Preservation of Primordial Follicles from Lions by Slow Freezing and Xenotransplantation of Ovarian Cortex into an Immunodeficient Mouse


Author's address (for correspondence): C Wiedemann, Leibniz Institute for Zoo and Wildlife Research, Berlin, Germany. E-mail:


Assisted reproductive technology (ART) is considered an important tool in the conservation of endangered species, but often the most limiting factor of ART is the availability of mature oocytes. The aim of the present study was to investigate the feasibility of preserving female germ cells from ovaries of female lions (Panthera leo). Good quality cumulus–oocyte complexes (COCs) were isolated and subjected to in vitro maturation (IVM). In addition, ovarian cortex was obtained and cut into pieces for culture and cryopreservation by slow freezing. The survival of ovarian follicles was assessed by histology. Frozen–thawed samples of ovarian cortex samples were xenotransplanted under the skin of ovariectomized immunodeficient mouse for 28 days. Overall, 178 intact COCs were obtained from 13 lions, but only 28.1% were matured in vitro indicating insufficient IVM conditions. In contrast, almost all follicles within the ovarian cortex survived culture when the original sample was from a young healthy lion collected immediately after euthanasia. Within the xenotransplants, the number of primordial follicles decreased after 28 days by 20%, but the relation between primordial and growing follicles changed in favour of follicular growth. Female gamete rescue from valuable felids may be performed by slow freeze cryopreservation of ovarian cortex. Although the IVM protocol for lions is not yet optimized, mature oocytes may be obtained after long-term xenotransplantation and IVM and could potentially represent one way of salvage of endangered felid species in the future.


As many other felid species, lions (Panthera leo) are listed on the IUCN Red list for endangered species mainly due to trophy hunting, habitat destruction and infectious diseases. In addition, infertility and hybrid production complicate captive breeding programmes. For overcoming these breeding problems, assisted reproduction techniques (ART) are employed for conservation (Lermen et al. 2009). Methods for ART consist of gamete collection, artificial insemination, embryo production, embryo transfer and cryopreservation of gametes and gonadal tissue. The rescue of gametes from deceased or castrated captive animals is used to develop protocols for feline species and to establish gamete cryobanks (Johnston et al. 1991; Jewgenow et al. 1997). Cryobanks allow movement of frozen gametes between living animal populations in captivity or the wild independently of time (Wildt 1992).

Mammalian oocytes are obtained from ovaries, but 90% are immature oocytes at the primordial follicular stage. In felid species, as in other mammals, less than 1% of oocytes are fully grown in antral follicles and capable for maturation (Johnston et al. 1991; Jewgenow et al. 1997). Based on the success in other species, the cryopreservation of oocytes within the ovarian cortex in conjunction with (xeno) transplantation was suggested for oocyte rescue in felids (Jewgenow et al. 2011). To date, live birth after cryopreservation and ovarian transplantation has been reported in mice, goats, ewes and macaques. In human medicine, cryopreservation of ovarian tissue has been proposed for cancer patients to preserve fertility, and since 2004, 18 human babies have been born after ovarian re-transplantation (Andersen et al. 2012).

The aim of the present study was to investigate the feasibility of preserving female germ cells from ovaries of lions. All isolated cumulus–oocyte complexes (COCs) were subjected to a standard feline in vitro protocol. In addition, ovarian cortex was obtained and frozen for xenotransplantation.

Materials and Methods

All chemicals were obtained from Sigma-Aldrich (Taufkirchen, Germany) unless otherwise stated.


Ovaries were obtained from ten euthanized, two ovariectomized and one recently deceased African, Angola and Asian lions kept in various European Zoos. Immediately after death or castration, the ovaries were placed in PBS, shipped at 4°C and processed within 5–40 h (Table S1).

Ovarian cortex preparation and oocyte isolation

One ovary was washed and cut into halves in petri dishes (Nunc IVF petri dish; Thermo Fischer Scientific, Braunschweig, Germany) containing PBS. Cumulus–oocyte complexes were collected by slicing from the medullary side to prevent dissecting the ovarian cortex. After removal of the medulla, a thin layer of ovarian cortex (approximately 200 μm thick) was obtained, placed into PBS and dissected with 2-mm biopsy punches (pfm medical ag; Köln, Germany) into similar size pieces. The pieces were allocated randomly for culture or freezing. The second ovary of each lion was sliced from outside for oocyte isolation as described previously (Ringleb et al. 2010).

In vitro maturation

Cumulus–oocyte complexes with a homogenous dark cytoplasm and intact cumulus layers were chosen for in vitro maturation (IVM) according to our protocol for domestic cats (Waurich et al. 2010). The maturity of oocytes was assessed by the presence of the first polar body. Fertilization was also performed as previously published (Ringleb et al. 2010), but the cleavage rate was insufficient (three cleaved oocytes from a total of 50).

Culture of ovarian cortex

Six cortex pieces were transferred into culture flasks (NunclonTM Surface, 25 cm2; Thermo Fisher Scientific, Braunschweig, Germany) with 7 ml McCoys 5a medium (20 mm HEPES, 3 mm glutamine, 0.1 mg/ml gentamicin, 10 ng/ml insulin, 2.5 μg/ml transferrin, 4 ng/ml sodium selenite, 20% FCS) and placed on a shaker at 5% CO2 and 5% O2 at 38.5°C. After 7 days, all pieces were removed and fixed in Bouins solution.

Cryopreservation and thawing

Slow freezing was performed in cryomedium consisting of PBS supplemented with 1.5 m ethylene glycol (Roth, Karlsruhe, Germany), 0.1 m sucrose and 20% FCS using a programmable freezer (Labotect Cryo Unit; LCU, Göttingen, Germany). Ovarian cortex samples were equilibrated in ice-cold cryomedium for 15 min, followed by transfer into cryo straws (0.25 ml, cut into halves, pre-filled with 150 μl cryomedium). The following cryoprotocol was used: start at 1°C, hold for 10 min, cooling with −2°C/min to −4°C, cooling with −0.2°C/min to −8°C, hold for 10 min, cooling with −0.3°C/min to −40°C and finally with −10°C/min to −140°C followed by plunging samples into liquid nitrogen.

For thawing, samples were shaken in water at 37°C for 10 s. The cryomedium was removed by transferring the ovarian pieces to PBS with 0.75 m ethylene glycol and 0.1 m sucrose (5 min), followed by PBS with 0.1 m sucrose (5 min) and PBS (5 min).


One female (NMRI-nu/nu) immunodeficient mouse was obtained from Taconic (Hudson, NY, USA). The animal was kept in a ventilated cage with free access to food and water. The murine part of the study was approved by the Danish Animal Experiments Inspectorate (2009/561-1590). At 7 weeks of age, the immunodeficient mouse was ovariectomized. At 9 weeks of age, the mouse was anaesthetized with isofluran (Baxter, Allerod, Denmark) and two small pockets were created on each side of the back. Nine thawed ovarian pieces were transplanted to the pockets. After 28 days, the mouse was euthanized. The pieces of ovarian tissue were removed and fixed in Bouins solution.

Histology of ovarian cortex pieces

The fixed ovarian pieces were embedded in paraffin by a standard histological procedure. They were placed vertically in paraffin wax, and serial cuttings of 3 μm were performed. Every tenth section was mounted and stained with haematoxylin and eosin (Merck, Darmstadt, Germany).

Follicle measurements

For each cortex piece, the number of consecutive sections needed to find 30 primordial follicles (PF) was noted and each follicle was measured with software cell^d (analysis; Olympus, Hamburg, Germany). Maximum follicular diameter was used to divide follicles: <30 μm degenerated follicles, 30–40 μm PF, 40–50 μm primary follicles and >50 μm secondary follicles. At least two ovarian pieces and 40 follicles per piece were analysed. In addition, the noted number of slices was used to estimate the ‘calculated follicles number’ (CFN) by extrapolation the number of sections (nSections) needed for identification of 30 PF (nPF) to the whole dimension of an ovarian piece:

display math

Statistical analysis

Statistical analysis was performed by using instat3 (GraphPad Software, Inc., California, USA). The mean number (±SD) of follicles per piece was determined. An unpaired t-test was applied for comparisons of primordial rates before and after culture. The follicle distribution before and after culture was analysed by one-way anova. The follicular abundance and size proportion between day 7 culture and day 28 xenotransplantation was compared by Mann–Whitney test. p values < 0.05 were considered significant.


Maturation rate of lion oocytes

From 10 lions, 178 COCs of good quality were isolated for IVM. From three remaining animals, no suitable oocytes were found (lion 7–9; Table S1). The overall maturation rate was 28.1% (50 of 178).

Culture of lion ovarian cortex

Table 1 presents the data on primordial follicle rates and CFN from six lions before and after 1 week in culture. The percentage of PF varied between 43% (lion 1) and 59% (lion 5–6). The CFN for these lions was between 201 and 344. After culture, both parameters did not change. One animal (lion 7) was characterized by a lower percentage of PF (37%) and very small CFN before and after culture (d0: CFN 18 and d7: CFN 35).

Table 1. Calculated follicle number (CFN) and percentage of viable primordial follicles (%PF40) determined in ovarian cortex pieces of six lions before (day 0) and after culture (day 7)
LionDay 0Day 7
143% (120)29349% (200)413
256% (80)34443% (80)224
355% (120)25753% (80)252
5–659% (320)20151% (80)285
737% (17)1833% (31)35

Ovarian cortex tissue from lion 8 and 9 (Table S1) could not be obtained, because of several pathological changes (tumour, cysts). Samples from the other five lions (lion 4, 10–13) are still frozen in liquid nitrogen.


Table 2 represents percentage of different follicle size groups before (day 0) and after culture (day 7) in comparison with the xenotransplanted ovarian tissue. Before and after culture, the follicle population consisted mainly of PF (53%). One-third started to grow (primary and early secondary follicles). This decrease was accompanied by an increase in degenerating follicles from 16% to 24% during culture. Within the frozen–thawed cortex pieces, a different follicular pattern was found after 4 weeks xenotransplantation (Fig. 1). Although the proportion of PF decreased by about one half (from 53% to 32%), the whole number and percentage of growing follicles increased within the xenotransplants. After culture, 27% follicles began to grow, whereas within the xenotransplants, 54% of all follicles were primary or more advanced follicles (p < 0.05). The secondary stage (multiple layers of granulosa cells) reached 10% (2–3 per cortex piece, p < 0.05).

Figure 1.

Histological section with 3 μm thickness of ovarian cortex from lion after 28 days xenotransplantation under the skin of immunodeficient mouse. Shown are primordial (30–40 μm) and primary follicles (40–50 μm)

Table 2. Follicle distribution within the ovarian cortex of lions before and after culture and after 28 days xenotransplantation in an immunodeficient mouse
Follicle groupMean number (%)
Day 0Day 7Day 28
  1. CFN, calculated follicles number; n, number of ovarian pieces were analysed and shown together with the number of all measured follicles. For fresh and cultured samples, only 40 follicles per piece were assessed for their stage, whereas in the xenotransplants, the sizes of all follicles were determined. The follicles were divided by size in shrunken/truncated follicles (<30 μm), PF (30–40 μm), primary follicles (40–50 μm) and secondary follicles (>50 μm).

  2. a

    Nine pieces were analysed together, because during xenotransplantation, they were grown tightly to each other.

  3. b

    p < 0.05, day 7 vs day 28.

n pieces19119a
Measured follicles640440235
CFN274 ± 60294 ± 8449
Shrunken/truncated follicles16%24%4%
Primordial follicles53%49%32%
Primary follicles30%25%54%b
Secondary follicles1%2%10%b


The study demonstrates that it is possible to preserve female germ cells from captive lions that had to be castrated or euthanized. Several strategies appear to function: 1) isolating fully grown oocytes from antral follicles followed by IVM and 2) undertaking cryopreservation of the isolated ovarian cortex. The present results suggest that both strategies may be developed to provide a successful method for saving lions and other endangered felid gametes.

The by far bigger oocyte reservoir is represented by the resting pool of oocytes in PF within the ovarian cortex. Both COCs from antral follicles as well as PF within the ovarian cortex were isolated and processed in this study.

From three lions (lions 7–9), no suitable COCs and no or only a small number of ovarian oocytes could be obtained because of ovarian pathology (Table S1). From the remaining ten lions, a total of 178 good quality COCs were isolated. This number is comparable to other studies in felids (Johnston et al. 1991; Jewgenow et al. 1997). However, the maturation rate (28.1%) was dramatically lower compared with that observed in the domestic cat under equal culture conditions (Waurich et al. 2010). This confirms results from other studies where low maturation rates of 16–23% in lions were described (Johnston et al. 1991; Ringleb et al. 2010). We conclude that in the case of the lion, the standard feline protocol used is not sufficient and requires modifications and optimization.

The mean number of COCs per lion was 17–18. Based on data from the domestic cat, after optimal maturation (Waurich et al. 2010) about 60%, thus 11 of the original 17–18 oocytes might reach the metaphase II stage of maturity. With a fertilization rate of 55% (Ringleb et al., 2010), six of those 11 oocytes would start to cleave. If 25% (Waurich et al. 2010) of them reach the blastocyst stage, in summary only 1–2 embryos per lion are attainable for embryo transfer (ET). In addition, the impact of spermatozoa quality and availability must be included into this calculation. For our study, neither unrelated or closely related semen nor a recipient for ET was available. Therefore, cryopreservation of oocytes and embryos is inevitable, but does not guarantee reproductive success. Taken together, the number of viable embryos from animals dying in captivity will be low. A potentially better approach will be to use all COCs to establish the necessary methods for oocytes cryopreservation and embryo production, as a prerequisite for handling oocytes obtained from ovarian tissue after xenotransplantation (Fassbender et al. 2007).

The successful preservation of viable oocytes within the ovarian cortex is essential to perform xenotransplantation. We showed that our protocol established for domestic cats is also applicable to lions. Culture of ovarian tissue for 1 week ensured the survival of follicles, as shown by CFN and percentage of PF (Table 2). This contrasts results from other species, where culture is used to induce follicular growth and obtain pre-antral follicles (Picton et al. 2008). Xenotransplantation on the other hand showed that a grafting period of 4 weeks resulted in high percentage of growing follicles (10%) in comparison with follicle population of fresh pieces (1–2%).

Cryopreservation of ovarian tissue is a prerequisite if conditions for xenotransplantation are unavailable at the time of tissue collection. In domestic cats, only a few freezing trials have been performed and it is still not clear whether slow freezing or vitrification is preferable. Slow freezing of cat ovarian cortex in conjunction with xenotransplantation was first reported by Bosch et al. (Bosch et al. 2004). Our group (Jewgenow et al. 2011 and this study) and Lima et al. (Lima et al. 2006) have also successfully applied slow freezing, but to date the survival of oocytes within the ovarian tissue was demonstrated by cytological and/or histomorphological parameters only. Recently, Luvoni et al. (Luvoni et al. 2012) demonstrated the integrity of feline ovarian follicles after vitrification, but a final proof of vitrification is still missing.

In other species, grafting of frozen–thawed ovarian tissue into the original ovarian donor has shown renewed ovarian activity (Baird et al. 1999). This is not possible for wild species, and xenotransplantation into immunodeficient recipients has been suggested. The survival of cryopreserved feline ovarian tissue in nude mice has been shown (Bosch et al. 2004), and retrieval of fully mature oocytes after gonadotropin treatment has been achieved (Fassbender et al. 2007).

In the present study, the xenotransplantation period lasted 4 weeks. The analysis of different follicle stages in the xenotransplants not only demonstrated the survival of lion follicles, but also showed clear indication for follicular growth initiation. The immunodeficient mouse was ovariectomized before xenotransplantation and was expected to have high levels of gonadotropins. In non-rodent mammals, oocyte growth from early primordial to antral follicles is suggested to last about 3 months. For the domestic cat, we already could show that follicular development to early antral stage requires more than 4 weeks (Fassbender et al. 2007) with permanent external gonadotropin treatment. The formation of antral follicles was monitored by high-resolution ultrasound, allowing a timely retrieval of fully grown oocytes capable for IVM. For lion ovarian tissue, our data suggest that a longer time period is required to allow antral follicles to develop.

In conclusion, ovaries obtained from genetically valuable lions (or other felid species) may be cryopreserved and used for gamete rescue. In the future, viable oocytes may be obtained from the cryopreserved tissue by xenotransplantation into immunodeficient recipients. It is recommended that isolated COCs should be used to optimize IVM, IVF and/or oocyte cryopreservation protocols. Reliable in vitro and freezing protocols are the essential basis to apply ART for conservation purposes.


We thank the Zoo of Valencia, Berlin, Hannover and Zoologico de Santillana del Mar for providing lion ovaries, Sandra Bernhardt and Sigrid Holz for their technical assistance. We thank Tine Greve for performing the xenotransplantation of feline ovarian tissue. The study was partly funded by BMBF 033L046.

Conflicts of interest

None of the authors have any conflicts of interest to declare.

Author contributions

R Hribal and J Ringleb performed IVM/IVF. MF Bertelsen and K Rasmusen supplied the gonads. CY Andersen and SG Kristensen performed xenotransplantation together with C Wiedemann. C Wiedemann performed the experiments and analyzed the data. C Wiedemann and K Jewgenow contributed to the study design and drafted the paper, with contributions of the other coauthors.