Biophysical, cellular, and mouse model approaches to investigate the mechanical regulation of folliculogenesis

Abstract

Folliculogenesis, which is the process by which ovarian follicles develop to support oogenesis and hormone production, is essential for female fertility. Although hormonal and biochemical signaling pathways regulating folliculogenesis have been extensively studied, increasing evidence suggests that mechanical cues within the ovary also play a critical role. The ovary is composed of follicles, corpora lutea, and stroma, each contributing to a biomechanical microenvironment that might change across the reproductive lifespan. Additionally, the spatial organization of the ovary, with a collagen-rich cortex and a softer medulla, may influence follicle activation and growth. This review explores the hypothesis that mechanical properties of the ovary regulate folliculogenesis, integrating current knowledge on ovarian architecture, extracellular matrix composition, and mechanotransduction pathways. We highlight recent findings supporting mechanical regulation of folliculogenesis, discuss contradictory data, and describe the tools and models used to investigate this concept. By considering mechanical forces alongside hormonal and biochemical signals, we propose a more integrated view of the factors governing follicle development, with implications for understanding ovarian physiology and pathology.

1. Introduction

The mammalian ovary has two essential roles: to generate developmentally competent oocytes that have all the components to support early embryogenesis, and to produce hormones that support functions of both the ovary and other tissues, including the bones, breast, and uterus. Central to these roles is the follicle, the functional unit of the ovary. Within each follicle, an oocyte is surrounded by granulosa and theca cells. During folliculogenesis, which takes approximately 300 days to complete in humans [1] and around four weeks in mice [2–4], these somatic cells support oocyte growth and maturation by providing nutrients and structural support. This process occurs within a highly dynamic ovary, which continuously undergoes follicle activation, growth, and degradation (atresia), ovulation, wound healing, and blood vessel remodeling. Additionally, folliculogenesis is likely affected by the surrounding ovarian structures, which can be categorized functionally or spatially.

Functionally, the ovary contains corpora lutea and stroma. Corpora lutea are temporal structures that form after ovulation and result from differentiation of the granulosa and theca cells from the ovulated oocyte. One of their main functions is to produce the hormone progesterone, which is essential to support embryo implantation and early pregnancy. The stroma is the connective tissue that surrounds the corpora lutea and the follicles. In humans, the stroma is the most abundant multicellular structure in the ovary and is composed of fibroblasts, endothelial cells, immune cells, and other cell types [5].

Spatially, the ovary can be divided into the outer cortex and the inner medulla. Primordial follicles reside in a quiescent state in the cortex and largely develop in the medulla [6]. In humans, the cortex has higher collagen content than the medulla and has long been thought to be stiffer. Thus, cortical stiffness has been proposed to maintain primordial follicle quiescence, whereas the softer medulla has been proposed to be permissive to follicle growth. In this review, we will discuss both supporting and contradictory evidence regarding this idea.

All primordial follicles form before birth in humans and right after birth in mice. In both species, follicles form when germ cell cysts break down and single germ cells become enclosed by a single layer of squamous granulosa cells. The oocytes of primordial follicles are arrested at diplotene in prophase I and remain arrested in this meiotic stage until ovulation, which can be months later in mice and decades later in humans. Once the primordial follicles form, they can undergo one of three fates: remain quiescent, go through atresia, or activate and start folliculogenesis. Primordial follicle activation is likely controlled by a combination of biomechanical and biochemical communication within the ovary. When a follicle is activated to become a primary follicle, squamous granulosa cells transition to a cuboidal shape, the oocyte starts to grow in size, and the zona pellucida (ZP) forms. The ZP, an extracellular matrix (ECM) coat located between the oocyte and the granulosa cells, plays an important role in preventing polyspermy. From this point on, the oocyte also starts to produce and store mRNA, proteins, and other components that will support early embryo development before embryonic genome activation. As follicles enter the secondary stage, the granulosa cells continue to proliferate, the oocyte grows, and the nearby stromal cells differentiate into theca cells and encapsulate the follicle. They form a highly vascularized region that receives cholesterol from the blood vessels and converts it to androgen, which will be secreted and transformed to estradiol in the granulosa cells. Next, a fluid-filled cavity, the antrum, is formed as granulosa cells differentiate into mural and cumulus cells. Finally, ovulation occurs when the ovary surface ruptures to release an oocyte, which resumes meiosis so that it is ready for potential fertilization in the oviduct in rodents or the fallopian tube in humans [7].

Folliculogenesis is tightly regulated at multiple levels. First, in the hypothalamic-pituitary-ovarian axis, the hypothalamus releases gonadotropin-releasing hormone, which induces the anterior pituitary gland to release luteinizing hormone and follicle-stimulating hormone into the bloodstream. In the ovary, these hormones regulate the last stages of folliculogenesis and stimulate production of estradiol and progesterone [8]. Second, complex biochemical bidirectional communication occurs between the oocyte and the somatic cells. For example, the oocyte-secreted growth differentiation factor-9 stimulates granulosa cells to produce metabolites. These metabolites are transferred to the oocyte by transzonal projections, filopodia-like structures that connect granulosa cells with the oocyte [9]. In this review, we propose a third level of regulation of folliculogenesis – mechanical inputs from the ovary – and provide evidence from the literature to support this idea. We focus on the current knowledge and techniques available to examine regulation of folliculogenesis by ovarian mechanical signals. Although numerous animal models have been used to address this issue, such as Drosophila, we primarily focus on three mammalian species in which most work in this area has been conducted: mice, bovines, and humans.

2. Principles of mechanotransduction

Cells can modulate their behavior in response to mechanical cues such as hydrostatic pressure, tensile forces, fluid shear stress, and ECM stiffness. Some of these forces are evident in the ovary. For example, hydrostatic pressure is generated during antrum formation [10], where fluid filled cavity is formed. As mechanotransduction has been reviewed multiple times [11–18], we only provide a brief summary of key elements involved in this process.

2.1. Mechanosensing

Cells detect mechanical cues at the plasma membrane, where the cell contacts the environment [19]. One primary driver of mechanosensing is integrins, transmembrane receptors that respond to various mechanical cues, especially ECM stiffness. In most cases, integrins mediate the formation of focal adhesions, which connect the ECM to the cytoskeleton to propagate the external forces intracellularly [20]. Ion channels can also act as mechanosensing components. For instance, the calcium channel transient receptor potential vanilloid type 4 (TRPV4) senses mechanical loading in chondrocytes [21].

2.2. Signaling pathways

Mechanical cues can directly or indirectly modulate intracellular signaling through several signaling pathways. In one pathway, the Rho-family small GTPase RhoA signals through the serine/threonine kinase ROCK to mediate cytoskeleton remodeling and cell contractility [22]. In another pathway, signaling through phosphatidylinositol 3-kinase (PI3K) leads to activation of AKT/mTOR, which promotes cell growth, survival, and metabolism by regulating protein synthesis and inhibiting apoptotic pathways [23]. Finally, several pathways, such as the Hippo signaling pathway and F-actin dynamics, can regulate the YAP/TAZ transcription factors, causing them to enter the nucleus or remain in the cytoplasm. These transcription factors regulate genes such as connective tissue growth factor (CTGF) and Baculoviral IAP repeat Containing 5 (BIRC5) [20]. Although many other pathways regulate mechano - signaling, we focus on RhoA-ROCK, PI3K-AKT/mTOR, and YAP/TAZ, which play important roles in the ovary [12,24–26].

2.3. Cellular response

Cells can respond to mechanical cues by modulating their differentiation, proliferation, or migration in cell- and tissue-dependent manners. During development, mechanical forces are important for cell specification (e.g., inner cell mass vs. trophectoderm) [27,28]. In wound healing, mechanical cues can stimulate ECM production [29,30]. In a disease context, increased ECM stiffness promotes cancer cell migration and invasion through enhanced mechanotransduction signaling [31–33]. Although most studies on cellular mechanotransduction and response have focused on cells cultured in two dimensions (2D), recent developments are allowing investigators to probe three-dimensional (3D) cultures. Given that we have techniques to culture ovarian follicles in vitro, the follicle is an ideal system to investigate novel mechanisms of and responses to mechanotransduction in 3D multicellular structures.

3. Biophysical tools to measure mechanical properties of the ovary

Understanding ovarian function, including folliculogenesis and oogenesis, requires measuring and perturbing forces to comprehend how mechanical forces are integrated and modulate ovarian function. In recent years, several researchers have measured the stiffness, elasticity, and viscoelasticity of the ovary, follicle, and oocyte. To do so, they have used various biophysical tools, each with strengths and limitations (Table 1), to provide complementary information. Below, we summarize the most significant discoveries made with these tools and highlight consistencies and discrepancies in the data.

Table 1.

Biophysical tools currently used to measure the mechanical properties of the ovary.

Biophysical tool	Description	Biomechanical properties analyzed	Advantages	Limitations
Indentation Atomic force microscopy Nanoindentation	Indenting tissue with a probe, where the deflection of the cantilever can be used to extract tissue’s mechanical properties	Elasticity Viscosity	High-resolution mapping Widely available	Requires ex vivo analysis and direct contact with the sample, which might damage specimens No access to tissue interior unless the tissue is sectioned, which can alter its mechanical properties Elasticity model dependent
Elastography Shear wave elastography Magnetic resonance elastography Microelastography	Mechanical stimulation applied, and mechanical properties inferred by matching the resulting deformation to a predefined model	Elasticity Viscosity	Shear wave and Magnetic resonance elastography: non-invasive. Optical microelastography: compartment and subcellular resolution	Shear wave and Magnetic resonance elastography: low resolution Optical microelastography: ex vivo, might damage the sample. Alginate embedding needed. Requires highly specialized and expensive systems
Micropipette aspiration	Aspiration of a cell or tissue into a glass pipette. The distance the cell travels inside the pipette is used to determine its mechanical properties	Viscoelasticity Cortical tension	Mechanical characterization of living cells Single cell analysis Widely available, affordable, and easy to perform Can be used to measure tissue surface tension	Lack of subcellular resolution May damage cells if too much aspiration is applied
Brillouin microscopy	Mechanical properties are determined by the changes in light scattering induced by the different densities of the same specimen	Bulk modulus Micro-viscosity	High resolution analysis of subcellular structure 3D quantification of tissue viscoelasticity Contact-free	Ex vivo analysis Can be phototoxic Complex system and not widely available

3.1. Indentation

In this approach, a tissue is indented with a cantilever probe, the deflection of which provides information about the mechanical properties of the tissue. We used this method to measure stiffness of the cortex of isolated ovaries from young mice. These experiments demonstrated that the ovary has an average stiffness of 1.79 ± 0.08 kPa, which is much softer than most other tissues in the body (Fig. 1). We also showed that mouse ovaries become stiffer with age, reaching 4.56 ± 2.03 kPa. This increase in stiffness was dependent on deposition of collagen in the ovarian stroma [34]. The age-associated increase in ovarian stiffness was later shown to also occur in cats [35] and humans (reproductively young ovaries: 3.2 ± 0.25 kPa, menopausal ovaries: 7.1 ± 0.71 kPa). In prepuberty, human ovaries have similar stiffness (6.5 ± 0.35 kPa) as menopausal ovaries [36], suggesting that prepubertal and menopausal ovaries — both of which are reproductively inactive — share biophysical features.

Fig. 1.

Stiffness of the ovary and other tissues. Mouse and human ovaries are softer than tissues such as the breast, skin, and muscle. Values from other tissues were obtained from Guimarães et al. [42].

An important open question to address is whether the ovarian cortex and medulla have different mechanical properties and whether this is an evolutionary conserved phenotype. Indentation studies have yielded contradictory results. In bovine ovaries, the medulla appears to be softer than the cortex [37,38]. In the cat, the cortex is significantly more elastic than the medulla [35]. However, in mouse ovaries, both the cortex and medulla are extremely soft, whereas the intermediate region, where large follicles are located, is stiffer [39]. In humans, the ovarian cortex is highly enriched in collagen, but this is not the case in mice [34,40,41]. These observations are inconsistent with the idea that large follicles are located in soft environments to allow them to expand.

Indentation has also been used to demonstrate that cumulus oocyte complexes (COCs), oocytes, and ZP from mice, bovines, and humans are extremely soft and elastic [43–47]. Mouse COCs matured in vitro are stiffer (higher Young’s Modulus) than those matured in vivo [47]. This difference in biomechanical properties might affect the ability of sperm to reach the oocyte. The ZP elasticity of human oocytes has also been correlated with oocyte maturation, quality, and fertility rates. After fertilization, the ZP hardens to prevent polyspermy [48]. For example, in mice, the ZP surrounding two-cell embryos is 2.9-fold stiffer than the ZP surrounding oocytes [49]. Stiffness values reported for mouse embryos are variable, ranging from 0.5 kPa [50] to 40 kPa [48,51–53]. These values may represent differences in oocyte maturation stage, mouse strains, biophysical tools, and mathematical models used to analyze the data.

Indentation has also been used to analyze whether in vitro fertilization procedures alter oocyte biomechanics. For example, one study showed that the open-vitrification protocol, which is commonly used to cryopreserve oocytes, does not affect the mechanical properties of retrieved human healthy oocytes [54]. Another study addressed whether postovulatory aging, which is defined as delayed fertilization and can result in lower fertilization and embryo development rates [55, 56], affects oocyte stiffness. The authors reported that postovulatory aging did not affect the mechanical properties of morphologically healthy oocytes, though stiffness did increase over time in oocytes that degraded after 6 h of culture [57].

3.2. Shear wave elastography

Elastography consists of mechanically stimulating a tissue and fitting its deformation to a parameterized model to obtain the mechanical properties of the tissue [58]. Because elastography damages tissue, shear wave elastography (SWE) was developed as a non-destructive elastography tool. In SWE, an ultrasound probe generates shear waves that propagate perpendicular to the target tissue. Stiffness is determined by measuring the time it takes for these waves to travel through the tissue. In soft tissues, the waves penetrate more quickly than in stiffer tissues [59]. This method has been used clinically to determine the stiffness of several tissues, including the liver [60,61]. Gargus and colleagues were the first to use SWE to determine the mechanical properties of the bovine ovary. In ex vivo experiments, they showed that the medulla was significantly stiffer than the cortex [62], which contradicts conclusions from indentation experiments [35,37,38]. This discrepancy may reflect differences in mechanical testing systems (atomic force microscopy vs. SWE), sample preparation (e.g., fresh vs. thawed, isolated vs. encapsulated in gel beads), and animal model tested (e.g., mouse vs. bovine), highlighting the need for additional rigorous research.

SWE is often used to measure the mechanical properties of human ovaries in patients with polycystic ovarian syndrome (PCOS). Such ovaries have higher collagen content and greater elasticity than ovaries from women without PCOS [63–67]. Although SWE has not been used in mice, which have much smaller ovaries than humans, He et al. used this method to define the mechanical properties of ovaries in a rat model of PCOS [68], suggesting that this approach is feasible in larger rodent models.

Three novel elastography modalities have been used recently to probe the ovary. First, multi-modal magnetic resonance elastography was used to measure the rigidity of the ovary from women with or without PCOS [69]. This method can be used in vivo without damaging the tissue, but obtaining tissue compartment-specific measurements is challenging. Second, optical microelastography has been used to measure the viscoelastic properties of individual mouse oocytes [70]. Although this analysis must be performed ex vivo, it offers subcellular resolution. However, whether the long exposure to direct light affects oocyte viability has not been determined. Third, elastography has been adapted to measure the mechanical properties of the entire ovary ex vivo at high resolution. This method revealed that the mechanical properties of the ovary differ in different sub-compartments [71].

3.3. Micropipette aspiration

In this method, a micropipette attached to a microfluidic vacuum pump is applied to the surface of a cell to non-invasively measure the stiffness, viscoelastic properties, and cortical tension of individual cells. Mechanical properties can be estimated from the distance the cell travels inside the micropipette in response to specific aspiration pressures [72, 73]. The micropipette aspiration system has been widely used to study the mechanical properties of various cell types, including leukocytes, blood cells, endothelial cells, and others [72,74–76]. To date, only one group has used this system to measure the mechanical properties of the ovarian cortex. They did not find significant differences in stiffness values between ovaries from dogs and mice, nor between pre-pubertal and post-pubertal animals [77].

Micropipette aspiration has been more commonly used to determine the mechanical properties of single oocytes [78]. For example, this system was used to examine the role of oocyte cortex tension in regulating oocyte quality in young and aged mice. Oocyte surface tension is regulated by actin nucleation and myosin-II localization beneath the plasma membrane, and alterations in oocyte surface tension result in chromosome segregation defects and embryo development arrest [79–83]. The micropipette system has also been used to assess the mechanical properties of oocytes and the ZP and to use this information to predict oocyte quality, fertilization, and embryo development rates in mammals [51,84–86]. Notably, the mechanical properties of cryopreserved fertilized human oocytes could be used to predict embryo viability and blastocyst formation [50]. Given that the micropipette aspiration equipment can be easily integrated into the microinjectors routinely used in in vitro fertilization laboratories, this method has great potential as a non-invasive way (if performed by properly trained personnel) to analyze oocyte quality in assisted reproduction.

3.4. Brillouin microscopy

This optical non-invasive system measures the balance between elasticity and viscosity in cells and tissues in 3D, and maps mechanical properties at micron-scale precision. It is based on the interaction and inelastic scattering of monochromatic laser light from thermally driven acoustic phonons at high frequencies, where the scattered light spectrum indicates the targeted sample’s sound velocity and thus its spatial mechanical properties [87–89]. Given its 3D imaging capability, high spatial resolution, and non-invasiveness, Brillouin microscopy is well-suited for cell and tissue biology, development, and delicate biological samples including ovarian tissue. Using this system, Chan et al., found that the ovarian cortex of postnatal day 7 and 14 mice is more elastic than the ovarian medulla. Additionally, the authors reported spatial mechanical compartmentalization within the follicle, such as differences between the theca layer and granulosa cells [90].

3.5. The potential for new approaches to describe mechanical properties of the ovary

There are two significant knowledge gaps in the characterization of the biomechanical properties of the ovary. First, different tools have yielded contradictory data regarding the mechanical properties of the ovarian cortex and medulla. Clarifying whether the cortex and medulla exhibit different stiffness and viscoelastic properties is crucial for defining the effects of mechanical properties on follicle dormancy and activation. Second, the mechanical properties of ovary-specific cell types, beyond the oocyte, are largely unknown. Recently, Biswas et al. [91] used traction force microscopy to show that theca cells exert more traction stress than granulosa cells. Another tool that may prove useful is optical tweezers, which were used to show that the mouse oocyte cytoplasm has a homogeneous viscosity [92]. Other potentially useful tools include microconstriction assays [93] and particle image velocimetry [94,95]. We are optimistic that these new tools, as well as improvements to existing tools, will yield detailed information about the biomechanical properties of each cell type in the ovary during folliculogenesis.

4. Three-dimensional culture methods to investigate mechanical regulation of folliculogenesis and oocyte quality

In 1996, Eppig and colleagues cultured mouse ovarian follicles in vitro for the first time, resulting in production of oocytes that were fertilized in vitro and transferred into mice, giving rise to viable offspring [96]. Twelve years later, Telfer and colleagues cultured human primordial follicles to produce antral follicles [97]. Both studies represented important advances in the field. However, although the oocytes and granulosa cells were functional in these 2D systems, the follicles lacked their characteristic 3D structure. In 2003, Pangas et al. developed the first method of 3D culture of follicles embedded in a hydrogel [98]. Since then, 3D in vitro culture of follicles has become a powerful tool to recreate the ovarian environment in a dish and to investigate how biomechanical cues and components of the ECM regulate folliculogenesis (Fig. 2). Several culture systems have been developed, as described next.

Fig. 2.

Hydrogels can be used to determine how mechanical properties of the ovary affect folliculogenesis and oocyte quality. Follicles can be cultured in hydrogelencapsulated structures to preserve their 3D structure in vitro. (A,B) With aging or in pathological conditions such as PCOS, the ovaries become stiffer and folliculogenesis is compromised. This results in less ovulation and fewer corpora lutea. (C) High-quality follicles can be cultured in hygrogels that recapitulate the mechanical properties of physiologic (A) and pathologic (B) conditions. Culture in stiff hydrogels results in reduced follicle growth, oocyte quality, and antrum formation, and generates a fibroinflammatory environment. Similar phenotypes are observed in ovarian follicles in humans with PCOS and in ovaries from reproductively old mice.

4.1. Alginate

The hydrogel alginate extracted from brown algae has been widely used to culture ovarian follicles because of its easy manipulation, stability, and reproducible results. Follicles from mice, bovines, sheep, pigs, dogs, goats, baboons, non-human primates, and humans have all been cultured in alginate, yielding matured oocytes [99–115]. A limitation of alginate hydrogel is that it cannot be degraded. This limits follicle expansion, especially during the last stages of folliculogenesis when granulosa proliferation and follicle size increase exponentially. To overcome this limitation, some researchers have conjugated alginate with ECM components such as fibrin or Matrigel. Such fibrin-alginate hydrogels yielded higher mouse follicle development rates and a higher percentage of matured oocytes than alginate hydrogels [116]. Similar results were obtained with macaque and caprine follicles [117, 118]. Matrigel-alginate 3D scaffolds have also proven to be an effective environment for follicle development [119,120]. Other researchers have created alginate hydrogels containing the glycosaminoglycan hyaluronan, which is abundant in the follicle and provides a soft and hydrated environment [34,121]. In one study, follicle development and oocyte meiotic maturation were more successful in hyaluronan-alginate than in fibrin-alginate or alginate alone [122]. Such systems will allow us to investigate how specific ECM components regulate folliculogenesis.

A significant advantage of using alginate is that its mechanical properties are easily manipulated by altering the alginate concentration, presenting an excellent system to investigate how changes in the mechanical properties of the environment influence folliculogenesis and oocyte quality (Fig. 2). In early experiments, researchers tested alginate concentrations between 0.25 % and 3 % and found that lower alginate concentrations favored follicle expansion, antral formation, steroid production, and oocyte development [123–127]. However, one alginate concentration is likely not ideal for follicles from all species. For example, primate primordial follicles require a more rigid environment to grow in vitro [128] than do primordial follicles from cows [114]. Additionally, a single alginate concentration does not mimic the different physical cues that a follicle might encounter during folliculogenesis. In an attempt to create a heterogenous environment, Choi et al. encapsulated follicles in beads containing an alginate gel that mimicked the stiffer region of the ovary. The beads also contained a collagen core that mimicked the soft environment of the medulla. In this system, the follicles could move from the alginate to the collagen during growth. However, fewer than 35 % of the follicles reached the antral stage, suggesting that more sophisticated systems might be needed to better mimic the heterogeneous physical properties of the ovary [129].

Recently, we cultured mouse follicles in alginate gels that mimicked the soft and stiff environments of ovaries from reproductively young and old mice, respectively. We found that follicles are highly mechanosensitive, as their transcriptome changed within three hours of transfer to stiff alginate, generating a fibro-inflammatory environment. Longterm culture resulted in impaired follicle development and oocyte quality. These experiments also revealed dynamic crosstalk between the follicle and the environment [130] (Fig. 2).

4.2. Polyethylene glycol (PEG)

Several laboratories have cultured mouse and bovine follicles in the synthetic hydrogel PEG [131]. To create a gel that can be degraded, PEG was crosslinked with peptides that can be degraded by proteases secreted by the follicle during culture. In this system, follicles could fully expand [132]. PEG hydrogels have been functionalized with integrin-binding peptides such as Arg-Gly-Asp, which allows cell-matrix interaction via integrins [133,134]. Likewise, PEG was functionalized with Activin A to investigate the role of members of the transforming growth factor β family in folliculogenesis [133]. Recently, a new PEG biomaterial that can retain the ECM components secreted by the follicle was synthesized by crosslinking the synthetic hydrogel with basement membrane binder peptides [135]. This system offers a unique opportunity to determine how the follicle-produced ECM modulates its surrounding environment and how that ECM affects follicle development and oocyte quality.

4.3. ECM-based hydrogels

Culture of follicles with ECM-derived hydrogels allows researchers to examine the contribution of the ovary-specific matrix in regulating folliculogenesis. An advantage of this system is the ability to create ECM hydrogels of varying compositions and fiber orientations and widths to match the different ovarian compartments. Such a system will allow investigators to study the effects of regional differences in ECM composition on folliculogenesis. McDowell and colleagues demonstrated that this approach is feasible and can be used to obtain information regarding how biochemical and biomechanical cues of the ovary crosstalk to regulate folliculogenesis and oocyte quality [136].

4.4. Tissue pressure chamber

To mimic the compressive environment of the ovarian cortex where primordial follicles are located, Nagamatsu cultured ovaries in a pressure culture chamber. This work revealed that mechanical stress from the ECM regulates follicle activation [137]. With such an approach, it should be possible to define the effects of compression on folliculogenesis in isolated follicles.

4.5. Future methods

One of the main limitations of the systems described in 4.1 through 4.4 is that they largely only mimic differences in stiffness. In the future, new technologies should be developed that allow researchers to vary other important biophysical parameters, such as viscoelasticity, porosity, and permeability. Recently developed biomaterials might allow us to better determine how biophysical cues modulate folliculogenesis. For example, granular and nanofibrillar alginate hydrogels have been used to culture mouse follicles [138]. Conductive hydrogels with tunable stiffness have recently been developed [139,140]. These novel biomaterials offer a promising tool to precisely mimic the ovarian microenvironment and enable high-throughput quantifications. Such advances will allow researchers to gain unprecedented insights regarding biophysical regulation of folliculogenesis.

5. Mouse models to investigate pathways that regulate follicle mechanotransduction

The PI3K/Akt, Hippo/YAP, and RhoA/ROCK signaling pathways have been linked to mechanotransduction across multiple tissues. To investigate how follicles sense and integrate mechanical cues from their environment, researchers are turning to mouse models originally developed to disrupt key mechanotransduction pathways in other systems. Yet, in the ovary, a direct connection between mechanical cues and the regulation of these pathways remains to be rigourously establishled. Here, we highlight selected models used to elucidate three key pathways: PI3K/Akt, Hippo/YAP, and RhoA/ROCK.

5.1. PI3K/Akt pathway leading to modulation of FOXO3 and mTOR

The PI3K/Akt pathway is a hub for mechanotransduction in many cell types. Given that this pathway mediates primordial follicle dormancy and activation, it may also contribute to mechanotransduction in follicles [25,26,141,142]. PI3K is activated in oocytes upon binding of the ligand Kit, secreted by granulosa cells, to the c-Kit receptor on oocytes. Upon this interaction, PI3K is activated and phosphorylates phosphatidylinositol-4,5-biphosphate (PIP2) to produce phosphatidylinosityol-3,4,5-triphosphte (PIP3) [143,144]. This conversion can be inhibited by PTEN (Phosphatase and Tensin Homolog). PIP3 synthesis can promote Akt phosphorylation, which has many downstream targets including the transcription factor Forkhead Box O3 (FOXO3). FOXO3 promotes expression of genes related to cell cycle arrest and thus mediates follicle dormancy. Akt-mediated phosphorylation causes FOXO3 to translocate from the nucleus to the cytoplasm, thereby allowing primordial follicle activation [145]. PI3K/Akt can also regulate the activity of mTOR in the oocyte by phosphorylating the TSC1/TSC2 complex. mTOR is a highly conserved serine/threonine kinase that regulates cell growth, metabolism, survival and migration. Therefore, mTOR activation results in primordial follicle activation [12].

Several in vitro approaches and mouse models have been used to investigate the role of the PI3K/AKT – FOXO3/mTOR mechanosensitive pathway in follicle activation, development, and oocyte quality (Table 2). Treatment of mouse and human ovaries in vitro with PI3K activators led to primordial follicle activation [146–148]. Similarly, oocyte-specific deletion of Pten in mice (Pten^flox/flox; Gdf9-Cre) caused premature activation of the entire primordial follicle pool [149].

Table 2.

Mouse models to study follicle mechanotransduction and associated phenotypes.

Mouse model	Target cell type	Gene(s) Target	Key findings	Reference
Ptenflox/flox; Gdf9-Cre	Oocyte	Pten	Primordial follicle pool activation	Reddy et al., 2008
FOXO3g−/−	Global KO	Foxo3a	Primordial follicle pool activation	Castrillon et al., 2003
Transgenic mouse with active Foxo3a	Oocyte	Foxo3a	Follicle development retardation	Liu et al., 2007
Mtor flox/flox ; Gdf9-Cre	Oocyte	Mtor	Compromised oocyte developmental competence and reduced ovulation	Guo et al., 2018
Mtor flox/flox ; Zp3-Cre	Oocyte	Mtor	Slightly compromised oocyte developmental competence	Guo et al., 2018
Yap flox/flox ; Zp3-Cre	Oocyte	Yap/Taz	Compromised germ cell development	Kagiwada et al., 2021
Yap flox/flox ; Gdf9-Cre	Oocyte	Yap/Taz	Impaired early embryogenesis	Yu et al., 2016
Yap flox/flox ; Foxl2-Cre	Granulosa cells	Yap/Taz	Compromised granulosa cell proliferation and follicle development	Lv et al., 2019
Yap flox/flox ; Cyp19-Cre	Granulosa cells	Yap/Taz	Compromised GC proliferation	Lv et al., 2019
Lats1/2 flox/flox ; Cyp19-Cre	Granulosa cells	Lats1/Lats2	Ovarian structural abnormalities	Tsoi et al., 2019
Cdh2 tm1Glr /; Amhr2 tm3 (cre)Bhr	Granulosa cells	Cdh2	Compromised follicle development and ovulation	Emery et al., 2024

To examine the link between the ECM and the PI3K/AKT pathway, Nagamatsu et al. treated mouse ovaries with collagenase. The decrease in collagen abundance led to FOXO3 translocation to the cytoplasm in primordial follicles. Conversely, application of mechanical stress by external pressure restored nuclear FOXO3 and follicle dormancy [137]. This work links ECM cues with a mechanosensitive pathway in the oocyte. Mouse models have also been generated to study FOXO3 in the ovary. Consistent with the idea that FOXO3 regulates follicle dormancy, global Foxo3a knockout (FOXO3a^−/−) mice had uncontrolled massive follicular activation and early infertility [150] but had normal follicular growth after activation [145]. In other work, mice with constitutively active FOXO3 in oocytes showed slower follicle development and infertility [151]. These mouse models demonstrate that FOXO3 is essential for follicle activation.

To examine the role of mTOR, researchers cultured mouse and human ovaries with the mTOR inhibitor rapamycin, leading to primordial follicle dormancy [152]. In other work, researchers used Gdf9-and Zp3-cre to delete candidate genes at the primordial and primary follicle stages, respectively. Depletion of Mtor specifically in primordial follicles led to complete infertility (Mtor^flox/flox; Gdf9-Cre), a dramatic reduction in ovulation, and compromised oocyte developmental competence. Depletion of Mtor in primary follicles (Mtor^flox/flox; Zp3-Cre) led to compromised oocyte developmental competence and infertility, but folliculogenesis and ovulation were only slightly reduced [153]. Currently, no animal models exist to disrupt mTOR in granulosa cells.

5.2. Hippo signaling pathway

The highly conserved mechanosensitive Hippo signaling pathway acts through a phosphorylation/dephosphorylation cascade to regulate activity of the transcription factor YAP and its coactivator TAZ. Hippo is regulated by various signals such as ECM stiffness, cell density, substrate stiffness, and shear stress. Activation of the Hippo pathway leads to YAP phosphorylation and recruitment to the cytoplasm, leading to YAP/TAZ inactivity. Conversely, inhibition of the Hippo pathway leads to YAP dephosphorylation, nuclear translocation, and binding to the TEA domain family members (TEAD 1–4) to promote transcription of target genes [12]. In mouse and human ovaries, key Hippo pathway components, including YAP, phosphorylated YAP, and TAZ, are expressed in primordial, primary, secondary, and antral follicles, as well as in atretic follicles and corpus luteum [154,155]. Yap mRNA is more highly expressed in oocytes than in granulosa and theca cells, whereas YAP protein is more abundant in granulosa and theca cells than in oocytes. These observations suggest that YAP plays different roles in the germ cell and somatic compartments [12].

Several lines of evidence suggest that the Hippo pathway regulates both primordial follicle activation and development. In 2013, Kawamura and colleagues [154] isolated ovaries from prepubertal mice, fragmented them, and grafted them under kidney capsules. Tissue fragmentation increased the number of growing follicles, decreased the number of primordial follicles, and increased F-actin abundance and YAP nuclear phosphorylation [154]. In another study, Cheng et al. [156] found that inducing F-actin polymerization caused YAP localization to the oocyte nucleus and follicle growth. Overall, these studies demonstrated in vitro that the Hippo signaling pathway, and its downstream effector YAP, are major regulators of follicle activation and development, and suggest that mechanical signals from the environment regulate follicle activation and development via YAP [156].

5.2.1. Oocyte-specific models

Several Cre systems have been used to analyze the role of YAP in the oocyte (Table 2). In a Yap^flox/flox;Zp3-Cre mice, which lack Yap specifically in oocytes, early embryonic development and germ cell specification were compromised [157]. Yu et al. used Yap^flox/flox;Gdf9-Cre mice to deplete Yap from primary oocytes. Although follicles developed without any major problems, the resulting embryos arrested at the two-cell stage because YAP is essential for zygotic genome activation [158]. These mouse models suggest that Yap in the oocyte is essential during germ cell specification and formation but dispensable during follicle development.

5.2.2. Granulosa-specific models

In 2019, Lv et al. [159] used a tamoxifen-inducible Cre system, Yap^flox/flox;Foxl2-Cre, to delete Yap specifically in granulosa cells. This led to disrupted follicle development, characterized by increased follicular atresia, fewer antral follicles, decreased granulosa cell proliferation, and increased granulosa cell apoptosis. To delete Yap in differentiated granulosa cells, the authors used a Cre driven by the promoter for Cyp19. CYP19 encodes aromatase, the enzyme responsible for converting androgens to estrogens. CYP19 protein is absent or low in early stages of follicle development, rises in early antral follicles, and is abundant in granulosa cells of antral follicles and in the corpus luteum. In Yap^flox/flox;Cyp19-Cre mice, the authors expected that Yap would be specifically knocked out in granulosa cells of late-stage follicles and in corpora lutea. However, they instead observed that Yap was knocked out in some, but not all, granulosa cells of preovulatory follicles, and in the majority of corpus lutea. Nevertheless, the ovary was morphologically normal [159]. Deletion of the upstream Hippo pathway kinases Lats1 and Lats2 in granulosa with Cyp19-Cre led to constitutive activation of YAP/TAZ. This activation resulted in transdifferentiation of granulosa cells into Sertoli-like and osteoblast-like cells, causing ovarian structural abnormalities and infertility. These findings underscore the importance of regulated YAP/TAZ activity in maintaining granulosa cell identity and ovarian function [160]. Overall, these mouse models suggest that Yap in granulosa cells plays a much more important role in regulating folliculogenesis than does Yap in oocytes Table 2.

5.3. RhoA/ROCK signaling pathway

Thus far, only one paper has reported the role of RhoA/ROCK in mouse follicles. The authors showed that ROCK expression increased during primordial follicle activation in vitro. Furthermore, in vivo administration of the ROCK inhibitor Y-27632 significantly decreased the number of primordial follicles and granulosa cell proliferation [161]. In chickens, RhoA expression is inhibited by FOXL2, resulting in actin reorganization [162]. In cumulus cells, the RhoA/ROCK pathway promotes ECM assembly, making the ECM more resistant to sperm penetration [163]. Future work should be directed toward developing mouse models to define the role of the RhoA/ROCK mechanotransduction pathway in regulating follicle activation, development, and oocyte quality.

5.4. Other mouse models to study mechanotransduction in follicles

An open question in the field is which mechanotransduction pathways are activated when follicles are exposed to mechanical stimuli during primordial follicle activation or follicle development. The pathways described in this section have been linked to mechanotransduction in multiple tissues. Although recent mouse models and in vitro studies show that these pathways play an important role in regulating folliculogenesis and oocyte quality, it is unknown which, if any, of these mechanotransduction pathways are responsible for transducing mechanical signals from the environment to the follicle. We also do not know whether different cell types in the ovary use different pathways to integrate these mechanical signals, though it is likely that the pathways described here act together in multiple cell types. To gain a global understanding of follicle mechanotransduction, two main questions must be addressed: 1) Which molecules sense the environment and transmit this information between cells? Integrins are an attractive option, but we have limited information about which integrins are expressed in ovarian follicles. The cell adhesion molecule cadherin is highly abundant in the follicle and appears to play a mechanosensory role in the follicle [164]. 2) Which mechanotransduction pathways are active in thecal and stromal cells? Developing systems to manipulate mechanotransduction pathways in these cell types will allow us to obtain a holistic understanding of how mechanical cues are sensed and transmitted in the ovary.

6. Concluding remarks

In this review, we highlight recent data revealing that, in addition to hormonal and biochemical signaling, mechanical regulation plays an important role in regulating folliculogenesis. New tools to analyze ovarian biomechanics, in vitro systems that replicate mechanical environments, and genetically engineered mouse models are allowing researchers to dissect how follicles sense and respond to mechanical stimuli. Together, these developments promise to deepen our understanding of follicle biology and advance the field of ovarian mechanobiology.