period in humans. PGR expression rapidly increases as described in the transcription factor section, whereas the levels of mRNA for estrogen receptor (ESR2) and androgen receptor (AR) decline in preovulatory follicles after ovulatory stimulation (Choi et al., 2017b; Poulsen et al., 2020a).

In addition to the sex steroid hormone profile, a switch in the glucocorticoid profile was also observed during the ovulatory period in women. The follicular fluid level of cortisol, an active glucocorticoid, was increased, while cortisone, an inactive glucocorticoid, was decreased when collected after the LH surge or hCG administration compared to levels prior to ovulation induction (Harlow et al., 1997; Andersen et al., 1999; Yong et al., 2000; Kushnir et al., 2009; Johannsen et al., 2024). This switch is thought to be mainly due to changes in the expression profile of hydroxysteroid 11-beta dehydrogenase isozymes, HSD11B1 and HSD11B2, that catalyze the interconversion of cortisone to cortisol and vice versa. Tetsuka et al. (1997) showed that in non-luteinized human granulosa cells, HSD11B2 mRNA was readily detectable but not HSD11B1, whereas the expression pattern of these genes was reversed in granulosa/lutein cells collected from IVF patients. Our recent study further extended these findings by demonstrating that HSD11B1 expression (mRNA and protein) was markedly increased exclusively in granulosa cells after hCG administration (Fig. 6) (Jeon et al., 2023). The microarray data of Poulsen and colleagues also agreed with our findings (Fig. 5) (Poulsen et al., 2020a; Johannsen et al., 2024). Corresponding with the increase in follicular cortisol levels, the expression of the glucocorticoid receptor (NR3C1) in preovulatory follicles increased after hCG administration in women (Fig. 5) (Jeon et al., 2023; Johannsen et al., 2024). Similar to in vivo findings, hCG treatment increased the expression of HSD11B1 and NR3C1 in hGLCs (Jeon et al., 2023). Importantly, hCG also increased the levels of progesterone and cortisol in hGLC cultured media, although the fresh media had no detectable levels of progesterone, corticosterone, cortisol, and cortisone (Jeon et al., 2023). These data suggest that granulosa cells have the ability to produce cortisol. Therefore, follicular cortisol increased after the ovulatory hCG stimulation, likely through the conversion of circulating cortisone to cortisol and de novo synthesis of cortisol. This cortisol acts on its receptor, NR3C1, in the dominant follicle to mediate the ovulatory process in humans.

Figure 6.

HSD11B1 and MMP9 expression in the ovulatory follicle. A dominant follicle was collected at the late ovulatory phase (18–34 h after hCG administration) from a normally cycling woman. Follicle tissue sections were subjected to immunohistochemistry for HSD11B1 and MMP9. The positive staining for HSD11B1 (pink) was localized predominantly to the granulosa cell layer, whereas MMP9-positive cells (purple staining) were localized throughout the theca layer as well as in and outside of blood vessels, suggestive of leukocytes. GC; granulosa cells, TC; theca layer, BV; blood vessel. (*) was noted to mark the same location in serial sections. Bars represent 200 µm.

In addition to glucocorticoids, the levels of mineralocorticoids are increased in human follicular fluid during the ovulatory period. For example, the level of 11-deoxycorticosterone, an active mineralocorticoid, is transiently increased, peaking at 12–32 h and then declining at 36 h after ovulatory stimulation (Johannsen et al., 2024). The increase in mineralocorticoids is followed by the accumulation of follicular fluid corticosterone, a metabolite of 11-deoxycorticosterone, at 32- and 36-h post-hCG in the human ovulatory follicle (Johannsen et al., 2024). Corticosterone has weak glucocorticoid and mineralcorticoid potencies in humans. Different from the glucocorticoid receptor (NR3C1), levels of mRNA for the mineral receptor (NR3C2) are initially down-regulated until 17 h but then slightly increase at the late ovulatory period (34- to 36-h post-ovulation induction) (Johannsen et al., 2024). At present, little to nothing is known about whether the increase in mineralocorticoids plays a role in the ovulatory process and luteal transformation in any species.

These findings of steroid hormone levels and genes involved in steroidogenesis in humans are aligned with the data from monkey studies. For instance, the follicular fluid levels of progesterone, 11-deoxycorticosterone, corticosterone, and cortisol are increased, whereas estradiol and cortisone decline in ovulatory follicles after ovulatory hCG stimulation in monkeys (Fru et al., 2006; Ravisankar et al., 2021). These changes correspond to increases in STAR and HSD11B1 mRNA and decreases in CY17A1, CYP19A, and HSD11B2 mRNA in macaque periovulatory follicles (Xu et al., 2011).

Transcription factors

The LH surge causes rapid and extensive reprogramming of the gene expression profile in preovulatory follicles, bringing about intra- and extra-cellular changes required for ovulation and concurrent luteinization (Wissing et al., 2014; Poulsen et al., 2020a). The transcription factors induced by the ovulatory stimulation have been shown to play an essential role in mediating these cellular changes by directly controlling the expression pattern of a diverse array of genes in preovulatory follicular cells. Among those included are PGR, RUNX1/2, C/EBPA/B, AP-1 transcription factors, and NR3C1. Detailed information on these transcription factors is described below.

Progesterone receptor

The induction of Pgr expression in preovulatory follicles after the LH surge was first reported in rat ovaries (Park-Sarge and Mayo, 1994). Since then, the transient up-regulation of PGR mRNA levels in preovulatory follicles has been reported in many species, including rodents, ruminants, and non-human primates (Park-Sarge and Mayo, 1994; Chaffin et al., 1999b; Jo et al., 2002). Previous studies using transgenic mouse lines (e.g. Pgr^−/− and Pgr^flxo/flox; Esr2^cre/+) and pharmacological inhibitors of PGR such as RU486, CDB2914, and Ulipristal acetate have demonstrated that the preovulatory induction of PGR is essential for successful ovulation (Loutradis et al., 1991; Brannstrom 1993; Robker et al., 2000; Gaytan et al., 2004; Palanisamy et al., 2006; Nallasamy et al., 2013; Park et al., 2020). Since then, an array of PGR-downstream regulated genes has been identified in rodent studies (Kim et al., 2009b; Akison et al., 2018; Dinh et al., 2019). PGR-downstream genes found to be necessary for the ovulatory process include Adamts1, Pparg1, HIFs, Ptgs2, Rgcc, Ctsl, and Edn2 (Robker et al., 2000; Palanisamy et al., 2006; Kim et al., 2008, 2009b; Park et al., 2020; Dinh et al., 2023). In humans, similar to animal models, the level of mRNA for PGR was transiently increased at the early ovulatory period but began to decline during the late ovulatory period, then returned back to the preovulatory levels at the post-ovulatory phase (Choi et al., 2017b). Meanwhile, the protein staining for PGR appears to be somewhat delayed, as the most pronounced staining was detected in granulosa cells of late ovulatory follicles (Choi et al., 2017b). Consistent with these findings, the microarray data also showed the transient induction of PGR mRNA, a peak at 12–17 h after ovulation induction, and a decrease to the baseline at 32 h (Poulsen et al., 2020a). In hGLC, hCG treatment also induced transient increases in PGR expression (mRNA and protein) with a first peak between 6 and 12 h (∼6-fold increase over control), followed by a down-regulation to control levels at 24 h (Al-Alem et al., 2015; Choi et al., 2017b). Interestingly, our recent study showed a second increase in PGR expression at 36 h after hCG, albeit the fold changes were moderate (∼2-fold increase to control levels) (Jeon et al., 2023). These biphasic changes (first peak during the early ovulatory period, followed by a smaller up-regulation around the time of ovulation) were also observed in cattle (Jo et al., 2002) and non-human primates (Chaffin et al., 1999b). Furthermore, in hGLC, treatment with RU486 inhibited the expression of genes involved in PG synthesis and transport (PTGS2, PTGES, AKR1C1, SLCO2A1, and ABCC4) and EGF-like peptides (AREG and EREG) as well as FOS and CXCR4 (Choi et al., 2017a,b). Many of these genes are known to play a critical role in ovulation, as described above or below, indicating that PGR mediates the ovulatory process by regulating the expression of these key genes in human ovulatory follicles.

Clinically, pharmacological inhibitors of PGR, such as RU486 and Ulipristal acetate, have been employed as emergency contraception regimes due to their ability to delay or block ovulation (Luukkainen et al., 1988; Danforth et al., 1989). However, the inhibitory effects of these two drugs on ovulation differ. RU486 treatment inhibited ovulation by blocking the induction of the LH surge (Ledger et al., 1992; Escudero et al., 2005). Meanwhile, Ulipristal acetate delayed ovulation by at least 5 days when given before, as well as after, the onset of the LH surge [100% (8/8) and 78.6% (11/14), respectively], but not when given on the day of the LH peak (Brache et al., 2010). Using a non-human primate follicle injection model, Bishop et al. (2016) showed that silencing PGR expression in periovulatory follicles by short-hairpin RNAs specific for PGR (shPGR) inhibited hCG-induced ovulation and progesterone production. These data, taken together, indicate that PGR is a critical mediator of the ovulatory process and luteinization for all animal models studied thus far.

RUNX transcription factors

RUNXs are a small group of transcription factors composed of three structurally similar members of a DNA binding protein, RUNX1, RUNX2, and RUNX3, and their common partner, non-DNA binding protein, Core Binding Factor beta (CBFB). These proteins function as a heterodimeric complex (RUNX/CBFb) (Coffman 2003). The LH surge/ovulatory hCG administration induces rapid increases in Runx1 and Runx2 expression, but not Runx3, in preovulatory follicles of rodent ovaries (Hernandez-Gonzalez et al., 2006; Jo and Curry, 2006; Park et al., 2012; Wilson et al., 2016). Interestingly, their expression profile was different in that Runx1 expression was transient in preovulatory follicles, while the up-regulated expression of Runx2 was sustained in the CL (Park et al., 2012; Wilson et al., 2016). Mice deficient in Cbfb alone (Cbfb^flox/flox; Esr2^cre/+) (Lee-Thacker et al., 2018) or both Cbfb and Runx2 (Cbfb^flox/flox; Runx2^flox/flox; Esr2^cre/+) (Lee-Thacker et al., 2020) in granulosa cells exhibited profound defects in ovulation and luteinization, underscoring the functional significance of this family of transcription factors in the ovary. Interestingly, a recent study showed that RUNX1 interacts with PGR, and their interaction is important for PGR-derived ovulatory transcriptome changes (Dinh et al., 2023). Similar to mice and rats, the expression of both RUNX1 and RUNX2 was highly up-regulated in human preovulatory follicles during the early ovulatory phase after hCG administration, yet during the late ovulatory phase, RUNX1 mRNA levels were slightly down-regulated, whereas RUNX2 mRNA levels were sustained (Park et al., 2012). These data results are consistent with the microarray data using human granulosa cells from IVF patient samples (Fig. 5) (Wissing et al., 2014; Poulsen et al., 2020a).

C/EBP alpha/beta transcription factors

CCAT/enhancer-binding proteins (C/EBPs) are encoded by at least six different genes, Cebpa, Cebpb, Cebpd, Cebpg, Cebpz, and Cebpe, that contain a highly conserved basic leucine zipper domain that is required for dimerization and DNA binding (reviewed in Ramji and Foka (2002)). In the mouse ovary, the ovulatory hCG administration induces a rapid and transient increase in the expression of Cebpa and Cebpb. The levels of Cebpa and Cebpb are highest at 4 h in granulosa cells of preovulatory follicles but gradually decrease in the corpus luteum by 48 h after hCG administration (Fan et al., 2009, 2011). An early study by Sterneck et al. (1997) showed that Cebpb null mice (Cebpb^−/−) exhibited reproductive defects in the ovary, including ovulation and luteinization failure, indicating that C/EBPB plays an important role in the ovulatory process. A later study using the targeted disruption of the Cebpb gene in granulosa cells of preovulatory follicles (Cebpb^fl/fl; Cyp19^cre), but not the Cebpa gene (Cebpa^fl/fl; Cyp19^cre), showed reduced, partially impaired ovulation and luteinization, resulting in subfertility in female mice (Fan et al., 2009, 2011). Cebpa/b double-mutant female mice (Cebpa^fl/fl; Cebpb^fl/fl; Cyp19^cre)were sterile; follicles failed to ovulate, cumulus–oocyte complex (COC) expansion was compromised, and ovaries were devoid of corpora lutea (Fan et al., 2011). Microarray analyses identified numerous genes affected in the ovary of Cebpa/b double-mutant female mice, many of these genes were found to be associated with luteinization and vascularization (e.g. Akr1b7, Runx2, Star, Saa3, Abcb1b, Apln, Igfbp4, Prlr, Ptgfr, Timp4, and Bhmt), but not early ovulatory events (e.g. Areg, Ereg, Ptgs2, Tnfaip6, Ptx3, and Il6) (Fan et al., 2011). These findings indicate that C/EBPA/B has a prominent role in the late ovulatory process, including terminal differentiation of granulosa cells and vascularization during luteinization in mice (Fan et al., 2011). Cebpd expression is also induced by hCG administration in ovarian theca and interstitial cells but is not essential for mouse ovarian function (Huang et al., 2007). Similar to mice, Cebpb expression is rapidly and highly up-regulated in rat granulosa cells of preovulatory follicles after hCG administration (Sirois and Richards, 1993). An ex vivo perfusion study in rats showed that reduction of ovarian Cebpb expression by using Cebpb-specific antisense oligonucleotide inhibited ovulation (Pall et al., 1997). In vitro studies using rat granulosa cells show that C/EBPB regulates the transcription of genes known to be involved in ovulation and luteinization, such as Ptgs2, Cyp11a1, and Vegf (Sirois and Richards, 1993; Okada et al., 2016; Shinagawa et al., 2019). Interestingly, the Cebpa expression pattern in rats is different from mice in that the expression is highest at 24 h after PMSG administration, decreased by 10-h post-hCG, and then increased again in newly forming CL (Piontkewitz et al., 1993). A subsequent study by this group showed that Cebpa small interfering RNA-induced Cebpa knockdown resulted in impaired follicular development and reduced ovulation rate, indicating that CEBPA is an important regulator for follicular development in the rat ovary (Piontkewitz et al., 1996). These data indicate a critical role(s) of C/EBPA/B in follicular development and the ovulatory process in the rat ovary but also show a species difference in the expression of this family of transcription factors. This also appears to be true in the human ovary. Microarray data showed that the levels of CEBPA and CEBPZ mRNA was increased at 32 and 12 h after ovulatory hCG administration, respectively, yet the increase in CEBPA was only modest at best (∼1.3-fold) (Fig. 5) (Poulsen et al., 2020a). Surprisingly, CEBPB was not included in the list of differentially regulated genes in granulosa cells after hCG administration (Wissing et al., 2014; Poulsen et al., 2020a). Meanwhile, in human granulosa/lutein cells, both C/EBPA and C/EBPB were detected, and 8-Br-cAMP treatment increased C/EBPB levels (Christenson et al., 1999). Therefore, it will be worthwhile to further characterize the expression of C/EBPB and other C/EBP family members in preovulatory follicle samples in human ovaries.

FOS/AP-1 transcription factor

AP-1 is a group of transcription factors comprising four subfamilies: Fos (cFos, Fosb, Fosl1, Fosl2), Jun (cJun, JunB, JunD), ATF, and Maf (Hess et al., 2004). This family of proteins contains a basic leucine zipper (bZip) domain that is required for dimerization with other members of AP-1 and interaction with DNA at the specific sequences including 12-O-tetradecanoylphorbol-13-acetate (TPA) responsive elements (TRE; TGA(G/C)TCA) and cAMP responsive elements (CRE; TGACGTCA, reviewed in Garces de Los Fayos Alonso et al. (2018)). Fos and Jun family members are the most well-characterized members of the AP-1 transcription factors. To be functional as transcriptional regulators, AP-1 proteins need to form homo- or hetero-dimeric complexes with their own members. Fos proteins form a complex with one of the Jun family proteins, whereas Jun proteins can also form a homo- or hetero-dimer with their own family members (Nakabeppu et al., 1988; Mechta-Grigoriou et al., 2001). Different dimer complexes exert different DNA binding and transcriptional activities (Halazonetis et al., 1988; Mechta-Grigoriou et al., 2001), causing a diversity and complexity in the action of AP1 depending on what members are expressed in any given cell or tissue. In the ovary, previous studies have reported the rapid induction of Fos and Jun expression in granulosa cells after PMSG injection in rats (Delidow et al., 1990) and by FSH and LH in cultured rat granulosa cells (Ness and Kasson, 1992). Fos null mice were infertile, with the ovary displaying no CL and arrested follicular development at the secondary follicle stage (Xie et al., 2015). Although these ovarian defects were explained owing to reduced levels of FSH and LH release from the pituitary, it was also reported that Fos null mice failed to ovulate and form CL even when exogenous gonadotropin was administered (Xie et al., 2015). This indicated that ovarian expression of Fos is critical for ovulation and CL formation. In human ovaries, recent studies reported that the expression of FOS and its partners, JUN, JUNB, and JUND, was increased in preovulatory follicles after ovulatory hCG administration from regularly cycling women (Choi et al., 2018). Similarly, the microarray data listed FOS, FOSL1, FOSL2, JUN, JUNB, and JUND as differentially up-regulated genes in granulosa cells from IVF patient samples (Fig. 5) (Wissing et al., 2014; Poulsen et al., 2020a). Choi et al. (2021a) further determined the regulation and specific function of FOS/AP-1 using hGLCs. This study showed that hCG induced a biphasic increase in the expression of FOS, showing the first peak at 1–3 h after hCG stimulation, which was followed by a rapid down-regulation of FOS expression at 6 h, and then increases again at 12 h. With each increase in FOS mRNA expression, there were corresponding increases in protein accumulation at respective time points. The increase in mRNA and protein was completely down-regulated by 24 h after hCG stimulation. Meanwhile, hCG increased the expression of all three Jun proteins, although with varying expression profiles (Choi et al., 2021a). These studies together suggest the presence of multiple forms of FOS/AP-1 complexes (e.g. FOS/JUN, FOS/JUNB, FOS/JUND) in granulosa cells of preovulatory follicles, although the ratio of each FOS/AP-1 complex is likely different at the early and late ovulatory period depending on the availability of different Jun proteins present at those time points.

The specific function of the FOS/AP-1 was investigated by treating hGLCs without or with hCG±T5224, a specific inhibitor of FOS. The results of this study identified a number of genes affected by FOS inhibition (Choi et al., 2021a). Many of these genes are known to be involved in the ovulatory process and luteinization. These included genes involved in matrix remodeling (e.g. HAS2, PTX3, F3, ADAMTS1, ADAMTS9, TFPI2), PG synthesis and transport (e.g. CD24, SLCO2A1, PTGES), transcription factors (e.g. RUNX1, ID2), cell signaling molecules (e.g. MAPK14, RGS2), glucose uptake/glycolysis (SLC2A1/2, PFKFP3/4, and ACSS1/2), and cholesterol biosynthesis (HMGCS1, MVK, MVD, FDPS, SQLE, LSS, CYP51A1, EBP, DHCR24) (Choi et al., 2021a). These data demonstrate that the FOS/AP-1 plays a critical role in periovulatory changes by regulating the expression of a diverse array of genes in ovulatory follicles.

Glucocorticoid receptor (NR3C1)

NR3C1 functions as a ligand-dependent transcriptional regulator. Upon binding to glucocorticoids (e.g. cortisol), NR3C1 undergoes the activation process: the molecular co-chaperone, FKBP prolyl isomerase 5 (FKBP5) is dissociated from the receptor complex in exchange of another co-chaperon, FKBP prolyl isomerase 4 (FKBP4). This exchange promotes the translocation of the NR3C1 complex from the cytosol to the nucleus, where the complex binds to glucocorticoid response elements on the promoter or enhancer regions of its target genes (Binder 2009; Zannas et al., 2016). A recent study reported the increased expression of NR3C1, FKBP4, and FKBP5 (mRNA and protein) in granulosa cells of preovulatory follicles after hCG administration from normally cycling women (Jeon et al., 2023). The microarray data also showed an increase in NR3C1 and FKBP5 mRNA in granulosa cells obtained after ovulation stimulation (Fig. 5) (Wissing et al., 2014; Johannsen et al., 2024). Similar to humans, in the monkey ovary, the increased expression of NR3C1 was detected in granulosa cells, cumulus cells, and oocytes of preovulatory follicles after hCG administration (Ravisankar et al., 2021). In hGLCs, hCG treatment increased NR3C1 and FKPB5 expression, mimicking the in vivo expression pattern. Using this in vitro model, the specific function of NR3C1 was investigated by treating the cells in the absence or presence of hCG±CORT125286, a selective antagonist for NR3C1. The results showed that the inhibition of NR3C1 reduced progesterone and cortisol production, and decreased the expression of enzymes involved in steroidogenesis, such as STAR, HSD3B, CYP11A1, and HSD11B1. In addition, CORT125286 treatment decreased the levels of mRNA for several hCG-induced genes such as AREG, EREG, RGS2, and FKBP5. Together, these data indicate that NR3C1 functions as a critical transcriptional regulator involved in the periovulatory processes in humans (Jeon et al., 2023).

Prostaglandins

The increase in PG production in the preovulatory follicle after the LH surge or ovulatory hCG administration has been documented in various species, including rodents, cattle, and non-human primates (Bauminger and Lindner, 1975; Liu et al., 1997; Duffy and Stouffer, 2001; Park et al., 2020). These increases were mediated by LH/hCG-induced increases in the expression of genes involved in PG synthesis and transport. In humans, our recent study has documented an increase in the expression of PTGS2, a rate-limiting enzyme in PG synthesis in preovulatory follicles. The levels of PTGS2 mRNA were highest during the early ovulatory period, yet protein was localized to granulosa cells of late ovulatory follicles (Choi et al., 2017b). Interestingly, PTGS2 mRNA levels were again increased after ovulation in post-ovulatory follicles. Poulsen’s human microarray data also showed a biphasic increase in PTGS2 mRNA, the first peak at 12 h and a second rise at 36 h after hCG administration. The discrepancy in the timing of the second rise (at 36 h in Poulsen’s microarray vs post-ovulatory period in our data) could be that our late ovulatory period spans between 18 and 34 h after hCG, thus missing the increasing trend at 36 h. In hGLCs, hCG treatment induced a transient increase in PTGS2 expression: the levels were highest at 12 h (Al-Alem et al., 2015; Choi et al., 2017b), although the second rise seen in vivo was not observed in vitro even when the expression was measured at 36 h after hCG stimulation. We interpret this finding to mean that the second rise in PTGS2 expression likely requires other paracrine stimulators, possibly coming from leukocytes, or perhaps due to the biophysical difference between in vivo and in vitro systems. Meanwhile, other enzymes involved in the synthesis of PGs (e.g. PTGES and AKR1C1) and in the transport of PGs (SLCO2A1 and ABCC4) were increased, whereas the levels of mRNA for HPGD, an enzyme metabolizing PGs in hGLCs, were decreased (Choi et al., 2017b). The microarray data also showed genes involved in PG synthesis and transport (e.g. PLA2G4A, PTGS2, PTGES, AKR1C1, SLCO2A1, and ABCC4) were up-regulated in granulosa cells after ovulatory hCG administration (Fig. 5) (Poulsen et al., 2020a). These human data are well aligned with the data from animal models, which also showed increases in the expression of many of these same genes in preovulatory follicles during the ovulatory period in various species, including mice, rats, cattle, and monkeys. These findings were described in a comprehensive review by Duffy and coworkers (Duffy 2015). This indicates that the PG system is one of the most conserved ovulatory pathways across different species.

Similar to the in vivo changes, hCG stimulated PGE₂ and PGF_2a production in hGLCs (Choi et al., 2017b). Further studies showed that this rise in PGs and the expression of genes involved in PG synthesis and transport were reduced by the treatment with RU486, a dual antagonist for PGR and NR3C1, and by AG1478, EGF receptor tyrosine kinase inhibitor, indicating that the up-regulation of PGs was a result of collaborative actions of P4/PGR and EGF signaling (Choi et al., 2017b).

PGE₂ exerts its effect by binding to multiple E series of receptors, EP1 (PTGER1), EP2 (PTGER2), EP3 (PTGER3), and EP4 (PTGER4), and each is coupled to different intracellular signaling pathways (reviewed in Duffy (2015)). The microarray data of human granulosa cells showed regulated expression of genes encoding PGE₂ and PGF_2a receptors during the ovulatory period (Poulsen et al., 2020a). For example, PTGER2 mRNA levels were increased, a first peak at 12–17 h and then a second rise at 36 h, whereas PTGER3 mRNA levels were decreased after ovulation induction (Fig. 5). Additionally, PGF_2a receptor (PTGFR) mRNA levels were increased at 36 h after ovulation induction (Fig. 5) (Poulsen et al., 2020a). In vitro studies using specific inhibitors for these receptors showed the expression and presence of functional EP1, EP2, and EP4 in human granulosa/lutein cells (Harris et al., 2001; Narko et al., 2001). Similar to humans, multiple subtypes of EP receptors were also reported to be differentially expressed in granulosa and cumulus cells of monkey periovulatory follicles; EP1 protein levels increased in mural granulosa cells, and EP2 and EP3 protein levels increased in cumulus cells after hCG administration, whereas EP4 expression was low in both cell types (Harris et al., 2011).

PGs have been shown to function as critical inflammatory mediators in various systems. Key aspects of the ovulatory process resemble an acute inflammatory response, displaying swelling of the follicular apex, increased blood flow, and infiltration of large numbers of leukocytes (reviewed in Duffy et al. (2019)). Therefore, it has been suggested that the PGs produced in preovulatory follicles play an important role in ovulation by acting as critical inflammatory mediators. Indeed, Ptgs2 null mice (Ptgs2^−/−) mice showed defective ovulation, cumulus cell expansion, and fertilization (Lim et al., 1997; Davis et al., 1999). Additionally, mice deficient in EP2, a subtype of the PGE2 receptors (EP2^−/−), also displayed defective cumulus cell expansion and reduced ovulation rates and fertilization, resulting in reduced fertility (Hizaki et al., 1999). These data indicate that periovulatory Ptgs2-derived PGE₂ plays a critical role in ovulation and cumulus cell expansion by acting through EP2. In support of this concept, in rhesus monkeys, intrafollicular administration of indomethacin, a PG synthesis inhibitor, prevented ovulation, and the growth of capillaries around ovulatory follicles. Meanwhile, the replacement with PGE₂, but not PGF_2a, restored ovulation and follicular angiogenesis (Duffy and Stouffer, 2002; Kim et al., 2014). Similarly, in humans, the administration of PTGS2 selective inhibitors (e.g. rofecoxib, meloxicam, and celecoxib) delayed follicle collapse, a measure of ovulation in humans, compared to the placebo group (Pall et al., 2001) (Bata et al., 2006; Jesam et al., 2010). At present, the specific actions PGE₂ exerts on follicular cells and how PGE₂ controls ovulation and cumulus cell expansion remain elusive, particularly in humans. One clue came from recent studies showing that PGE₂ stimulated the migration of human and monkey ovarian vascular endothelial cells in cultures (Trau et al., 2015, 2016). These data indicate that PGE₂ contributes to ovulatory angiogenesis by acting directly on endothelial cells of ovulatory follicles.

Proteolytic systems

During the ovulatory period, the preovulatory follicle undergoes massive morphological changes that culminate in the rupture of the follicle wall at the apex and extensive remodeling of the collapsed follicle into a CL. These structural changes are mediated by the actions of an array of proteinases expressed in and around the periovulatory follicle. Among these proteinases increased during the ovulatory period include the matrix metalloproteinase (MMP) family, a disintegrin and metalloproteinase (ADAM), and ADAM with a thrombospondin motifs (ADAMTS) families and the plasminogen activator (PA)/plasmin system. These proteases are thought to be involved in breaking down the basement membrane and extra-cellular matrix surrounding the preovulatory follicle. Exogenous chemical inhibitors of MMPs or antibodies against MMPs blocked ovulation in rodents and non-human primates (Ichikawa et al., 1983; Reich et al., 1985; Brannstrom et al., 1988; Gottsch et al., 2002; Peluffo et al., 2011), demonstrating that these proteinases play a critical role in ovulation. The activity of these proteinases is controlled by proteinase inhibitors such as serum-born inhibitors (e.g. α2-macroglobulin and ovostatin), tissue inhibitors of metalloproteinases (TIMPs), and serine protease inhibitors, including tissue factor pathway inhibitors 2 (TFPI2). Previous studies have shown that the LH surge/ovulatory hCG stimulation increases the expression and activity of various proteinases and their inhibitors in preovulatory follicles, although the type of proteinases and inhibitors up-regulated varies among different species. Since previous reviews on this topic provide comprehensive information generated from studies using animal models and humans (Curry and Osteen, 2003; Curry and Smith, 2006; Duffy et al., 2019), the following sections will mainly focus on the expression of these proteinases and their inhibitors generated using human ovarian samples obtained throughout the periovulatory period. In addition, the expression of genes found to be up-regulated in human ovulatory follicles was compared with those in monkeys and mice (Table 1).

Matrix metalloproteinases

The MMP family is composed of at least 25 related proteolytic enzymes (reviewed in Nagase and Woessner (1999), Curry and Osteen (2003)). These enzymes are divided into six classes: collagenases, gelatinases, stromelysins, matrilysins, membrane-type enzymes (MT-MMPs), and other non-classified MMPs, depending on substrate recognition site and specificity (Kleiner and Stetler-Stevenson, 1993). Collagenases (MMP1, MMP8, and MMP13) can cleave various types of collagens, resulting in the denaturation of the collagen molecule into gelatin. Among these collagenases, MMP1 was the only proteinase that was highly up-regulated after ovulatory hCG administration in both granulosa and theca cell compartments (Rosewell et al., 2015). MMP1 expression was highest during the late ovulatory phase, whereas MMP8 and MMP13 mRNA expression was extremely low and did not change throughout the periovulatory period in either granulosa or theca cells (Rosewell et al., 2015). As for gelatinases (MMP2 and MMP9), Western-blot analysis of the entire dominant follicle showed no change in either MMP2 or MMP9 levels throughout four different ovulatory phases (Lind et al., 2006). Recent single-cell RNAseq data generated using follicular aspirates from four IVF patients indicated the presence of MMP2 mRNA only in theca cells and macrophages and MMP9 mRNA in leukocytes (Choi et al., 2023). Immunohistochemical analysis using late ovulatory follicles and post-ovulatory follicles also localized MMP9 predominantly in leukocytes (Fig. 6). In addition to these data, follicular fluid of IVF patients was shown to have both MMP2 and MMP9 activity (Nikolettos et al., 2003). These data indicate that the gelatinases are present in the ovulatory follicle, and the sources of these enzymes are theca cells and leukocytes. Besides gelatinases, stromelysins (MMP3, MMP7, MMP10, MMP11) are capable of degrading gelatin, type IV collagen, lamin, and fibronectin that are all major constituents of basement membranes between granulosa and theca layers and underlying the ovarian surface epithelium (Birkedal-Hansen et al., 1993; Nagase and Woessner, 1999). In the human ovary, MMP10 mRNA was transiently up-regulated in preovulatory follicles after hCG administration (McCord et al., 2012), whereas expression of other stromelysins was either extremely low or down-regulated during the ovulatory period. Another group of MMPs, the membrane type (MT)-MMPs (MMP14-17 and MMP24) were either undetectable (low) or appeared to be unchanged in periovulatory follicles throughout the ovulatory period (Rosewell et al., 2015). Lastly, among other non-classified MMPs, MMP19 expression was up-regulated both in granulosa and theca cells of preovulatory follicles collected at the late ovulatory phase (Rosewell et al., 2015). Similar to these data, Poulsen’s microarray data also showed a transient increase in MMP1 and MMP10 mRNA at 12 h in granulosa cells. Meanwhile, the increase in MMP19 mRNA in granulosa cells was continued and highest 36 h after ovulatory stimulation (Poulsen et al., 2020a). In addition to these MMPs, MMP3, MMP8, MMP14, and MMP15 mRNA were detected in the microarray data, yet their increases were relatively minor (<2-fold change, Fig. 5) (Poulsen et al., 2020a). Together, these data indicate that hCG administration induces a substantial increase in MMP1, MMP10, and MMP19 expression. Although not increased in granulosa cells after ovulatory induction, MMP2 and MMP9 are also present in the follicular fluid of ovulatory follicles, and these two gelatinases are likely derived from theca cells and leukocytes. In the monkey ovary, MMP1, MMP10, and MMP19 mRNA levels were also up-regulated. However, the expression patterns were slightly different in that the first peak was detected at 12 h after hCG administration, followed by a secondary increase in 36-h post-ovulatory follicles (Peluffo et al., 2011). In addition, MMP9 mRNA levels were also transiently increased at 12-h post-hCG in monkeys, although no secondary rise was observed at later time points (Peluffo et al., 2011). In the rodent ovary, the expression of mRNA for Mmp1, Mmp2, Mmp9, Mmp13, Mmp14, and Mmp19 were increased in the preovulatory ovaries after hCG administration (Reich et al., 1991; Hagglund et al., 1999; Robker et al., 2000; Curry et al., 2001; Jo and Curry 2004; Jo et al., 2004). Notably, the increase in Mmp2, Mmp9, Mmp13, and Mmp14 was limited to the residual tissue in both mouse and rat ovaries, whereas the expression of Mmp1 in rats and Mmp19 in both mice and rats was in both granulosa cells and residual tissues (Reich et al., 1991; Hagglund et al., 1999; Jo and Curry 2004). Together, these data indicate the conserved up-regulation of specific MMPs after hCG administration but also showed species-specific differences in the type and expression pattern of these MMPs in the ovulatory follicle.

In hGLC, similar to the in vivo expression pattern, MMP10 mRNA was rapidly and transiently up-regulated by hCG stimulation (the levels were highest at 3 h after hCG and gradually decreased), whereas MMP19 mRNA levels were increased at 36 h after hCG. Yet, different from the in vivo expression pattern, hCG stimulation had no effect on MMP1 expression in hGLC (Rosewell et al., 2015). These data suggest that the LH/hCG-activated signaling cascade is the primary stimulator of MMP10 and MMP19 expression, whereas the induction of MMP1 requires additional factor(s) in human granulosa cells.

A disintegrin and metalloproteinase (ADAM) and ADAM with thrombospondin motifs (ADAMTS)

Similar to MMPs, the ADAM and ADAMTS families are composed of multiple members. So far, 21 ADAM proteases and nineteen ADAMTS proteases have been reported in humans (Porter et al., 2005; Edwards et al., 2008). They also exert diverse actions; degradation of proteoglycans (e.g. aggrecan, brevican, and versican), processing of procollagens to collagen, antiangiogenic activity, basal lamina remodeling, and shedding of diverse growth factors (Porter et al., 2005; Edwards et al., 2008; Stanton et al., 2011). Several members of the ADAM and ADAMTS family were found to be expressed in the ovary, and some of them were up-regulated during the ovulatory period in various species including rodents, cows, and primates (reviewed in Duffy et al. (2019)). In normally cycling women, ADAMTS1 and ADAMTS9 mRNA levels were transiently up-regulated in granulosa cells, but not in theca cells of dominant follicles collected after ovulatory hCG administration; the levels were highest at the early ovulatory phase (Rosewell et al., 2015). Poulsen’s microarray data also showed marked increases in the level of mRNA for ADAMTS1 and ADAMTS9 in granulosa cells, although their expression pattern differs in that the levels were increased as early as 12 h, but continued to increase and were highest at 36 h after hCG stimulation (Poulsen et al., 2020a). In addition, the microarray data showed that the levels of mRNA for ADAMTS4, 7, and 15 were increased after ovulatory induction, whereas ADAMTS3 mRNA was decreased (Poulsen et al., 2020a). Similarly, studies in the monkey ovary showed an increase in mRNA for ADAMTS1, 4, 9, and 15 in ovulatory follicles after hCG administration (Peluffo et al., 2011). Increases in Adamts1 and Adamts4 were also reported after hCG administration in the mouse ovary (Richards et al., 2005). Besides ADAMTS, levels of mRNA for several ADAMs (ADAM9, 10, 12, and 17) were up-regulated after ovulatory stimulation in the microarray data and the level of induction varies among different ADAMs (Fig. 5) (Poulsen et al., 2020a). In the experiments with hGLCs, hCG treatment increased ADAMTS1 mRNA levels but not those of ADAMTS9 (Rosewell et al., 2015). Together, these data indicate that various members of the ADAMTS (ADAMTS1, 4, 7, 9, and 15) and ADAMs (ADAM9, 10, 12, and 17) are up-regulated in granulosa cells of preovulatory follicles after ovulatory hCG stimulation.

Tissue inhibitors of metalloproteinases

There are four members of the TIMP family (TIMP1, 2, 3, 4), and each of them differs in their selectivity toward different MMPs and ADAMTS (Gomez et al., 1997; Brew and Nagase, 2010; Arpino et al., 2015). In the human ovary, in vivo data of dominant follicles showed that TIMP1 expression (mRNA and protein) is increased in all ovulatory phases compared to that of preovulatory phases, with immunopositive staining localizing to both granulosa and theca cell layers (Lind et al., 2006). Consistent with these data, both Poulsen’s and Wissing’s microarray data listed TIMP1 and TIMP2 as up-regulated, whereas TIMP3 as down-regulated, in granulosa cells collected after ovulatory stimulation (Fig. 5) (Wissing et al., 2014; Poulsen et al., 2020a). scRNA-seq data from follicular aspirates of IVF patients showed the presence of mRNA for TIMP1, TIMP2, and TIMP3 in granulosa cell clusters, but TIMP4 mRNA was barely detectable (Choi et al., 2023). Similarly to humans, the increase in Timp1 expression in preovulatory follicles after ovulatory hCG stimulation was also documented in monkeys and rodents (Mann et al., 1993; Chaffin and Stouffer 1999; Hagglund et al., 1999). Together, these studies indicate that TIMP1, TIMP2, and TIMP3 are expressed in human preovulatory follicles; among these, TIMP1 and TIMP2 likely play a role in ovulatory matrix remodeling by controlling proteinase activities.

Tissue factor pathway inhibitor 2

TFPI2 belongs to a Kunitz family of serine protease inhibitors and can inhibit multiple proteases including trypsin, plasmin, certain MMPs, plasma kallikrein, factor Xia (FXIa), and FVIIa/tissue factor (TF) complex (Chand et al., 2005; Kobayashi and Imanaka, 2021). In the ovary, TFPI2 was first detected in follicular fluid obtained from women undergoing in vitro fertilization (Seppala et al., 1985). During the preovulatory period, ovulatory hCG administration increased TFPI2 expression in preovulatory follicles at both early and late ovulatory phases, with more robust up-regulation in granulosa cells (∼2000-fold increases in mRNA levels compared to before hCG stimulation) than in theca cells (∼50-fold increases) (Puttabyatappa et al., 2017). Poulsen’s microarray data showed peak induction at 12–17 h, but the mRNA levels were on a downward trend at 34 and 36 h after hCG administration (Fig. 5) (Poulsen et al., 2020a). Similar to humans, hCG injection increased the expression of Tfpi2 mRNA in rat periovulatory follicles and cultured granulosa cells (Puttabyatappa et al., 2017). Knockdown of hCG-induced Tfpi2 expression resulted in increased plasmin activity but reduced expression of key ovulatory genes, including Areg and IL6, in cultured rat granulosa cells (Puttabyatappa et al., 2017). Together, these data suggest that TFPI2 is likely involved in ECM remodeling by moderating plasmin activity and regulating granulosa cell gene expression during the periovulatory period in both humans and rats.

Other autocrine/paracrine mediators

Studies from animal models and human samples have accumulated a list of intrafollicular factors that serve as critical mediators of the ovulatory process (reviewed in Duffy et al. (2019)). Besides the factors mentioned above, we recently identified new intrafollicular factors that are induced by an ovulatory stimulus in human preovulatory follicles (Hannon et al., 2018; Al-Alem et al., 2021; Choi et al., 2021b) and are likely involved in ovulatory processes, such as angiogenesis and immune responses. These factors include neurotensin (NTS), secretogranin II (SCG2), angiotensin converting enzyme 2 (ACE2), and CC motif chemokine ligand 20 (CCL20). In addition, our recent study identified several leukocyte-derived factors that can directly act on follicular cells of ovulatory follicles. The following section will discuss these factors and their potential role(s) in human ovulation.

Angiogenesis mediators

After the LH surge, rapid angiogenesis occurs in/around the ovulatory follicle in the human ovary (Trau et al., 2016). Recent studies using non-human primates showed that this process is mediated by the LH surge-induced intrafollicular factors, such as PGE₂, VEGFA, placental growth factor (PGF), and thrombospondin 1 (THBS1) (Trau et al., 2015) (Bender et al., 2018, 2019). In human microarray data, ovulation induction increased levels of mRNA for PGF at 36 h and TBHS1 at both 12–17 and 36 h, whereas VEGFA was listed as a down-regulated gene (Wissing et al., 2014; Poulsen et al., 2020a). In addition to these genes, we recently identified new intrafollicular factors that are induced by an ovulatory stimulus in human preovulatory follicles and are likely involved in angiogenesis. These factors include neurotensin (NTS), secretogranin II (SCG2), angiotensin-converting enzyme 2 (ACE2) (Hannon et al., 2018; Al-Alem et al., 2021; Choi et al., 2021b).

NTS is a neuropeptide that was originally identified in neuronal cells of the hypothalamus and acts as a neurotransmitter (Boules et al., 2013). Since its discovery, NTS has been found throughout the body (Kalafatakis and Triantafyllou, 2011; Osadchii, 2015; Li et al., 2016). Recently, our in vivo studies showed a massive induction of NTS mRNA levels in dominant follicles; 15 000-fold in granulosa cells and 700-fold in theca cells collected after ovulatory hCG stimulation compared to those obtained before the endogenous LH surge in the human ovary (Al-Alem et al., 2021). Consistent with the mRNA data, the immunopositive staining for NTS was also localized to granulosa cells and the thecal layer of dominant follicles collected after hCG administration. Similar to our data, the microarray data showed that NTS mRNA levels were increased in granulosa cells after hCG administration: the levels were up-regulated at 12 h, and this increase was sustained until 36 h (Fig. 5) (Poulsen et al., 2020a). Notably, NTS mRNA levels were also found to be much higher in mural granulosa cells than cumulus cells in individual follicles collected at the time of oocyte retrieval in IVF patients in microarray analyses (Grondahl et al., 2012). In hGLC, hCG also increased NTS mRNA levels. This increase was partially reduced by AG1478, suggesting the involvement of EGF signaling in the hCG-induced NTS expression. Three different receptors for NTS have been so far identified; NTSR1, NTSR2, and NTSR3 (a.k.a, SORT1). Different from the NTS expression profile, the levels of mRNA for NTR receptors were either very low (NTSR1), diminished (NTSR2), or unchanged (NTSR3) in granulosa cells of dominant follicles obtained after hCG administration (Al-Alem et al., 2021). Nonetheless, these data show the presence of the NTS receptor in the human dominant follicle. In addition to humans, NTS expression increases in preovulatory follicles after hCG injection in monkey, rat, and mouse ovaries (Hernandez-Gonzalez et al., 2006; Al-Alem et al., 2021; Campbell et al., 2021). Functional studies demonstrated that NTS treatment enhanced the migration of monkey ovarian microvascular endothelial cells in vitro (Campbell et al., 2021). Moreover, the injection of NTS antibody into preovulatory follicles at the time of ovulatory hCG administration disrupted ovulation in non-human primates (Campbell et al., 2021). In cultured mouse granulosa cells, siRNA-mediated NTS knockdown, followed by RNA-seq analysis, revealed a list of differentially regulated genes (e.g. Ell2, Rsad2, Vps37a, and Amtnl20), identifying these genes as downstream of NTS (Shrestha et al., 2023). Together, these studies indicate that NTS plays an important role in ovulation, in part by modulating the expression of genes in granulosa cells of ovulatory follicles and endothelial cell migration.

SCG2 is a member of the chromogranin family of acidic secretory proteins (Beuret et al., 2004). The full-length protein, secretograin II, was shown to be involved in the packaging or sorting of peptide hormones into secretory vesicles (Beuret et al., 2004). Secretogranin II can be cleaved to produce the active peptide, secretoneurin (SN), which exerts chemotactic effects on specific cell types (Kirchmair et al., 1993; Schneitler et al., 1998). In the human ovary, the expression of SCG2 is increased in granulosa cells of dominant follicles after hCG administration, which was sustained throughout the periovulatory period (Hannon et al., 2018). Similar to the in vivo expression profile, hCG increased the levels of SCG2 mRNA in hGLC cultures, and this increase was partially reduced by an inhibitor of EGF receptor activation (Hannon et al., 2018). Furthermore, SN treatment stimulated human ovarian microvascular endothelial cell migration and new sprout formation in vitro (Hannon et al., 2018). These data suggest that SCG2 induced by ovulatory LH/hCG stimulation plays a role in the ovulatory process and development of the highly vascularized CL, in part by promoting ovarian angiogenesis.

ACE2 is a zinc metalloprotease that is best known for its ability to catalyze the conversion of angiotensin I (Ang I) and angiotensin II (Ang II) to angiotensin-(1–9) and angiotensin-(1–7), respectively (Donoghue et al., 2000; Vickers et al., 2002), thereby playing a key role in the renin angiotensin system (RAS) (Santos et al., 2013). Besides its enzymatic function, ACE2 was identified as a binding site for the human coronavirus (HCoV-NL63) and human severe acute respiratory syndrome coronaviruses, SARS-CoV and SARS-CoV-2 (Li et al., 2005). In an effort to understand the potential impact of COVID-19 on women’s reproductive health, the expression of ACE2 was examined in dominant follicles collected throughout the periovulatory period. The levels of ACE2 mRNA were markedly increased in granulosa cells of dominant follicles at the early ovulatory phase, which was sustained through the post-ovulatory phase (Choi et al., 2021b). The increase in ACE2 mRNA level was also observed at 12–17 h after ovulatory stimulation in Poulsen et al.’s (2020a) microarray data (Fig. 5). Corresponding to these changes in mRNA levels, staining for ACE2 was negligible in dominant follicles obtained before the endogenous LH surge. ACE2 staining became evident in both granulosa and theca cell layers of early ovulatory follicles and the most intensive staining was seen in late ovulatory follicles (Choi et al., 2021b). After ovulation, intense staining of ACE2 was detected in the stroma layer of late- and post-ovulatory follicles, specifically around blood vessels (Choi et al., 2021b). In addition to granulosa and theca cells, ACE2 mRNA was detected in cumulus cells obtained at the time of oocyte retrieval from IVF patients (Grondahl et al., 2012; Choi et al., 2021b). ACE mRNA and protein expression was also detected in human oocytes and embryos (Rajput et al., 2021). Meanwhile, neither mouse nor monkey ovaries showed up-regulation of ACE2 expression after ovulatory hCG stimulation (Choi et al., 2021b). In in vitro studies using hGLCs, hCG increased ACE2 expression (mRNA and protein), similar to the in vivo observations. This increase was completely blocked by RU486, a dual antagonist for the PGR and glucocorticoid receptor, and partially blocked by CORT125281, a selective antagonist for glucocorticoid receptors (Poulsen et al., 2020a). These data indicate that ACE2 expression is mediated by gonadotropin (hCG) and steroid hormones (progesterone and glucocorticoid) in human ovulatory follicles. Considering the well-described role of ACE2 in both angiogenesis and inflammatory responses in other systems, it is likely that this enzyme is expressed and involved in the ovulation-associated acute inflammatory response and angiogenesis in the human ovary. Being a primary entry receptor for SARS-CoV-2, our findings also indicate that the ovary can be a target organ for SARS-CoV-2, being particularly vulnerable during the periovulatory period.

In summary, our studies revealed the highly up-regulated expression of NTS, SCG2, and ACE2 in human ovulatory follicles. In vitro function studies further demonstrated the action of NTS and SCG2 in human endothelial cell migration and/or sprout formation. Together with the implicated role of ACE2 in angiogenesis, these three intrafollicular factors likely play a key role in ovulatory angiogenesis, which is necessary for collapsed follicles to transform into the highly vascularized CL.

Immune response mediators

Accumulating evidence has shown a rapid increase in the number of leukocytes infiltrating into the ovary or preovulatory follicles after the LH surge or ovulatory hCG stimulation in various species including humans (Brannstrom et al., 1994; Oakley et al., 2010). It has been hypothesized that the LH surge triggers follicular cells of dominant follicles to secret specific factors that attract leukocytes to preovulatory follicles. In return, the leukocytes act on/around the ovulatory follicle to promote ovulation and CL formation.

The chemokines are best known for their ability to direct leukocyte migration, adhesion, and activation in various systems (Hughes and Murphy, 2021). Typically, the chemokines are divided into four subfamilies (CXC, CC, CX3C, and XC) based on the configuration of cysteine motifs to the animo terminus (Hughes and Murphy, 2021). In the human ovary, high levels of chemokines were found in the follicular fluid after ovulatory hCG stimulation, including CXCL1 (Karstrom-Encrantz et al., 1998), CXCL8 (Runesson et al., 1996) and CCL2 (Arici et al., 1997). Immunohistochemical studies further confirmed the expression of CXCL1 and CXCL8 in granulosa cells and the theca layer of dominant follicles collected during the preovulatory period from naturally cycling women (Karstrom-Encrantz et al., 1998; Runesson et al., 2000). In both studies, staining in the theca layer was stronger than the staining in granulosa cells, indicating that theca cells are likely the major source of these chemokines in ovulatory follicles (Karstrom-Encrantz et al., 1998; Runesson et al., 2000). More recently, CCL20 was also identified to be highly increased in granulosa and theca cells in cycling women after hCG administration (Al-Alem et al., 2015). This increase in CCL20 was sustained during the late ovulatory phase. Strong immunopositive staining for CCL20 in both granulosa cells and the theca layer of dominant follicles further confirmed the source of this chemokine (Al-Alem et al., 2015). Importantly, its receptor, CCR6 was detected in subsets of leukocytes in follicular aspirates of IVF patients, and the addition of CCL20 stimulated the migration of these leukocytes in Boyden transwell chambers (Al-Alem et al., 2015). The microarray data also showed an increase of CCL20 mRNA, along with CXC3, CXCL8, CXCL16, CCL25, and CXCL27 mRNA in granulosa cells after hCG stimulation, but CXCL1 and CCL2 mRNA were not listed as differentially regulated in granulosa cells (Fig. 5) (Poulsen et al., 2020a). Importantly, Ccl20 expression has not been recognized as differentially regulated in ovulatory follicles in any other species, indicating human-specific up-regulation. Nevertheless, these data provide compelling evidence that in humans, the LH surge/ovulatory hCG stimulates preovulatory follicles to produce and secret these chemokines as specific signals to recruit leukocytes from circulation.

Leukocytes recruited to the ovulatory follicle are believed to function as an essential in situ regulator of ovulation and luteal formation. Historically, due to the limited access to in vivo human samples, only immunohistochemical detection was used to identify the types of leukocytes (e.g. macrophages, neutrophils, and T cells) present around human ovulatory follicles (Brannstrom et al., 1994). Recently, single cell(sc)-RNA sequencing technology has allowed us to identify the subpopulations of cells and the gene expression profiles of each subpopulation present in follicular aspirates obtained from women undergoing the IVF procedure due to non-ovarian infertility etiologies (Choi et al., 2023). RNA-seq analyses revealed the presence of at least ten different subtypes of leukocytes, including M1-macrophages, M2-marcrophages, helper T cells, cytotoxic T cells, NK cells, NKT cells, neutrophils, baso/eosinophils, B cells, and dendritic cells, in addition to granulosa and theca cells (Choi et al., 2023). These leukocytes express a diverse array of factors that can either directly impact follicular cell function (e.g. cytokines: CXCR1, IL1B, TNF, IL16, IFNG, LTA, LTB, TNFSF10; and secretory ligands: AREG, EREG, NRG1, PLAU, etc.) by activating their receptors present in follicular cells or are involved in tissue remodeling (e.g. MMPs, ADAMs, ADAMTSs, and TIMPs) and angiogenesis (e.g. VEGFs, PGF, FGF, IGF, and THBS1) (Choi et al., 2023). These data suggest that leukocytes likely play an important role in ovulation in humans by secreting these and potentially other factors.

Ovarian disorders associated with ovulatory dysfunction

Successful ovulation requires precisely controlled actions of many factors expressed in and around dominant follicles. Therefore, defects in any factor and/or process described above can potentially lead to anovulation. Anovulation is a major cause of infertility in women, affecting up to 40% of women of reproductive age (Mosher and Pratt, 1991). Clinically, the most common and well-described disorders associated with anovulation are polycystic ovarian syndrome and luteinized unruptured follicle syndrome (LUFS) (Franks et al., 1985).

Polycystic ovarian syndrome (PCOS) is the most common cause of infertility in women of reproductive age, with a prevalence of 5–13% according to diagnostic criteria of the NIH, AP-PCOS, and Rotterdam (Bozdag et al., 2016). The main characteristics of this disorder include oligo/anovulation, polycystic ovarian morphology, and hyperandrogenism (Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004). In PCOS women, multiple mid-size follicles are present in the ovary, but these follicles fail to mature to become a dominant follicle that can undergo the ovulatory process in response to the LH surge (Catteau-Jonard and Dewailly, 2013). Instead, these follicles produce high levels of androgens, which interfere further with the process of follicle maturation (Pan et al., 2015). The increase in androgen levels in PCOS is thought to be caused by multiple factors, including increased insulin levels and increased thecal production of androgen in response to tonic elevation of LH levels (Dumesic and Richards, 2013). To induce ovulation in PCOS women, several pharmacological strategies are employed, including insulin-sensitizing agents (e.g. metformin), estrogen receptor modulators (clomiphene citrate), aromatase inhibitors (letrozole), gonadotropins, or a combination of these drugs (Vyrides et al., 2022). Both clomiphene citrate and letrozole act to increase FSH levels in PCOS women so to induce ovulation. However, around 15–30% of anovulatory PCOS women are resistant to clomiphene citrate or gonadotropin treatments to induce ovulation (Bansal et al., 2021; Tsiami et al., 2021). Although the pathophysiological mechanisms underlying ovarian changes in PCOS are not fully delineated, recent studies using gene expression analyses of various compartments in the ovary including whole ovaries, granulosa cells, stroma, cultured theca cells, cumulus cells, and oocytes from PCOS women have revealed distinctly different gene expression patterns compared to those of non-PCOS women (Wood et al., 2003, 2004, 2007; Oksjoki et al., 2005; Kenigsberg et al., 2009; Schmidt et al., 2014; Adams et al., 2016). Of the differentially expressed genes identified, several inflammatory response genes were found to be overexpressed in cells from follicular aspirates of PCOS women. For instance, the expression of IL1B, CXCL8, LIF, NOS2, and PTGS2 were elevated in granulosa cells collected from PCOS patients compared to non-PCOS patients (Schmidt et al., 2014). These patients underwent immature oocyte retrieval on the day after a follicle of 10 mm in diameter was detected. Similar to these findings, Adams et al. (2016) also reported enhanced inflammatory transcriptome in granulosa/lutein cells from PCOS patients undergoing the IVF procedure collected at 36 h after hCG and/or GnRH agonist. These inflammatory response genes include CXCL1, CCL20, CXCL8, IL6, TNF, and CXCR1 (Adams et al., 2016). As mentioned in the sections above, the expression of PTGS2 and CCL20 is markedly up-regulated by ovulatory hCG stimulation in human periovulatory follicles in normal cycling women (Al-Alem et al., 2015; Choi et al., 2017b). PTGS2 and CCL20 function as critical inflammatory mediators by acting as rate-limiting enzymes in the synthesis of PGs and chemoattractants for leukocytes, respectively (Al-Alem et al., 2015; Choi et al., 2017b). These data, together, indicate that the increase in the expression of these genes promotes inflammation and leukocyte invasion into ovulatory follicles in women with PCOS. Indeed, in these patient samples, the levels of CD45, a leukocyte marker, were also elevated (Adams et al., 2016). According to recent scRNAseq data of human follicular aspirates collected at 36 h after hCG administration, CXCR1, IL1B, TNF, CXCL8, and IL6 are predominantly expressed in leukocytes (Choi et al., 2023), indicating that the increased number of leukocytes infiltrating into the ovary/follicles of PCOS patients are likely responsible for elevated levels of these cytokines. These findings are in agreement with the fact that PCOS is a low-grade chronic inflammation state (Diamanti-Kandarakis et al., 2006). PCOS patients show elevated levels of C-reactive protein, a nonspecific marker of inflammation (Escobar-Morreale et al., 2011), and increased leukocyte count and inflammatory cytokines (e.g. IL-1α, IL-1β, CX3CL1) in blood samples (Zafari Zangeneh et al., 2017; Demi et al., 2019). These data all point to the concept that aberrations in the timing or the extent of leukocyte invasion may disrupt the precisely controlled follicle maturation process, as well as the ovulatory process in PCOS women, ultimately leading to anovulation.

Unlike PCOS, the ovaries of women with LUFS show a normal follicular phase with the typical development of a dominant follicle(s), but the dominant follicle fails to release an oocyte after the LH surge (Etrusco et al., 2022). Clinically, LUFS was first diagnosed as a symptom where patients exhibited normal endocrinologic profiles of luteinization (e.g. increased progesterone levels, the presence of secretory endometrium, and elevated basal body temperature) despite the lack of signs of follicle rupture and release of an oocyte (Marik and Hulka, 1978). Later studies reported that LUFS was found even in natural menstrual cycles in fertile women (5–10%) (Kerin et al., 1983), but the incidence of LUFS was higher in women with unexplained infertility undergoing intrauterine insemination (e.g. 25% and 56.5% at the first and second cycle, respectively, and recurrence rate of patients with LUFS at the first cycle is 78% and 90%, respectively) (Qublan et al., 2006). Like in humans, naturally occurring LUFS or LUFS-like phenomena were also reported in baboons (D’Hooghe et al., 1996) and mares (Cuervo-Arango and Newcombe, 2012). However, considering normal follicle growth and the higher incidence of LUFS in patients treated with ovulation inducers, the failure of follicle rupture is likely due to inadequate or altered response to the LH surge in a dominant follicle, leading to the disruption in the precisely regulated ovulatory processes. With the view that ovulation is an acute inflammatory process, the dysregulation of the ovulatory inflammatory process has been proposed as one of the potential causes of LUFS. For instance, the administration of PG synthetase inhibitors induced LUFS in humans (Killick and Elstein, 1987; Pall et al., 2001; Bata et al., 2006; Jesam et al., 2010; Edelman et al., 2013). PGs are known to function as a key inflammatory mediator necessary for successful ovulation. Similar to humans, the blockade of PG synthesis via inhibitors and genetic deletions resulted in LUFS phenotypes in mice (Lim et al., 1997; Davis et al., 1999), rabbits (Salhab et al., 2003), mare (Cuervo-Arango and Newcombe, 2012), sheep (Murdoch and Dunn, 1983), and monkeys (Hizaki et al., 1999). Besides PGs, granulocyte-colony stimulating factor (G-CSF) has also been linked to LUFS. G-CSF is a multifunctioning cytokine that is known to have a key role in inflammatory responses in various systems (Franzke 2006). Serum levels of G-CSF are highest in the ovulatory phase during the natural menstrual cycle and at the time of ovulation induction with gonadotropin stimulation in infertile patients (Shinetugs et al., 1999; Salmassi et al., 2005). In those patients, G-CSF levels in the follicular fluid were higher than those in serum at the time of oocyte retrieval (Shinetugs et al., 1999; Salmassi et al., 2005), suggesting that this cytokine plays a role in ovulation. Importantly, in patients diagnosed with LUF syndrome, a single injection of G-CSF within 24–48 h before hCG administration significantly reduced the incidence of LUFS (4.4%) compared to that in subsequent G-CSF nontreated control cycles (19.1%) (Shibata et al., 2016). A recent study indicated that granulosa cells produce G-CSF, and the ovulatory increase in follicular G-CSF levels results from a paracrine interaction between granulosa cells and leukocytes in humans (Noel et al., 2020). This result suggests that defects in the recruitment and activation of leukocytes in ovulatory follicles may be a contributing factor leading to LUFS formation in women. In summary, the exact mechanism(s) and factors causing LUFS in humans remain elusive, and so far, only a few possible targets (e.g. PGs and G-CSF) are proposed as having clinical potential. It is conceivable that there exist additional factors or mechanisms yet to be discovered to be involved in the pathophysiology of LUFS. Several candidates can be found in mutant mouse studies where the deletion/knockdown of ovulatory genes resulted in anovulatory phenotypes that resemble LUFS symptoms. For example, Pgr KO mice (e.g. Pgr^−/− and Pgr^flox/flox; Esr2^cre/+) (Robker et al., 2000; Park et al., 2020) failed to ovulate but unruptured follicles with entrapped oocytes transformed into CL similar to that observed in women with LUFS. Similar anovulatory phenotypes were observed in other ovulatory gene knockout mice, including Cebpa/b (Fan et al., 2011), Cbfb/Runx2 (Lee-Thacker et al., 2020), Edn2 (Cacioppo et al., 2017), and EGF-like factors (e.g. Areg, Ereg, Nrg1) (Hsieh et al., 2007; Kawashima et al., 2014). The failure of ovulation was also observed when an NTS-neutralizing antibody was injected into the preovulatory follicles of the monkey (Campbell et al., 2021). Therefore, to determine whether these factors are truly associated with anovulatory disorders in women with LUFS, it will be important to study the expression of these factors and their roles in the human periovulatory follicles in normally cycling women and then to determine whether the expression of these genes is affected in women with LUFS. Such studies will provide fundamental information to identify potential targets that can be used as diagnostic factors and/or pharmacological treatments for anovulatory disorders of female infertility.

Conclusion and perspectives

During the past 30 years, with the advancement of sequencing technology and the generation of various mutant mouse lines, substantial progress has been made in understanding the basic mechanisms underlying ovulation and luteal transformation. Despite this progress and direct translational potential, the understanding of the ovulatory process in humans has lagged compared with animal models. The fundamental challenge has been to obtain ovarian tissue samples from normally cycling women at frequent time intervals throughout the periovulatory period, so as to establish the baseline information on the ovulatory process. Our group was the first to collect such tissue samples and has characterized the expression profile of a diverse array of genes in dominant follicles collected throughout the ovulatory period in humans (Brannstrom et al., 1994; McCord et al., 2012; Park et al., 2012; Al-Alem et al., 2015, 2021; Rosewell et al., 2015; Choi et al., 2017a,b, 2018, 2021a,b, 2023; Puttabyatappa et al., 2017; Hannon et al., 2018; Jeon et al., 2023). Similar efforts have been made by the group from Copenhagen, Denmark (Wissing et al., 2014; Poulsen et al., 2020a). This collaborative group of IVF clinicians and scientists has obtained granulosa cells of dominant follicles before or at different time points after hCG administration from women enrolled for IVF, and performed microarray analyses using these granulosa cell samples. Our findings and the microarray data agree on the expression profile of a majority of genes as described in this review, indicating that ovulatory stimulation induces similar responses in a single dominant follicle from normal cycling women as well as in multiple dominant follicles from IVF patients, at least in granulosa cells. The differences observed appear to be mainly associated with the timing of sample collection among different approaches as well as stimulation protocols. As described in this review, it is evident that the ovulatory induction of key follicular mediators such as progesterone, EGF signaling, and PGs are conserved among rodents, ruminants, monkeys, and humans. However, the expression pattern and regulation of these ovulatory genes differ slightly among different species. Even between monkeys and humans, the expression profiles of a substantial number of genes are not aligned when the microarray data from these two species were compared (Bishop et al., 2011; Wissing et al., 2014; Poulsen et al., 2020a). The species-specific differences could be due, in part, to the differences in the time interval between the LH surge and ovulation, and to the genetic make-up in each species, stressing the necessity of utilizing human samples in delineating the mechanisms underlying the ovulatory process in women. This is critically important as this knowledge can be translated directly to enhance the basic understanding of the ovulatory process and to aid in the diagnosis and/or amelioration of ovulation-associated pathological conditions in women. For this purpose, transcriptomic datasets generated using follicular cells from human ovulatory follicles have great utility for the development of novel non-hormonal contraceptives that target the ovulatory process, as well as for studying ovulatory dysfunction associated with advanced reproductive age and pathological conditions such as PCOS and LUFs in women (Hurwitz et al., 2010; Wissing et al., 2014; Adams et al., 2016; Poulsen et al., 2020a; Wu et al., 2022; Choi et al., 2023). Besides these available datasets, it is also important to generate spatial transcriptomics, single-cell proteomics, and transcriptomics data of human dominant follicle tissues collected through the periovulatory period from naturally cycling women as well as women with ovulation-associated pathological conditions. These new datasets will serve as invaluable resources for potential translational studies.

In addition to granulosa cells, cumulus cells in preovulatory follicles undergo dramatic changes in gene expression during the ovulatory period and were thought to play an important role in ovulation (Hernandez-Gonzalez et al., 2006). There are several reports on gene expression of cumulus cells in human ovulatory follicles. However, these studies were typically conducted using cumulus cells collected at the time of oocyte retrieval after ovulatory hCG induction or after different stimulation regimens (e.g. IVM vs IVF). In those studies, cumulus cell gene expression was compared to granulosa cell gene expression or used to identify markers for developmental changes of cumulus cells, quality of oocytes, and embryos, as well as IVF or pregnancy success (Grondahl et al., 2012; Ouandaogo et al., 2012; Borgbo et al., 2013; Yerushalmi et al., 2014; Richani et al., 2021; Venturas et al., 2021; Gao et al., 2022; Russo et al., 2022; Sachs et al., 2024). Recent review articles provide a comprehensive summary of the relationship between cumulus cell gene expression and these parameters (Hu et al., 2023; Massoud et al., 2024). However, there is, to our knowledge, no report profiling the changes in cumulus cell gene expression occurring throughout the ovulatory period in humans. Therefore, the collection and analysis of cumulus cells isolated from human preovulatory follicles throughout the periovulatory period would be invaluable in understanding the role of cumulus cells in the ovulatory process and oocyte maturation in the human ovary.

In delineating the regulatory mechanism and specific function of each ovulatory gene, it is vital to establish an in vitro model that can recapitulate in vivo changes. As described above, the primary human granulosa lutein cell culture model (hGLC) has been shown to mimic key aspects of ovulatory gene expression changes and hormone production. However, it is also worth noting that not all genes examined showed comparable expression patterns to those observed in vivo. These differences are in some ways inevitable considering the source of cells (granulosa/lutein cells from IVF patients), although a 6-day acclimation has reprogrammed these luteinized cells so as to become responsive to ovulatory hCG stimulation. Yet, other factors could have also contributed to these differences, such as biophysical changes occurring within the ovulatory follicle during the ovulatory period (e.g. pressure, oxygen levels), dimensional differences (e.g. monolayer vs multilayered and enclosed follicle), and interactions with other cell types (e.g. theca cells, endothelial cells, and leukocytes) and the extra-cellular matrix. For example, leukocytes recruited into the ovulatory follicle likely secrete cytokines or other secretory proteins, which in turn act on granulosa cells of ovulatory follicles to modulate gene expression. We have recently identified several leukocyte-derived secretory proteins and their receptor expression in granulosa cells of human ovulatory follicles (Choi et al., 2023). Therefore, ongoing efforts are directed toward improving the hGLC model to more closely mimic in vivo conditions by adding leukocyte-derived factors or endothelial cells to the culture system. Alternatively, to maintain the biophysical structures of in vivo preovulatory follicles, an encapsulated human follicle culture system can be employed for ovulation studies, similar to mouse follicle cultures. Skory et al. (2015) showed that cultured encapsulated human follicles simulated hormonal profiles of follicular and luteal phases, although the authors stated that ovulation was not assessed in those cultured human follicles due to the small number of human follicles obtained. Evidently, difficulties in obtaining fresh human ovaries will be a limitation of human follicle cultures as an experimental model, along with some technical issues, such as human follicles requiring a much longer culture period to grow than mouse follicles. Another option is to utilize human-induced pluripotent stem cell-derived granulosa-like cells or follicle-like organoids from cryopreserved ovarian tissues (Jung et al., 2017; Pierson Smela et al., 2023). However, these systems, though a burgeoning area of research, need thorough characterization and validation for the suitability of an ovulatory study model. For this purpose, human datasets of in vivo ovulatory follicles can be useful in determining and validating the suitability of these models as ovulatory experimental models. Nonetheless, these systems represent potential resources for human ovulatory studies.

In summary, recent progress in human studies from our and other laboratories has unveiled many key aspects of the ovulatory process by documenting the expression profiles of genes throughout the ovulatory period from normal cycling women or granulosa cells collected from women undergoing IVF (Figs 7 and 8). At the same time, using a hGLC culture model, we have begun to delineate the regulatory mechanisms underlying ovulatory gene expression and determine the functional contribution and the specific role of each ovulatory mediator. These data from in vivo and in vitro models will enhance our basic understanding of the ovulatory process in women. This knowledge will serve as a foundation for developing more effective treatments and diagnostic tools to help women with anovulatory infertility or etiology, or to better manage ovulation for contraceptive purposes.

Figure 7.

The LH surge/hCG induces temporal and cell-type specific ovulatory changes in the morphology, hormones, and gene expression in human ovulatory follicles. Image created with BioRender.com.

Figure 8.

Hypothetical model of the ovulatory events occurring in granulosa cells and leukocytes of the dominant follicle in humans. The preovulatory LH surge, upon binding to the LH receptor (LHGCR), triggers the induction of transcription factors and the activation of EGR signaling (1 and 2), leading to an increase in the expression of a diverse array of autocrine/paracrine mediators (3). These mediators include EGF-like factors, steroidogenic enzymes, transcription factors, prostaglandin (PG) synthases and transporters, proteolytic enzymes and their inhibitors, angiogenic mediators, and immune response mediators. Granulosa cell-secreted chemokines play a role in the recruitment of leukocytes (4), which in turn secrete cytokines and secretory ligands (5) that can impact granulosa cell function (6). These inter- and extra-cellular factors coordinate the complex ovulatory process and the subsequent luteinization. The numbers represent the hypothetical order of events occurring in the dominant follicle. P4, progesterone; ERBBs, EGF receptors. Image created with BioRender.com.

Data availability

There are no new data associated with this article.

Authors’ roles

M.J. and T.E.C. outlined the manuscript and wrote the first draft of the manuscript. M.J., T.E.C., and M.B. generated the figures and diagrams. M.J., T.E.C., M.B., and J.W.A. reviewed and edited the manuscript.

Funding

This work was supported by the National Institutes of Health P01HD71875 to T.E.C., M.J., and M.B.; R01HD096077 to M.J.; R01HD097675 to T.E.C.

Conflict of interest

None declared.