Intercellular Cyclic Nucleotide Dynamics Mediate Oocyte Meiosis in Mammalian Preovulatory Follicles

Abstract

The cell cycle of the mammalian oocyte arrests in prophase around the time of birth and remains in meiotic arrest as it grows interdependently with increasing layers of surrounding somatic cells as a structure called an ovarian follicle. Within preovulatory follicles, gap junction communication between somatic cells and the oocyte allow cyclic nucleotides to maintain equilibrium concentrations that prevent premature meiotic resumption. Intercellular communication is also required for the mid-cycle surge in luteinizing hormone to alter cyclic nucleotide dynamics and cause the oocyte cell cycle to resume. In this review, the increasingly well-understood mechanisms by which meiotic arrest and meiotic resumption occur will be summarized, along with recent developments made possible by an improved cyclic nucleotide sensor and imaging techniques. These studies have uncovered new aspects of this process and helped clarify the required role of epidermal growth factor receptor signaling in meiotic resumption. Growing evidence that the cyclic nucleotide-associated participants in oocyte cell cycle regulation, first described in rodents, appear to be universal across mammals will also be summarized. This understanding has fostered new approaches to assisted reproductive technologies (ARTs) in domesticated animals and humans.

1. Introduction

The mammalian ovarian follicle is a strikingly elegant, relatively self-contained tissue in which an individual oocyte develops into a mature egg capable of fertilization. The largest follicles, called preovulatory (or Graafian) follicles, can comprise more than 10 concentric layers of somatic cells, called granulosa cells, around the oocyte (Baena et al., 2020; Owen & Jaffe, 2024). Outside of the granulosa cells is a basal lamina separating them from the layers of theca and stromal cells that harbor the vasculature and participate in steroidogenesis (Fig. 1). Crucially, all cells within the basal lamina, including the oocyte, are connected to each other by gap junctions, forming a functional syncytium for molecules less than ~1 kDa (Anderson & Albertini, 1976; Baena & Terasaki, 2019; Beyer, Kistler, Paul, & Goodenough, 1989; Norris et al., 2008). Because they are easily isolated from the ovary at various stages and can be cultured extensively in vitro in several different ways (Converse, Zaniker, Amargant, & Duncan, 2023; Nayudu & Osborn, 1992; Vigone et al., 2018), ovarian follicles represent one of the most tractable models to study intercellular communication and signaling across an intact tissue. One of the clearest examples of this intercellular signaling is the maintenance of oocyte meiotic arrest within preovulatory follicles by gap junction-mediated cyclic nucleotide diffusion, and, conversely, the mid-cycle surge in luteinizing hormone (LH) that alters cyclic nucleotide dynamics to initiate meiotic resumption.

Figure 1: Preovulatory follicle structure.

Isolated mouse preovulatory follicle imaged as described in (Nakashima et al., 2025). The prominent nucleolus of an oocyte in meiotic arrest is often easier to visualize than the larger nuclear envelope. Disappearance of the nucleolus is a marker of meiotic resumption. Scale bar = 100 μm.

Since our last review on this topic (Jaffe & Egbert, 2017), significant progress has been made in understanding both facets of this process. In particular, regulatory mechanisms of cyclic nucleotide dynamics in both the granulosa cell and oocyte compartments have been clarified. Recent studies have also demonstrated that the molecular participants behind meiotic arrest and resumption that were first described in rodents are fundamentally the same across all mammals examined to date, including humans. This has led to exciting advances in assisted reproductive technologies applied to domesticated animals and humans, with implications for conservation of endangered mammals. Furthermore, understanding intercellular signaling in ovarian follicles may provide further insights into gap junction-mediated signaling in other tissues. This review focuses on the regulation of oocyte meiosis by intercellular cyclic nucleotide dynamics in the preovulatory follicle and will first provide a brief background on the development of the follicle to this stage.

2. Overview of folliculogenesis

Folliculogenesis begins during embryonic development of the female, when the full lifetime complement of oocytes is produced (Grive & Freiman, 2015). Shortly after formation and following DNA replication, oocyte meiosis is initiated to the point of homologous chromosome recombination before arresting in late prophase (Wang & Pepling, 2021). Individual oocytes then become surrounded by squamous pre-granulosa cells to comprise a primordial follicle (Hummitzsch et al., 2013; Niu & Spradling, 2022; Pepling & Spradling, 2001), which is enclosed by a basal lamina. Primordial follicles remain dormant until cohorts are activated (Chen et al., 2022; Fortune, Cushman, Wahl, & Kito, 2000; Granados-Aparici et al., 2019; Hardy et al., 2018; Telfer, Grosbois, Odey, Rosario, & Anderson, 2023) into a slow growth phase covering weeks in mice and months in humans (Hirschfield, 1991; Hsueh, Kawamura, Cheng, & Fauser, 2015; McGee & Hsueh, 2000). Throughout this time, the granulosa cells around the oocyte become cuboidal and proliferate to first form a single layer (called a primary follicle), then into two and more layers (secondary and preantral follicles)(Chen et al., 2022). At the initiation of growth, a glycoprotein coat called the zona pellucida forms between the oocyte and granulosa cells (Wassarman & Litscher, 2013). Granulosa cells extend multiple actin- or tubulin-containing processes called transzonal projections (TZPs) through the zona pellucida that terminate on the oocyte plasma membrane (Anderson & Albertini, 1976; El-Hayek & Clarke, 2016; El-Hayek, Yang, Abbassi, FitzHarris, & Clarke, 2018; Granados-Aparici et al., 2022; Marchais et al., 2025). Whether individual TZPs are stable or dynamic in vivo remains an open question (Clarke, 2023), but they clearly possess dynamic potential. When oocytes are separated from the surrounding somatic cells and then reaggregated, TZPs reform and connect to the oocyte (El-Hayek, Yang, Abbassi, FitzHarris, & Clarke, 2018; Fushii, Yamada, Lee, & Miyano, 2021; Herta et al., 2021). Gap junctions present at TZP termini physically connect the somatic cell and oocyte cytosols, allowing the oocyte to acquire nucleotides, amino acids, pyruvate, and other small molecules required for growth (Clarke, 2023). During granulosa cell proliferation, the first two or even three layers surrounding the oocyte elongate toward the zona pellucida and extend TZPs, forming a pseudostratified epithelium with direct oocyte contact (Baena & Terasaki, 2019). Throughout the growth process, oocyte-secreted factors, especially TGFβ family members GDF9 and BMP15, form diffusion gradients that cause the surrounding somatic cells to proliferate, guide their TZP formation, and alter their metabolism for the purpose of nourishing the oocyte (Clarke et al., 2018; Richani, Dunning, Thompson, & Gilchrist, 2021; Sanfins, Rodrigues, & Albertini, 2018; Su, Sugiura, & Eppig, 2009).

Theca cells become associated with follicles outside the basal lamina when there are 2–3 layers of granulosa cells (Hirschfield, 1991). The theca layer grows with the rest of the follicle and harbors an increasingly complex vascular network that provides nutrients, hormones, and gas exchange to the avascular granulosa cells and oocyte (Brown & Russell, 2014). Though theca cells never make a physical connection with granulosa cells in preovulatory follicles, a steroidogenic subset provide androgens as a substrate for conversion to estrogens by granulosa cells (Jamnongjit & Hammes, 2006). When the follicle has grown to several layers of granulosa cells, a fluid-filled antrum rich in glycoproteins (Rodgers & Irving-Rodgers, 2010) begins to form, which will eventually separate the somatic cells into two broad transcriptionally distinct subtypes: a pseudostratified epithelium of cumulus cells that are directly connected to the oocyte (Baena & Terasaki, 2019), and the 5–10 layers of mural granulosa cells inside the basal lamina (Morris et al., 2022; Wigglesworth, Lee, Emori, Sugiura, & Eppig, 2015). Since growth beyond this point in follicle development is dependent on gonadotropins, all follicles recruited into the initial growth phase before puberty will be lost to atresia. In fact, atresia is by far the most common outcome of follicle development even after puberty (Hirschfield, 1991; McGee & Hsueh, 2000, Telfer et al., 2023).

During each reproductive cycle following puberty, follicle-stimulating hormone (FSH) from the anterior pituitary rescues one (in humans) or a cohort (in rodents) of the early antral follicles from atresia and recruits them into a rapid growth phase (~14 days in humans, 4–6 days in mice). Granulosa cells continue proliferating (Richard, Zhou, Jasoni, & Pankhurst, 2024) as the antrum enlarges to fully separate the cumulus cells from the rest of the granulosa except for a “stalk” of cells between the cumulus and mural granulosa that maintain gap junction communication between the somatic compartment and the oocyte. As the follicle reaches the preovulatory stage, FSH and estradiol synergize to induce expression of the luteinizing hormone receptor (LHR) (Segaloff et al., 1992) in a subset of the outermost mural granulosa cells next to the basal lamina (Baena et al., 2020). At the mid-cycle LH surge, these receptors are activated to induce several crucial events at different time scales: oocyte meiotic resumption ~2–4 hours after LH in mice (Nakashima et al., 2025), cumulus expansion ~4–8 hours after LH (Suebthawinkul, Babayev, Zhou, Lee, & Duncan, 2022), and ovulation ~11–16 hours after LH (Duffy, Ko, Jo, Brannstrom, & Curry, 2019; Owen & Jaffe, 2024; Thomas, 2025; Thomas, Marx, Penir, & Schuh, 2024). The remainder of this review will focus on the mechanisms of oocyte meiotic arrest in preovulatory follicles, and the LH-induced release of this arrest through modulation of cyclic nucleotide levels and gap junction permeability.

3. Cyclic nucleotide-mediated meiotic arrest

3.1. Background

Throughout a female’s lifetime and during all stages of oocyte and follicle growth, oocytes remain arrested in prophase until a select few within preovulatory follicles are stimulated to resume meiosis by the mid-cycle LH surge. There are two broad mechanisms for meiotic arrest, depending on the developmental stage of the oocyte. The first mechanism is intrinsic to the oocyte, occurring within follicles up to about the late preantral to early antral stage in mice (Mehlmann et al., 2004; Vaccari, Horner, Mehlmann, & Conti, 2008). Although oocytes have nearly reached their full size at this stage, they do not yet possess sufficient cell cycle proteins to resume meiosis (Chesnel & Eppig, 1995; de Vantery, Stutz, Vassalli, & Schorderet-Slatkine, 1997; Firmani, Uliasz, & Mehlmann, 2018; Nishimura, Shimoaka, Kano, & Naito, 2009;), at least partly due to translational repression of accumulated transcripts (Cheng & Schuh, 2024). The second mechanism occurs as the oocytes become capable of resuming meiosis but are inhibited from doing so by signals from the surrounding somatic cells. This has been appreciated for many decades, when it was shown that oocytes removed from mature follicles of several mammalian species spontaneously resume meiosis without any gonadotropin stimulation (Edwards, 1965; Pincus & Enzmann, 1935). The involvement of cyclic nucleotides in this process was demonstrated by the inhibition of spontaneous meiotic resumption when isolated oocytes were exposed to membrane-permeable cAMP (Cho, Stern, & Biggers, 1974) or an inhibitor of phosphodiesterase (PDE) activity (Magnussen & Hillensjo, 1977). Through several decades of research, there were experimental hints of the now-established mechanisms (Davis, Weakland, West, & Farese, 1986; Hubbard & Greenwald, 1982), and a nearly complete model for somatic cell-mediated oocyte meiotic arrest by cyclic nucleotides was proposed (Törnell, Billig, & Hillensjö, 1991) long before the techniques to test it existed. While more historical context on this topic is provided in Jaffe & Egbert (2017), the well-established mechanisms that have been elucidated since 2002 will now be summarized.

3.2. Established mechanisms of meiotic arrest

Meiotic arrest of mature oocytes is ultimately maintained by high levels of cAMP that activate oocyte protein kinase A (PKA), which phosphorylates regulatory proteins to inhibit the cell cycle (Han, Chen, Paronetto, & Conti, 2005; Pirino, Wescott, & Donovan, 2009). This cAMP is produced within the oocyte by a constitutively active orphan G-protein-coupled receptor (GPCR) called GPR3 (Mehlmann et al., 2004; Mehlmann, 2005) that acts through Gs (Freudzon et al., 2005; Kalinowski et al., 2004; Mehlmann, Jones, & Jaffe, 2002) to stimulate oocyte adenylyl cyclase activity (Horner et al., 2003). GPR12, another orphan receptor closely related to GPR3, is responsible for oocyte cAMP production in rats (Hinckley, Vaccari, Horner, Chen, & Conti, 2005). Because oocytes in mice lacking Gpr3 cannot maintain high cAMP levels, meiosis spontaneously resumes long before the ovulatory signal (Mehlmann et al., 2004), leading to subfertility that worsens with age (Ledent et al., 2005). To balance the constitutive cAMP production, mammalian oocytes also express phosphodiesterase 3A (PDE3A)(Tsafriri, Chun, Zhang, Hsueh, & Conti, 1996), which hydrolyzes both cAMP and cGMP (Bender & Beavo, 2006). Mice lacking Pde3a have no mechanism to sufficiently lower oocyte cAMP levels, so ovulated oocytes remain in meiotic arrest, causing complete infertility (Masciarelli et al., 2004). PDE3A has similar affinity for both cyclic nucleotides but hydrolyzes cGMP with several-fold lower velocity compared to cAMP (Bender & Beavo, 2006; Degerman, Belfrage, & Manganiello, 1997). These qualities allow cAMP hydrolysis by PDE3A to be competitively inhibited by elevated cGMP levels. Indeed, meiotically arrested mature oocytes have cGMP levels in the micromolar range (Norris et al., 2009). When oocytes within preovulatory follicles were injected with a cGMP-specific PDE, oocyte cAMP fell and meiosis spontaneously resumed (Norris et al., 2009), confirming that meiotic arrest is maintained by cGMP-mediated inhibition of PDE3A activity toward cAMP (Fig. 2).

Figure 2: Preovulatory follicle cyclic nucleotide dynamics that maintain oocyte meiotic arrest.

Prior to LH receptor (LHR) activation, adenylyl cyclase (AC) activity and stable cAMP (green circles) levels are present in somatic cells, generated by follicle-stimulating hormone (FSH) receptor signaling as well as other GPCRs (not shown). Secretion of CNP from the mural granulosa maintains activity of phosphorylated (red dots) NPR2 in all somatic cells to generate cGMP (orange squares), which diffuses into the oocyte to competitively inhibit PDE3A activity toward oocyte cAMP that is constitutively produced by GPR3 signaling. Oocyte cAMP ultimately maintains meiotic arrest. Cyclic nucleotide diffusion is mediated by gap junctions containing connexin 43 (Cx43) between somatic cells and connexin 37 (Cx37) between cumulus cells and the oocyte. Bold labels indicate high activity/predominance; non-bold labels indicate low or unstimulated activity.

However, the oocyte does not produce cGMP; rather, it diffuses into the oocyte from the surrounding somatic cells through gap junctions (Norris et al., 2009; Shuhaibar et al., 2015). Follicle cGMP is produced in both mural granulosa and cumulus cells by the membrane guanylyl cyclase natriuretic peptide receptor 2 (NPR2, also called GC-B)(Zhang, Su, Sugiura, Xia, & Eppig, 2010). NPR2 requires binding of its agonist, C-type natriuretic peptide (CNP, from the gene Nppc) and NPR2 activity is also regulated by phosphorylation of juxtamembrane sites (Potter, 2024). In ovaries deficient in either Npr2 or Nppc, oocytes within antral follicles spontaneously mature and mice are subfertile (Zhang, Su, Sugiura, Xia, & Eppig, 2010), similar to the Gpr3-knockout phenotype (Mehlmann et al., 2004; Ledent et al., 2005), though a different Npr2 mutant exhibits infertility (Geister et al., 2013). NPR2 is expressed in all somatic cells of the follicle, though it becomes increasingly enriched with proximity to the oocyte (Zhang, Su, Sugiura, Xia, & Eppig, 2010; Baena et al., 2020). Stable NPR2 expression requires estradiol (Zhang et al., 2011), and oocyte-derived factors like GDF9, BMP15, and FGF8 also increase Npr2 expression, explaining its enrichment in cumulus cells (Zhang, Su, Sugiura, Xia, & Eppig, 2010). Ovarian CNP is almost exclusively produced by the mural granulosa cells (Zhang, Su, Sugiura, Xia, & Eppig, 2010) and diffuses extracellularly to activate NPR2 throughout the follicle. Importantly, when isolated mouse cumulus-oocyte complexes (COCs) are treated with exogenous CNP, spontaneous meiotic resumption is prevented (Zhang, Su, Sugiura, Xia, & Eppig, 2010). Although the cumulus cells are likely the direct source of oocyte cGMP, elevated cGMP levels throughout the rest of the somatic cells are likely necessary to maintain equilibrium concentrations (Shuhaibar et al., 2015).

3.3. Minimal contribution of other guanylyl cyclases to meiotic arrest

At least in mice, other membrane or soluble guanylyl cyclases (sGCs) contribute little, if any, to the elevated cGMP levels that maintain meiotic arrest. This is supported by studies showing that lack of Npr2 leads to undetectable levels of cGMP in preovulatory follicles (Geister et al., 2013) and spontaneous meiotic resumption (Geister et al., 2013; Zhang, Su, Sugiura, Xia, & Eppig, 2010). Additionally, compared to Npr2 in mural granulosa cells, other guanylyl cyclase genes exhibit undetectable or low levels of mRNA expression (Robinson et al., 2012). Although a small amount natriuretic peptide 1 (Npr1) expression or activity in response to its agonist atrial natriuretic peptide (ANP) has been reported in mouse and rat preovulatory follicles (Gutkowska et al., 1999; Robinson et al., 2012), exogenous ANP cannot maintain meiotic arrest in isolated mouse COCs (Zhang, Su, Sugiura, Xia, & Eppig, 2010).

Pharmacological modulation of sGC activity has been reported to influence gonadotropin-induced meiotic resumption in follicle-enclosed rat oocytes (Nakamura, Yamagata, Sugino, Takayama, & Kato, 2002; Sela-Abramovich, Galiani, Nevo, & Dekel, 2008), and LH receptor activation has been reported to reduce inducible nitric oxide synthase (iNOS) mRNA expression in rat granulosa cells (Yamagata et al., 2002). However, there has been no clear demonstration that endogenous sGC activity contributes to the maintenance of meiotic arrest, or that LH signaling modulates sGC activity in mouse or rat follicles. While it remains possible that sGC regulation plays a minor role in this process, NPR2 activity is clearly the primary mechanism for maintaining meiotic arrest in rodents (Egbert et al., 2014; Zhang, Su, Sugiura, Xia, & Eppig, 2010).

3.4. Other potential roles of follicle cGMP

Whether follicle cGMP has a function beyond maintaining meiotic arrest is an open question. The primary signaling effector for cGMP is protein kinase G (PKG), but its protein expression in preovulatory follicles prior to LH exposure has not been conclusively demonstrated. PKG2 (Prkg2) mRNA expression was nearly undetectable in COCs and isolated granulosa cells from mouse preovulatory follicles but became highly expressed ~8 hours after LH receptor activation (Sriraman, Rudd, Lohmann, Mulders, & Richards, 2006). Similarly, PKG2 protein levels in the whole ovary were dramatically upregulated 8–12 hours after LH receptor activation, which could indicate a role for cGMP signaling in ovulation or luteinization (Sriraman, Rudd, Lohmann, Mulders, & Richards, 2006). The same study reported that PKG1 protein was restricted to ovarian cells outside the follicle and was not regulated by gonadotropins. Although some studies have reported evidence of PKG-mediated effects in granulosa cells isolated and cultured from preovulatory follicles using a combination of inhibitors (Peluso & Pappalardo, 2004; Tian et al., 2018), many PKG inhibitors have off-target effects (Gambaryan, Butt, Geiger, Lohmann, & Walter, 2011) and aberrant PKG expression due to cell culture (18–48+ hours) cannot be ruled out. Nevertheless, in isolated mouse COCs, cGMP itself has been shown to be required for the maintenance of TZPs connecting cumulus cells to the oocyte (Abbassi et al., 2021). TZPs are maintained by cGMP even in the presence of a gap junction inhibitor, indicating that cGMP is acting directly within the cumulus cells by an unknown mechanism (Abbassi et al., 2021). Perhaps low levels of functional PKG expression could be restricted to cumulus cells, which might have been below the limit of detection in earlier studies. It is also possible that PKG signaling from cGMP produced by NPR2 or sGCs could occur in a specific subcellular compartment, which has yet to be investigated in this context.

3.5. Minimal contribution of somatic cell cAMP to meiotic arrest

Because early studies showed that spontaneous meiotic resumption following oocyte isolation could be prevented by incubation with a membrane-permeable form of cAMP (Cho, Stern, & Biggers, 1974), it was long presumed that the somatic cells supplied the oocyte with cAMP to maintain meiotic arrest. It is now understood that the oocyte is the source of cAMP (Mehlmann, Jones, & Jaffe, 2002), while cGMP is the somatic cell contributor to meiotic arrest by competitively inhibiting oocyte PDE3A (Norris et al., 2009; Vaccari, Weeks, Hsieh, Menniti, & Conti, 2009). However, the idea that cAMP diffusion from the somatic cells to the oocyte contributes to meiotic arrest has been proposed (Gershon et al., 2019; Gilchrist et al., 2024). Evidence against this concept is that when oocyte cAMP production is negated through Gpr3 deletion (Mehlmann et al., 2004) or injection into the oocyte of either an anti-Gs antibody (Mehlmann, Jones, & Jaffe, 2002) or a dominant negative form of Gs (Kalinowski et al., 2004), meiosis spontaneously resumes even though the oocyte remains connected to the somatic cells. It could be suggested that because oocyte PDE3A is still present in these models, any cAMP from the somatic cells would be more readily hydrolyzed, masking its potential contribution. However, oocytes that lack both Gpr3 and Pde3a still have low oocyte cAMP levels and spontaneously resume meiosis at the same late preantral/early antral stage as those that express Pde3a (Vaccari, Horner, Mehlmann, & Conti, 2008). These studies indicate that while some cAMP may stochastically diffuse from the somatic cells to the oocyte through gap junctions, it is negligible for the normal maintenance of meiotic arrest.

4. Cyclic nucleotide-mediated meiotic resumption in response to luteinizing hormone

4.1. Overview

With an understanding of how meiotic arrest is maintained, it becomes clear that the mid-cycle surge in luteinizing hormone must lower follicle and oocyte cGMP levels to relieve the competitive inhibition of oocyte PDE3A, allowing enough cAMP hydrolysis to trigger meiotic resumption. Because the expression of LH receptors in preovulatory follicles is restricted to the outer half of the mural granulosa (Baena et al., 2020; Owen & Jaffe, 2024), communication across several layers of cells is required to transmit the meiotic resumption signal to the oocyte. In general, this communication could occur through cleaved ligands that diffuse extracellularly, or by diffusion of signals within cells through gap junctions (Jaffe & Egbert, 2017). As will be discussed below, both mechanisms are required in concert for the normal timing of meiotic resumption.

4.2. Mechanisms contributing to the rapid cGMP decrease in response to LH

4.2.1. Regulation of the NPR2 guanylyl cyclase

When isolated mouse or rat follicles are treated with LH, NPR2 guanylyl cyclase activity decreased to about half within 20–30 min (Egbert et al., 2014; Robinson et al., 2012) and follicle cGMP levels fell precipitously over the same time (Egbert et al., 2014; Liu, Xie, Zamah, Cao, & Conti, 2014; Norris, Freudzon, Nikolaev, & Jaffe, 2010), declining to ~3 % of baseline by 20 min (Shuhaibar et al., 2015) and remaining low for at least 5 h (Nakashima et al., 2025; Norris, Freudzon, Nikolaev, & Jaffe, 2010). During this time, NPR2 protein levels in crude membrane preparations remained unchanged, though potential internalization was not examined (Egbert et al., 2014). While follicle levels of the NPR2 agonist CNP in follicles were also initially stable in response to LH receptor activation, CNP declined by 2–4 h in mice and 4–6 h in rats—long after the rapid fall in cGMP levels (Egbert et al., 2014; Liu, Xie, Zamah, Cao, & Conti, 2014; Robinson et al., 2012; Tsuji, Kiyosu, Akiyama, & Kunieda, 2012). Thus, while the decline in CNP levels could help maintain low follicle cGMP levels in the hours after LH exposure, it cannot explain the initial fall in cGMP that permits oocyte meiotic resumption.

In addition to CNP, full NPR2 guanylyl cyclase activity requires phosphorylation on at least some of 7 closely-spaced serine and threonine residues just inside the plasma membrane (Potter, 2024). This region of NPR2 is a kinase homology domain but was found to lack functional autophosphorylation activity; the identity of the kinase or kinases that maintain NPR2 phosphorylation remain unknown (Edmund, Walseth, Levinson, & Potter, 2019; Potter, 2024). Since NPR2 phospho-specific antibodies are lacking, initial tests of whether LH signaling rapidly inactivates NPR2 by dephosphorylation were conducted by Phos-tag gel electrophoresis (Kinoshita, Kinoshita-Kikuta, Takiyama, & Koike, 2006; Kinoshita-Kikuta, Kinoshita, & Koike, 2015) of rat follicle membranes following immunoprecipitation with an anti-NPR2 rabbit serum and immunoblotting with the same antiserum (Egbert et al., 2014). This method conclusively showed that NPR2 is rapidly dephosphorylated in response to LH with similar timing to the decrease in guanylyl cyclase activity and the fall in follicle cGMP. The function of this dephosphorylation was confirmed using mice in which 7 regulatory serines and threonines were mutated to glutamate on both alleles (Npr2-7E) to constitutively mimic the negative charge of phosphorylation (Shuhaibar et al., 2016). At the time these mice were generated, evidence pointed to 7 residues that altered NPR2 activity when site-specific phosphorylation was prevented (Yoder et al., 2012). However, later research found that only 6 of these sites are likely to be regulated in vivo (Otto & Potter, 2022; Potter, 2024). When preovulatory follicles from Npr2-7E mice were treated with LH, NPR2 guanylyl cyclase activity remained high and cGMP levels in the follicle and oocyte stayed elevated for 3–4 hours, causing oocyte meiotic resumption to be delayed by ~5 hours (Nakashima et al., 2025; Shuhaibar et al., 2016). The eventual decline in cGMP levels and the resumption of meiosis are likely due to LH-induced decreases in follicle CNP levels, which is supported by the observation that supplementation with CNP further delayed the cGMP decrease (Nakashima et al., 2025).

The signaling that results in NPR2 dephosphorylation remains unclear. However, it has recently come into better focus partly due to the major advance of mice expressing NPR2 tagged with a 9 amino acid sequence from the hemagglutinin (HA) protein of human influenza virus (Baena et al., 2020; Egbert et al., 2022). Using these mice, small amounts of isolated follicles can be solubilized and electrophoresed on a Phos-tag gel without immunoprecipitation, followed by detection with commercial anti-HA tag antibodies. With HA-NPR2 mice, it was shown that follicle treatment with the adenylyl cyclase activator forskolin led to NPR2 dephosphorylation and inactivation similar to LH (Egbert et al., 2021). Furthermore, LH-induced NPR2 dephosphorylation was prevented in follicles treated with a specific inhibitor of protein kinase A (PKA) (Egbert et al., 2024). In response to LH, the initial cAMP elevation that occurs in the subset of outer mural granulosa cells expressing the LH receptor (Baena et al., 2020; Owen & Jaffe, 2024) has been shown to rapidly diffuse throughout the entire granulosa layer and deeper into the follicle through gap junctions (Lyga et al., 2016; Nakashima et al., 2025). Thus, NPR2 appears to become dephosphorylated and inactivated in all somatic cells by intercellular cAMP/PKA signaling. This signaling could either phosphorylate and activate a phosphatase that acts on NPR2, or phosphorylate and inactivate a kinase acting on NPR2 (or possibly both).

In isolated rat follicles, cantharidin, an inhibitor of several PPP-family phosphatases (notably not PPP3/calcineurin; Swingle & Honkanen, 2019) completely prevented the LH-induced dephosphorylation of NPR2 and the decrease in guanylyl cyclase activity (Egbert et al., 2014). To identify PPP-family candidates that could rapidly dephosphorylate NPR2 after LH, phosphopeptides were enriched from isolated rat follicles treated for 30 min with or without LH and quantified by mass spectrometry (Egbert et al., 2024). Previous work in isolated rat granulosa cells showed that LH receptor/PKA signaling induces phosphorylation of serine 566 of the PPP2 regulatory protein PPP2R5D (also known as PP2A B56δ) (Flynn et al., 2008), which increases PPP2 catalytic activity (Ahn et al., 2007). In the mass spectrometry analysis, the PPP2R5D phosphopeptide containing phosphorylated S566 had 4-fold higher intensity in LH-treated follicles (Egbert et al., 2024). This screen also identified 3 other serines on PPP2R5D that were significantly or marginally more phosphorylated after LH, all of which were found to be dependent on PKA signaling. Additionally, serine 507 of the PPP1 regulatory protein PPP1R12A (also known as MYPT1), which has been shown to regulate phosphatase activity (Samson et al., 2019), exhibited increased phosphorylation after LH that was dependent on PKA activity (Egbert et al., 2024). In response to these candidates, mutant mice were made in which either the four identified serines in PPP2R5D, or S507 of PPP1R12A, were mutated to alanine to test whether LH-induced NPR2 dephosphorylation would be prevented. However, follicles from these mice, either individually or when bred together to combine the five S-A mutations, had normal LH-induced NPR2 dephosphorylation and unaltered timing of meiotic resumption (Egbert et al., 2024; 2025). Perhaps other regulatory phosphosites on PPP-family member proteins are responsible for effecting NPR2 dephosphorylation, or other such sites can at least compensate for the mutations in PPP2R5D and PPP1R12A. Another possibility is that phosphatase activity before and after LH remains stable, while LH/PKA signaling acts to decrease the activity of the kinase or kinases that phosphorylate NPR2. Identification of the mechanisms regulating NPR2 phosphorylation before and after LH is one of the biggest remaining questions surrounding oocyte meiotic arrest and resumption.

4.2.2. Dynamics of the cGMP decrease in response to LH-induced NPR2 inactivation

Regardless of the mechanism, the LH-induced dephosphorylation and inactivation of NPR2 causes cGMP levels to fall rapidly throughout the follicle and oocyte. By imaging isolated follicles from mice that express a fluorescent sensor for cGMP, the dynamics of cGMP changes after LH were studied by Shuhaibar et al. (2015). After LH addition, the cGMP decrease was first detected in the outer mural granulosa cells, which uniquely express the LH receptor, with 50 % of the decrease occurring within 3 min (Fig. 3). The cGMP decrease then propagated inward, with levels reaching a minimum in the cumulus cells and oocyte within 20 min after LH (Shuhaibar et al., 2015). These changes result in a ~40-fold decrease in cGMP levels, as follicle concentrations of cGMP before LH and 20 min after LH were measured by ELISA to be ~4 μM and ~100 nM, respectively (Shuhaibar et al., 2015). The mechanism by which the rapid cGMP decrease occurs is probably multi-factorial, but the initial inactivation of NPR2 in the outer mural granulosa cells would allow PDE activity to predominate. This would perturb the cGMP equilibrium between the follicle and oocyte, resulting in a concentration gradient that causes cGMP to diffuse through gap junctions away from the follicle interior and out of the oocyte (Shuhaibar et al., 2015). Simultaneously, the LH-induced cAMP elevation diffuses through gap junctions toward the follicle interior (Lyga et al., 2016; Nakashima et al., 2025), presumably inactivating NPR2 located in the inner mural granulosa and cumulus cells (Fig. 4). The outward diffusion of cGMP and inward diffusion of cAMP likely synergize to cause the fall in cGMP levels that rapidly equilibrates at a low level throughout the follicle and oocyte (Nakashima et al., 2025; Shuhaibar et al., 2015). Blocking all gap junction communication in the follicle with the inhibitor carbenoxolone (CBX) prevented both the LH-induced cGMP decrease in the cumulus cells (Shuhaibar et al., 2015), demonstrating that gap junction communication is critical for the cGMP decrease that permits oocyte meiotic resumption.

Figure 3: Cyclic nucleotide dynamics ~3 min after LH receptor activation.

LH signaling increases cAMP levels, which activates PKA to phosphorylate PDE5A, increasing its activity. Through a multi-step process (dashed line), PKA activation also leads to rapid NPR2 dephosphorylation and inactivation. These initial events cause cGMP to fall in outer mural granulosa cells, while dynamics in the cumulus cells and oocyte remain unchanged at this time point.

Figure 4: Cyclic nucleotide dynamics ~15 min after LH receptor activation.

The LH-stimulated cAMP increase has diffused into the cumulus cells to cause the PKA-mediated dephosphorylation and inactivation of NPR2 and phosphorylation of PDE5A throughout the follicle. In addition to hydrolysis by PDE3A, cGMP diffuses down its concentration gradient out of the oocyte. At the same time, elevated levels of cAMP diffuse into the oocyte, creating a barrier to meiotic resumption even though cGMP levels have fallen.

4.2.3. Contribution of phosphodiesterases to the rapid cGMP decrease

Although dephosphorylation and inactivation of NPR2 is required to lower cGMP to levels that allow meiotic resumption, LH stimulation of rat or mouse follicles still caused a partial rapid decrease in cGMP when NPR2 dephosphorylation was prevented (Egbert et al., 2014; Shuhaibar et al., 2016). Granulosa cells of both rat and mouse follicles express the cGMP-specific PDE5A (Egbert et al., 2014; Vaccari, Weeks, Hsieh, Menniti, & Conti, 2009), which exhibits increased catalytic activity when phosphorylated on serine 92 by PKA or PKG (Rybalkin, Rybalkina, Feil, Hofmann, & Beavo, 2002). Experiments in rat follicles showed that LH/PKA signaling rapidly phosphorylated PDE5A on serine 92, increasing its activity (Egbert et al., 2016) (Figs. 3, 4). To test the contribution of this phosphorylation on cGMP regulation, mice were generated in which serine 92 of PDE5A was mutated to alanine (Pde5a-S92A). These experiments initially showed that either alone or in combination with the Npr2-7E mutations, the Pde5a-S92A mutation prevented only a small part of the LH-induced rapid cGMP decrease (Egbert, Yee, & Jaffe, 2018). However, a more recent study found that cGMP levels in mice with the combined Npr2-7E and Pde5a-S92A mutations remained stable in the mural granulosa and oocyte for 2–3 hours after LH, with the eventual cGMP decrease likely occurring in response to the fall in CNP levels (Nakashima et al., 2025). This discrepancy is likely due to LH being applied at a saturating concentration (~300 nM) in the earlier experiment (Egbert, Yee, & Jaffe, 2018), compared to a more physiological concentration (~10 nM) in the later experiment (Nakashima et al., 2025) that is close to the measured LH surge peak in mice (Owen, Zhou, Bernard, & Jaffe, 2021). Thus, the most recent data suggest that the rapid fall in follicle and oocyte cGMP levels is explained by the simultaneous inactivation of NPR2 and activation of PDE5A by PKA signaling in response to LH.

There are clearly other contributors to follicle cGMP dynamics before and after LH. PDE1A, a dual-specificity PDE with higher affinity for cGMP (Bender & Beavo, 2006), and PDE5A each account for at least 40 % of cGMP-hydrolyzing activity in rat follicles (Egbert et al., 2016). Pharmacologically inhibiting each PDE individually had at most a partial effect on meiotic resumption, while inhibiting both enzymes completely prevented oocyte meiotic resumption by 8 h after LH in rats (Egbert et al., 2016). This suggests that PDE1A has a role in lowering cGMP in response to LH. PDE1A activity is regulated by calcium-calmodulin signaling (Goraya & Cooper, 2005), but whether LH increases this activity is unclear. Although LH-induced calcium oscillations have been reported in mouse follicles, preventing these oscillations with a Gq inhibitor had no effect on the timing of meiotic resumption (Egbert et al., 2019). Taken together, it appears PDE1A activity—whether increased or stable after LH—contributes to the decrease in follicle cGMP, along with the increased PDE5A activity.

Another potential cGMP-regulating factor is sphingosine-1-phosphate (S1P), which has been shown to inhibit NPR2 activity independently of phosphorylation (Abbey-Hosch, Cody, & Potter, 2004; Abbey-Hosch, Smirnov, & Potter, 2005). It has been reported that S1P is an intermediate of epidermal growth factor receptor (EGFR)-induced calcium elevation in the cumulus cells of isolated mouse COCs (Hao et al., 2016; Wang et al., 2013). This elevated calcium was reported to decrease the affinity of NPR2 for CNP, thereby reducing NPR2 activity. There is evidence that LH receptor signaling elevates S1P levels in cumulus cells (Yuan et al., 2022); therefore, this mechanism could act together with the LH-induced decrease in CNP to ensure NPR2 activity remains low after it is dephosphorylated.

5. New insights into cAMP dynamics after LH

5.1. Limitations of previous approaches

Previous work using follicles from mice expressing an early Förster resonance energy transfer (FRET) sensor for cAMP called Epac1-camps (Calebiro et al., 2009; Nikolaev, Bünemann, Hein, Hannawacker, & Lohse, 2004) had shown that the cAMP elevation in response to LH receptor activation rapidly diffuses through gap junctions to the follicle interior (Lyga et al., 2016). There are likely multiple functions of this follicle-wide cAMP increase, including the dephosphorylation and inactivation of NPR2 (Egbert et al., 2024; Egbert, Robinson, Uliasz, Potter, & Jaffe 2021). However, it remained unclear what, if anything, prevented the elevated levels of cAMP from continuing to diffuse through gap junctions into the oocyte, since oocyte cAMP levels must fall for meiosis to resume. Previous cAMP sensors were not well-suited to investigate these dynamics because of the small relative FRET changes to physiological stimuli combined with the relative dimness of the sensors that precluded effective measurement at the depth of the oocyte (Lyga et al., 2016). Additionally, earlier methods for stably imaging isolated follicles were not optimized for maintaining tissue health for longer than 30–40 min (Shuhaibar et al., 2015).

5.2. Improved imaging of cAMP dynamics reveals a transient oocyte cAMP peak after LH

Recently, a new cAMP FRET sensor called cAMPFIRE-M was developed with increased brightness and a larger response to physiological stimuli (Massengill et al., 2022). Using mice generated to globally express this sensor, as well as refined long-term imaging methods, the spatiotemporal dynamics of cAMP in isolated follicles were investigated (Nakashima et al., 2025). In the outer mural granulosa cells, cAMP remained elevated for at least 6 hours after LH, while the cumulus cells exhibited complex and variable cAMP dynamics in the hours after LH. This study also revealed that the progression of the LH-induced cAMP increase does transiently continue through gap junctions into the oocyte, where cAMP remains elevated above the initial baseline for ~10–60 minutes (see Fig. 4). Thus, LH signaling creates a temporary barrier to meiotic resumption in the oocyte that must be overcome. When the cAMPFIRE-M sensor was introduced into Npr2-7E;Pde5a-S92A mice in which follicle and oocyte cGMP levels remain elevated for several hours after LH, the initial oocyte cAMP increase also persisted for longer (60–120 min) before returning to baseline levels for another 1–2 hours. Oocyte cAMP levels did not fall below baseline until after cGMP levels had fallen due to the LH-induced decrease in CNP levels (Nakashima et al., 2025), which inactivates NPR2 regardless of its phosphorylation state (Potter, 2024). This supports the concept that oocyte cGMP levels must decrease first to relieve the competitive inhibition of PDE3A and permit the fall in oocyte cAMP below baseline levels, triggering meiotic resumption. However, the timing of the oocyte cAMP decrease in wild-type mice is much more variable than the consistently rapid cGMP decrease (Nakashima et al., 2025). This suggests that at least one additional mechanism besides the cGMP decrease is required to cause oocyte cAMP to fall from its LH-induced peak and below baseline.

5.3. Contribution of gap junction regulation by EGFR signaling to the oocyte cAMP decrease and meiotic resumption

Gap junctions between the somatic cells of the follicle are known to be transiently closed by LH signaling (Norris et al., 2008; Norris, Freudzon, Nikolaev, & Jaffe, 2010), and blocking gap junctions in isolated rat follicles with CBX caused oocyte cAMP levels to fall (through loss of cGMP inhibition of PDE3A) and meiosis to resume (Sela-Abramovich, Edry, Galiani, Nevo, & Dekel, 2006). Though different species may also express other types of gap junctions at various follicle stages, preovulatory follicles in most mammals express two primary types with distinct localizations (Gershon, Plaks, & Dekel, 2008). The granulosa and cumulus cells are connected to each other by gap junctions composed of connexin 43 (Cx43; Gja1), while cumulus-oocyte gap junctions are composed of connexin 37 (Cx37; Gja4) (Ackert et al., 2001; Gittens & Kidder, 2005; Juneja, Barr, Enders, & Kidder, 1999; Simon, Chen, & Jackson, 2006; Veitch, Gittens, Shao, Laird, & Kidder, 2004). Both gap junctions have been shown to be required for the cyclic nucleotide-mediated maintenance of meiotic arrest (Richard & Baltz, 2014). Cx43 is highly regulated by phosphorylation, and its permeability can be decreased by phosphorylation of several C-terminal serine residues by mitogen-activated protein kinase (MAPK) in many cell types (Solan & Lampe, 2009). In isolated ovarian follicles, gap junction permeability was assessed by microinjection of a small fluorescent molecule (Alexa 350) into the oocyte, which diffused to the outermost mural granulosa in the absence of LH (Norris et al., 2008). LH signaling was demonstrated to rapidly increase MAPK-dependent phosphorylation of the Cx43 C-terminal residues, which closed Cx43 junctions by 30–60 minutes after LH such that very little fluorescence reached the mural granulosa cells from the oocyte (Norris et al., 2008). These events were found to be mediated by activation of EGFR, and pharmacological inhibition of EGFR activity with AG1478 prevented the LH-induced closure of somatic cell gap junctions in isolated follicles, which delayed oocyte meiotic resumption (Hsieh, Thao, & Conti, 2011; Norris, Freudzon, Nikolaev, & Jaffe, 2010). When follicles from cAMPFIRE-M mice were treated with AG1478, the LH-induced fall in oocyte cGMP was unaffected, but the oocyte cAMP peak rose to a higher amplitude and persisted for an average of 90 minutes, compared to ~30 minutes in controls (Nakashima et al., 2025). Furthermore, once oocyte cAMP was back to baseline, cAMP declined much more gradually than in control oocytes, taking over 4 hours longer to reach the same level. As a result, meiotic resumption in response to LH was delayed between 4–5 hours in follicles where Cx43 gap junction closure was inhibited (Nakashima et al., 2025).

When combined with previous work, these studies lead to a new model of how cyclic nucleotide dynamics are altered by LH signaling to cause oocyte meiotic resumption. Following the discovery of the role and regulation of NPR2 in granulosa cells (Egbert et al., 2014; Robinson et al., 2012; Shuhaibar et al., 2015; Zhang, Su, Sugiura, Xia, & Eppig, 2010), it was presumed that the fall in oocyte cGMP resulted in an immediate oocyte cAMP decrease that triggered meiotic resumption, while the specific function of the LH-induced gap junction closure was not known (Jaffe & Egbert, 2017). It now appears that the oocyte cAMP decrease results from a two-step process both temporally and mechanistically, requiring both intracellular and extracellular communication between distant cells. These data suggest that in response to LH, the follicle and oocyte cGMP decrease occurs prior to gap junction closure and is required, but not sufficient, for the timely fall in oocyte cAMP. Closure of somatic cell Cx43 junctions in response to LH-induced EGFR signaling is then required to accomplish the full oocyte cAMP decrease that restarts the cell cycle with a normal time course (Nakashima et al., 2025) (Fig. 5).

Figure 5: Cyclic nucleotide dynamics ~30–60 min after LH receptor activation.

Elevated cAMP levels in the oocyte decline due to cessation of competitive PDE3A inhibition by cGMP, but this is not sufficient to cause meiotic resumption with a normal time course without gap junction closure. It has been proposed that LH signaling in granulosa cells stimulates matrix metalloproteinase (MMP) activity to cleave EGFR ligands (EL) that are already present in the membrane. PKA has been implicated in this process. This moderate level of EGFR ligand release causes partial autophosphorylation and activation of EGFR signaling, resulting in MAPK-dependent closure of Cx43 gap junctions between somatic cells. This event interrupts the flow of cAMP to the follicle interior and oocyte. Note that CNP levels remain elevated, but NPR2 activity is low due to dephosphorylation.

While regulation of Cx43 permeability by phosphorylation in follicles is well established, other emerging aspects of gap junction regulation could also be explored. One remaining question is whether LH signaling induces any changes in Cx37 permeability to cAMP that would isolate the oocyte from the elevated cAMP in cumulus cells that was observed in the hours after LH (Nakashima et al., 2025). Although Cx37 permeability can be regulated by phosphorylation in some tissues (Jacobsen et al., 2017), Alexa 350 injected into the oocyte diffused into the cumulus cells at all measured time points after LH up to 5 h, suggesting that Cx37 permeability was unchanged by LH signaling (Norris et al., 2008). However, diffusion of a fluorescent dye may not fully reflect a channel’s permeability to other molecules (Kumar, Harris, & Luo, 2025). The characteristics that mediate differential selectivity of molecules, including cyclic nucleotides, by different gap junction proteins (Brink, Valiunas, & White, 2020; Kanaporis et al., 2008; Valiunas & White, 2020) are currently being investigated with cutting-edge ultrastructural and computational methods (Kumar, Harris, & Luo, 2025).

5.4. Potential functions of elevated oocyte cAMP after LH

The presence of a transient LH-induced elevation of oocyte cAMP certainly raises new questions about the mechanisms by which cAMP decreases to allow meiotic resumption. There is also the question of whether the transient cAMP peak has any signaling functions in the oocyte or is merely a physiological consequence that must be overcome. One potential function could be the modulation of PDE3A activity. It has been reported that PKA can directly phosphorylate and increase the catalytic activity of PDE3A (Degerman, Belfrage, & Manganiello, 1997; Murthy, Zhou, & Makhlouf, 2002). If the elevated oocyte cAMP further increases PKA activity, then PDE3A activity could also increase to accelerate cAMP hydrolysis in the oocyte. This mechanism has yet to be tested, but it has been reported that LH receptor stimulation activated Akt/protein kinase B in mouse oocytes, which also phosphorylates PDE3A (Han et al., 2006). This caused a measurable increase in PDE3A activity and was proposed to partly mediate the fall in oocyte cAMP (Han et al., 2006). It would be interesting to explore the potential role of increasing PDE3A activity through either PKA or Akt (or both) in the context of LH-induced gap junction closure to accomplish the cAMP changes that lead to meiotic resumption. It is also possible that the oocyte cAMP increase may facilitate some aspects of cytoplasmic maturation. For instance, in addition to other kinases, phosphorylation by PKA can sensitize the inositol 1,4,5-trisphosphate receptor (IP3R) (Taylor, 2017), which could help regulate the response to the fertilization-induced elevation in IP3 that causes calcium release (Stein, Savy, Williams, & Williams, 2020; Wakai et al., 2011).

6. Summary of the intercellular mechanisms that alter cyclic nucleotide dynamics to restart oocyte meiosis

To summarize our current understanding of how changes in cyclic nucleotide dynamics within a preovulatory follicle mediate the oocyte’s reentry into meiosis to prepare for fertilization, we start with the equilibrium state before LH. Perhaps unsurprisingly given the functional syncytium, relative levels of both cGMP and cAMP are similar throughout the follicle and oocyte when measured with endogenously expressed FRET sensors (Nakashima et al., 2025; Shuhaibar et al., 2015). Phosphorylated NPR2 in the somatic cells, stimulated by its agonist CNP, generates the cGMP that diffuses into the oocyte to competitively inhibit PDE3A. This allows the oocyte to maintain elevated cAMP that is produced in the oocyte in response to the constitutively active GPR3. PKA activity in response to oocyte cAMP phosphorylates regulators of the cell cycle to maintain meiotic arrest.

LH receptor stimulation activates two separate pathways that are both required for meiotic resumption: one immediate and another with a delay of a few tens of minutes. First, cAMP immediately rises in the outermost granulosa cells, activating PKA to cause both the dephosphorylation and inactivation of NPR2, and the phosphorylation and activation of PDE5A. These events cause cGMP levels to fall rapidly. The elevated cAMP then diffuses through gap junctions to cells deeper in the follicle that lack LHR to trigger the same PKA-dependent responses that lower cGMP. While cAMP diffuses down its concentration gradient toward the oocyte, a new cGMP concentration gradient is established that likely “pulls” cGMP out of the oocyte, in addition to hydrolysis by PDE3A. cGMP levels consistently reach their minimum in the follicle and oocyte by 30 min after LH. As oocyte cGMP levels are falling, the LH-induced cAMP wave reaches the oocyte via Cx37 gap junctions. The transient elevation of oocyte cAMP could potentially boost PDE3A activity by PKA-dependent phosphorylation and help turn cAMP levels back toward the initial baseline. However, it is also possible that this is accomplished entirely or in part by the fall in cGMP that allows PDE3A to have unrestricted activity toward cAMP. Regardless, these most rapid events are not sufficient to lower cAMP below the threshold for meiotic resumption with a normal time course.

The second pathway is also initiated rapidly after LH exposure, but occurs through EGFR signaling, explaining the slower kinetics. It is not fully understood how LH signaling leads to EGFR activation, but it has been proposed to occur via cAMP-dependent activation of matrix metalloproteinase (MMP) activity that results in shedding of EGFR ligands already present in the membrane (Panigone, Hsieh, Fu, Persani, & Conti, 2008). This results in autophosphorylation and activation of EGFR that is detected within 10–15 min (Egbert et al., 2016; Liu, Xie, Zamah, Cao, & Conti, 2014). Downstream activation of MAPK leads to phosphorylation and closure of Cx43 gap junctions between somatic cells, apparently halting most of the cAMP diffusion toward the oocyte. Depending on the timing, upregulation of EGFR ligand transcription (which is increased but not maximal at 1 h after LH receptor stimulation) could also contribute to this signaling. EGFR-mediated gap junction closure appears to be the final step required for oocyte PDE3A to fully overcome GPR3 activity and lower cAMP enough for meiosis to resume with a normal time course (Fig. 5). The precise cAMP threshold for triggering oocyte meiotic resumption is not clear. However, the threshold is reached before cAMP levels have fallen to their minimum, because most oocytes had already resumed meiosis before the minimum was observed (Nakashima et al., 2025). It is also possible that the threshold is different for each oocyte depending on the expression levels of its cell cycle machinery, PKA, or the regulatory proteins on which PKA acts.

7. Clarifying the role of EGFR signaling in meiotic resumption

EGFR signaling plays a large role in mediating many crucial events in the follicle initiated by LH (Ashkenazi et al., 2005; Conti, Hsieh, Park, & Su, 2006; Richani & Gilchrist, 2018). Among the later LH-induced events known to require EGFR/MAPK signaling are: the decline in CNP levels (Tsuji et al., 2012), cumulus expansion (Park et al., 2004; Reizel, Elbaz, & Dekel, 2010; Su, Wigglesworth, Pendola, O’Brien, & Eppig, 2002), downregulation of aromatase (Andric, Thomas, & Ascoli, 2010), retraction of TZPs to fully decouple the cumulus and oocyte (Abbassi et al., 2021), and ovulatory gene expression (Ashkenazi et al., 2005). These events are presumably driven by the dramatically increased mRNA expression of EGFR ligands such as epiregulin, amphiregulin, and betacellulin that peaks around 3 h after LH exposure (Ashkenazi et al., 2005; Park et al., 2004).

EGFR stimulation has also long been known to independently cause meiotic resumption without LH exposure (Dekel & Sherizly, 1985; Norris, Freudzon, Nikolaev, & Jaffe, 2010; Park et al., 2004; Shuhaibar et al., 2016). Administration of EGFR ligands also lowered follicle cGMP, either to levels similar to those after LH (Hsieh, Thao, & Conti, 2011; Liu, Xie, Zamah, Cao, & Conti, 2014; Vaccari, Weeks, Hsieh, Menniti, & Conti, 2009), or induced a partial cGMP decrease (Shuhaibar et al., 2015). Furthermore, pharmacological or genetic inhibition of EGFR/MAPK signaling delays or prevents meiotic resumption in response to LH (Ashkenazi et al., 2005; Hsieh, Thao, & Conti, 2011; Liu, Xie, Zamah, Cao, & Conti, 2014; Nakashima et al., 2025; Norris et al., 2008; Norris, Freudzon, Nikolaev, & Jaffe, 2010; Su et al., 2003). This result is likely explained by LH signaling rapidly transactivating the EGFR by stimulating MMP activity that cleaves EGFR ligands from the cell surface (Fischer, Hart, Gschwind, & Ullrich, 2003; George, Hannan, & Thomas, 2013; Liu, Xie, Zamah, Cao, & Conti, 2014; Panigone, Hsieh, Fu, Persani, & Conti, 2008). It is through this mechanism that EGFR has been proposed to mediate the rapid decrease in cGMP after LH, leading to meiotic resumption (Liu, Xie, Zamah, Cao, & Conti, 2014). Indeed, some studies report that inhibition of EGFR activity prevented the LH-induced cGMP decrease (Liu, Xie, Zamah, Cao, & Conti, 2014; Vaccari, Weeks, Hsieh, Menniti, & Conti, 2009). However, other reports have found that inhibiting EGFR had little to no effect on the rapid cGMP decrease (Hsieh, Thao, & Conti, 2011; Nakashima et al., 2025; Norris, Freudzon, Nikolaev, & Jaffe, 2010; Shuhaibar et al., 2015). The reason for this discrepancy is not known, but as described in section 5.3, the role of EGFR signaling in meiotic resumption appears to be largely clarified in a recent study (Nakashima et al., 2025). Namely, transactivation of EGFR by LH signaling closes Cx43 gap junctions shortly after cGMP levels have already fallen. Gap junction closure is proposed to be required to block the flow of the somatic cell cAMP elevation into the oocyte, allowing for the full decrease in oocyte cAMP that causes timely meiotic resumption (Nakashima et al., 2025).

However, this model still leaves some previous results unexplained. For instance, how does direct stimulation of EGFR by treating follicles with EGFR ligands at least partially lower cGMP levels (Shuhaibar et al., 2015; Vaccari, Weeks, Hsieh, Menniti, & Conti, 2009) when live imaging studies have found the LH-induced cGMP decrease to be unaffected by EGFR inhibition (Nakashima et al., 2025; Shuhaibar et al., 2015)? It has been demonstrated that GPCRs and EGFRs can physically interact with each other (Gekle et al., 2024; Patwa et al., 2021) and there is some evidence that EGFR signaling can phosphorylate and activate heterotrimeric G-proteins, leading to cAMP production (Marty & Ye, 2010; Nair, Parikh, Milligan, & Patel, 1990; Nair, Rashed, & Patel, 1989; Poppleton, Sun, Fulgham, Bertics, & Patel, 1996). Though speculative, it is possible that treatment of follicles with EGFR ligands could activate Gs and cause a cAMP elevation sufficient to dephosphorylate at least some NPR2, leading to a fall in cGMP. This could also potentially explain meiotic resumption observed in response to EGFR ligands, since sustained elevation of follicle cAMP by any mechanism can restart meiosis (Vigone et al., 2018). Furthermore, if LH-stimulated activation of EGFR further activates Gs in granulosa cells, it could be a normal mechanism to amplify and extend the cAMP increase initially induced by LH. These possibilities would be exciting to explore, though there is also a second potential explanation for how meiotic resumption might be stimulated by EGFR ligands. The initial EGFR activation between 10–60 min after LH is relatively modest compared to EGFR activity beyond 2 h after LH (Fig. 6), likely due to the time required to upregulate the expression of EGFR ligands (Ashkenazi et al., 2005; Egbert et al., 2016; Panigone, Hsieh, Fu, Persani, & Conti, 2008; Park et al., 2004). If the amount of EGFR ligands applied in previous follicle experiments is at or beyond the concentration normally seen after several hours of LH-induced upregulation, then possibly some EGFR-required events that could affect oocyte meiosis are initiated much earlier than they would occur in vivo. Some examples would be the decrease in follicle CNP (thus lowering cGMP levels), initiation of cumulus expansion, and perhaps even disrupting cumulus-oocyte communication by stimulating TZP retraction. Testing some of these possibilities could help further clarify the role of EGFR in the resumption of meiosis.

Figure 6: Cyclic nucleotide dynamics and other events starting ~2 h after LH receptor activation.

CNP levels begin to fall, likely further lowering NPR2 activity. Upregulation of EGFR ligand (EL) expression is occurring, which will further activate EGFR and mediate later events in the follicle. Closure of Cx43 gap junctions has permitted PDE3A to hydrolyze cAMP to levels that permit meiotic resumption, despite the persistent GPR3 activity.

8. Ubiquity of cyclic nucleotide-mediated meiotic arrest and resumption in mammals

There is strong evidence that the mechanisms regulating the oocyte’s transition to the cell cycle that were first identified in rodents are universal among mammals. Inhibition of PDE3A in isolated COCs was shown to elevate cAMP and prevent spontaneous meiotic resumption in bovine (Franciosi et al., 2014; Li et al., 2016; Bian et al., 2022), feline (Lee, Wildt, & Comizzoli, 2018), and human (Nogueira et al., 2006; Vanhoutte et al., 2007) oocytes. Porcine oocytes express GPR3, and its knockdown caused spontaneous meiotic resumption, while overexpressing GPR3 elevated cAMP to inhibit meiotic resumption (Yang et al., 2012). Human oocytes also express GPR3, and injection of a Gs-blocking antibody caused meiotic resumption (DiLuigi et al., 2008). Two human patients with primary ovarian insufficiency (POI) were found to have missense mutations in GPR3 (Ren et al., 2023), suggesting that as in mice, human oocytes with inactive GPR3 spontaneously resume meiosis as soon as sufficient cell cycle proteins are expressed.

Meiotic arrest mediated via NPR2-produced cGMP in somatic cells has also been demonstrated in other species. Application of CNP or a CNP source to COCs prevents spontaneous meiotic resumption and elevates oocyte cGMP and cAMP in both porcine (Hiradate, Hoshino, Tanemura, & Sato, 2014; Santiquet, Papillon-Dion, Djender, Guillemette, & Richard, 2014; Zhang W et al., 2014) and bovine (Bezerra et al., 2016; Franciosi et al., 2014; Jia & Wang, 2020; Xi et al., 2018) models. Localization of CNP and NPR2 expression in feline follicles was demonstrated to be identical to rodents, and CNP also maintained cGMP levels and meiotic arrest in feline COCs (Zhong et al., 2015). Similar mechanisms in humans were first suggested by the observation that CNP levels fall in follicular fluid following stimulation with human chorionic gonadotropin (hCG) (Kawamura et al., 2011). CNP expression was found to be positively associated with antral follicle count and the number of preovulatory follicles and negatively associated with oocytes that had resumed meiosis (Casalechi et al., 2019). Similarly, NPR2 mRNA was significantly downregulated in somatic cells surrounding oocytes that had resumed meiosis (Casalechi et al., 2019). A different study also found that human NPR2 mRNA decreases following hCG stimulation, though CNP levels remained unchanged (Cadenas et al., 2023). While the molecular components for cGMP-mediated meiotic arrest are present in humans and CNP is being used in protocols for human in vitro maturation (IVM; see below), it remains to be systematically tested whether exogenous CNP can delay or prevent spontaneous meiotic resumption in isolated human COCs.

Understanding the mechanisms of cyclic nucleotide-mediated regulation of meiosis has led to several improvements in protocols for IVM of both large animal and human oocytes from preovulatory and small antral follicles (Gilchrist et al., 2024; Lodde et al., 2025). Especially in COCs from smaller follicles, it has long been appreciated that transiently maintaining cumulus cell function and gap junction communication (pre-IVM) allows for more complete oocyte cytoplasmic maturation prior to meiotic resumption, with measurable benefits for IVM outcomes (Gilchrist et al., 2024; Morato, Verruma, Furtado, & Dos Reis, 2025). Initially, this was accomplished through pharmacological modulation of oocyte cAMP with PDE inhibitors, activation of adenylyl cyclase, or cell-permeable cAMP. However, knowledge of cGMP-mediated meiotic arrest has enabled a more physiological approach to pre-IVM intervention by providing exogenous CNP, either alone or with other hormones, often termed capacitation-IVM (CAPA-IVM) (Gilchrist et al., 2024; Morato, Verruma, Furtado, & Dos Reis, 2025). CAPA-IVM has been tested in bovine (Franciosi et al., 2014; Zhang, Zhang, Fan, Li, & Zhang, 2017), porcine (Zhang et al., 2017), ovine (Zhang, Fan, Li, Zhang, & Zhang, 2018), caprine (Assareh et al., 2022; Zhang, Wei, Cai, Zhao, & Ma, 2015), equine (Fakhar-I-Adil et al., 2025), and murine (Gong et al., 2023; Romero, Sanchez, Lolicato, Van Ranst, & Smitz, 2016; Zhao et al., 2020) models. Additionally, CAPA-IVM has been applied to human IVM (Sanchez et al., 2017), with the first live births from such a protocol being reported in 2020 (Vuong et al., 2020). Overall, these studies have demonstrated clear benefits of CAPA-IVM compared to conventional IVM, although a relatively small randomized controlled trial in humans did not reveal an increase in live births following CAPA-IVM (Gilchrist et al., 2024; Vuong et al., 2020). In addition to driving improvements in domesticated animal, and possibly human, fertility interventions, cyclic nucleotide-modulating IVM protocols could also be applied to conservation of endangered species (Hildebrandt & Holtze, 2024; Mastromonaco, 2023; Thongphakdee, Sukparangsi, Comizzoli, & Chatdarong, 2020).

9. Conclusions and unresolved questions

Considerable progress has been made in our understanding of how intercellular cyclic nucleotide signaling between the somatic cells and the oocyte controls meiotic arrest and then mediates meiotic resumption in response to the mid-cycle LH surge. The most recent work has demonstrated that cyclic nucleotides in the follicle and oocyte are regulated in a two-step process after LH. First, cAMP generated in the outer mural granulosa cells diffuses through gap junctions to all somatic cells (Nakashima et al., 2025), activating PKA to cause the dephosphorylation and inactivation of NPR2 (Egbert et al, 2024). This drives the rapid fall in follicle and oocyte cGMP levels which permits, but is not solely sufficient for, meiotic resumption with a normal time course. Initially, cAMP generated by LH signaling in the granulosa cells diffuses into the oocyte, creating a transient barrier to meiotic resumption. In the second step, LH signaling causes the release of extracellular EGFR ligands that partially activate EGFR/MAPK signaling to phosphorylate and decrease the permeability of Cx43 gap junctions. This step, which occurs with variable timing between follicles, interrupts the diffusion of cAMP into the oocyte and allows cAMP to fall below a threshold for meiosis to restart (Nakashima et al., 2025). Evidence that these mechanisms are likely to be universal among mammals continues to emerge. As a result, knowledge from rodent models of the cyclic nucleotide-mediated regulation of meiosis has led to tangible improvements in IVM protocols for domesticated animals and humans (Gilchrist et al., 2024; Morato, Verruma, Furtado, & Dos Reis, 2025). Finally, intercellular and gap junction-mediated signaling in ovarian follicles might apply to other aspects of oocyte maturation (Tscherner & Baltz, 2025) or to related signaling in other tissues (Aasen, Mensil, Naus, Lampe, & Laird, 2016; Sedovy et al., 2023).

Nevertheless, important questions remain unanswered. Those below would be particularly exciting and fruitful avenues for future study:

• What are the kinases and phosphatases responsible for regulating NPR2 activity by phosphorylation/dephosphorylation?

• Besides maintaining meiotic arrest and number of TZPs, do the elevated levels of cGMP play other roles in the follicle, and by what mechanisms?

• Can CNP delay meiotic spontaneous resumption in isolated human COCs?

• What is the full scope of events by which LH signaling transactivates the EGFR? What are the early kinetics of EGFR ligand transcription? Does LH signaling increase MMP activity?

• Does EGFR stimulation cause an additional increase in cAMP in the follicle?

• Does the LH-induced transient elevation of cAMP in the oocyte have any functions, such as to increase oocyte PDE3A activity, accelerating cAMP hydrolysis?

• Do gap junctions have differential permeability to cGMP and cAMP either before or after LH?

Abbreviations:

ANP

Atrial natriuretic peptide

ART

Assisted reproductive technologies

BMP15

Bone morphogenic protein 15

cAMP

Cyclic adenosine monophosphate

CAPA-IVM

Capacitation IVM

CBX

Carbenoxolone

cGMP

Cyclic guanosine monophosphate

CNP

C-type natriuretic peptide

COC

Cumulus-oocyte complex

Cx

Connexin

EGFR

Epidermal growth factor receptor

ELISA

Enzyme-linked immunosorbent assay

FGF8

Fibroblast growth factor 8

FRET

Förster resonance energy transfer

GDF9

Growth differentiation factor 9

GPCR

G protein-coupled receptor

GPR3

G protein-coupled receptor 3

GPR12

G protein-coupled receptor 12

HA

Hemagglutinin

hCG

Human chorionic gonadotropin

iNOS

Inducible nitric oxide synthase

IP3

Inositol 1,4,5-trisphosphate

IP3R

Inositol 1,4,5-trisphosphate receptor

IVM

In vitro maturation

LH

Luteinizing hormone

LHR

LH receptor

MAPK

Mitogen-activated protein kinase

MMP

Matrix metalloproteinase

NO

Nitric oxide

NPR1

Natriuretic peptide receptor 1

NPR2

Natriuretic peptide receptor 2

PDE

Phosphodiesterase

PKA

Protein kinase A

PKG

Protein kinase G

POI

Primary ovarian insufficiency

PPP

Phosphoprotein phosphatase

S1P

Sphingosine-1-phosphate

sGC

Soluble guanylyl cyclase

TGFβ

Transforming growth factor-β

TZP

Trans-zonal projection