Human Reproduction, Vol.31, No.2 pp. 427–435, 2016 Advanced Access publication on November 30, 2015 doi:10.1093/humrep/dev299

ORIGINAL ARTICLE Reproductive biology Influence of follicle stage on artificial outcome using fibrin as a matrix

M.C. Chiti1, M.M. Dolmans1,2,*, R. Orellana1, M. Soares1,2, F. Paulini1, J. Donnez3, and C.A. Amorim1 Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 1Poˆle de Recherche en Gyne´cologie, Institut de Recherche Expe´rimentale et Clinique, Universite´ Catholique de Louvain, Avenue Mounier 52, bte. B1.52.02, 1200 Brussels, Belgium 2Gynecology Department, Cliniques Universitaires Saint-Luc, 1200 Brussels, Belgium 3Society for Research into Infertility, Brussels, Belgium

*Correspondence address. Poˆle de Recherche en Gyne´cologie, Institut de Recherche Expe´rimentale et Clinique, Universite´ Catholique de Louvain, Avenue Mounier 52, bte B1.52.02, 1200 Brussels, Belgium. Tel: +32-2-764-5237; Fax: +32-2-764-9507; E-mail: marie-madeleine. [email protected]

Submitted on July 15, 2015; resubmitted on October 19, 2015; accepted on November 6, 2015

study question: Do primordial-primary versus secondary follicles embedded inside a fibrin matrix have different capabilities to survive and grow after isolation and transplantation? summaryanswer: Mouse primordial-primary follicles showed a lower recovery rate than secondary follicles, but both were able to grow. what is known already: Fresh isolated mouse follicles and ovarian stromal cells embedded in a fibrin matrix are capable of surviving and developing after short-term autografting. study design, size, duration: In vivo experimental model using 11 donor Naval Medical Research Institute (NMRI) mice and 11 recipient severe combined immunodeficiency (SCID) mice. Both from all NMRI mice were mechanically disrupted and primordial- primary and secondary follicles were isolated with ovarian stromal cells. They were then encapsulated in a fibrin matrix composed of 12.5 mg/ml of fibrinogen (F12.5) and 1 IU/ml of thrombin (T1) (F12.5/T1), and grafted to the inner part of the peritoneum of SCID mice for 2 and 7 days. participants/materials, setting, methods: This study was conducted at the Gynecology Research Unit, Universite´ Catholique de Louvain. All materials were used to conduct histological (H-E staining) and immunohistochemical (Ki67, TUNEL) analyses. mainresultsandtheroleofchance:Although all grafted fibrin clots were recovered, the follicle recovery rate on day 2 was 16 and 40% for primordial-primary and secondary follicles respectively, while on day 7, it was 6 and 28%. The secondary group showed a significantly higher recovery rate than the primordial-primary group (23%, P-value ,0.001). Follicles found in both groups were viable, as demonstrated by live/dead assays, and no difference was observed in the apoptosis rate between groups, as evidenced by TUNEL. Their growth to further stages was confirmed by Ki67 immunostaining. limitations, reasons for caution: As demonstrated by our results, secondary follicles appear to be more likely to survive and develop than primordial-primary follicles in a fibrin matrix after both periods of grafting. These findings may also be attributed to the specific features of the fibrin matrix, which could benefit larger follicles, but not smaller follicles. wider implications of the findings: This study is essential to understanding possible impairment caused by factors such as the isolation procedure or fibrin matrix composition to the survival and development of different follicle stages. It therefore provides the basis for further investigations with longer periods of grafting. study funding/competing interest(s): This study was supported by grants from the Fonds National de la Recherche Scien- tifique de Belgique (grant Te´le´vie No. 7.4578.14 and 7.4627.13, grant 5/4/150/5 awarded to Marie-Madeleine Dolmans), Fonds Spe´ciaux de Recherche, Fondation St Luc, the Foundation Against Cancer, and the Region Wallone (Convention N86519-OVART) and donations from Mr Pietro Ferrero, Baron Fre`re and Viscount Philippe de Spoelberch. None of the authors have any competing interests to declare.

Key words: artificial ovary / isolated pre-antral follicles / allografting / mouse / fibrin matrix

& The Author 2015. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: [email protected] 428 Chiti et al.

Introduction In recent years, different research teams have been developing the concept of an artificial ovary (Amorim et al., 2009; Shikanov et al., 2009, 2011; Hornick et al., 2012; Vanacker et al., 2012, 2014; Donnez and Dolmans, 2013; Luyckx et al., 2013a, 2014; Laronda et al., 2014; Smith et al., 2014; Rajabzadeh et al., 2015) with the aim of restoring fer- tility in patients who cannot undergo transplantation of frozen-thawed ovarian tissue due to the risk of reintroducing malignant cells (Dolmans et al., 2010; Luyckx et al., 2013b; Donnez and Dolmans, 2015). To this end, a murine model was used to test the feasibility of encapsulating isolated murine pre-antral follicles in different natural polymers. A previ- Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 ous study by our group showed that 20% of murine pre-antral follicles could be detected after isolation, encapsulation in an alginate matrix and autografting for 1 week (Vanacker et al., 2014). Despite these en- Figure 1 Different developmental stages of isolated follicles. Light microscopy pictures of isolated follicles at 10 × magnification: (a) prim- couraging results, alginate was found to have a number of limitations, ordial-primary (PP) follicles; (b) secondary (S) follicles. such as a low degradation rate, poor bioadhesion to encapsulate ovarian stromal cells, and decreased vascularization inside the matrix. To overcome these drawbacks, Luyckx et al. (2013a) developed a Ethical approval fibrin scaffold after testing different combinations of fibrinogen and Guidelines for animal welfare were approved by the Committee on Animal thrombin. Using isolated mouse follicles encapsulated in a fibrin formula- Research of the Universite´ Catholique de Louvain. tion with low concentrations of fibrinogen and thrombin, Luyckx et al. (2014) reported a higher degradation rate of the matrix and a more active process of revascularization inside this scaffold. Probably Ovariectomy procedure because of this, they recovered a greater number of follicles after auto- Eleven female NMRI mice aged 6–25 weeks were used for this study. Animal grafting (32%) (Luyckx et al., 2014). Similarly to Vanacker et al. (2014), housing conditions were previously reported by Vanacker et al. (2012). All they also found a higher proportion of secondary follicles and even the procedures were carried out under laminar flow hoods using sterilized some antral follicles after 1 week of transplantation (Luyckx et al., materials. The ovaries were removed through a dorsal incision in anesthe- 2014). However, it is important to stress that before transplantation, tized mice (ketamine 75 mg/kg; Anesketin, Eurovet, Heusden Zolder, Belgium). Buprenorphine (0.1 mg/kg; Temgesic, Schering Plough, Kenil- Luyckx et al. (2014) observed that more than 50% of the follicles were worth, NJ, USA) was administered for analgesia. Soon after the ovariectomy at the secondary stage, which raised an important question: did primor- procedure, the mice were euthanized by CO2 asphyxiation. The ovaries dial and primary follicles grow to the secondary stage in 7 days, or were were kept in minimum essential medium + GlutamaxTM (MEM, Gibco, these secondary follicles found after grafting already at the secondary Belgium) at 48C until the isolation procedure. stage atthe moment of transplantation? If they were originally primordial- primary follicles, did we encounter faster growth, as suggested by Telfer and Zelinski (2013)? If they were secondary follicles, what happened to Isolation of pre-antral follicles and ovarian the population of primordial-primary follicles? These are very pertinent stromal cells questions if wetakeinto account the factthat the large majority of follicles Fatty tissuesurrounding theovaries wasremoved with surgicalscissors.In order isolated from human ovarian tissue are at the primordial-primary stage to isolate pre-antral follicles and ovarian stromal cells, we applied the protocol (Amorim et al., 2009). To address these crucial points, we investigated described by Vanacker et al. (2014). The ovaries were initially minced into frag- × the impact of the developmental stage of isolated follicles, primordial- ments of 0.5 0.5 mm with a tissue chopper (McIlwain Tissue Chopper, Mickle Laboratory, Guildford, UK) and transferred to 50-ml conical tubes primary or secondary, on their survival and growth after encapsulation containing Dulbecco’s phosphate-buffered saline medium (PBS 1X) without in a fibrin matrix and grafting. calcium and magnesium (Gibco, Life Technologies, Merelbeke, Belgium), sup- plemented with 10% fetal bovine serum (FBS; Sigma-Aldrich, Bornem, Belgium). The suspension was pipetted several times to mechanically disrupt the tissue fragments and then transferredtoPetri dishesand investigated for fol- Materials and Methods licles under a stereomicroscope (Leica, Van Hopplynus Instruments, Brussels, Belgium). Over the course of one hour, isolated follicles were picked up Experimental design by two operators using a 130-mm micropipette (Flexipet, Cook, Limerick, Pre-antral follicles and ovarian stromal cells were isolated from adult Ireland)and divided intotwo different groups, primordial-primary or secondary female Naval Medical Research Institute (NMRI) mouse ovaries. Around (Fig. 1), on the basis of their size. In order to confirm this classification, all the 50 primordial-primary and 50 secondary follicles were separately embedded follicles were measured under an inverted microscope (Leica DMNI) using in fibrin clots, together with 50 000 ovarian stromal cells in each case, and re- the internal ocular micrometer at 10× magnification, and then identified as spectively allografted to right and left pockets specially created in the inner primordial-primary or secondary follicles according to their diameter (Griffin wall of the peritoneum of adult cycling female severe combined immunodefi- et al.,2006). ciency (SCID) mice (n ¼ 2 clots per recipient) for 2 or 7 days. Thereafter, the The remaining ovarian suspension was then successively fil- animals were euthanized and the clots were removed, fixed and processed tered through sterilized 80- and 11-mm nylon net filters (Millipore, Overijse, for histology and immunohistochemistry. Belgium). A volume of 10 ml of filtered solution was analyzed using Trypan Grafting of different follicle stages in a fibrin matrix 429 blue (Sigma-Aldrich), and a Bu¨rker chamber (VWR, Leuven, Belgium) was polymerization, the droplet was incubated at 378C for 45 min. Thereafter, utilized to count cell numbers. Thereafter, the filtered solution was centri- the fibrin clot was gently detached and grafted to a SCID mouse. fuged at 260 g for 5 min and the pellet resuspended in PBS supplemented with 10% FBS to obtain a concentration of 50 000 cells/ml. Allotransplantation to SCID mice To graft primordial-primary and secondary follicles to SCID mice, the animals Follicle viability were anesthetized and treated with analgesic, as described earlier for the Shortly after isolation, a fresh follicle sample was used to evaluate follicle ovariectomy procedure. viability and identify any deleterious effect of the isolation procedure on For the surgical procedure, a ventral midline incision was made and a cir- the quality of pre-antral follicles. For this purpose, fresh isolated follicles cular pocket created on the right inner peritoneal wall with non-absorbable were incubated for 30 min at 378C in a droplet of calcein-AM and ethidium 6/0 Prolene suture (Ethicon, Johnson & Johnson Medical, Livingston, UK) homodimer (Life Technologies, Ghent, Belgium), applying the protocol (Fig. 2). The peritoneal surface was scratched with a scalpel blade to described by Luyckx et al. (2014). Green fluorescence was visualized in live induce . The fibrin clot containing primordial-primary follicles cells (excitation/emission 495/515 nm) and red fluorescence in dead was then gently slid into the pocket, which was closed with a knot (Fig. 2). Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 cells (excitation/emission 495/635 nm) using two different filters. Folli- The same procedure was repeated on the left side of the peritoneum and cles were then classified into four categories as previously described a fibrin clot containing secondary follicles was inserted. In total, 11 mice (Martinez-Madrid et al., 2004), based on oocyte and granulosa cell (GC) via- were allografted: five for 2 days and six for 7 days. Finally, the abdominal bility. V1 and V2 represented follicles with high viability (,10% dead GCs), wall and skin were closed. and V3 and V4, follicles with poor viability (10–100% dead GCs or a dead After surgery, anesthesiawas reversed by injection of atipamezole (1 mg/kg; oocyte). Antisedan, Pfizer, New York, USA). On Day 2 (D2) or Day 7 (D7), the animals

were euthanized by CO2 asphyxiation and the grafts were recovered and fixed Fibrin clot formation in 4% formalin. Tissucol (Baxter Belgium SPRL, Braine-l’Alleud, Belgium) is a two-component fibrin sealant. When combined, the two components, namely sealer protein Analyses after allografting and thrombin, mimic the final stage of the blood coagulation cascade. The After these two periods of grafting to SCID mice, the follicle recovery rate, active ingredient in sealer protein (human) is fibrinogen. Both constituents maturation stage, survival and development were evaluated. were reconstituted and diluted as previously described by Luyckx et al. (2013a). Fibrinogen (100 mg/ml) was reconstituted in a solution containing 3 000 KIU/ml bovine aprotinin, a fibrinolysis inhibitor, at 378C. Thrombin Histological analysis

(4 IU/ml) wasreconstituted in40 mmol/mlcalciumchloride(CaCl2) solution. Following fixation in formalin, the grafts were dehydrated and embedded in The reconstituted fibrinogen was diluted in saline solution (0.9% NaCl) con- paraffin for histological analysis. They were cut into 5-mm serial sections taining 9 g/l NaCl to obtain a final concentration of 12.5 mg/ml, and the and every fourth section was stained with hematoxylin-eosin to first identify thrombin in 40 mmol/l CaCl2 solution to achieve a concentration of 1 IU/ the fibrin clot and then to calculate the number of follicles present in each ml. A droplet of 12.5 ml fibrinogen was placed on a glass Petri dish and group. To detect and count all follicles in grafts, after immunohistochemical around 50 isolated murine ovarian follicles (primordial-primary or secondary) analysis, the remaining intermediate sections were also stained with in 4 ml of medium were added and mixed with the fibrinogen droplet. Fifty hematoxylin-eosin. Follicles characterized by a surrounding basal lamina thousand isolated ovarian stromal cells contained in 1 ml of suspension were and GCs with an oocyte were counted and evaluated blindly by two obser- added to the fibrinogen droplet with the follicles. This droplet, incorporating vers. They could be easily identified in clots due to their scarcity and position. both isolated follicles (primordial-primary or secondary) and ovarian stromal Follicle quality was appraised by the integrity of the basement membrane, cel- cells, was then mixed with 12.5 ml thrombin on a plastic Petri dish. For fibrin lular density, detachment between GCs and between GCs and the oocyte,

Figure 2 Grafting of afibrin clot containing isolated pre-antral follicles and ovarian cells. (a) A circleis stitched with non-absorbable 6/0 Prolene thread on the visceral peritoneal wall. (b) A fibrin clot is slid into the peritoneum. (c) After pulling both ends of the thread, the pocket is closed and the fibrin matrix is captured inside. 430 Chiti et al. presence or absence of pyknotic bodies and condensed chromatin, and the Results aspect of the oocyte (Amorim et al., 2000; Myers et al., 2004). Primordial follicles contained an oocyte surrounded by a partial or complete layer of Follicle and ovarian stromal cell analysis flat GCs, primary follicles had one layer of cuboidal GCs surrounding the oocyte, while secondary follicles showed two or more layers of cuboidal before grafting GCs around the oocyte and no visible antrum, which characterizes the Follicle diameter, population and viability presence of antral follicles (Myers et al., 2004). All isolated follicles were classified as primordial-primary or secondary follicles according to their diameter. Before grafting, mean (+SD) follicle Follicle apoptosis diameters in the primordial-primary and secondary groups were 40.6 + Three slides per graft were analyzed by terminal deoxynucleotidyl 4.4 mm and 90.9 + 9.5 mm respectively (Fig. 1a and b). Compared with transferase-mediated dUTP nick end labeling (TUNEL) to detect DNA the primordial-primary group, follicles from the secondary group fragmentation using the In Situ Cell Death Detection Kit, TMR red (Roche, showed a significantly greater diameter (P-value ,0.001). Vilvoorde, Belgium). The complete protocol was previously reported by Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 Among follicles to be grafted for 2 days, 286 were primordial-primary Vanacker et al. (2012). Slides were scanned with the Mirax Scan apparatus and 273 secondary. More specifically, the primordial-primary pool com- (Zeiss, Go¨ttingen, Germany) and analyzed with Mirax Viewer (Zeiss, prised 18% (52/286) primordial follicles, 71% (203/286) primary folli- Go¨ttingen, Germany) and Image J (National Institutes of Health, USA) soft- ware. TUNEL-positive and -negative GCs were counted for each follicle. cles, and 11% (31/286) secondary follicles, which were encapsulated They were then classified into live or dead categories following criteria into 5 clots in total. In the secondary group, there were no (0/273) defined in the follicle viability section. primordial follicles, 8% were primary (21/273), and 92% (252/273) secondary, and these were also encapsulated into 5 clots. Follicle proliferation Among follicles to be grafted for 7 days, 289 were primordial-primary Ki67 is a nuclear protein associated with cellular proliferation and is always and 291 secondary. More specifically, the primordial-primary pool present during all active phases of the cell cycle (G1,S,G2, M), but absent comprised 36% (104/289) primordial follicles, 58% (168/289) primary from resting cells (G0). follicles, and 6% (17/289) secondary follicles, which were encapsulated Three slides per graft were subjected to Ki67 immunohistochemical stain- into 6 clots in total. In the secondary group, 2% (5/291) of follicles were ing, as previously described by Luyckx et al. (2013c). The primary antibody primordial, 12% (35/291) primary, and 86% (251/291) secondary, and used was rabbit anti-mouse Ki67 (1:100 dilution; Abcam, Cambridge, UK) these were also encapsulated into 6 clots. incubated overnight at 48C. Follicles were classified into two categories: non- Despite apparent differences in stage between follicles assigned to the growing, when all GCs were Ki67-negative, or growing, when at least one GC two grafting periods (D2 or D7), therewasno difference in the numberof was Ki67-positive. transplanted follicles in either of the two follicle populations (primordial- Statistical analysis primary or secondary). Shortly after isolation from donor ovaries, a total of 55 follicles (prim- The t-test was applied to compare average follicle diameter between ordial-primary n ¼ 8, secondary n ¼ 47) were analyzed for viability primordial-primary and secondary follicles before grafting. Two-way (Fig. 3a). The percentage of viable and minimally damaged follicles ANOVA (Graphpad Prism 4 software) was used to analyze follicle viability (V1+V2) was significantly higher than the percentage of moderately before grafting and evaluate the recovery rate and immunohistochemical + + results (Ki67 and TUNEL) for both follicle stages after grafting. Statistical ana- damaged and dead follicles (V3 V4) (72 + 32% V1 V2 versus 28 + + lysis was performed to compare the mean (+SD) percentage between the 32% V3 V4). Moreover, no significant difference was found between primordial-primary and secondary groups. A P-value ,0.05 was considered viable (V1+V2) primordial-primary follicles and viable secondary statistically significant. follicles (Fig. 3b).

Figure 3 Viability testing of isolated mouse ovarian follicles before grafting. (a) A live/dead kit was used: viable granulosa cells (GCs) in fresh follicles were stained with calcein-AM, emitting green light. (b) Viability before grafting: follicles were classified into two categories depending on the percentage of dead GCs, namely follicles with high viability (,10% dead GCs), and follicles with poor viability (10–100% dead GCs or a dead oocyte). The percentage of highly viable primordial-primary (PP) follicles (white column) was not significantly different from the percentage of highly viable secondary (S) follicles (black column). Grafting of different follicle stages in a fibrin matrix 431

Follicle evaluation after grafting observed in the recovery rate in either follicle group (Fig. 4c). At histological Follicle recovery rate analysis, it was possible to detect primordial, primary, secondary and antral follicles. These follicles showed normal GCs, minimal amounts of con- All (22/22) grafted fibrin clots were easily recognizable in the peritoneal densed chromatin in a few GCs, regular oocytes, a linear basement wall of the mice (Fig. 4a) on D2 and D7 and identified by histological membrane, and even the presence of a homogeneous zona pellucida in analysis (Fig. 4b). larger follicles. Follicle distribution on D2 and D7 according to group is Of the 22 recovered fibrin clots, 21 showed the presence of follicles, shown in Fig. 5. On D2, 4% (2/46) of follicles were primordial, 89% (41/ but one clot contained only ovarian stromal cells. After careful analysis 46) primary and 7% (3/46) secondary in the primordial-primary group, of each histological section, the mean (+SD) of the follicle recovery compared with the secondary group, where 6% (5/78) of follicles were rate was 16 + 9% (between 8 and 24%) for the primordial-primary primary and 94% (73/78) secondary. On D7, 12% (2/17) of follicles group and 40 + 11% (between 29 and 56%) for the secondary group were primordial, 82% (14/17) primary and 6% (1/17) secondary in the on D2, and 6 + 7% (between 0 and 20%) for the primordial-primary primordial-primary group, compared with the secondary group, where group and 28 + 20% (between 8 and 59%) for the secondary group Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 3% (2/78) of follicles were primary, 92% (72/78) secondary and 5% on D7. Hence, the total number of follicles still present after 2 days (4/78) at the antral stage. was 46/286 in the primordial-primary group and 107/273 in the sec- ondary group. After 7 days, the total number of follicles found was 17/289 in the primordial-primary group and 78/291 in the secondary Follicle apoptosis group. Follicle apoptosis was evaluated using TUNEL assays. Seventy and 36 fol- Taking into account the effect of follicle size, the secondary group licles were analyzed in D2 and D7 grafts respectively, with 70% of D2 showed a significantly superior recovery rate (23%, P-value ,0.001) com- follicles and 83% of D7 follicles showing high viability (V1+V2). The per- pared with the primordial-primary group. Nevertheless, when we analyzed centage of live follicles found on D2 (V1+V2) was 77% in the primordial- the effect of the grafting period (D2 versus D7), no significant decrease was primary group and 63% in the secondary group. On D7, the proportion of

Figure 4 Allografting of murine ovarian follicles in fibrin clots and follicle recovery rates. (a) Fibrin clots were macroscopically identified in the peritoneal wall of mice, as shown by black arrows. (b) A fibrin clot with follicles and ovarian stromal cells can be identified in a histology section after hematoxylin-eosin staining. (c) Primordial-primary (PP) follicles arerepresentedby thewhite column and secondary (S) follicles by the blackcolumn. On D2 and D7,wefound a higher recovery rate for S follicles than PP follicles (***P-value ,0.001). Similar and not significantly decreasing recovery rates werealso noted in both groups (PP and S) with increased grafting time (D2 versus D7). 432 Chiti et al. Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020

Figure 5 Follicle developmentin murine ovariangrafts. Proportions of primordial, primary,secondaryand antralfollicles areshownbeforeand after 2 and 7 days of grafting in the primordial-primary (PP) and secondary (S) groups.

live follicles was 100% in the primordial-primary group and 92% in the mostly at the primary stage. Initial follicle loss was expected due to secondary group. Thus, no significant difference was found between live ischemia-reperfusion damage that occurs in the first few days following follicle percentages in the two groups (primordial-primary and secondary) transplantation (Nisolle et al., 2000; Van Eyck et al., 2009, 2010). More- after either period of grafting (D2 or D7). over, a decrease in the follicle population was already described in our previous publications (Luyckx et al., 2014; Vanacker et al., 2014). Follicle development However, we did not anticipate the massive drop in the population of Ki67 immunostaining was used to evaluate the proliferative status of GCs primordial and primary follicles as early as D2. Due to the small size of in follicles after both periods of grafting (Fig. 6a and b). The mean (+SD) these follicles, it is possible that some of the structures classified as folli- rate of follicles showing Ki67-positive GCs was 78 + 18% in the cles under the stereomicroscope could actually have been round groups primordial-primary group and 100 + 0% in the secondary group on of cells randomly stuck together, which would partially explain the follicle D2, compared with 100 + 0% in the primordial-primary group and loss encountered after transplantation. To prove this hypothesis, we will 98 + 48% in the secondary group on D7 (Fig. 6c). After 2 days of grafting, need to add a control group (non-grafted fibrin clots containing isolated follicles from the secondary group showed a significantly higher propor- follicles and ovarian cells) to our next experiments. Although one could tion of Ki67-positive GCs than primordial-primary follicles (P-value suggest that such massive loss may be due to large-scale activation and ,0.01), but after7daysof grafting, nosignificant differencewasobserved burnout of the primordial follicle population (Dolmans et al., 2007; in the proportion of growing follicles between the primordial-primary Roness et al., 2013; Gavish et al., 2014), we do not consider that this ad- and secondary groups. However, when comparing D2 and D7, a signifi- equately explains our findings, since we grafted more primary than prim- cantly higher proportion of growing follicles was found on D7 than on D2 ordial follicles and we lost both populations at the same rate. On the (P-value ,0.05) (Fig. 6c). other hand, the lower recovery rate of primordial and primary follicles is consistent with findings reported in the literature on isolation and in vitro culture of small follicles, demonstrating that they are not as resistant Discussion to isolation and in vitro culture as secondary follicles (Liu et al., 2000; Jin The artificial ovary, namely a biocompatible and biodegradable matrix in et al., 2010). According to Lenie et al. (2004), small follicles are very vul- which isolated follicles and ovarian stromal cells can be encapsulated and nerable to damage during the isolation procedure. Even if large numbers transplanted to patients, may be a promising alternative to restore fertility of follicles are viable soon after isolation, as observed in our experiments, in women who cannot undergo transplantation of ovarian tissue after gona- they can rapidly deteriorate and lose their 3D structure, leading to dotoxic treatment due to the high risk of reseeding cancer cells (Amorim reduced contact between GCs, and GCs and oocytes (Picton et al., et al.,2009; Shikanov et al.,2009, 2011; Hornick et al.,2012; Vanacker 2008), which is one of the features of apoptosis and follicle death et al.,2012, 2014; Donnez and Dolmans, 2013; Luyckx et al., 2013a, (Wyllie, 1997; Laronda et al., 2014). An alternative could be applying a 2014; Laronda et al.,2014; Smith et al.,2014; Rajabzadeh et al.,2015). less aggressive isolation procedure, which may allow better preservation In the present study, we were aiming to assess if the developmental of follicles. For instance, in order tooptimize the qualityofisolatedhuman stage could have an impact on the recovery rate of follicles after grafting. follicles, Dolmans et al. (2006), Kristensen et al. (2011) and Vanacker Our findings revealed that larger follicles had a significantly higher recov- et al. (2011) tested a purified enzyme blend (Liberase) to replace collage- ery rate than smaller follicles after both periods of grafting. Even after nase for isolation of human pre-antral follicles. When Liberase was 7 days of grafting, follicles from the primordial-primary group were used, greater numbers of viable follicles were found with unaltered Grafting of different follicle stages in a fibrin matrix 433 Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020

Figure 6 Percentage of growing follicles analyzed by Ki67 immunostaining. Follicles were considered to be growing when at least one Ki67-positive gran- ulosa cell (GC) was identified by brown staining: (a) Ki67-positive GCs in primordial-primary (PP) follicles; (b) Ki67-positive GCs in secondary (S) follicles. (c) The percentage of growing PP follicles (white column) was significantly different from the percentage of growing S follicles (black column) on D2. The percentage of growing follicles was significantly higher on D7 compared with D2 (**P-value ,0.01; *P-value ,0.05).

morphology and ultrastructure (Dolmans et al., 2006), even after in vitro Anotherplausibleexplanationcouldbealackoffactorsneededforinitial culture (Vanacker et al., 2011; Lierman et al., 2015). However, even if an follicular development. While we do not fully understand early optimal isolation protocol were to be applied, it is important to bear in , some authors (Osborn et al.,1997; Hovatta et al.,1999; mind that the ovary contains a heterogeneous primordial follicle popula- Liu et al.,2000; Abir et al.,2001; Jin et al.,2010; Dath et al.,2011; tion; while a few of them will reach the Graafian stage, some will grow to Laronda et al., 2014; Soares et al.,2015) have suggested that small follicles the dominant stage, but most will die. Unfortunately, since there is a lack require surrounding stromal cells/tissue for their initial developmental of adequate markers of degeneration in primordial follicles (Hirshfield, steps, as they provide support and produce a number of factors involved 1994; De Felici et al., 2008; Kerret al., 2013), wecannot identifyand elim- in follicle activation, growth and vascularization(Parrott and Skinner, 2000; inate those that are programmed to die. The loss of intercellular commu- Nilsson and Skinner, 2004; Dath et al.,2011; Soares et al.,2015). Indeed, nication observed in our study could also have been caused by the poor complete in vitro development of mouse follicles, starting from the primor- stiffness of the fibrin matrix, which was not able to maintain the 3D struc- dial stage and culminating in production of ovulatory follicles and embryos ture of isolated follicles. Indeed, in a similar study using higher concentra- orlive births,has onlybeenachieved whenfollicles are surroundedbytheir tions of fibrinogen (40 mg/ml) and thrombin (50 IU/ml), Smith et al. natural microenvironment during the first steps of activation and growth (2014) recovered around 32 and 14% of isolated primordial follicles (Eppig and O’Brien, 1996; O’Brien et al.,2003; Jin et al.,2010). Only after respectively 3 and 9 days of grafting inside a fibrin matrix. Although after this initial growth can secondary follicles be isolated for further in these authors also observed a decrease in the follicle population, it was vitro development. Although this technique may be considered feasible, not as dramatic as that seen in our study. Since there is a positive correl- there is still a lack of information on the potential risk of altering imprinting ation between fibrin stiffness and fibrinogen and thrombin concentra- patterns in in vitro-cultured oocytes (Anckaert et al.,2013; Telfer and tions (Duong et al., 2009), it is likely that the more rigid fibrin matrix Zelinski, 2013; Anckaert and Fair, 2015), but reports to date are some- described by Smith et al. (2014) could have provided better structural what contradictory. Nevertheless, it has been clearly demonstrated that support for isolated follicles. after a few days of in vitro culture (Eppig and O’Brien, 1996; O’Brien 434 Chiti et al. et al.,2003; Jin et al., 2010) or transplantation (Luyckx et al.,2014; 7.4627.13, grant 5/4/150/5 awarded to Marie-Madeleine Dolmans), Vanacker et al.,2014), primordial-primary follicles are able to reach the Fonds Spe´ciaux de Recherche, Fondation St Luc, the Foundation secondary stage. In our study, the marked difference in the number of fol- Against Cancer, and the Region Wallone (Convention N86519- licles recovered from the primordial-primary group and the secondary OVART) and donations from Mr Pietro Ferrero, Baron Fre`re and group may have been due to the lack of factors needed for initial follicular Viscount Philippe de Spoelberch. development. Knowing the importance of stromal cells (Dath et al.,2011; Soares et al.,2015), we added them to our fibrin matrix. However, it is probable that the number of grafted cells was not sufficient. Indeed, in Conflict of interest the first days after transplantation, they are not in close enough contact None declared. with isolated follicles, and crosstalk between cells and follicles is not there- fore possible. After 7 days of grafting, the recovery rate of follicles from the second- Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 ary group was similar to that reported by Luyckx et al. (2014) and Rajab- References zadeh et al. (2015), also showing their capacity to develop to the antral Abir R, Fisch B, Nitke S, Okon E, Raz A, Ben Rafael Z. Morphological study stage. Combined with our findings on the primordial-primary group, of fully and partially isolated early human follicles. Fertil Steril 2001; we can confirm that the large majority of follicles that survived and devel- 75:141–146. oped in the previous study (Luyckx et al., 2014) were indeed at the sec- Amorim CA, Lucci CM, Rodrigues AP, Carvalho FC, Figueiredo JR, Rondina D, Cecchi R, Giorgetti A, Martini A, Gonc¸alves PB. Quantitative and qualitative ondary stage. Despite significant follicle loss after grafting, our findings analysis of the effectiveness of a mechanical method for the isolation of appear to suggest that the remaining follicles were in good condition preantral follicles from ovine ovaries. Theriogenology 2000;53:1251–1262. and able to resume their development, as demonstrated by the presence Amorim CA, Van Langendonckt A, David A, Dolmans MM, Donnez J. of antral follicles, as well as TUNEL and Ki67 analyses. Survival of human pre-antral follicles after cryopreservation of ovarian In conclusion, our results show that isolated secondary follicles encap- tissue, follicular isolation and in vitro culture in a calcium alginate matrix. sulated in a fibrin matrix with low concentrations of fibrinogen and Hum Reprod 2009;24:92–99. thrombin are able to survive and grow to the antral stage after a short Anckaert E, Fair T. DNA methylation reprogramming during oogenesis and period of grafting. On the other hand, primordial and primary follicles interference by reproductive technologies: studies in mouse and bovine appear to be more sensitive to the isolation procedure and/or grafting models. Reprod Fertil Dev 2015;27:739–754. with this fibrin formulation. Our next goals are to extend the grafting Anckaert E, Sa´nchez F, Billooye K, Smitz J. Dynamics of imprinted DNA period in order to evaluate complete development of secondary follicles methylation and gene transcription for imprinting establishment in mouse oocytes in relation to culture duration variability. Biol Reprod and their ability to produce healthy oocytes, improve our identification 2013;89:130. procedure during isolation of primordial and primary follicles, and Dath C, Dethy A, Van Langendonckt A, Van Eyck AS, Amorim CA, Luyckx V, enhance survival rates of these follicles by optimizing the composition Donnez J, Dolmans MM. Endothelial cells are essential for ovarian stromal of the matrix. tissue restructuring after xenotransplantation of isolated ovarian stromal cells. Hum Reprod 2011;26:1431–1439. Acknowledgements De Felici M, Lobascio AM, Klinger FG. Cell death in fetal oocytes: many players for multiple pathways. Autophagy 2008;4:240–242. The authors thank Patricia Meijers for her collaboration and scientific Dolmans MM, Michaux N, Camboni A, Martinez-Madrid B, Van Langendonckt A, advice on fibrin in the project. Theyalso thank Mira Hryniuk for reviewing NottolaSA,DonnezJ.EvaluationofLiberase, a purified enzyme blend, for the the English language of the manuscript, Dolores Gonzales and Olivier isolation of human primordial and primary ovarian follicles. Hum Reprod 2006; Van Kerk for their technical assistance, Guillame Courtoy for his essential 21:413–420. Dolmans MM, Martinez-Madrid B, Gadisseux E, Guiot Y, Yuan WY, Torre A, role creating high resolution pictures and Jessica Azevedo for her Camboni A, Van Langendonckt A, Donnez J. Short-term transplantation of precious help checking results. isolated human ovarian follicles and cortical tissue into nude mice. Reproduction 2007;132:253–262. Authors’ roles Dolmans MM, Marinescu C, Saussoy P, Van Langendonckt A, Amorim C, Donnez J. Reimplantation of cryopreserved ovarian tissue from patients M.C.C.: Study design, experimental procedures, analyses, interpretation with acute lymphoblastic is potentially unsafe. Blood 2010; of results and manuscript preparation. R.O.: Statistical analysis, participa- 116:2908–2914. tion in manuscript preparation. M.S.: Experimental procedures, partici- Donnez J, Dolmans MM. in women. Nat Rev Endocrinol pation in manuscript preparation. F.P.: Experimental procedures, 2013;9:735–749. participation in manuscript preparation. J.D.: Manuscript preparation Donnez J, Dolmans MM. Ovarian tissue freezing: current status. Curr Opin and revision. M.M.D.: Manuscript preparation and revision. C.A.A.: Ex- Obstet Gynecol 2015;27:222–230. Duong H, Wu B, Tawil B. Modulation of 3D fibrin matrix stiffness by intrinsic perimental design, experimental procedures, interpretation of results fibrinogen-thrombin compositions and by extrinsic cellular activity. Tissue and manuscript revision. Eng Part A 2009;15:1865–1876. Eppig JJ, O’Brien MJ. Development in vitro of mouse oocytes from primordial Funding follicles. Biol Reprod 1996;54:197–207. Gavish Z, Peer G, Roness H, Cohen Y, Meirow D. Follicle activation and This study was supported by grants from the Fonds National de la ‘burn-out’ contribute to post-transplantation follicle loss in ovarian Recherche Scientifique de Belgique (grant Te´le´vie No. 7.4578.14 and tissue grafts: the effect of graft thickness. Hum Reprod 2014;29:989–996. Grafting of different follicle stages in a fibrin matrix 435

Griffin J, Emery BR, Huang I, Peterson CM, Carrell DT. Comparative analysis Nilsson EE, Skinner MK. Kit ligand and basic fibroblast of follicle morphology and oocyte diameter in four mammalian species interactions in the induction of ovarian primordial to primary follicle (mouse, hamster, pig, and human). J Exp Clin Assist Reprod 2006;3:2. transition. Mol Cell Endocrinol 2004;214:19–25. Hirshfield AN. Relationship between the supply of primordial follicles and the Nisolle M, Casanas-Roux F, Qu J, Motta P, Donnez J. Histologic and onset of follicular growth in rats. Biol Reprod 1994;50:421–428. ultrastructural evaluation of fresh and frozen-thawed human ovarian Hornick JE, Duncan FE, Shea LD, Woodruff TK. Isolated primate primordial xenografts in nude mice. Fertil Steril 2000;74:122–129. follicles require a rigid physical environment to survive and grow in vitro. O’Brien MJ, Pendola JK, Eppig JJ. A revised protocol for in vitro development Hum Reprod 2012;27:1801–1810. of mouse oocytes from primordial follicles dramatically improves their Hovatta O, Wright C, Krausz T, Hardy K, Winston RM. Human primordial, developmental competence. Biol Reprod 2003;68:1682–1686. primary and secondary ovarian follicles in long-term culture: effect of Osborn S, Gook D, Stern K, Speirs A. The isolation and culture of human partial isolation. Hum Reprod 1999;14:2519–2524. primordial follicles from fresh ovarian tissue. Hum Reprod 1997;12-1:153. Jin SY, Lei L, Shikanov A, Shea LD, Woodruff TK. A novel two-step strategy Parrott JA, Skinner MK. Kit ligand actions on ovarian stromal cells: effects on for in vitro culture of early-stage ovarian follicles in the mouse. Fertil Steril theca cell recruitment and steroid production. Mol Reprod Dev 2000; Downloaded from https://academic.oup.com/humrep/article/31/2/427/2379968 by guest on 13 December 2020 2010;938:2633–2639. 55:55–64. Kerr JB, Myers M, Anderson RA. The dynamics of the primordial follicle Picton HM, Harris SE, Muruvi W, Chambers EL. The in vitro growth and reserve. Reproduction 2013;146:205–215. maturation of follicles. Reproduction 2008;1366:703–715. Kristensen SG, Rasmussen A, Byskov AG, Andersen CY. Isolation of Rajabzadeh AR, Eimani H, Mohseni Koochesfahani H, Shahvardi AH, Fathi R. pre-antral follicles from human ovarian medulla tissue. Hum Reprod Morphological study of isolated ovarian preantral follicles using fibrin gel 2011;26:157–166. plus platelet lysate after subcutaneous transplantation. Cell J 2015; Laronda MM, Duncan FE, Hornick JE, Xu M, Pahnke JE, Whelan KA, Shea LD, 17:145–152. Woodruff TK. Alginate encapsulation supports the growth and Roness H, GavishZ, Cohen Y, Meirow D. burnout: a universal differentiation of human primordial follicles within ovarian cortical tissue. phenomenon?. Cell Cycle 2013;12:3245–3246. J Assist Reprod Genet 2014;31:1013–1028. Shikanov S, Woo J, Al-Ahmadie H, Katz MH, Zagaja GP, Shalhav AL, Lenie S, Cortvrindt R, Adriaenssens T, Smitz J. A reproducible two-step Zorn KC. Interpenetrating fibrin-alginate matrices for in vitro ovarian culture system for isolated primary mouse ovarian follicles as single follicle development. Biomaterials 2009;30:5476–5485. functional units. Biol Reprod 2004;71:1730–1738. Shikanov A, Xu M, Woodruff TK, Shea LD. A method for ovarian follicle Lierman S, Tilleman K, Cornelissen M, De Vos WH, Weyers S, T’Sjoen G, encapsulation and culture in a proteolytically degradable 3 dimensional Cuvelier CA, De Sutter P. Follicles of various maturation stages react system. J Vis Exp 2011;49:e2659. differently to enzymatic isolation: a comparison of different isolation Smith RM, Shikanov A, Kniazeva E, Ramadurai D, Woodruff TK, Shea LD. protocols. Reprod Biomed Online 2015;30:181–190. Fibrin-mediated delivery of an ovarian follicle pool in a mouse model of Liu J, Van Der Elst J, Van Den Broecke R, Dumortier F, Dhont M. Maturation infertility. Tissue Eng Part A 2014;20:3021–3030. of mouse primordial follicles by combination of grafting and in vitro culture. Soares M, Sahrari K, Chiti MC, Amorim CA, Ambroise J, Donnez J, Biol Reprod 2000;62:1218–1223. Dolmans MM. The best source of isolated stromal cells for the artificial Luyckx V, Dolmans MM, Vanacker J, Scalercio SR, Donnez J, Amorim CA. ovary: medulla or cortex, cryopreserved or fresh? Hum Reprod 2015; First step in developing a 3D biodegradable scaffold for an artificial 30:1589–1598. ovary. J Ovarian Res 2013a;6:83. Telfer EE, Zelinski MB. Ovarian follicle culture: advances and challenges for LuyckxV, Durant JF, Camboni A,Gilliaux S, AmorimCA,VanLangendonckt A, human and nonhuman primates. Fertil Steril 2013;99:1523–1533. Irenge LM, Gala JL, Donnez J, Dolmans MM. Is transplantation of Van Eyck AS, Jordan BF, Gallez B, Heilier JF, Van Langendonckt A, Donnez J. cryopreserved ovarian tissue from patients with advanced-stage breast Electron paramagnetic resonance as a tool to evaluate human safe? A pilot study. JAssistReprodGenet2013b;30:1289–1299. tissue reoxygenation after xenografting. Fertil Steril 2009;92:374–381. LuyckxV, Scalercio S, JadoulP,AmorimCA,SoaresM, DonnezJ, Dolmans MM. Van Eyck AS, Bouzin C, Feron O, Romeu L, Van Langendonckt A, Donnez J, Evaluation of cryopreserved ovarian tissue from prepubertal patients after Dolmans MM. Both hostand graftvesselscontribute torevascularization of long-term xenografting and exogenous stimulation. Fertil Steril 2013c; xenografted human ovarian tissue in a murine model. Fertil Steril 2010; 100:1350–1357. 93:1676–1685. Luyckx V, Dolmans MM, Vanacker J, Legat C, Fortun˜o Moya C, Donnez J, Vanacker J, Camboni A, Dath C, Van Langendonckt A, Dolmans MM, Amorim CA. A new step toward the artificial ovary: survival and Donnez J, Amorim CA. Enzymatic isolation of human primordial and proliferation of isolated murine follicles after autologous transplantation primary ovarian follicles with Liberase DH: protocol for application in a in a fibrin scaffold. Fertil Steril 2014;101:1149–1156. clinical setting. Fertil Steril 2011;96:379–383. Martinez-Madrid B, Dolmans MM, Van Langendonckt A,DefrereS, Donnez J. Vanacker J, Luyckx V, Dolmans MM, Des Rieux A, Jaeger J, Van Langendonckt A, Freeze-thawing intact human ovary with its vascular pedicle with a passive Donnez J, Amorim CA. Transplantation of an alginate-matrigel matrix cooling device. Fertil Steril 2004;82:1390–1394. containing isolated ovarian cells: first step in developing a biodegradable McLaughlin M, Telfer EE. Oocyte development in bovine primordial follicles is scaffold to transplant isolated preantral follicles and ovarian cells. promoted by activin and FSH within a two-step serum-free culture system. Biomaterials 2012;33:6079–6085. Reproduction 2010;139:971–978. Vanacker J, Dolmans MM, Luyckx V, Donnez J, Amorim CA. First Myers M, Britt KL, Wreford NG, Ebling FJ, Kerr JB. Methods for quantifying transplantation of isolated murine follicles in alginate. Regen Med 2014; follicular numbers within the mouse ovary. Reproduction 2004;127: 9:609–619. 569–580. Wyllie AH. Apoptosis: an overview. Br Med Bull 1997;53:451–465.