Bioengineering of Human Articial Ovary From Cryopreserved Tissue: Optimization of the Follicles Enzymatic Isolation.

Background: This study aims to evaluate the effectiveness of Tumor Dissociation Enzyme (TDE) for the isolation of cryopreserved human ovarian cortex tissues, proposing an optimized applicable for articial ovary. Methods: This is a prospective experimental study. We present a comparative analysis among the outputs of follicles isolated from cryopreserved ovarian biopsies of ten young women undergoing laparoscopy. Follicles were isolated by Tumor Dissociation Enzyme (TDE) or Liberase Dispase High (DH) enzymatic digestion. Follicles were assessed by evaluating the number, viability, morphology, oxidative stress. Moreover, the follicle growth and viability were analyzed after eight days of in vitro culture (IVC) between each protocol. Results: The recovery rate of follicles in Group 1 was signicantly higher by 78 follicles compare to Group 2 (p < 0.05), while the difference in weight and volume of biopsies in both groups was no signicant (p > 0.05). Group 1 allowed to isolate more primordial follicles (p < 0.05) and smaller diameter of follicles (p < 0.01) than Group 2. Group 1 allowed to isolate 7% more of bright red follicles than Group 2 (p < 0.01). We also found that Group 1 had isolated a signicantly higher percent of viable follicles (p < 0.05), more morphologically normal follicles (p < 0.05), and lower oxidative stress levels compared with Group 2 (p < 0.01). The velocity of follicles growth from Day 0 to Day 8 in Group 1 was signicantly higher (p < 0.05) than in Group 2. The viability of follicles on Day 8 of in vitro culture in Group 1 was signicantly higher than Group 2 (p < 0.05). Conclusions: TDE treatment can be for Liberase DH, allows the of highly viable follicles from the cryopreserved ovarian and


Background
Current cancer therapy improvements have remarkably raised cancer patients' survival rates and prolong patients' life expectancy. However, Anti-cancer treatments such as chemo/radiotherapy often lead to irreversible side effects to cancer survivors: premature ovarian failure and infertility [1]. Cryo-banking ovarian tissue precedes anti-cancer therapy supplies hope in terms of future fertility reestablishment for patients who have no su cient time to stimulate the ovary and cryopreserve embryo or oocyte [2].
However, reimplantation of cryopreserved ovarian tissue may reintroduce malignant cells and puts them at risk of disease recurrence [3,4]. However, it is known that the follicles surrounded by a layer of basement membrane can bar the blood capillaries, white blood cells and nerve processes, as a physiological barrier protecting follicles from contaminated by malignant cells of ovarian tissue [5]. Thus, embedded isolated preantral follicles in a bioengineered transplantable arti cial ovary can be a safe, alternative treatment and may be regarded as an extension of cancer care. In this case we have decided to establish a standard protocol for the follicle isolation and choose a scaffold to support these follicles growth.
Numerous mechanical and enzymatic methods and their combination have been developed to isolate ovarian preantral follicles [6][7][8]. The mechanical method combined with different types of collagenase (Ia, II, IX, XI) alone or together with bovine pancreatic deoxyribonuclease I (DNase I) [8,9], which avoids sticky DNA ends, may isolate more follicles in comparison with mechanical methods, However, it may damage the basement membranes and affect the viability of follicles [10]. Liberase, a mixture of puri ed enzymes, has been reported as a powerful tool to improve the isolated follicles' quality compared with collagenase [11]. Liberase DH (Dispase High) research grade belongs to the group of Liberase Research Grade with highly puri ed Collagenase I and Collagenase II blended together in a precise ratio and contains a high concentration of Dispase®, a non-clostridial neutral protease (Sigma-aldrich GmbH, Darmstadt, Germany). Moreover, others have shown that follicles isolated by Liberase DH had better morphology and high developmental competence compared with collagenase [12].
During enzyme isolation, increased proteolytic activity, cell lysis, and the liberate of DNA can hamper the isolated cells [13,14]. The release of DNA from dead cells induced the isolated cells cluster leading it challenging to digest the tissue and may also more comfortable to contaminate malignant cells in cancer patients. Besides DNase I can hydrolyze extracellular DNA released by degenerating cells without damaging intact cells, ensure the digestion not sticky and t to handle [15]. Telfer et al. rst introduced collagenase and DNase I to isolated many well-preserved preantral ovarian follicles of mice [16]. Xu et al. [17] retested using collagenase and DNase I to digest ovarian tissue of mice and culture isolated follicles in alginate bead and demonstrate average growth and development of follicles and oocytes. Wang et al. [18] using Liberase DH enzymatic solution containing DNase I to digest human cryopreserved ovarian tissue and get many good quality follicles.
Our prior study documented using "Tumor Dissociation Enzyme Reagent" (TDE, Innovative Diagnostic Systems Dr. Christian Sartori, Hamburg, Germany) to digested the cryopreserved ovarian tissue could yielded better viability follicles in contrast to Liberase TM [19]. Tumor Dissociation Enzyme (TDE) is a commercial enzyme-cocktail from the DCS ATP-Chemo-sensitivity Assay Kit (DCS Innovative Diagnostik-Systeme, Hamburg, Germany) for gentle enzymatic digestion [20,21]. Prior studies in our group reported that TDE contains the mixture of highly puri ed collagenase and other enzymes that enable the digested suspensions to be not sticky, easy to isolate, and yield a higher number of healthy cells [19,22]. However, we do not know what effect this enzyme will have on follicular growth and development after in vitro culture.
We hypothesized that using of Tumor Dissociation Enzyme (TDE) may isolate many human follicles with proper morphology, vitality, and development potential after in vitro culture. To test this, we apply TDE to digest our cryopreserved ovarian cortex in comparison with Liberase DH combine with DNase I on the quantity and quality of isolated follicles and culture in alginate gel for eight days to analyze follicular growth of both enzymes.

Number of retrieved follicles
After enzymatic digesting of the frozen ovarian cortex 1502 follicles in both treatment groups were recovered. The most of follicles in both groups were almost fully isolated from the stromal tissue ( Fig. 1A).
In treatment Group 1 790 follicles using TDE enzyme cocktail were recovered and in treatment Group 2 712 follicles using Liberase DH/DNase enzyme cocktail were recovered ( Table 1). As shown on Fig. 3A the recovery rate of follicles in Group 1 was signi cantly higher by 78 follicles compare to Group 2 (p < 0.05), while the difference on weight and volume of biopsies in both groups was no signi cant (p > 0.05, Table 1).  Grade composed of collagenase I, II and the neutral protease Dispase ® (Sigma-Aldrich Chemie, GmbH, Schnelldorf, Germany) and DNase I (Bovine pancreatic deoxyribonuclease I) which is a DNA minor grove-interacting nuclease with relatively low speci city (Sigma-Aldrich Chemie, GmbH, Schnelldorf, Germany).
On Fig. 3A has been presented the data of follicles recovery with different maturity. From these follicles distribution possible to see that the enzyme cocktail of treatment Group 1 (TDE) allowed to isolate more primordial follicles than enzyme cocktail of Group 2 (Liberase DH/DNase) as follow: 80.5 ± 5.4% vs 77.7 ± 5.2%, respectively (p < 0.05).
The comparison of follicles diameter in both treatment groups ( Fig. 3B) has shown that follicles of Group 1 has signi cantly smaller diameter than the follicles of Group 2: 44.5 ± 11.4 µm vs 45.8 ± 11.1 µm, respectively, p < 0.01.

Viability assessment of retrieved follicles
Neutral red dye A total of 1502 follicles were stained with neutral red dye. The viability of follicles using express staining with neutral red dye have presented on Figure 4A. The data have shown that TDE-cocktail enzyme of Group 1 allowed to isolate 7% more (p < 0.01) of bright red follicles than Liberase DH/DNase enzyme cocktail of Group 2 (70.9 ± 4.7% vs 63.9 ± 4.2%, respectively). It was also noted that percent of follicles stained in light red colour in treatment group Group 2 was 5% more (p < 0.01) than in Group 1 (17.9 ± 4.8% vs 23,0 ± 3.7%, respectively). However, the percent of non-stained follicles was not signi cantly different (p > 0.05) in both treatment groups (12.3 ± 5.5% of Group 1 vs 13.1 ± 3.3% of Group 2, respectively).
Calcein AM and ethidium homodimer-1 A total of 317 follicle were stained with Calcein AM in combination with ethidium homodimer-1. The viability of follicles with commonly used uorescence dye Calcein AM in combination with ethidium homodimer-1 presented on Figure 4B.
The data have shown that the percent of follicles of class V1 in Group 1 was for 9.6 signi cantly higher (p < 0.05) than in Group 2 (64.8 ± 5.2% vs 55.2 ± 6. 2%, respectively).
The percent of class V3 follicles in Group 2 was signi cantly higher (p < 0.05) than in Group 1 (14.5 ± 5.9% vs 9.7 ± 4.7%, respectively). However, the percent of class V2 and V4 follicles was not signi cantly different (p > 0.05) in both treatment groups.

ROS assay
To measure the amount of ROS in live cells a total of 20 isolated follicles in each treatment group were stained with CellROX ® . The data presented on Figure 4C have shown that generally the oxidative stress was very low in both treatment groups, but nevertheless the percent of oxidative level was signi cantly lower (p < 0.01) in Group 1 than in Group 2 (0.2 ± 0.2% vs 2.1 ± 1.9%, respectively).

Morphology assessment
The evaluation of follicles' morphology using uorescence dyes anti-α-Tubulin antibody in combination with DAPI presented on Figure 5.
In vitro culture of follicles in 3D system using alginate scaffold However, we have noted that the velocity of follicles growth (Fig. 6D) from Day 0 to Day 8 in Group 1 was signi cantly higher (p < 0.05) than in Group 2 (29.1 ± 9.2% vs 22.3 ± 8.6%, respectively). The data of viability of follicles on Day 8 after staining with Calcein AM and ethidium homodimer-1 for identifying of live and dead cells presented on Figure 6E. The data shown that the percent of follicles of class V1 + V2, which represented of viable follicles, was signi cantly higher (p < 0.05) in treatment Group 1 compare to Group 2 (68.9 ± 5.8% vs 49.5 ± 9.4%, respectively).

Discussion And Conclusion
Our results indicate that a new enzymatic digestion protocol using TDE, a commercial enzyme-cocktail, is initiated for human follicle isolation as the rst step to bioengineer a transplantable arti cial ovary. The e ciency, safety and productivity of the proposed protocol can be a suitable method for isolating plenty of intact and viable follicles with good growth potential. Our goal is to replace Liberase DH as enzyme digestion on the rst step in constructing arti cial ovaries for female cancer patients preventing disease recurrence and clinical application.
The ovarian tissue freezing and thawing protocol in our previous studies have shown promising results in viability and developmental and have a live birth after auto-seeding of cryopreserved ovarian tissue [22,23]. Therefore, we are con dent to use cryopreserved tissue. To assess enzyme TDE, we compared it with Liberase DH combined with DNase I which is widely used as the gold standard for human follicle digestion [18,24,25].

Yield of isolated follicles
In our study, considering the similar size of the digested ovarian cortex, the recovery rate of follicles in Group 1 was signi cantly higher by 78 follicles compare to Group 2 (p < 0.05,  [27]. As the ovarian cortex holds a vastly dense and brous structure, it is crucial using a suitable enzymatic digestion protocol to maximize the yield of isolated follicles from the neighboring tissue. Liberase DH Research Grade contains highly puri ed Collagenase I and Collagenase II blended in a precise ratio to each other and contains a high concentration of Dispase ® (www.Sigma-Aldrich.de). Compared to widely use of Liberase DH, we can assume that TDE not only contains highly puri ed collagenase, and may also include other enzymes: protease, dispase, or DNase can fully digested human ovarian tissues.

Quality of isolated follicles
Quality analysis of the isolated follicles in our project was based on follicle viability assessment by neutral red dyes, vital uorescent dyes, ROS assay, and on the general follicle morphology assessment using uorescence dyes.
Follicle viability assessment with neutral red dye The advantages of neutral red dye are that the dye can be reconstituted in physiological medium, is non-toxic and can be easy to visualized under a standard light microscopy, it is proved to be no long-term deleterious effects on the follicles [28]. Neutral Red (NR) is a vital dye that readily diffuses through the cell membrane and concentrates in lysosomes of viable cells [29] shows a deep red colour (Fig. 1A). It is thus easy to distinguish between viable, damaged or dead cells [30]. Our study presented the percentage of bright red follicle in Group 1 digested from cryopreserved ovarian tissue was found to be signi cantly higher (p < 0.01) than Group 2 (70.9 ± 4.7% vs 63.9 ± 4.2%, respectively), similar to the observations of intense red -stained follicles in the suspension of cryopreserved ovarian tissues made by Schmidt [31]. In contrast to our nding, mechanical isolation technique using a tissue chopper could isolate large numbers of morphologically normal preantral porcine follicles than scalpels no matter changing incubation time or size of the tissue pieces [32].
Viability assessment using Calcein AM and ethidium homodimer-1 While follicles stained with Calcein AM in combination with ethidium homodimer-1, our data have presented that Group 1 had isolated more viable follicles than Group 2 (class V1: 64.8 ± 5.2% vs 55.2 ± 6. 2%, respectively). Liberase DH Research Grade contains highly puri ed Collagenase I and Collagenase II, endotoxin is one of the detrimental components of collagenase extract, maybe reduce the vitality of follicles [33]. Indeed, for Liberase DH/DNase treatment, the percentage of class V1 follicles in our study was in accordance with previous reports: 65% (digested from cryopreserved ovarian cortex) [26] and 61% (digested from fresh ovarian cortex) [11]. Other observations shown by Dolmans et al. [11], observed other observations that early atresia were frequently found in follicles isolated by Liberase. This may infer that Liberase may alter the viability of isolated follicles by inducing apoptosis [34].
Viability assessment by ROS assay Reactive oxygen species make a vital part as mediators of various intracellular signaling pathways [35] regulating apoptosis and necrosis [36]. Oxidative stress is the result of an imbalance in the production and elimination of the intracellular reactive oxygen species (ROS). However, a high level of oxidative can damaging follicle [37] and affecting their in vitro culture [38]. A high concentration of oxidative may also gap mitochondrial membrane pores, emit cytochrome and trigger caspases, with following apoptosis [39] and affect oocyte maturation [40]. In our result, the percent of oxidative level was signi cantly lower (p < 0.01) in Group 1 than in Group 2 (0.2 ± 0.2% vs 2.1 ± 1.9%, respectively). Collectively, our results suggest that reactive oxygen species generation were triggered during Liberase isolation procedures. Tirmenstein et al. found that Liberase, contains highly puri ed Collagenase can induces oxidative stress after isolation of hepatocytes, leading to a loss of cytochrome enzyme activity [41]. Liberase, which contains very low levels of endotoxin, still produced nitrite. Nitric oxide (NO) which is produced by nitric oxide synthase (NOS) can react with superoxide anion to form the reactive oxidant species [42]. Similar observations were made by Forstermann et al. [43] that endotoxin has been shown to induce NOS in cells leading to the increase of ROS. Therefore, we can deduce that TDE may well preserve ROS balance and reduce oxidative stress.
Morphology Our result presented that on the percent of morphologically normal follicles of class M1 was signi cantly higher (p < 0.05) in Group 1 compared to Group 2 (72 ± 4.2% vs 65.7 ± 5.1%, respectively). It was indicated that using TDE digestion can well preserve the morphology and granulosa cell layer of follicle. Laminin and collagen IV which are essential components of the basement membrane of human primordial follicles [44]. Liberase DH presents a high concentration of dispase, although it does not cleave laminin, but it can cleave bronectin and collagen IV [45], and damage the junction and membrane of the follicle, that would probably explain our result of damaged follicles or extruded oocytes in Liberase DH/DNase group. This is in accordance with previous investigations were made by Dolmans et al. [11] and Oktay et al. [46] suggest that enzymatic isolation using Liberase may induce slight morphofunctional changes in the follicular cell compartment, such as alteration in shape or extensive disruption to the follicular wall.

Culture of isolated follicles
Culture of isolated preantral follicles has several potential applications, including in vitro culture and grafting or bioengineering a transplantable arti cial ovary or for molecular biology analyses. Therefore, ecsuring good viability, preserving well structure and maintaining follicular development is essential before attempting such approaches in humans. Observations in preliminary studies reported that follicles isolation using Liberase were degenerated in vitro culture [8]. Thus, to examine the effect of different types of enzymic digestion on the follicle, we isolated follicles from cryopreserved ovarian cortex embedded in 3D alginate scaffold for in vitro culture 8 days.
After 8 days of in vitro culture in our study, follicles isolated from both TDE and Liberase DH/DNase treatments can maintain good viability and preserved well 3D structure, and an increase in size in both groups. Our results are similar to those follicles obtained by Yin et al. [47] with 7 days in vitro culture. Such an increase in follicle diameter after IVC may be due to some products generated during ovarian tissue cryopreservation: hypoxia, increasing of intracellular Ca 2+ , osmotic disruption of cellular membranes, generation of reactive oxygen species (ROS) and lipid peroxidation. These products generate by cryopreserved or ovarian cortex chopped by a tissue sectioner and follicles isolated by enzyme may interrupt ovarian Hippo signaling and activate follicle in vitro culture [48]. Besides, our result noted a greater growth on the velocity of follicles from Day 0 to Day 8 in Group 1 was signi cantly higher (p < 0.05) than in Group 2 (29.1 ± 9.2% vs 22.3 ± 8.6%, respectively). The data of viability of follicles on Day 8 shown that the percent viable follicles, was signi cantly higher (p < 0.05) in treatment Group 1 compare to Group 2 (68.9 ± 5.8% vs 49.5 ± 9.4%, respectively). Liberase, which contains very low levels of endotoxin, still produced nitrite [42], can induce oxidative stress in follicle after enzymatic isolation [41], induce apoptosis in vitro culture [33], strongly delay and impair grafting outcomes [34]. These probably re ecting Liberase may weaken the initial health status of isolated follicles or a lack of factors essential to sustaining their survival and growth [49]. Based on these results presented after IVC, we can assume that TDE can well preserve follicles during digestion better than Liberase without impairing their ability to survive and grow in vitro.
To our knowledge, this study was the rst to provides insight into the relationships between morphology, viability, cell intact, oxidative stress and diameter of growing preantral follicles between TDE and Liberase DH digestion. Gap junction formation based on the ability of the oocyte to communicate with granulosa cells and signal for cell division between oocytes and granulosa cells may explain the differences in growth rate in both groups. Our nding supports the hypothesis that using TDE digestion can well preserve granulosa cell-oocyte contact which is important for cell proliferation in early stages. Our study had several limitations: First it did not evaluate the potential of isolated follicles contamination by cancer cell; Second, it did not examine the RNA expression levels of each oocyte and granulosa cell from a single follicle. Additional studies are needed to con rm the isolated follicle with the risk of cancer cell reseeding by enzymatic digestion and isolated human ovarian follicles' RNA levels.

Conclusion
In conclusion, this study using tumor dissociation enzyme (TDE) to isolated follicles can allow us to obtain a high number of viable follicles from the human ovarian cortex in contrast with the earlier described Liberase DH method, which is the initial and crucial step to construct an arti cial ovary for future clinical applications. With our isolation technique, will enable a foundation as research is increasingly focusing on preantral folliculogenesis.

Materials And Methods
All chemicals were purchased from Sigma (Sigma-Aldrich Chemie, GmbH, Schnelldorf, Germany) except where otherwise stated.

Collection, dissection and distribution into treatment groups
The presented study was approved by the Ethics Board of the Cologne University, Germany. Informed consent was given by 10 patients aged 18 to 35 years (23.8 ± 4.66) years. These patients were undergoing laparoscopic surgery within the FertiProtect program at the Department for Gynecology and Obstetrics of the University Hospital of Cologne.
All ovarian tissue used for this research were cryopreserved and stored at the cryo-bank of the Maternal Hospital of Cologne University.
According to protocol approved in our department, 10% of ovarian tissue collected from patients was used for 'patient-oriented' to assess the tissues' viability for re-transplantation.
The collection and transport of ovarian tissue were performed as follow: immediately after surgical recovery the ovarian tissue was located in a sterile 100 ml specimen container (Sarstedt, Nümbrecht, Germany) lled with 50 ml of a basal medium composed from Leibovitz L-15 supplemented with 5% Dextran Serum Substitute (Irvine Scienti c, Santa Ana, USA). A closed specimen container was transported within 3 minutes at room temperature to the laboratory. The dissection of the ovarian cortex from the medulla was done using sterile surgical tweezers and scalpels no. 22 at room temperature. A small part of the ovarian cortex piece (OCP, ~ 1 mm 3 ) was xed in Bouin's solution for later evaluation of quality and density of follicles. Then the medulla part of ovarian tissue pieces was immediately removed using sterile surgical tweezers with scalpel no. 22, and 10% of ovarian tissue was cut for patient-oriented research. The pieces were cooled to 5°C for 24 h in 50 ml of Leibovitz L-15 with 5% Dextran Serum Substitute before the freezing procedure was performed. The size of OCPs for patient-oriented research in this study depended on the amount of surgically recovered ovarian tissue and ranged from 37 to 364 mm 3 (Table 1). After this, the piece was frozen and stored in liquid nitrogen vapor until use.
Ovarian tissue freezing and thawing The cryopreservation of OCPs was performed according to our protocol as previously published [23]. On freezing day, the OCPs were placed for 30 min at room temperature in a 20-ml freezing medium composed of basal medium supplemented with 6% dimethyl sulfoxide, 6% ethylene glycol, and 0.15-M sucrose. After this, the OCPs were located into a standard 5-ml cryo-vials (Thermo Fisher Scienti c, Rochester, NY, USA) previously lled with 4.5 ml freezing medium and frozen using IceCube 14S programmable freezer (Sy-Lab, Neupurkersdorf, Austria). The slow cooling pro le started at − 6 °C, and the samples were then cooled from − 6 °C to − 34 °C at a rate of 0.3 °C/min. At − 34 °C, the cryo-vials were nally plunged into liquid nitrogen and stored until thawing. The freezing protocol for cryopreservation of ovarian tissue included an auto-seeding step at − 6 °C.
The thawing of frozen probes was performed as follow: the removed from liquid nitrogen cryo-vial was hold at room temperature in air for 30 seconds with subsequent, direct immersion in boiling water bath (100°C) for 60 seconds. The exposure time in the boiling water was visually controlled by ice apex in the medium. As soon as the ice apex became 2 to 1 mm thick, the cryo-vial was removed from the boiling water. At this point, the nal temperature of the medium was between 4°C and 10°C. The thawed OCPs were immediately transferred from the cryo-vials to a 10 ml thawing solution (basal medium containing 0.5-M sucrose) in a 100 ml specimen container (Sarstedt, Nümbrecht, Germany). For dilution of cryoprotectants from thawed tissue, the stepwise rehydration technique was applied due to drop by drop addition of basal medium to the solution of sucrose with OCPs. This rehydration step was performed in combination with continuous shaking of specimen containers with OCPs with 200 OSC/min during 15 minutes at room temperature. During these rehydration time, the 50-ml of a basal medium was dropped into the specimen container. The nal concentration of sucrose was 0.083-M, resulting in almost isotonic conditions. After the rehydration time was over, the three wash steps in a fresh basal medium for 10 minutes was performed.

Isolation and collection of follicles
After warming, the OCPs were cut into two equal parts. Each part of OCPs was weighted using an analytical balance (Ohaus EP64C Explorer Pro Analytical Balance, Ohaus Corporation, USA) and the volume of the piece was measured (Table 1). A total of 10 OCPs in each treatment group were evaluated.
Our research compared two enzyme-cocktails' in uence on viability and ability to in vitro growth in a 3dimensional system (3D) of preantral follicles recovered from the cryopreserved ovarian cortex.
The tested enzyme-cocktails were: 1. "Tumor Dissociation Enzyme Reagent" (TDE) from the DCS ATP-Chemo-sensitivity Assay Kit (DCS Innovative Diagnostik-Systeme, Dr. Christian Sartori Labor, Hamburg, Germany) which composed of various types of highly puri ed collagenase (personal communication with Dr. Christian Sartori Labor, Hamburg, Germany) and is a commercial enzyme-cocktail for enzymatic digestion of any solid tumors [20,21,50].
For this research, 10 frozen OCPs immediately after warming were cut into two equal parts and randomly distributed in two treatment groups as follows: Group 1 -10 OCPs for the enzymatic digestion with TDE enzyme-cocktail and Group 2 -10 OCPs for the enzymatic digestion with Liberase DH/DNase enzymecocktail.
The procedure of follicle isolation was performed as described previously [18,19] with some modi cation. Brie y, the OCPs were mechanically cut into ~ 0.5 × 1 to 1 × 1 mm 3 pieces by quick simultaneous movements of two scalpels no. 22 on a sterile 100 mm Falcon Petri dish (Falcon, Corning Incorporated-Life Science, Corning, USA) under a stable temperature cooling on ice to + 4 °C. Then the minced ovarian cortical tissue was suspended in basal medium containing 16% of TDE-enzymatic solution (Group 1) or in basal medium containing 0.28 Wünsch units/mL Liberase DH in basal medium plus 10μg/ml DNase I (Group 2). The 50μg/ml of neutral red (NR) dye was added to digesting each treatment group's solutions for quick and roughly visualization of isolated follicles [29]. This vitality dye shows a deep red color in acid cell structures and is normally used for the express estimation of viable follicles (Fig. 1A). The incubation time for both treatment groups was 75 minutes at 37°C on KS 260 basic shaker (IKA, Staufen, Germany) with 120 rotations/minute with additional every 15 minutes of gentle pipetting with a 1 ml pipette. Enzymatic digestion was terminated by adding an equal volume of 4°C cold Leibovitz L-15 medium supplemented with 20% fetal calf serum.
The follicles were collected with 125 μm V-denuded capillaries (Vitromed GmbH, Jena, Germany) one by one (no more than 45min) by two experienced operators under a light inverted stereomicroscope Nikon SMZ1270 (Nikon, Düsseldorf, Germany) and washed three times in 4 °C basal medium. The follicle diameter was measured by the basement membrane surrounding the granulosa cell layer(s) as the follicular boundary under Program Zen (Nikon, Düsseldorf, Germany), two perpendicular measurements were made at right angles to each other and the mean diameter was calculated. The preantral stages of follicles were classi ed according to Gougeon and Fortune [51,52] as follow: primordial follicle (< 60 μm), oocytes surrounded by a single layer of attened pre-granulosa cells; primary follicle (> 60 μm to ≤75 μm), oocytes with a single layer of cuboidal granulosa cells; secondary follicle (> 75 μm to < 200 μm), with at least two complete layers of granulosa cells.
Follicle viability assessment with uorescence dye Calcein AM and ethidium homodimer-1 The stained isolated with neutral red follicles were subjected to uorescent staining with Calcein AM and ethidium homodimer-1 [53]: Brie y: the follicles were exposed to 2 μM of Calcein AM and 5 μM of ethidium homodimer-1 in Dulbecco's phosphate buffer saline (DPBS) and incubated for 45 min at 37 °C in the dark. Non-uorescent cell-permeant Calcein AM enters the cell and cleaved by esterase in living cell giving an intense green uorescence (excitation/emission, 495 nm/515 nm). Ethidium homodimer-1 enters cells with damaged membranes binding to nucleic acids of dead cells and shows a red uorescence (excitation/emission, 528 nm/617 nm). After exposure to uorescence dyes, follicles were washed in DPBS and observed under a confocal microscope (Olympus Fluoview FV 1000, Hamburg, Germany). The viability of follicles was classi ed into four categories depending on the percentage of dead granulosa cells [11]: V1, live follicles with the oocytes and all granulose cells (GCs) viable; V2, minimally damaged follicles with <10% of dead GCs; V3, moderately damaged follicles with 10-50% dead GCs; V4, dead follicles with both the oocyte and/or > 50% GCs dead (Fig. 1B -E).
Viability assessment due to measurement of Oxidative stress level of isolated follicles The isolated follicles were staining with CellROX ® Deep Red uorogenic probe (Life Technologies, New York, USA), which was designed to reliably measure reactive oxygen species (ROS) in live cells. This dye applies specially to detect cellular superoxide anion (O ·−) and hydroxyl radical (OH•) and produces deep red uorescence (excitation/emission, 640 nm/665 nm). The staining procedure was performed according to Korkmaz [54]. Brie y: the follicles were xed in formalin, washed twice with DPBS, incubated in CellROX ® Deep Red Reagent and anti-α-Tubulin antibody for 30 minutes at 37°C in the dark. After incubation, follicles were cover by Fluoromount-G mounting medium containing 4',6-diamidino-2phenylindole (DAPI, Fluoromount-G ™ with DAPI; Thermo Fisher Scienti c), and then observed under a confocal microscope (Olympus Fluoview FV 1000, Hamburg, Germany). To measure the level of oxidative stress, follicles stained by DAPI show the deep blue uorescence of cells' DNA. The incubation of follicles with the anti-α-tubulin antibody produces a green uorescence. The incubation of follicles with CellROX ® Deep Red Reagent shows a deep red uorescence of reactive oxygen species in cells (Fig. 1F). Semiquanti cation of the oxidative level was calculated by using Fiji software (version 1.40; National Institute of Health, Bethesda, MD) and the percentage of oxidative level (%) was calculated in these two digestion groups.
Follicle morphology assessment with uorescence dye The morphological evaluation of follicles were performed according to their morphology and integrity of granulosa cell layer [11] by staining with the anti-α-Tubulin antibody in combination with DAPI. To contrast the cytoskeletal α-tubulin the speci c to the anti-α-tubulin antibody producing green uorescence (excitation/emission, 500 nm/520 nm) was applied. To contrast the nuclei of follicular cells the DAPI was applied. This dye binds to DNA and RNA of nuclei of follicular cells, producing blue uorescence (excitation/emission, 360 nm/460 nm). The uorescent-stain procedure was performed as follow: isolated follicles were xed in formalin, washed twice with DPBS, and then stained with anti-α-Tubulin antibody for 30 minutes at 37°C in the dark. After incubation, follicles were cover by Fluoromount-G mounting medium containing DAPI and observed under a confocal microscope (Olympus Fluoview FV 1000, Hamburg, Germany). According to performed staining the follicles were classi ed into four categories as follow: M1, spherical form with complete granulosa cell layer; M2, irregular form with complete granulosa cell layer; M3, irregular form with <10% granulosa cell loss; and M4, totally atypical form with 10-50% granulosa cell loss (Fig. 2).

Follicle embedding and in vitro culture
Immediately after isolation, only well morphologically preserved follicles stained bright red were collected for embedding in alginate scaffold and in vitro culture (Fig. 1A, white asterisk).
For embedding of isolated follicles in a sodium alginate scaffold, we have performed the technique previously described by the research group of Donnez [55] with some modi cation. Shortly: the isolated neutral red stained follicles were randomly distributed in two groups and transferred into 1.5% (w/v) sodium alginate solution. Then the follicles using 20µl Eppendorf pipette were gently released into a plastic Petri dish containing 0.1M CaCl 2 solution for 2 minutes to cross-link. To wash out the CaCl 2 solution each alginate scaffold contained 5 -10 follicles/bead was three times washed in serum-free culture medium (AIM V medium; Thermo Fisher Scienti c) supplemented with 100 IU/ml penicillin, 100 mg/ml streptomycin, 5% dextran sulfate sodium (DSS), 1% insulin-transferrin-selenium (ITS), 50 μg/ml ascorbic acid, 2mM Glutamax-1, 50 mIU/ml FSH). After washing, each alginate scaffold transferred into 500μl of culture medium/well of 24-well plate (Cell culture plate, Sigma -Aldrich, USA). The culture was performed at 37.5°C under 5% CO 2 for up to 8 days (Fig. 6A). Every second day the half of the culture medium with a fresh one was replaced. Every 2 day the follicles were observed and photographed under a standard light microscope. The follicles' diameter was calculated by averaging two perpendicular measurements from the basement membrane surrounding the granulosa cell layers as the follicular boundary under Program Zen (Nikon, Düsseldorf, Germany).
After 8 days of in vitro culture, all scaffolds with follicles were washed with PBS, then lysed using the alginate lyase buffer (Alginate Lyase, Sigma -Aldrich, Chemie, GmbH, Schnelldorf, Germany). For viability evaluation, all recovered from the sodium alginate scaffold follicles were incubated with Calcein AM and ethidium homodimer-1.

Statistical analysis
For statistical analysis, the confocal images were processed and analyzed using program Fiji, Statistical