Effects of Alginate Concentration and Ovarian Cells on In Vitro Development of Mouse Preantral Follicles: A Factorial Study

In the first step of this experimental study, preantral follicles were isolated from mice ovaries in five independent replicates, randomly allocated to encapsulate in 0.5, 0.75 and 1% ALG hydrogels in the absence or presence of OCs, and cultured for 13 days. The diameter and morphological appearance of developing follicles were analyzed on days 1, 6 and 13 of culture. Additionally, on day 13, the survival rate of the follicles was calculated and healthy follicles were evaluated with regard to their antrum formation rate. Then, antral follicles were induced by 2.25 IU/ml human chorionic gonadotropin (hCG, Choriomon, Switzerland) and 20-22 hours later, on day 14 of culture, the developmental stages of the obtained oocytes were assessed. After determining the best concentration of ALG based on the larger diameter, higher survival, antrum formation, and maturation rates, in the second step of the study this concentration was used for culturing preantral follicles in the absence or presence of OCs. On day 13 of culture, antral follicles were fixed for investigation of DNA fragmentation and assessment of Cx37 and Cx43 protein levels. Conditioned media from the follicle cultures were collected in three replicates for the measurement of estradiol, progesterone, and androstenedione secretions. Finally, after hCG induction, MII oocytes were collected and evaluated in terms of their cortical granule distribution, meiotic spindle organization, and chromosomal alignment.

Female NMRI mice (Pasteur Institute, Iran) were housed in the animal facility of Royan Institute under standard housing conditions, with controlled temperature (20-25°C) and lighting (12 hours light: 12 hours dark). They were handled pursuant to the ethical guidelines set by Royan Institute (ethical permission number: IR.ACECR.ROYAN.REC.1395.93).

Twenty three-four-week-old immature mice were sacrificed by cervical displacement, and their ovaries were isolated in an aseptic condition and placed in ice-cold base medium containing Dulbecco's Modified Eagle’s medium (DMEM, Gibco, UK), penicillin (Gibco, UK), streptomycin sulfate (Gibco, UK), sodium bicarbonate (NaHCO3, Sigma, USA) and 10% fetal bovine serum (FBS, Gibco, UK). Next, the ovaries were cleaned of the bursa and adipose tissue under a stereomicroscope (SZ61, Olympus, Japan). Oocytes and granulosa cells were removed from the ovaries by puncturing follicles with two 29G insulin syringes and then discarded. The remnants were chopped and incubated for 45 minutes at 37°C in 200 µl per ovary of collagenase solution containing 4 mg/ml collagenase IV (Gibco, UK) in serum-free base medium. During this time, the ovarian tissue pieces were pipetted up and down at least 20 times every 10-15 minutes to mechanically disrupt them. To stop the enzymatic activity an equivalent volume of base medium was added to the samples. The isolated cell solution was then filtered through a sterilized 40 µm filter mesh (Falcon, Mexico) and centrifuged at 1800 rpm for 5 minutes. The obtained cells were washed and the final pellet was re-suspended in a known volume of base medium. The cells were transferred to T25 culture flasks containing 4 ml of base medium supplemented with 1% insulin-transferrin-selenium (ITS, Gibco, UK), 1% L-glutamine (Sigma, USA), 1% non-essential amino acids (Gibco, UK) and 0.1% β-mercaptoethanol (Sigma, USA), and then incubated at 37°C in a water-saturated atmosphere of 95% air and 5% CO2 until they reach full confluency. Next, OCs were trypsinized and washed and the viable cells were counted with a trypan blue staining and a Neubauer chamber. Afterward, 1ml aliquots of the cells (5×10⁵ cells/ml) were stored in 10% DMSO (Sigma, USA)/FBS at -80°C for later use.

A total of thirty 12-14-day-old mice were sacrificed by cervical displacement. Mouse ovaries were mechanically dissected under a stereomicroscope at 37°C, using two 29G needles attached to 1ml insulin syringes, and placed in alpha minimum essential medium (α-MEM, Gibco, UK) supplemented with penicillin, streptomycin, NaHCO₃ , and 10% FBS. Only intact preantral follicles with 2-3 layers of granulosa cells and 100-130 µm in diameter were chosen and divided randomly into experimental groups.

To make a 1.0% (w/v) ALG solution, 10 mg/ml alginic acid sodium salt (Sigma, USA), 25 mM 4-(2-hydroxyethyl)- 1-piperazineethanesulfonic acid (HEPES, Sigma, USA) and 150 mM sodium chloride (NaCl, Sigma, USA) were dissolved in deionized water, filtered through a sterilized 0.22 µm filter (Millipore, USA) (22), then diluted with sterile 1X phosphate buffered saline (PBS, Takara, Japan) without calcium and magnesium to reach the final concentrations of 0.75 and 0.5% (w/v). In order to prepare hydrogels, cross-linking solution (50 mM calcium chloride (CaCl₂, Sigma, USA)/140 mM NaCl in deionized water) was mixed with hydrogel solutions.

In the first step of the study, groups of 109.83 ± 7.59 preantral follicles were individually encapsulated in 0.5, 0.75 and 1% ALG solutions and in the absence or presence of OCs, in five independent replicates. For cell encapsulation, about 5×103 OCs per follicle were mixed with hydrogel solutions and pipetted in 5-µl droplets on sterile ultra-low attachment culture dishes (Dow Corning, USA). The concentration of OCs in each droplet was determined based on the best results obtained in our pilot study. Afterwards, follicles were individually placed in the 5-µl droplets, cross-linking solution was gently pipetted on top of each droplet, and then incubated at 37°C for 2 minutes. After incubation, the beads were washed with α-MEM medium and then placed into 96-well plates (TPP, Switzerland). Each well contained one bead in 100 µl culture medium [α-MEM supplemented with 5% FBS, 1% ITS, 10 mIU/ml follicle stimulating hormone (FSH, Merck, Germany)]. Lastly, plates were incubated in a 5% CO₂ incubator at 37°C for 13 days and 50 µl of the medium was replenished every 3-4 days.

Morphological features and the diameters of developing follicles were assessed on days 1, 6, 13 of culture. The diameters were determined as the mean of two perpendicular measurements of each follicle using ImageJ software (U.S. National Institutes of Health) (23). Moreover, on day 13 of culture, the survival rate of the cultured follicles and antrum formation rate of the survived follicles were evaluated observationally based on the morphological appearances of the follicles and their oocytes. In this regards, extrusion of the oocytes, their dark appearance and surrounding granulosa cells were considered the indications of degeneration. Also, antrum formation was determined as an observable transparent cavity within the granulosa cell masses.

In vitro maturation and ovulation of antral follicles were induced by 2.25 IU/ml hCG, on day 13 of culture. For evaluation of meiotic maturation of the oocytes, at 20-22 hours after induction the extruded cumulus-oocyte complexes (COCs) were removed from the culture wells and denuded by gentle pipetting, then the number of germinal vesicle (GV), GV breakdown/metaphase II (GVBD/MII), and degenerated oocytes were determined.

In the second step of the study, some preantral follicles were encapsulated in the ultimate concentration of ALG hydrogel from the first step, either in the absence or presence of OCs, and were cultured similar to the first step. On day 13 of culture, survived antral follicles were fixed in 4% paraformaldehyde overnight at 4°C; then the follicles were rinsed twice in PBS, dehydrated in increasing concentrations of ethanol and embedded in paraffin. Next, 5 µm thick slices were prepared and mounted on adhesion slides for DNA and protein analyses. Three sections per group, taken from the middle of three random follicles, were selected for either DNA fragmentation or Cx37 and Cx43 protein expression detection. To prepare the sections, they were first deparaffinized at 60°C for 30-40 minutes, washed in xylene solution for 20 minutes, and rehydrated by rinsing in serially diluted ethanol and water bath.

Strand breaks of DNA in apoptotic cells were detected by terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay utilizing the In Situ Cell Death Detection Kit, TMR Red (Roche Diagnostics GmbH, Mannheim, Germany). The procedure was done according to the manufacturer's protocol and previous descriptions (24). In brief, after deparaffinization and rehydration, sections were pretreated with freshly prepared permeabilization solution [0.1% Triton X-100, 0.1% sodium citrate (Sigma, USA)] for 8 minutes at room temperature. After rinsing with PBS/0.05% Tween 20 (PBS-T, Sigma, USA), sections were incubated in TUNEL reaction mixture containing 50 µl enzyme solution (terminal deoxynucleotidyl transferase) and 450 µl label solution (nucleotide mixture in reaction buffer) at 37°C in a humid incubator for 1 hour. Finally, the sections were washed in PBS-T, counterstained in 40,6-diamino-2-phenylindole (DAPI, Sigma, USA), and examined under an inverted fluorescence microscope (Eclipse 50i; Nikon, Japan). All images were processed for publication by the Adobe Photoshop software (CS5.1, Adobe Systems Inc., San Jose, USA).

Red fluorescence was visualized in TUNEL-positive cells. Sections from mouse ovarian tissue were incubated with DNase I recombinant [5 U/ml in 50 mM Tris- HCl, 1mg/ ml bovine serum albumin (BSA), pH=7.5] and were used as a positive control. Negative control sections were incubated with the label solution rather than TUNEL reaction mixture.

To quantify the DNA fragmented follicular cells, all cells found in the sections, either for TUNEL staining or DAPI counterstaining, were counted using ImageJ cell counter plugin; then, the percentage of TUNEL-positive cells was computed.

After deparaffinization and rehydration, antigen retrieval was performed by incubating the sections of the antral follicles in 0.01 M sodium citrate buffer (pH=6.0, Sigma, USA) for 1 hour in a 96°C oven. The sections were washed two times in PBS-T, and incubated in 10% goat and donkey serums (Sigma, USA) diluted in PBS for 1 hour at 37°C to block non-specific protein bindings in Cx37 and Cx43 immunostainings, respectively. After two PBS washes, the sections were incubated overnight at 4°C with primary antibodies against Cx37 [Primary rabbit polyclonal antibody (ab181701, Abcam, UK)] and Cx43 [primary mouse monoclonal antibody (C8092, Sigma, USA)]. Both primary antibodies were diluted 1:400 in related blocking solutions. Then, the sections were rinsed with PBS carefully and incubated with secondary antibodies [(goat anti-rabbit IgG (H+L) crossadsorbed (Alexa Fluor 488, A11034, Thermo Fisher Scientific, USA) and donkey anti-mouse IgG H&L (Alexa Fluor 488, ab150105, Abcam, UK)] for 1 hour at 37°C. Both secondary antibodies were diluted 1:1000 in related blocking solutions. Lastly, the sections were washed, counterstained with DAPI for 1 minute, and inspected under an inverted fluorescence microscope (Eclipse 50i, Nikon, Japan). Sections from rat heart tissue were used as a positive control. For the negative control, ovarian sections were processed without the primary antibodies.

On day 13 of culture, the level of estradiol (E2), progesterone (P4) and androstenedione (A4) hormones were measured in conditioned media collected from 30 cultured antral follicles per group, in three replicates. The hormonal measurement was conducted using mouse ELISA kits (Bioassay Technology Laboratory, China) and according to the kits’ instruction. Data were adjusted for every follicle by dividing each of the measured hormonal secretions by the number of the follicles. According to the kits’ datasheets, the sensitivity of the assay for E2, P4 and A4 were 1.51 ng/L, 0.28 ng/ml and 0.022 ng/ml, respectively.

Following the assessment of DNA fragmentation and protein expressions, cultured antral follicles were induced by hCG as was explained in detail in the determination of oocytes meiotic maturation section, and then a total number of 20 MII oocytes (10 oocytes per group) were collected in three replicates. Also, 10 in vivo matured MII oocytes were gathered as the control group. To get in vivodeveloped oocytes, three 6-8-week-old NMRI mice were injected intraperitoneally by 7.5 IU of pregnant mare’s serum gonadotropin (PMSG, Sigma, USA) followed 48 hous later by 7.5 IU hCG. After 18 hours, the mice were sacrificed, the COCs were isolated from their oviduct ampulla and then denuded. Afterwards, by using pronase (0.5 mg/ml in PBS, Sigma, USA) the zona pellucida of in vivo- and in vitro-developed oocytes were removed at 37°C and then oocytes were fixed in 4% paraformaldehyde for at least 1 hour. After washing in PBS with 0.01% Tween 20, oocytes were permeabilized in PBS containing 0.3% BSA (Gibco, UK) and 0.1% Triton X-100 (Sigma, USA) for 15 minutes at room temperature and blocked in PBS containing 0.3% BSA for 1 hour at 37°C. To stain the meiosis spindle and cortical granules, oocytes were incubated in the blocking solution containing anti-alpha tubulin antibody-microtubule marker (FITC) (1:100; Abcam, UK) and rhodamine-labeled Lens Culinaris Agglutinin (LCA) (1:500, Vector Laboratories, Burlingame, CA, USA) for 1 hour at 37°C. Finally, the stained oocytes were washed in PBS-T, counterstained with Hoechst 33342 (1 mg/ml in 1X PBS, Sigma, USA) for 5 minutes at 37°C and mounted on adhesion slides. Fluorescence labeling was detected using an inverted fluorescence microscope (Eclipse 50i, Nikon, Japan) and images were processed by Adobe Photoshop software. The lack of a cortical distribution was considered as the sign of cortical granule abnormality, and disorganized spindle or misaligned chromosomes were considered as an indicator of spindle abnormality.

Statistical differences in follicle survival rates, antrum formation rates, oocyte maturation rates, oocyte abnormality rates and DNA fragmentation were analyzed using GENMOD procedure including function link logit in the model. GENMOD procedure produced odds ratio (OR) as the strength of difference between the groups. Data associated with follicle diameters were analyzed by MIXED procedure including RANDOM and REPEATED statements in the model to identify between and within covariances, respectively. Data pertaining to hormonal secretions were analyzed using GLM procedure. In addition, LSMEANS statement was included in the model to perform multiple comparisons. All analyses were conducted in SAS version 9.4 (SAS Institute Inc., NC, USA). Differences were considered significant at P<0.05.

Assessment of the morphologies and diameters of the follicles on days 1, 6 and 13 of culture indicated no significant difference in follicles encapsulated in different concentrations of ALG, either in the absence or presence of OCs. On the other hand, follicles which were co-cultured with OCs had a more spherical shape and relatively larger diameters than the non-co-cultured ones. On day 13 of culture, for instance, the difference in the diameters of 0.5%-OCs and 0.5%+OCs groups was significant (347.18 ± 10.63 vs. 418.14 ± 21.89 µm, respectively, P<0.05, Fig .1A, B).

Assessment of follicle survival rates on day 13 of culture indicated that there was a linear trend towards a better survival rate with lowering ALG concentration, in both the absence and presence of the OCs. However, the difference between 0.5%- OCs group and both 0.75%-OCs and 1%-OCs groups reached statistical significance (71.87 vs. 52.52 and 40%, respectively, P<0.05, Table 1). Also, the comparison of -OCs and +OCs groups revealed that adding the OCs to all hydrogel beads had an affirmative effect on the follicle survival rate, and the difference between 1%-OCs and 1%+OCs groups was significant (40 vs. 63.91%, respectively, P<0.05, Table 1).

Surprisingly, in the absence of OCs, the proportion of the follicles developed to antral stage was higher in 1% group as compared to the 0.75 % group (75 vs. 59.61%, respectively, P<0.05); while in the presence of OCs there was no significant difference between the groups. On the other hand, all -OCs groups had a relatively lower antrum formation rate than the +OCs ones. Nevertheless, the difference reached statistical difference in 0.5 and 0.75% groups, only (69.59 vs. 88.57% and 59.61 vs. 77.65% for-OCs and +OCs groups, respectively, P<0.05, Table 1).

The evaluation of oocytes obtained from 0.5, 0.75 and 1% ALG-cultured antral follicles showed that there was no significant difference between groups regarding the rates of GV and GVBD/MII oocytes, either in the absence or presence of OCs. However, 0.5% ± OCs groups had a higher rate of degenerated oocytes than the 1% ± OCs ones (P<0.05). Also, it was clear that the oocytes of 0.5%+OCs and 0.75%+OCs groups were more likely to break down their GVs and develop to GVBD or MII stages as compared to the -OCs groups (P<0.05, Table 1).

Based on the first step results, 0.5% ALG hydrogel is potentially best suited for preantral follicle culture, either in the absence or presence of the OCs. In the second step, the evaluation of DNA fragmentation in 0.5% ALG ± OCcultured antral follicles revealed that although a negligible percentage of the follicular cells was TUNEL-positive in both groups, follicles that were co-cultured with OCs demonstrated a relatively lower percentage of DNA fragmentation compared to the non-co-cultured ones (2.2 ± 0.7 vs. 3.9 ± 0.7%, respectively, Fig .2A).

Immunofluorescence staining for Cx37 and Cx43 are displayed in Figure 2B and C as shown, the strong immuno-labeling of Cx37 and Cx43 were observed on granulosa cells of both -OCs and +OCs groups, while qualitatively, no remarkable difference was observed between the two groups.

Hormonal secretion data indicated that there was no significant difference between the groups in the levels of E2 and A4 hormones; however, the level of P4 in the +OCs group was significantly higher than that in the -OCs group (3.2 ± 0.4 vs. 1.8 ± 0.2 ng/ml, respectively, P<0.05, Fig .3A-C).

The normal and abnormal MII oocytes in terms of cortical granule distribution, meiotic spindle organization, and chromosomal alignment are shown in Figure 4. Our data revealed that all in vivo-developed oocytes had a normal cortical granule distribution; whereas 100 and 40% of the oocytes that were developed in 0.5%-OCs and 0.5%+OCs groups, respectively, were abnormal (lack of a cortical distribution of cortical granules) (P<0.05). Concerning meiotic spindle organization, under both in vivo and in vitro conditions, oocytes with abnormal spindle or chromosomal alignments were observed (disorganized spindle or misaligned chromosomes were considered as an indicator of spindle abnormality) (in vivo: 20%; 0.5%-OCs: 40%; 0.5%+OCs: 50%). Interestingly, the presence of OCs had no significant effect on the rate of abnormalities.