FSH and GDF-9 modulate extracellular matrix remodeling in human preantral follicles in a 3D hydrogel culture

Abstract

Background

Human preantral follicle development requires finely tuned signaling between oocytes and granulosa cells, supported by a structured extracellular matrix (ECM). This study investigates how follicle-stimulating hormone (FSH) and growth differentiation factor 9 (GDF-9), alone or combined, influence follicular growth and ECM organization in a 3D cell culture system.

Results

All culture conditions maintained high follicle viability on Days 1 and 7 with no significant intergroup differences; Day-7 viability ranged from ~ 69–76% across conditions. On Day 1, follicle diameters were comparable among groups. By Day 7, follicles cultured without added FSH or GDF-9 reached a significantly larger mean diameter (171.98 ± 61.05 μm) than those cultured with both supplements (130.22 ± 56.75 μm; p < 0.05); diameters in single-supplement groups were intermediate and not significantly different from the unsupplemented group. Immunofluorescence revealed that laminin expression at the follicle peripheral and across the whole follicle was significantly higher in the FSH-only and GDF-9-only groups than in the unsupplemented and combined-supplement groups. In contrast, collagen IV exhibited a distinct pattern: total collagen IV was lowest in the GDF-9-only group, whereas combined-supplemented group resulted in the higher collagen IV fluorescent intensity in both peripheral and total compared to GDF-9-only group, highlighting the importance of synchronized proliferative and organizational cues.

Conclusion

These findings underscore the critical need to evaluate follicle quality not only by growth but also by ECM integrity. They also reinforce the translational value of serum-free systems enriched with physiological regulators to support oocyte-somatic coordination and preserve follicular architecture for reproductive engineering and fertility preservation applications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13048-026-02080-3.

Background

The extracellular matrix (ECM) is a fundamental component of the ovarian follicle, providing structural integrity and regulating cell-cell communication between the oocyte and its surrounding somatic cells. Within the follicle, the basal lamina, composed primarily of laminin and collagen IV, separates the granulosa cell layer from the surrounding stroma and plays a central role in maintaining follicular architecture, granulosa cell polarity, and mechanotransduction [1, 2]. In addition, an intrafollicular ECM (iECM) is present within the granulosa cell compartment, where it facilitates cytoskeletal organization and modulates paracrine signaling [3].

In vitro culture systems for human follicles aim to replicate the complex in vivo microenvironment, but often fall short in preserving the iECM’s spatial organization, particularly in serum-rich conditions. Our recent findings demonstrated that fetal bovine serum (FBS) concentration strongly affects both granulosa cell proliferation and iECM compartmentalization, where higher FBS levels promoted growth at the expense of iECM integrity, suggesting a structural asynchrony between cellular and matrix development. These results highlight the need for defined culture conditions that support both somatic cell function and iECM preservation.

In vivo, iECM remodeling during folliculogenesis is driven by a tightly regulated interplay of hormonal and local signals. Among them, follicle-stimulating hormone (FSH) and growth differentiation factor 9 (GDF-9) are key regulators. FSH, secreted by the pituitary, binds to receptors on granulosa cells to promote their proliferation, differentiation, and ECM-related gene expression [4, 5]. GDF-9, an oocyte-derived TGF-β family member, is indispensable for early follicular development and has been shown to modulate granulosa cell function, basement membrane composition, and theca cell recruitment [6–8]. Importantly, both factors are involved in pathways, such as SMAD2/3 and PI3K/Akt, that intersect with ECM production and remodeling.

Despite their known roles in folliculogenesis, the specific influence of FSH and GDF-9 on follicle-intrinsic ECM secretion and localization remains poorly understood, especially in human 3D culture systems. This knowledge is critical for advancing ovarian tissue engineering strategies aimed at preserving not just follicle survival, but also structural fidelity and functional competence.

In this study, we investigated the effects of FSH and GDF-9, alone or in combination, on the development of human preantral follicles cultured in a PEGylated fibrin hydrogel, previously shown to mimic the biomechanical properties of the ovarian cortex [9]. We focused on their ability to modulate laminin and collagen IV deposition at the follicular boundary and within the granulosa cell compartment, as markers of basal lamina and intrafollicular ECM integrity. Our aim was to determine whether these physiological signals could provide a more targeted and controlled alternative to serum-based systems for supporting follicle development in vitro.

Methods

Experimental design

The aim of this study was to examine how GDF-9 and FSH influence the in vitro survival, growth, and ECM remodeling of isolated human preantral follicles and encapsulated in tailored PEGylated fibrin hydrogels. For this purpose, culture media were prepared with 10% FBS, either alone or supplemented with FSH and/or GDF-9. Each experimental group was performed in four independent repetitions. Follicle survival and diameter were evaluated on Days 1 and 7, and the total number of follicles analyzed for each parameter is presented in Table 1. For immunostaining analyses, two follicles were randomly selected from each group for laminin and collagen IV staining. A confocal z-stack comprising five optical sections was obtained for each follicle. Fluorescence intensity was measured in two defined regions: the peripheral region, representing the outer granulosa layers adjacent to the basal lamina, and the whole follicle, representing the entire follicular area. The total number of quantified peripheral regions and whole-follicle fluorescence measurements for both laminin and collagen IV is summarized in Table 1.

Table 1.

Experimental groups and number of follicles analyzed in each condition. The four culture conditions included: no supplementation (− F−G), FSH only (+ F−G), GDF-9 only (− F + G), and combined supplementation (+ F+G)

	-F-G	+F-G	-F + G	+F + G
Total number of human follicles for viability analysis	50	39	43	46
Total number of follicles for diameter measurements on day 1	13	12	17	13
Total number of follicles for diameter measurements on day 7	29	19	30	25
Total number of follicles for analyzing laminin staining	2	2	2	2
Total number of z-stack sections	10	10	10	10
Analyzed peripheral region’s fluorescence	40	40	40	40
Analyzed whole follicle’s fluorescence	10	10	10	10
Total number of follicles for analyzing collagen IV staining	2	2	2	2
Total Number of z-stack sections	10	10	10	10
Analyzed peripheral region’s fluorescence	40	40	40	40
Analyzed whole follicle’s fluorescence	10	10	10	10

Human ovarian cortex collection

Ovarian tissue was obtained from deceased multi-organ donors following approval from the Université Catholique de Louvain’s Institutional Review Board (IRB reference 2018/19DEC/475, approval date: May 25, 2019). Ovaries from 32-year-old organ donor were transported in PBS at 4 °C to the laboratory, where the cortex was separated from the medulla and cut into small fragments (maximum size: 5 × 5 mm²). Tissue cryopreservation was carried out following Lieman et al. [10] protocol.

Human preantral follicle isolation

Preantral follicles were isolated using Chiti et al. [11] protocol. Initially, the thawed tissue fragments were mechanically minced using a tissue chopper (McIlwain, Campden Instruments, Loughborough, UK) and then digested in Dulbecco’s phosphate-buffered saline (with Ca²⁺/Mg²⁺) (DPBS; 14040091, ThermoFisher Scientific, Dilbeek, Belgium) containing Liberase DH (5401089001, Sigma-Aldrich, Overijse, Belgium) and DNase I (11284932001, Sigma-Aldrich, Overijse, Belgium) for two 30-minute cycles until complete digestion was achieved. After each cycle, the suspension was filtered through 100 μm and 20 μm cell strainers. Fragments retained on the 100 μm filter were re-digested, while follicles captured on the 20 μm filter were washed in DPBS (without Ca²⁺/Mg²⁺) (14190144, ThermoFisher Scientific, Dilbeek, Belgium) containing 10% heat-inactivated fetal bovine serum (FBS; 16140071, ThermoFisher Scientific, Dilbeek, Belgium). Follicles were collected using a 130-µm micropipette (Flexipet, Cook, Belgium) under a stereomicroscope and maintained at 4 °C until encapsulation. To verify that the isolation process yielded viable and morphologically intact follicles, we performed a viability assay using LIVE/DEAD™ Viability/Cytotoxicity Kit on a small fraction of the isolated follicles for 30 min at 37 °C.

PEGylated fibrin hydrogel preparation

A PEGylated fibrin hydrogel was prepared as previously described [9] to mimic the mechanical properties of ovarian cortex tissue at reproductive age. Briefly, human fibrinogen (F3879, Sigma-Aldrich) was dissolved in PBS (10010-015, ThermoFisher Scientific) at a concentration of 85 mg/mL and then incubated at 37 °C for 2 h before filtration (0.20 μm). In the meantime, O,O′-Bis[2-(N-Succinimidyl-succinylamino)ethyl]polyethylene glycol (NHS-PEG-NHS; 2000 Da; 713783, Sigma-Aldrich) was dissolved in PBS (5 mg/mL), filtered, and mixed with fibrinogen solution in a 1:1 volume ratio, resulting in a 10:1 PEGylation molar ratio. Follicles were added to this mixture (final fibrinogen concentration: 39 mg/mL), and crosslinking was induced using human thrombin (T7009, Sigma-Aldrich) reconstituted in 40 mM CaCl₂ (final concentration: 50.3 IU/mL).

Follicle encapsulation and in vitro culture

Approximately 10 follicles (2 µL suspension) were mixed with 13 µl of PEGylated fibrinogen and crosslinked with 15 µl of thrombin. The droplets were polymerized by inverting the culture plates and incubating at 37 °C for 15 min. Following completion of the gelling incubation, 600 µL of culture medium was gently added to fully cover the hydrogel constructs. The culture medium consisted of MEM⁺ supplemented with 10% FBS,1% antibiotic-antimycotic (5240096, ThermoFisher Scientific), 1% insulin-transferrin-selenium (41400-045, ThermoFisher Scientific), and 2 mM sodium pyruvate (s8636, Sigma-Aldrich). Four conditions were tested: no supplementation (-F-G), FSH only (+ F-G), GDF-9 only (-F + G), or both (+ F+G). FSH (10 mIU/mL; 2242 − 659, Merck) and human recombinant GDF-9 (200 ng/mL; SRP4872, Sigma-Aldrich) were added as appropriate.

Follicles were cultured for 7 days at 37 °C in a humidified incubator containing 5% CO₂, with half of the medium replaced every other day. Four replications were considered for each condition.

Analysis of follicle survival and growth

Follicle viability and growth were evaluated on days 1 and 7 using an inverted microscope, following Chiti et al. [12]. Follicles were considered non-viable if they displayed extruded oocytes, dark granulosa cells or oocytes, incomplete granulosa-oocyte association, or marked shrinkage. The diameter of viable follicles was measured in two perpendicular axes using Image J software (https://imagej.nih.gov/ij/), and the mean value was recorded. Measurements were taken on days 1 and 7 of in vitro culture.

Immunofluorescence analysis of follicular ECM

Hydrogels were washed in PBS and fixed in 4% paraformaldehyde (J61899,ThermoFisher Scientific) at room temperature for 1 h [9]. Non-specific background was blocked overnight at 4 °C under gentle agitation using PBS containing 0.05% Triton X-100 (T8787, Sigma-Aldrich), 5% normal goat serum (31872, ThermoFisher Scientific), and 0.2% sodium azide (s2002, Sigma-Aldrich). The hydrogels were then incubated for 48 h at 4 °C with primary antibodies diluted in blocking solution: rabbit polyclonal anti-collagen IV (ab6586, Abcam, Cambridge, UK) or rabbit polyclonal anti-laminin 1 + 2 antibody (ab7463, Abcam), both at 1:200. After PBS-Triton X-100 washes, samples were incubated with goat anti-rabbit secondary antibodies conjugated to Alexa Fluor™ 488 (a11034, ThermoFisher Scientific) or 647 (A-21245, ThermoFisher Scientific), both at 1:250, for another 48 h at 4 °C under gentle shaking. Nuclei were counterstained with Hoechst 33,342 (1:100, H3570, ThermoFisher Scientific) solution, and samples were cleared with Rapiclear for 1 h at 37 °C before imaging. Confocal images were acquired using a ZEISS LSM800 microscope with a 20× objective and identical settings across all samples. Z-stacks included 5 optical slices at 5 μm intervals, following Tomaszewski et al. [13] procedure. All samples were exposed to the same image acquisition settings, including scan speed, resolution, laser power, pinhole, and master gain. Background reduction was uniformly applied.

Mean fluorescent intensities were quantified using Zen 3.5 software (blue edition). For each slice of a follicle, four rectangular regions (10 × 4 μm²) at the follicle edge were used to calculate peripheral signal intensity (Fig. 1a). The whole-follicle signal was measured by outlining its contour (Fig. 1b). Negative controls were processed in parallel without primary antibodies. The average fluorescent intensities of the follicular basal lamina, the whole follicle, and the ratio of peripheral ECM to the whole follicle ECM were calculated for all conditions.

Fig. 1.

Expression analyses in supplemented media. a Peripheral region (10 × 4 µm²) used for measuring edge fluorescence. b Whole follicle intensity measured via freeform contour selection

Statistical analysis

Data were analyzed using one-way ANOVA, with statistical significance offset at p < 0.05. Results are expressed as mean ± standard deviation (SD), representing central tendency and variability, respectively. Error bars in graphs represent ± 1 SD from the group mean.

Results

Similar follicle survival and comparable growth across supplementation groups

The viability assessment on a small subset of freshly isolated follicles indicated viable follicles on day zero before encapsulation in the PEGylated fibrin hydrogel. The representative image from this assay, showing the overall health status of the isolated follicles, as provided in Fig. 2.

Fig. 2.

Representative live/dead staining of isolated follicles. Viable cells are indicated by green fluorescence, and non-viable cells are indicated by red fluorescence. Scale bar: 50 μm

All culture conditions supported the survival of isolated human preantral follicles throughout the 7-day culture period, with no significant differences in viability observed between groups on either day 1 or day 7 (Fig. 3a-c). On day 7, follicle viability remained high across all groups, ranging from 69% to 76%, and was not significantly influenced by the addition of FSH or GDF-9.

Fig. 3.

Follicle survival after in vitro culture under different supplementation conditions. a Representative images on day 7. Scale bar: 200 μm. b Viability during culture (day 1 to day 7). c Viability comparison among conditions on day 7. Data are presented as mean ± SD. A total of 178 follicles were used for survival analysis. Asterisks indicate significant differences (p < 0.05)

Follicle diameter measurements on day 1 showed no significant differences between groups, indicating comparable initial sizes (Fig. 4). The average diameters of follicles on Day 1 were 51.53 ± 13.33 μm, 47.92 ± 7.36 μm, 41.89 ± 9.7 μm, and 42.54 ± 11.56 μm in -F-G, +F-G, -F + G, and + F + G groups, respectively. By day 7, follicles in -F-G group reached the largest average diameter (171.98 ± 61.05 μm), which was significantly larger than those cultured with both FSH and GDF-9 (130.22 ± 56.75 μm, p < 0.05) (Fig. 4). However, no significant differences were observed between -F-G group and the + F-G (152.26 ± 62.66 μm) or -F + G (146.32 ± 41.08 μm) groups (Fig. 4a, b).

Fig. 4.

Follicle growth during culture. a Diameter distributions on day 1. b Diameter distribution on day 7. Lines indicate medians. A total of 158 follicles were measured: 17, 13, 12, and 13 follicles on Day 1 and 30, 29, 19, 25 follicles on Day 7 for -F + G, -F-G, +F-G, and + F + G groups, respectively. Asterisks denote significant differences (p < 0.05)

FSH and GDF-9 differentially modulate ECM remodeling, with FSH and GDF-9 alone enhancing laminin and the combined treatment increasing collagen IV

Immunostaining for ECM components and subsequent confocal microscopy analysis were performed on the preantral follicles cultured in various media (Figs. 5a and 6a). The mean laminin fluorescence on the follicle surface was significantly higher in the FBS + F (38.58 ± 20.6) and FBS-F + G (37.86 ± 17.05) groups compared to FBS-F-G (23.69 ± 14.85) and FBS + F+G (23.63 ± 12.86) groups (p < 0.01) (Fig. 5b).

Fig. 5.

Laminin expression after 7 days in culture. a Immunostaining images: Laminin (green), nuclei (blue). Scale bar: 50 μm. b Peripheral laminin fluorescent (n = 40). c Whole follicle laminin fluorescence (n = 10). d Peripheral-to-total laminin fluorescence ratio. Lines in graphs represent the medians of values. Asterisks indicate significant differences (**p < 0.01; ***p < 0.001; ****p < 0.0001)

Fig. 6.

Collagen IV expression after 7 days in culture. a Representative immunostaining images: collagen IV (green), nuclei (blue). Scale bar: 50 μm. b Peripheral collagen IV fluorescence intensity (n = 40). c Total collagen IV fluorescence across the follicle (n = 10). d Ratio of peripheral to total collagen IV. Lines show medians. Asterisks indicate significant differences (*p < 0.05; **p < 0.01; ***p < 0.001)

A similar significant difference was observed in whole-follicle laminin deposition, with + F-G (110.37 ± 25.15) showing greater expression than -F-G (63.32 ± 12.48, p < 0.0001) and + F + G (65.91 ± 14.77, p < 0.001) (Fig. 5c).

Laminin expression on the follicle surface and throughout the whole follicle was similar in the + F-G and FBS-F + G groups. The proportion of surface laminin expression relative to total laminin expression across the entire follicle did not differ significantly between the groups, with averages around 43%, 35%, 38%, and 36% for -F-G, +F-G, -F + G, and + F + G, respectively. Notably, -F-G demonstrated a slightly higher, yet non-significant, proportion of laminin expression on the surface of follicles compared to the other groups (Fig. 5d).

In contrast to laminin, collagen IV expression was significantly reduced in follicles treated with GDF-9 alone (-F + G), both at the periphery and across the entire follicle (Fig. 6). The total collagen IV signal in this group was significantly lower than in all other conditions, including the control (Fig. 6). Interestingly, despite this decrease, the proportion of collagen IV localized at the follicle edge was relatively higher in the -F + G group (34%) compared to + F-G (27%; p < 0.05), suggesting altered ECM organization (Fig. 6d).

Discussion

Within in vitro follicle culture systems, FSH is widely incorporated due to its well-established role in granulosa cell proliferation, estrogen production, and follicle growth [14–16]. GDF-9, though less commonly included, is a key oocyte-derived factor critical for early folliculogenesis [6, 17–19]. Several studies have demonstrated that GDF-9 not only promotes follicular progression [20–22] and protects against apoptosis [23], but also enhances granulosa cell responsiveness to FSH by upregulating FSHR expression [23, 24]. Despite the frequent application of these factors to support growth and survival, few studies have investigated their effects on ECM remodeling.

This study explored the effects of FSH and GDF-9 supplementation, alone or in combination, on the growth and ECM organization of human preantral follicles. Our findings demonstrate that while FSH stimulated follicle expansion, ECM integrity at the follicular boundary was not preserved unless GDF-9 was also present. Notably, GDF-9 alone enhanced laminin and collagen IV localization without promoting follicular growth, indicating a role in ECM remodeling independent of somatic proliferation. These results suggest that the coordination between growth and ECM deposition depends on the precise molecular cues provided in the microenvironment.

It is important to highlight that in this study, the term ECM encompasses both the follicular basal lamina, a specialized matrix encasing the granulosa cells, and the iECM found between granulosa cells, as both compartments were assessed through laminin and collagen IV localization.

GDF-9 is a well-established oocyte-derived factor essential for early folliculogenesis, acting on granulosa cells to regulate ECM-related gene expression (including HAS2, PTGS2) via SMAD2/3 signaling, and to promote oocyte–somatic coordination through PI3K/AKT pathways [25, 26]. In our study, GDF-9 supplementation preserved the spatial organization of laminin, at the follicular basal lamina and within the intrafollicular compartment. This suggests that GDF-9 may promote laminin synthesis or retention through paracrine signaling between the oocyte and granulosa cells, supporting a balanced microenvironment necessary for follicular homeostasis. Moreover, GDF-9 supplemented medium yielded lower expression of collagen IV, suggesting that GDF-9 altered the follicle microenvironment in distinct ways, promoting a laminin-rich, less collagen-dense matrix. Laminin and collagen IV may play opposing roles in follicle development. As a study on the culture of mouse ovaries in the presence of activin A showed that laminin substrates tend to enhance follicle expansion and multilayer formation, whereas type IV collagen can impose suppressed multilayer growth [27]. Therefore, the increased laminin and reduced collagen IV observed in our study may represent an adaptive ECM remodeling in response to GDF-9 stimuli, potentially supporting follicular development at later stages. This statement can align with the expression disappearance of different chains of collagen IV by follicle development [3, 28, 29].

FSH alone led to enhanced laminin expression at both the follicular periphery and throughout the entire follicle and moderate collagen IV level, relative to the unsupplemented group, without significantly impacting follicle viability or diameter. FSH promotes granulosa cell proliferation and steroidogenesis through activation of the cAMP-PKA pathway, and downstream PI3K/AKT and MAPK/ERK signaling [30, 31]. The elevated laminin levels observed in the group supplemented with FSH-alone suggest that FSH may directly or indirectly promote ECM assembly. Previous studies have been shown that FSH can promote ECM deposition in cumulus cells-oocyte complexes (COC) during antral follicle development [32–34], potentially through enhanced glucose uptake in cumulus cells, as proposed by Sutton-McDowall et al. [35]. In our study, the FSH-driven ECM enhancement was not accompanied by significant changes in follicle growth, which may reflect the FSH-independency of early preantral follicles’ growth. Nevertheless, the marked increase in laminin under FSH treatment alone suggests a potential priming effect for later stages of follicular development, when FSH responsiveness becomes more pronounced.

The combined treatment with FSH and GDF-9 yielded intermediate results: follicles displayed moderate growth while partially retaining ECM structure. The fact that the combination did not result in additive effects on follicular diameter implies that ECM integrity and volumetric growth can be decoupled, and that structural maturation may be a more sensitive and relevant marker of developmental potential. These findings align with earlier reports that emphasized the necessity of bidirectional communication between oocytes and granulosa cells to orchestrate follicle growth and differentiation in a temporally coordinated manner [36, 37]. Moreover, this outcome aligns with reports suggesting that GDF-9 may modulate FSH-mediated signaling. Although GDF-9 is known to support preantral follicle development in several species [38, 39], its interaction with FSH can be more complex. GDF-9 has been shown to suppress FSH-induced granulosa cell differentiation and cAMP accumulation [40], and to influence FSH receptor expression in developing follicles [23], indicating potential cross-regulatory effects when both factors are given simultaneously. Supporting this, Alves et al. [41] demonstrated that caprine preantral follicles benefit from sequential rather than concurrent exposure to GDF-9 and FSH. In our culture system, the combined treatment was also associated with the highest collagen IV deposition and reduced laminin compared with the single-factor groups. Given that collagen-rich environments can restrict granulosa multilayer formation, while laminin promotes expansion, this ECM shift may partially explain the moderate growth observed in the combined group.

Importantly, the preservation of ECM organization, mainly at the follicular basal lamina, is critical for maintaining cell polarity, mechanical stability, and bidirectional signaling between granulosa cells and the oocyte [1, 42, 43]. The basal lamina acts as a scaffold that not only supports cell adhesion but also mediates growth factor localization and diffusion, all of which are essential for follicle development. Disruption of this architecture could compromise oocyte competence, even when follicular expansion appears adequate. Thus, structural ECM markers such as laminin and collagen IV localization may provide a more nuanced and predictive readout of follicle quality than size alone.

From a tissue engineering perspective, these findings hold significant translational value. They support the development of defined, serum-free culture systems that incorporate physiologically relevant signaling factors to preserve ECM integrity. Unlike serum-containing media, which often introduce undefined mitogens and proteases that may disrupt ECM assembly, serum-free systems offer greater control over the biochemical environment and may better preserve follicular architecture. This is especially important for fertility preservation strategies where long-term follicle viability and oocyte competence are critical endpoints. Additionally, these results advocate for the integration of ECM readouts, both molecular and structural, as part of the quality assessment criteria in in vitro follicle culture systems. Such metrics could help refine culture conditions to better emulate the native ovarian niche, ultimately improving the success of bioengineered ovarian constructs.

The concentrations of FSH (10 mIU/mL) and GDF-9 (200 ng/mL) were selected based on prior studies demonstrating their efficacy in supporting preantral follicle growth and functional activity in vitro. FSH at 10 mIU/mL has been widely used to promote granulosa cell proliferation and oocyte maturation [44–47], and GDF-9 at 200 ng/mL has shown positive effects on early follicle development and oocyte–somatic coordination [21, 48, 49]. Both factors may exhibit dose-dependent influences on ECM production and follicular architecture, suggesting that future studies incorporating various concentrations would be valuable for refining the optimal conditions for ECM remodeling. Moreover, Future directions should include quantification of ECM-related gene and protein expression in response to FSH and GDF-9, assessment of MMP activity, and time-course studies to elucidate the temporal dynamics of ECM remodeling. The use of proliferation markers such as Ki-67, in parallel with ECM visualization, could also clarify the interplay between somatic cell expansion and matrix deposition. Finally, extending this work to multi-factorial culture systems, including other oocyte-derived growth factors like BMP15, may help define the optimal biochemical environment for supporting structurally and functionally competent folliculogenesis in vitro.

On the other hand, the basal lamina surrounding early human follicles is composed of collagen IV and laminin in vivo, which have been demonstrated by immunofluorescence in primary and secondary follicles in the human ovarian cortex [50]. During folliculogenesis, the composition of this ECM dynamically remodels in animal models where the relative amount of collagen IV isoforms changes and laminin chains increase [51]. For instance, in mouse follicles, immunohistochemical analyses reveal that specific types of laminin (α1, β1, γ1) and collagen IV (α1, α2) persist through preantral and antral stages, while specialized ECM structures such as focimatrix arise between granulosa cells [52]. In our PEGylated-fibrin 3D culture system, we observe deposition of both laminin and collagen IV around and within the follicle structure, indicating that granulosa cells retain the capacity to produce a basal-lamina like ECM. However, differences in matrix density and spatial distribution persist, likely due to the absence of stromal cell-derived cues and native tissue stiffness. Thus, while our model recapitulates key ECM components, future optimization, including co-culture with theca or stromal cells or ECM templating, may further enhance recapitulation of the in vivo tissue architecture [53, 54].

Conclusions

In summary, our findings underscore the importance of balancing mitogenic stimulation with ECM preservation to support structurally sound folliculogenesis in vitro. The differential effects of FSH and GDF-9 on follicle expansion and ECM integrity highlight the need for defined culture systems that mimic the tightly regulated oocyte-somatic signaling of the native ovary. These insights will inform the refinement of bioengineered ovarian models and enhance strategies for fertility preservation.

Abbreviations

ECM

Extracellular matrix

FSH

Follicle-stimulating hormone

GDF-9

Growth differentiation factor 9

iECM

Intrafollicular ECM

DPBS

Dulbecco’s phosphate-buffered saline

FBS

Fetal bovine serum

NHS-PEG-NHS

O, O′-Bis[2-(N-Succinimidyl-succinylamino)ethyl]polyethylene glycol

-F-G

No supplementation with FSH and GDF-9

+F-G

Supplementation with FSH only

-F + G

Supplementation with GDF-9 only

+ F + G

Supplementation with both FSH and GDF-9

SD

Standard deviation

Authors’ contributions

**Arezoo Dadashzadeh** : Conceptualization, Methodology, Validation, Investigation, Data curation, Writing – original draft, Visualization. **Saeid Moghassemi** : Conceptualization, Formal analysis, Investigation, Writing – original draft, Visualization. **Christiani A. Amorim** : Conceptualization, methodology, resources, writing—review and editing, supervision, project administration, and funding acquisition.

Funding

This study was supported by grants from the Fonds de la Recherche Scientifique (grant FNRS-PDR T.0004.20 awarded to C.A.A. and FC 46825 awarded to S.M.) and the Belgian Foundation against Cancer (grant OST-2022/2169 awarded to C.A.A.).

Data availability

The data supporting the findings of this study are available from the corresponding author, Christiani A. Amorim, upon reasonable request.

Declarations

Ethics approval and consent to participate

Ovarian tissue was obtained from deceased multi-organ donors following approval from the Université Catholique de Louvain’s Institutional Review Board (IRB reference 2018/19DEC/475, approval date: May 25, 2019).

Competing interests

The authors declare no competing interests.