A Novel Three-Dimensional Follicle Culture System Decreases Oxidative Stress and Promotes the Prolonged Culture of Human Granulosa Cells.

Secondary follicles were cultured in 2D and 3D systems to observe the follicle development and oocyte competence. Partial follicular morphological changes were presented in Figure 1 A. A portion of the 2D cultured follicular GCs adhered to the culture plate ( Figure 1 A, a–c). Follicles remained intact in hydrogel and gradually increased in size ( Figure 1 A, d–i). Follicular growth was identified by follicle diameters and antrum formation rates. There were no significant differences in follicle diameters among the three groups on day 2 based on the same size follicles collected initially ( p > 0.05, Figure 1 B). On day 6, the diameters of follicles in the 2D and the RGD and IYR hydrogel encapsulation groups had grown to 191 ± 11 μm, 214 ± 22 μm, and 207 ± 22 μm, respectively. The follicles in the RGD hydrogel encapsulation group were significantly larger than those in the 2D group ( p 0.05, Figure 1 B). On day 10, the follicles in the RGD hydrogel encapsulation group showed the largest diameters (338 ± 15 μm, p < 0.05), significantly larger than the 2D (305 ± 16 μm) and the IYR hydrogel encapsulation group (318 ± 12 μm, p < 0.001, Figure 1 B). Ovarian follicle growth and viability. (A) The typical images of the follicles in different culture systems. (B) Follicle diameters (μm) of the ovarian follicles in different culture systems at days 2, 6, and 10. (C) Fluorescence staining for viability (green: live cells, red: dead cells, n = 10, original magnification: 200×). (D) Percentage of living cells (%) of ovarian follicles in different culture systems at day 10. Data are represented as the mean ± SD * p < 0.05, ** p < 0.01, and *** p < 0.001. In addition, we examined the viability of ovarian follicles by live/dead staining in each group on day 10. There were no significant differences in the percentage of live cells among the three groups ( n = 5, p > 0.05, Figure 1 C and 1 D). As shown in Table S1 , the antrum formation rates were observed on day 10. The antrum formation rates in the RGD (78.0%) and IYR (75.1%) hydrogel encapsulation groups were higher than in the 2D group (64.3%, p < 0.05). We then assessed the survival rate of the follicles in each group on day 12. The survival rates in the RGD (66.5%) and IYR(68.4%) hydrogel encapsulation groups were higher than in the 2D group (48.8%, p < 0.05). These results indicated that 3D culture significantly improved follicle growth and survival. The RGD hydrogel encapsulation group was able to obtain the largest follicles, and there was no significant difference in follicle development and survival between the RGD and IYR hydrogel encapsulation groups. A total of 193 cumulus–oocyte complexes (COCs) including 56 in the 2D group, 72 in the RGD hydrogel encapsulation group, and 65 in the IYR hydrogel encapsulation group were obtained and cultured in the subsequent IVM experiments. As shown in Table S2 and Figure 2 , the germinal vesicle (GV) rates in the RGD (16.7%) and IYR (4.6%) hydrogel encapsulation groups were significantly lower than in the 2D group (33.9%, p < 0.05). The maturation rates (MII, metaphase II) in the RGD (52.8%) and IYR (72.3%) hydrogel encapsulation groups were significantly higher than in the 2D group (30.4%, p < 0.05). In the 3D culture systems, the IYR (72.3%) hydrogel encapsulation group accessed a higher maturation rate than the RGD encapsulation group. Images on oocyte maturation of the follicles in different culture systems (original magnification: 200×). In the 2D culture, hGCs were attached to the cell plate after seeding for 2–4 h. HGCs were polygonal or spindle-shaped and interconnected with pseudopods, with large and round nucleoli, distinct nucleoli, and uniformly particle-rich cytoplasms ( Figure S1 ). Immunofluorescence confirmed that the hGCs protein markers, AMH and FSHR, were expressed in the cytoplasm of the collected hGCs ( Figure S2 ). The viability of hGCs in the 2D and 3D cultures was assessed at different time points ( n = 3). HGCs in the 2D culture were stretched well and grew vigorously on day 2, and there was no significant difference in the percentage of live hGCs in the 2D and 3D cultures (98.6 ± 0.3% vs 95.7 ± 1.2%, p > 0.05, Figure 3 A). On day 4, hGCs began to degenerate with increased apoptosis and partial loss of protruding pseudopods, and the percentage of live hGCs in the 2D culture was significantly lower than that in the 3D culture (60.7 ± 5.2% vs 91.1 ± 1.2%, p < 0.01, Figure 3 A). On day 6, most of the hGCs in the 2D culture were apoptotic, and the percentage of live cells decreased significantly. However, hGCs in the 3D culture were able to maintain their viability (43.0 ± 8.3% vs 85.8 ± 2.0%, p < 0.01, Figure 3 A). The morphology of the hGCs in the 3D culture did not change and remained spherical in shape on days 2, 4, and 6 (Videos 1 – 3 ). 3D culture improved viability, regulated the cell cycle, and inhibited apoptosis of hGCs. (A) Fluorescence staining for viability (green: live cells, red: dead cells, n = 3, magnification: 100×) and percentage of living cells (%) of hGCs in 2D and 3D cultures on days 2, 4, and 6. (B) Flow cytometry analysis of the cell cycle ( n = 5) of hGCs in 2D and 3D cultures on day 2. (C) Flow cytometry detection of cell apoptosis by Annexin V/PI staining. HGCs ( n = 5) cultured in 2D and 3D systems for 2 and 4 days were tested. (D) Relative expression of apoptosis-associated genes ( caspase3 , caspase8 , BCL2 , BAX , n = 8). Data are represented as the mean ± SD * p < 0.05, ** p < 0.01, and *** p < 0.001. To observe the effect of 3D culture on hGCs proliferation, we detected the cell cycle of hGCs by flow cytometry ( n = 5). After culture in vitro for 2 days, the percentage of hGCs in the S ( p < 0.001) and G2/M ( p < 0.05) phases was significantly increased in the 3D culture compared with the 2D culture ( Figure 3 B). Furthermore, we examined the apoptosis of 2D- and 3D-cultured hGCs on days 2 and 4 by flow cytometry ( n = 5) and RT–PCR ( n = 8). Flow cytometry showed that the percentages of early (Annexin V+PI-) apoptotic cells were significantly decreased in the 3D culture on day 2 ( p < 0.001) and day 4 ( p 0.05, Figure 3 C). RT–PCR showed that the expression levels of the apoptosis-related genes caspase3, caspase8 , and BAX were significantly decreased in the 3D-cultured hGCs on day 2 and day 4 ( p 0.05) ( Figure 3 D). Regarding hormone production, AMH and progesterone are produced by hGCs in the ovary. We thus measured the concentrations of AMH and progesterone in the cell supernatant of hGCs collected from the 2D and 3D culture systems ( n = 3, Figure 4 ). There were no significant differences in the AMH concentrations between the two groups on days 1, 2, and 6 ( p > 0.05, Figure 4 A). However, on day 4, the average concentration of AMH secreted by hGCs in the 3D culture (85.5 ± 5.7 pg/mL) was significantly higher than that in the 2D culture (43.5 ± 3.3 pg/mL, p < 0.05, Figure 4 A). Additionally, the average concentrations of progesterone secreted by the 3D-cultured hGCs were lower than the 2D-cultured hGCs on day 1 and day 2 (D1: 3.3 ± 0.6 vs 6.4 ± 1.4, p < 0.05; D2: 10.8 ± 1.3 vs 14.5 ± 1.7, p < 0.05, Figure 4 B). On day 4 and day 6, the 3D-cultured hGCs secreted much more progesterone than the 2D-cultured hGCs (D4: 15.4 ± 1.10 vs 8.3 ± 1.2 ng/mL, p < 0.05; D6: 12.8 ± 0.4 vs 1.2 ± 0.2 ng/mL, p < 0.001, Figure 4 B). ELISA of AMH and progesterone in the cell supernatant. (A) AMH (pg/mL, n = 3) and (B) progesterone (ng/mL, n = 3). Data are represented as the mean ± SD * p < 0.05, ** p < 0.01, and *** p < 0.001. Given that the 3D culture system was able to maintain hGCs viability, regulate cell cycle, and inhibit apoptosis. We further investigated whether the 3D culture affected the senescence of hGCs ( n = 3). The results obtained from the senescence-associated β-galactosidase (SA β-gal) assay showed that the SA β-gal activity and the senescent cell ratio were significantly reduced in a 3D culture on day 2 ( p < 0.05, Figure 5 A). Delayed senescence in the 3D-cultured hGCs was also demonstrated at the protein level ( n = 3), as senescence-related proteins P16 and P21 were significantly increased compared with the 2D-cultured hGCs ( p < 0.05, Figure 5 B). Delayed senescence, decreased oxidative stress, elevated mitochondrial membrane potential, and decreased abnormal mitochondria ratio in 3D-cultured hGCs. (A) SA β-gal assay of hGCs in 2D and 3D cultures on day 2 ( n = 3, magnification: 100×). (B) Western blot of senescence-related proteins (P16 and P21) in 2D- and 3D-cultured hGCs. (C) DCFH-DA fluorescence staining for intracellular ROS levels of hGCs in 2D and 3D cultures on day 2 (fluorescence intensity represents the level of ROS, magnification: 200×). (D) Western blot of antioxidant proteins (GPX4 and SOD1) in 2D- and 3D-cultured hGCs. (E) Relative expression of antioxidant genes ( GPX4 and SOD1 , n = 10). (F) JC-1-based immunofluorescence analysis of hGCs in 2D and 3D cultures (red: JC-1 aggregate signal, green: JC-1 monomer signal, n = 5, magnification: 200×). (G) JC-1-based flow cytometry analysis of hGCs in 2D and 3D cultures ( n = 5). (H) Ultrastructure of hGCs in 2D and 3D cultures by a transmission electron microscope (magnification: 1000×; M, normal mitochondria; AM, abnormal mitochondria). Data are represented as the mean ± SD * p < 0.05, ** p < 0.01, and *** p < 0.001. Excessive reactive oxygen species (ROS) production can trigger oxidative stress and impair mitochondrial function, which are the most common inducers for cell senescence. 33 , 34 In this regard, the mean fluorescence intensity of ROS was significantly lower in the 3D-cultured hGCs ( n = 5, p < 0.05, Figure 5 C). At the protein level, the expression of the antioxidant proteins GPX4 and SOD1 was significantly decreased in the 3D-cultured hGCs ( n = 3, p < 0.05, Figure 5 D). RT–PCR results showed that the expression of the antioxidant genes GPX4 and SOD1 was remarkably lower in the 3D-cultured hGCs ( n = 10, p < 0.05, Figure 5 E). These results indicated that oxidative stress in hGCs significantly decreased after 3D culture. Excessive oxidative stress has been shown to damage mitochondrial function and reduce mitochondrial membrane potential (MMP), leading to an energy crisis. 35 We thus examined the MMP of hGCs using JC-1 staining and flow cytometry ( n = 5). Immunofluorescence showed a weaker green JC-1 monomer signal in the 3D-cultured hGCs, while the fluorescence intensity of the red JC-1 aggregate was stronger than that in the 2D-cultured hGCs. The relative fluorescence ratio of hGCs was significantly higher in the 3D culture than in the 2D culture ( p < 0.001, Figure 5 F). Moreover, flow cytometry showed that the 3D-cultured hGCs had weaker JC-1 monomer signals and stronger JC-1 aggregate signals, and the relative MMP ratios were dramatically higher in the 3D-cultured hGCs than in the 2D-cultured hGCs ( p < 0.05, Figure 5 G). Transmission electron microscopy demonstrated that the percentage of abnormal mitochondria in the 2D-cultured hGCs reached 74.5 ± 21.3%, and the 3D culture significantly reduced the percentage of abnormal mitochondria to 28.54 ± 11.4% ( n = 5, p < 0.01, Figure 5 H). These results revealed that 3D culture reduced oxidative stress, improved mitochondrial function, and decreased the percentage of abnormal mitochondria, which in turn delayed senescence in hGCs. Our findings suggested that the 3D culture promoted follicular development and survival. Cumulus expansion is a critical event that is required for ovulation. 35 Our results showed that the expression of cumulus expansion-related genes ( n = 10) and proteins ( n = 3) COX2, HAS2, and PTX3 were significantly higher in the 3D-cultured hGCs than in the 2D-cultured hGCs ( p < 0.05, Figures 6 A and 6 B). Studies have indicated that integrin α6β1 is the target receptor for IKVAV and YIGSR. 28 We observed that the expression of integrin α6β1 was significantly increased in 3D-cultured hGCs ( Figure 6 C). Accordingly, the 3D culture increased the expression levels of COX2, HAS2, and PTX3 and upregulated integrin α6β1 expression. 3D culture promoted the expression of the cumulus expansion-related genes and proteins and upregulated integrin α6β1 expression in hGCs. (A) Relative expression of cumulus expansion-related genes ( COX2 , HAS2 , and PTX3 , n = 10). (B) Western blot of cumulus expansion-related proteins (COX2, HAS2, and PTX3) in 2D- and 3D-cultured hGCs. (C) Western blot of the protein levels of integrin α6 and integrin β1 in 2D- and 3D-cultured hGCs. Data are represented as the mean ± SD * p < 0.05, ** p < 0.01, and *** p < 0.001.

This study was approved by the ethics committee of Shengjing Hospital, China Medical University (2021PS017F), and was performed following the principles of the Declaration of Helsinki. Female ICR mice were raised in accordance with the regulations and guidelines of China Medical University institutional animal care, and the experiments were approved by the Committee of Experimental Animal Ethics, China Medical University (2021PS586K). Intact two-layered secondary follicles (follicle diameter: 100–130 μm) were mechanically isolated using insulin gauge needles in Leibovitz’s L-15 Medium (Invitrogen, Gibco, MA, USA) with 5% FBS from day 11–13 ICR female mice. The hydrogels (VitroGelRGD and LDP1, Well, NJ, USA) used for 3D culture in this study are tunable, xeno-free (animal origin-free), permeable, viscoelastic, biocompatible, and ion-cross-linking systems. VitroGelRGD was modified with the integrin-binding peptide RGD (arginine-glycine-aspartic). VitroGel LDP1 was modified with RGD and the laminin-derived peptides IKVAV (isoleucine-lysine-valine-alanine-valine) and YIGSR (tyrosine-isoleucine-glycine-serine-arginine) at a 1:1:1 ratio. IKVAV is a bioactive sequence located at the C-terminus of the long arm of the laminin α1 chain, and YIGSR is located on the laminin β1 chain. The follicles were randomly divided into the following three groups: the 2D culture group and the RGD and IYR (IKVAV+YIGSR+RGD) modified hydrogel encapsulation groups. The 2D cultured follicles were placed in a 96-well plate and 100 μL of in vitro follicle culture media composed of glutaMAX α-minimal essential medium (a-MEM; Gibco, Waltham, MA, USA), 10% FBS, 1% ITS (5 mg/mL insulin, 5 mg/mL transferrin, and 5 ng/mL selenium; Sigma-Aldrich), 50 ng/mL activin A (Sigma-Aldrich, USA), 1% penicillin and streptomycin (PS; Gibco), and 10 mIU/mL follicle-stimulating hormone (FSH; Merk-serono, Darmstadt, Germany). 31 In 3D culture, hydrogel beads (50 μL) were diluted at ratios of 1:5. Ten follicles were pipetted into each bead in a minimal amount of media and then mixed with 12.5 μL of α-MEM medium with 50% FBS. The gel was placed in the incubator for 30 min for solidification and then rinsed in 100 μL of culture media. Follicles were isolated, encapsulated, and cultured at a constant temperature of 37 °C and pH 7. Follicles were cultured at 37 °C in 5% CO 2 for 12 days. Every other day, half of the media was exchanged. Follicles were considered dead if there was no GCs layer around the oocytes or GCs had become dark and fragmented. Follicle images were assessed using an inverted microscope (Multizone ART workstation, ESCO) to evaluate the follicle diameters, antrum formation, and follicle survival rates. Two diameters of follicles were measured from the outer layer of theca cells using ImageJ 1.33U (NIH, USA). On the 12th day of culture, the medium was exchanged with maturation media (a-MEM supplemented with 10 mIU/mL FSH, 1% ITS, 50 ng/mL activin A, 1.5 IU/mL human chorionic gonadotropin (hCG; Merck, Germany), and 5 ng/mL epidermal growth factor (Sigma-Aldrich)). The oocyte maturation (MII) rate was evaluated 14–16 h later by checking for the presence of the first polar body under an inverted microscope. Follicular fluid (FF) was collected from 30 women aged 20–35 years with tubal infertility or male factor infertility under their first in vitro fertilization (IVF)/intracytoplasmic sperm injection cycle at the reproductive center of Shengjing Hospital in Shenyang, China. Informed consent for participation was not required for this study as FF is a clinically discarded specimen. The demographics, clinical characteristics, and outcomes of the patients were listed in Table S1 . The controlled ovarian stimulation protocol for patients was a long GnRH agonist or a short GnRH antagonist protocol. At 36 to 38 h after the hCG trigger, transvaginal ultrasound-guided oocyte aspiration was performed and the FF samples were collected. The protocol for isolating human granulosa cells (hGCs) was based on a previous study. 32 Briefly, the FF was immediately centrifuged at 400 g for 10 min. The cell pellet was resuspended in DMEM/F12 medium containing 10% FBS. The hGCs were purified by density gradient centrifugation with Ficoll-Paque Plus (GE Health Life Sciences) at 600 g for 20 min. Purified hGCs were plated in conventional 2D or 3D culture vessels. For 2D culture, hGCs were seeded at 10 6 cells per well in a six-well culture plate and cultured in DMEM/F12 medium containing 10% FBS and 1% PS. For 3D culture, IYR hydrogel and dilution solution were mixed at a ratio of 1:5 at room temperature. The diluted IYR hydrogel was then combined with hGCs and DMEM/F12 medium containing 50% FBS at 1 × 10 6 cells/mL to initiate the cross-bonding process. The IYR hydrogel mixture was transferred to a 24-well plate and maintained for 20 min at 37 °C for soft gel formation. Then, the culture medium was added to cover the hydrogel carefully. The medium was changed every other day. HGCs were cultured on coverslips (14 mm, NEST) at 5 × 10 4 cells/mL in 6-well plates for 2 days. Then, the coverslips were washed twice with PBS and fixed in 4% paraformaldehyde (PFA). After protein blocking (C0265, Beyotime, China) for 30 min at 37 °C, the coverslips were incubated with an antibody against follicle-stimulating hormone receptor (FSHR, 1:100, 22665-1-AP, Proteintech, China) and an antibody against anti-Müllerian hormone (AMH, 1:100, 23479-1-AP, Proteintech) diluted in PBS overnight at 4 °C. The next day, all the coverslips were washed and incubated with secondary antibodies (1:500, Cy3-labeled goat antirabbit IgG, A0516, Beyotime) at room temperature for 2 h. 4′,6′-Diamidino-2-phenylindole (DAPI, C1005, Beyotime) was used to visualize nuclei. Images were observed and captured using an Olympus IX73 inverted microscope (Olympus, China). The viability of follicles from the 2D and 3D culture systems was determined using a Meilun Calcein-AM/PI Double Staining Kit (MA0361, Meilunbio, China). On days 2, 6, and 10, the cover medium was removed, and follicles were incubated with 100 μL of staining solution per well in the dark for 8 min and then imaged with an Olympus IX73 inverted microscope (Olympus). Quantification of the percentage of live and dead cells was performed by ImageJ software (ImageJ, National Institutes of Health, Bethesda, MD, USA). For hGCs, the experimental protocols were the same as before. The viability of hGCs was examined on days 2, 4, and 6 of in vitro culture. For cell cycle analysis of DNA content, the cells were cultured for 48 h and 96 h in each group before being collected, washed with PBS, and resuspended with 2 mL of precooled 75% ethanol for 12 h at 4 °C. The cells were then washed and resuspended in 300 μL of PBS. Then, 3 μL of RNase A (10 mg/mL, Solarbio, China) was added to the suspension which was incubated at 37 °C for 30 min. Before the test, 3 μL of PI (10 mg/mL, Solarbio) was added to each tube at room temperature for 30 min. These cells were analyzed by fluorescence-activated cell sorting using a flow cytometer (BD FACSCalibur; BD Biosciences, China) and quantified by ModFit LT for Mac v3.0 (BD Biosciences). The hGCs were cultured in vitro for 2 and 4 days. HGC apoptosis was observed using Annexin V-FITC/propidium iodide (PI) apoptosis detection kits (A211-01, Vazyme, China). Briefly, 1 × 10 5 hGCs per test were collected by trypsin digestion and washed 2 times using prechilled PBS. hGCs were resuspended in 100 μL of binding buffer and then labeled with Annexin V-FITC (5 μL) and PI (5 μL) for 10 min in the dark at room temperature. After incubation, 400 μL of binding buffer was added to the cell mixture, and samples were kept on ice. Green (Annexin V-FITC) and red (PI) fluorescence were detected by flow cytometry (BD FACSCalibur; BD Biosciences). The excitation wavelength was 488 nm. Freshly collected hGCs were seeded in 24-well plates at 5 × 10 5 /well for 48 h. The 2D-cultured hGCs were fixed for 15 min at room temperature and washed twice with PBS. Then, hGCs were stained in freshly prepared SA-β-gal solution overnight at 37 °C (G1580, Solarbio). For the 3D-cultured hGCs, 1 mL of preheated VitroGel Cell Recovery Solution (Well, NJ, USA) was added to each well and thoroughly mixed, and the cells were collected after centrifuging at 1500 rpm for 5 min. The cells were then fixed and washed twice with PBS and centrifuged at 1500 rpm for 5 min. Finally, the cell pellets were resuspended with SA-β-gal solution overnight at 37 °C. Positive cells were stained blue and counted in three randomly selected fields under a microscope (IX73, Olympus). The concentrations of AMH and progesterone in the collected culture media per 20,000 cells were measured on days 1, 2, 4, and 6 of in vitro culture. To measure the hormone levels, ELISA kits (CSB-E12756h and CSB-E07283h, Cusabio, China) were used following the manufacturer’s instructions. The absorbance was measured at 450 nm by a microtiter plate reader (Synergy HTX, USA). The mitochondria of hGCs from the 2D and 3D culture systems were evaluated in at least 5 random sections for each sample, and the average percentage of abnormal mitochondria in each group was calculated. hGCs were fixed in 2.5% glutaraldehyde at room temperature for 30 min and then incubated overnight at 4 °C. Samples were dehydrated in a graded series of ethanol and embedded in epoxy resin; ultrafine sections were obtained using an ultramicrotome. Subcellular organelles were observed with a TecnaiG2 Spirit120kV electron microscope (Thermo Fisher Scientific). We examined the mitochondrial membrane potential (MMP) of hGCs using JC-1 staining and flow cytometry with a JC-1 Assay Kit (M8650, Solarbio) as described in the product manual. Briefly, 1 mL of cell culture medium and 1 mL of JC-1 staining solution was added to the cells cultured for 2 days in the 2D or 3D system (the method of cell collection in the 3D culture system was the same as Section 2.9 ), incubated for 20 min at 37 °C, washed twice with buffer, and observed by fluorescence microscopy (IX73, Olympus) or analyzed by flow cytometry. At high MMP, JC-1 aggregated in the mitochondrial matrix and formed polymers (JC-1 aggregates), which can produce red fluorescence. At low MMP, JC-1 cannot aggregate in the mitochondrial matrix, when JC-1 was a monomer, which can produce green fluorescence. The relative MMP ratio was calculated as red fluorescence intensity/green fluorescence intensity. Intracellular ROS levels in hGCs cultured for 2 days in the 2D or 3D system were examined using 2′-7′-dichlorodihydrofluorescein diacetate (DCFH-DA)-based fluorescence imaging (CA1410, Solarbio). HGCs collected from each group (the method of cell collection in the 3D culture system was the same as Section 2.9 ) were washed twice with DMEM and immediately incubated in 10 mM DCFH-DA at 37 °C for 20 min in the dark, after which the hGCs were washed with DMEM to remove the surface fluorescence. The fluorescence intensity was observed by fluorescence microscopy (IX73, Olympus). Total RNA was extracted from hGCs cultured for 48 h in the 2D and 3D systems with RNAiso Plus (9108, Takara, Tokyo, Japan), and the expression of caspase3 , caspase8 , BCL2 , BAX , GPX4 , SOD1 , COX2 , HAS2 , and PTX3 was analyzed. Complementary DNA was synthesized using a HiScript III RT Reagent Kit (R323-01, Vazyme) according to the manufacturer’s protocol at 37 °C for 15 min, 85 °C for 5 s, and 4 °C. Quantitative PCR was performed using a ChamQ Universal SYBR qPCR Master Mix (Q711-02/03, Vazyme). Quantitative RT–PCR was conducted at 95 °C for 30 s followed by 40 cycles at 95 °C for 5 s and 60 °C for 30 s and final extension at 95 °C for 15 s, 60 °C for 60 s, and 95 °C for 15 s using a Roche LightCycler 480 Instrument I (Roche Molecular Systems, Inc., Budapest, Hungary). The relative mRNA expression was normalized to that of GAPDH . The primers used in this study are listed in Table S2 . Next, 150 μL of a radioimmunoprecipitation assay (RIPA) buffer combined with 1% protease inhibitor (PMSF, P0013B, ST506, Beyotime) was added to lyse the hGCs. The lysates were collected by centrifugation at 12,000 g for 20 min at 4 °C. The total protein concentrations were determined using a BCA kit (P0010S, Beyotime), and the samples were standardized to 1.5 μg/μL. Protein samples were separated by 15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (P0012AC, Beyotime) and subsequently blotted onto 0.22 μm PVDF membranes (EMD Millipore, Billerica, MA, USA). Then, 5% skimmed milk powder dissolved in TBST was used for blocking. Then, the membranes were incubated with primary antibodies at 4 °C overnight. Specific primary antibodies included P16 (1:1000, Cell Signaling Technology, USA, 92803), P21 (1:1000, Proteintech, 10355-1-Ap), SOD1 (1:1000, Proteintech, 10269-1-Ap), GPX4 (1:1000, Proteintech, 67763-1-Ig), PTX3 (1:1000, Abcam, UK, ab190838), COX2 (1:1500, Proteintech, 66351-1-Ig), HAS2 (1:1000, Bioss, bs-11290R), integrin α6 (1:1000, Proteintech, 27189-1-AP), integrin β1 (1:1000, Proteintech, 12594-1-AP), and GAP-DH (1:1000, Cell Signaling Technology, 5174). The next day, the membranes were incubated with secondary antibodies (1:2000, peroxidase-labeled goat antirabbit/mouse IgG, Beyotime, A0208/A0216) for 2 h at room temperature and visualized with an Amersham Imagequant 800. All experiments were repeated at least in triplicate. Statistical analysis was conducted using GraphPad Prism 8 (San Diego, CA, USA). Values are presented as the mean ± SD. One-way ANOVA was used in the comparison of continuous variables among groups. A two-tailed paired t -test was used in the comparison between the two groups after confirming the normal distribution of the data by the Kolmogorov–Smirnov test. Significance is indicated by * p < 0.05, ** p < 0.01, and *** p < 0.001.

The emerging field of tissue regeneration has driven the evolution of biomaterials. Bioactive peptides derived from recognition motifs of specific proteins and targeting stimulation of follicular granulosa cells by modified biomaterials are under-researched. In this study, we established a 3D follicle culture system utilizing LYR-modified hydrogel that was able to provide a biophysical microenvironment for follicle development and promote oocyte meiosis. In this culture system, hGCs showed prolonged survival, increased viability, decreased apoptosis, enhanced hormone secretion, reduced oxidative stress, improved mitochondrial membrane potential, delayed cellular senescence, increased expression of cumulus-expansion-related genes and proteins (COX2, HAS2, and PTX3), and upregulated integrin α6β1. The extracellular matrix of the follicle is constantly changing as the follicle progresses, and it has been found that different stages of follicles have distinct responses to the ECM. 3 Collagen I and the RGD peptide can affect two-layer secondary follicle growth but not multilayered follicle growth compared with the other ECM. Fibronectin and laminin delayed the growth of multilayered follicles to antral follicles and improved oocyte meiotic competence. 36 In terms of GCs, laminin was able to retard GCs differentiation and regulate the production of steroid hormones. 37 Thus, laminin may synchronize oocyte and GCs development by regulating GCs to avoid overspeed development. Laminins are large heterotrimeric glycoproteins that consist of α, β, and γ chain and interact with a variety of cell-surface receptors such as integrins and syndecans. 38 Integrin α6β1 is mostly expressed on medium to large follicles and luteinized GCs, consistent with the stages of laminin action on follicles. 10 Blocking integrin α6β1 was found to result in an elevated response to gonadotropins and increased progesterone production in GCs, consistent with our results. 39 Thus, the use of laminin mimetic peptides YIGSR and IKVAV was able to achieve effects similar to those of laminin. On the other hand, the intrinsic mechanical property of the ECM is considered as an important mechanical cue in a native microenvironment. 21 In detail, the mechanical links between the cellular cytoskeleton and focal adhesions allow the cells to generate cytoskeletal tensional states which are transmitted to the cell nucleus via intracellular pathways, and these mechanical cues are converted to biochemical responses that may affect cell fate. 40 Hydrogels have been extensively used as synthetic ECMs for 3D cell culture, and their elastic modulus can influence cell fate. 41 Alginate is the most widely available biomaterial for the 3D culture of follicles and does not require exposure to high temperature or UV radiation; however, controlling the degradation rate of the alginate hydrogels to match with the follicle growth is challenging, and the rigidity can negatively affect further development of the follicles. 42 − 44 Mouse preantral follicles encapsulated in 0.5% alginate had better follicle survival and antral formation than those encapsulated in 1.5% alginate. 45 Another study also showed that 0.25% and 0.5% alginate supported the more rapid growth of follicles and antrum formation compared with 1.5% and 1.0% alginate, which indicated that softer materials may be more suitable for the development of mouse follicles. 46 Moreover, alginate has no cell binding sites, so cell adhesion peptides, such as the RGD sequence, can be used to enhance cell adhesion. 47 Kreeger et al. modified alginate with ECM proteins or RGD to encapsulate secondary follicles and improved follicle development with a follicle survival rate of 71.7% and matured oocyte rate of 13.3%. 6 Combining alginate with other polymers such as fibrin and matrigel is another way to enhance its cell attachment properties and biodegradation rate. 48 , 49 Jin et al. encapsulated isolated mouse secondary follicles in a fibrin–alginate matrix and showed larger oocyte diameter, more antrum formation, and theca cell differentiation compared with the alginate matrix. 48 Follicle survival rates reported in the literature for secondary follicles encapsuled in the fibrin–alginate matrix were 70–81%, with a wide variation in oocyte maturation rates of 22.6–88%. 6 , 46 , 48 − 50 In fact, fibrin is the main material for capillary formation, which may improve oxygen transport under long-term culture conditions. The polysaccharide hydrogel used in this study has good oxygen permeability, and the mouse secondary follicles were able to obtain good survival rate, cell activity, and oocyte maturation rate. The results of this study showed that 3D-cultured hGCs in LYR-modified hydrogel exhibited significantly higher viability and longer survival time compared to those grown in the 2D culture. In addition, the proportion of early apoptosis was significantly lower in 3D-cultured hGCs, but there was no difference in late apoptosis. These results may be attributed to part of the apoptotic hGCs being encapsulated in the hydrogel as the 3D culture system is established, resulting in apoptotic hGCs in the 3D culture not being removed by culture medium exchange, while part of the apoptotic cells in the 2D culture floated away and was lost after culture medium exchange, thus not allowing comparison of late apoptosis due to the different culture systems. In terms of hormone secretion, the AMH on day 1 and progesterone on days 1 and 2 detected in the culture supernatant of 3D-cultured hGCs were lower compared to those cultured in 2D. The result is consistent with previous studies that laminin mimetic peptides are able to inhibit GCs flavinization by specifically binding integrin α6β1. 6 Another possible reason includes the retardation effect of the hydrogel, resulting in the production of hormones not being immediately released from the hydrogel. Excessive oxidative stress in hGCs may induce impaired development of the cumulus–oocyte complex, follicular apoptosis, and disturbance in follicular metabolism. 51 , 52 One study observed that hGCs from patients with ovarian endometriosis had significantly increased oxidative stress, reduced MMP, and decreased ATP production, leading to hGCs senescence, similar to women of advanced age. 53 , 54 Our study verified that 3D-cultured hGCs had significantly reduced oxidative stress, increased MMP, and enhanced integrity of mitochondria relative to the 2D-cultured hGCs, retarding the rapid senescence of hGCs after external stimulation in vitro. Cumulus expansion is a vital step in the process of follicle maturation, during which GCs produce hyaluronic acid to deposit outside, allowing oocytes and GCs to bind tightly and promoting oocyte meiosis and maturation. 55 In this study, the expression of cumulus-expansion-related genes and proteins (COX2, HAS2, and PTX3) was significantly higher in hGCs cultured in 3D than those cultured in 2D. Several studies have demonstrated that integrin α6β1 can regulate GCs proliferation and steroid formation, but the underlying mechanisms have not been elucidated. 8 , 56 A shortcoming of this study is that the fertilization and embryonic development capabilities were not assessed in this culture system. In addition, the hydrogel utilized in this study lacks exploration of its long-term safety. Moreover, the molecular mechanisms underlying the regulation of integrin α6β1 on hGCs biological behavior need to be further explored. Future studies aimed at addressing these issues will be of great interest. In summary, we have demonstrated that a novel 3D culture system using laminin-peptide-modified hydrogels can potentially establish a biophysical environment suitable for follicular development, which may improve the biological function of hGCs and reduce the biological changes caused by the in vitro culture.

The ovarian follicle is a separate multicellular functional unit that contains a central oocyte, surrounding supporting cells including cumulus cells and granulosa cells (GCs), as well as a basement membrane composed of the extracellular matrix (ECM). Under the stimulation of a variety of hormones (endocrine effects) together with locally produced cytokines (paracrine and autocrine effects), ovarian follicles undergo multiple stages of development from primordial, primary, secondary, antral, to periovulatory follicles. Individual immature follicle culture is an important technique for investigating follicle development, allowing researchers to explore the intimately regulated interaction between oocytes and GCs, the synthesis and secretion of hormones, and the role of the ECM. 1 In addition, radiotherapy and chemotherapy are known to impair the ovarian function of young women suffering from cancer. Therefore, maintaining ovarian tissue function in vitro is essential for fertility preservation. Moreover, follicle culture for obtaining fertilizable oocytes can diminish the risk of reintroduction of cancer cells by direct transplantation of ovarian tissue. 2 Follicular development is regulated by the ECM which provides the mechanical structure and recognition motif for the initiation of signaling cascades. By adhering to the cell surface integrin receptors, the ECM regulates cell proliferation, survival, differentiation, and steroid formation and promotes follicular development and maturation. 3 − 5 Both the components of the ECM and the expression of integrin are dynamic. By simulating the 3D environment of the follicle in vitro, it was discovered that the major components of ECM, fibronectin and laminin can delay GC differentiation, synchronize the development of GCs and oocytes, inhibit the premature development of oocytes in vitro, and improve the meiotic capacity of oocytes. 6 LAMA1, LAMA3, LAMB1, and LAMB2 are highly expressed in GCs according to studies of the expression profile of laminin in the ovary. 7 In the human ovary, integrin α6 is highly expressed on GCs of medium to large follicles and luteinized GCs in the early luteal phase, forming a heterodimer with integrin β1. 8 Integrin α6β1 has been shown to promote preovulatory follicle development, ovulation, oocyte maturation, and fertilization. 9 With regard to GCs, integrin α6β1 contributes to promoting cell proliferation, regulating steroid formation, and inhibiting luteinization to prevent premature luteinization resulting in oocyte dysplasia. 10 However, to date, the effect and the underlying mechanism of activating integrin α6β1 with a defined ECM molecule on follicle development, meiosis of oocytes, and the regulation of GCs remain largely unknown. A three-dimensional (3D) culture system is able to provide a close resemblance to the in vivo microenvironment, maintain the integrity of follicles, and avoid follicle damage by keeping GCs and theca cells from adhering to the bottom of the culture plate in a two-dimensional (2D) culture. 11 Researchers have devoted themselves to creating 3D culture models of follicles utilizing different materials, such as gelatin, alginate, hyaluronic acid, and matrigel. 11 − 13 However, these materials do not simultaneously provide appropriate biological properties, degradability, and adequate stiffness. 14 , 15 In this regard, hydrogels are water-swollen polymers with tunable physicochemical properties and advanced fabrication approaches that can be gelated without altering the external environment and have been used for the 3D culture of a variety of cell types, such as stem cells, colorectal cancer cells, breast cancer cells, and osteosarcoma cells. 16 − 19 Thus, hydrogels can be manipulated in vitro with biophysical cues to recapitulate the three-dimensional microenvironment of the follicle. 20 , 21 Bioadhesive ligands such as full-length proteins or short peptides have been shown to functionalize biomaterials to promote cell adhesion and modulate cell behavior. 22 − 25 The utilization of short peptides showed a number of apparent advantages: small peptides are more stable, cheaper, and capable of mass production. 26 Full-length laminin proteins have multiple bioactive amino acid sequences that may lead to uncontrollable intercellular interactions. 27 IKVAV and YIGSR are laminin-mimetic peptides located in the Laminin α1 and β1 chains, respectively. They are recognized by the cell surface receptor integrin α6β1, which mimics the function of laminin. 28 − 30 The hydrogel loaded with IKVAV and YIGSR provides the possibility to investigate the effect of defined ECM molecules on follicles and GCs in 3D structures. In this study, we established a novel 3D culture system using a synthetic polysaccharide hydrogel modified with laminin-mimetic peptides and RGD to simulate the developmental environment of follicles in the ovary and investigate the effects of specific ECM molecules on follicles and granulosa cells.