This study used established histological and culture methods to develop a successful cryopreservation protocol for felid species to be used for gamete rescue under Zoo settings.
Very few studies address cryopreservation of the ovarian cortex in cats. Bosch et al.  reported the survival of cat ovarian tissue after xenotransplantation into nude mice, whereas Lima et al.  and Luvoni et al.  assessed the post-thawed survival of cat ovarian tissue by histological evaluation of follicular structure. Our group used BrdU-labeling and Hoechst staining for integrity assessment of mechanical isolated preantral follicles to control culture impacts , but it was shown that the procedure of isolation from the ovarian cortex itself can cause a loss of integrity in primordial follicles . We also suggested the application of high-resolution ultrasound to monitor the survival of fresh ovarian xenotransplants in immunodeficient recipients as a prerequisite for oocytes retrieval . The freezing of isolated preantral follicles as well as COCs is not well established yet for felid species. In the present study we focused on preserving ovarian tissue by cryopreservation in conjunction with integrity assessment after culture of frozen-thawed pieces by applying standard histology and a novel aspect of follicle counting (40 follicles per piece) for time saving.
To standardize the histology procedure, similar sized ovarian cortex pieces were obtained using biopsy punches. Within these pieces the follicular number and size distribution was determined during maximal 14 days of in vitro culture. In the domestic cat, the number of preantral follicles per piece was quite variable with mean values of 654 ± 427 (Table 2). Histology of ovarian cortex before culture depicted nicely structured PF (Figure 1, Figure 3) that were not reflective of the dramatic loss of total number that occurred after two days. This change was accompanied by an increased percentage of small (< 30 μm) follicles compared to those of normal size (30 to 40 μm). We suggest that the shrunken follicles were already dead and underwent apoptosis during the culture period. After two days of culture, the overall follicle number dropped down to 419 ± 240, with the percentage of PF reduced by one half. After seven days, about 39% of the overall number were still detectable by histology with 40% intact PF. Therefore, it appears that follicles of different size undergo a similar rate of attrition during culture. The high standard deviation in all follicle groups demonstrates the variation in the number of follicles between ovarian cortex pieces. Culture of ovarian cortex pieces in an agitating system and in addition the use of a large volume of culture medium resulted in healthy follicles after culture, as also described for human tissue .
To determine the overall follicle number in each piece, serial sections were performed with at least 240 cuts. But in summary only every tenth cut was used for follicles measurements to avoid a repeat measurement of the same PF. To reduce this work-load for morphological analysis, the percentage of structural intact PF (30 to 40 μm) was determined in a representative number of follicles (40 follicles per piece, 2 to 6 pieces per data point). As shown for the domestic cat pool and the Amur leopard the percentage of PF determined was only 40 per piece and was not different from the percentage of PF from all follicles (Tables 2, 3). If only 40 follicles per piece were measured (all other specimens) the CFN was extrapolated by considering the number of sections to view 30 primordial follicles. As shown in Tables 2, 3 and Figure 2 this methodological simplification (focus count) still reflects the follicle survival during culture and was successfully used to optimize the culture conditions.
In contrast to reports on other species where the growth activation of PF in vitro occurred spontaneously in whole ovary culture of rodents, in cortical tissue culture of cattle , baboons  and humans , we never observed an initiation of follicular growth from PF to more advanced stages. In some species, follicle activation was initiated by supplementation of gonadotropins like FSH and growth factors in caprine  and human ovarian cortex . In others, the primary to secondary transition might be increased by Testosterone addition such as in sheep cortical tissue in ovo  and in bovine ovarian tissue in vitro . For further investigations we plan to stimulate the follicular growth within the ovarian tissue of feline species by supplementation of gonadotropins or to apply xenotransplantation to circumvent long-term in vitro culture.
In vitro maturation and fertilization in the domestic cat is usually performed by the addition of BSA as the main protein source. Immature cat oocytes do not tolerate the presence of FCS in the culture medium  and small preantral follicles prefer BSA . In contrast to isolated feline oocytes, intraovarian PF might depend on the addition of a more complex protein source like FCS during culture. At least we could show for the ovarian cortex pieces of an Amur leopard, the relative PF number was constant during the first seven days of culture in FCS supplemented medium, whereas BSA supplementation resulted in a significant decrease by day 7. The Amur leopard tissue was used for this experiment because we consider it as a representative species for all felids. Experiment 1 and 2 showed that both feline species behave similarly in culture, and differences in survival between all examined species can be mostly related to age, temperature and duration of transportation (Table 4). Due to the large number of ovarian cortex pieces per ovary (~ 100) obtained for the Amur leopard, it is possible to compare different culture conditions with samples collected under the same conditions. In the case of smaller cats (especially for the domestic cat as model organism) only 10 to 16 pieces per ovary are available. For this reason comparative culture or freezing studies can be performed only on a pool of samples, introducing individual variation into the experiment.
Vitamin C was found to be essential for ovarian tissue culture in the goat . It influenced the maintenance of follicular integrity and promoted follicular activation and growth in combination with FSH and/or growth factors. Also in mice  and cattle  ascorbic acid treatment was involved in follicular growth initiation. For feline immunodeficiency virus-infected cells supplementation of Vitamin C resulted in an inhibition of apoptosis and virus replication . Furthermore, the determination of ill cats showed a high ascorbate concentration in plasma which function as a potential species response to oxidative stress . Our results, however, suggested the detrimental effect of Vitamin C supplementation on cat follicle survival (Figure 2) in the ovarian cortex. Therefore, we omitted this component from our standard post-thawed culture medium.
Based on the results of experiment 1 and 2 we established a standard protocol for post-thawed culture and integrity assessment of intra-ovarian feline follicles which has been applied to ten feline species (Table 4). The culture medium composition was identical to Telfer et al.  supplemented with FCS and omitting Vitamin C. No structural differences are seen in histology on day 0 before and after thawing (Figure 3). Therefore, prolonged culture for at least seven days is necessary to reveal any possible impact of cryopreservation on the survival of PF.
The percentage of PF (30 to 40 μm) in tissue samples on day seven of culture in both, fresh and frozen-thawed, ovarian tissue samples was used to assess the cryopreservation outcome. In addition, the initial proportion of PF before culture was determined to correct for the (possible) impact of culture. Only in domestic cat, Northern Chinese leopard and African Lion ovarian samples a significantly decrease of PF between day 0 and 7 in fresh samples was observed. We indicate that loss, as an effect of tissue culture conditions, are still not sufficient to be used for these three feline species.
In all other examined samples, culture had no negative effect on PF survival as shown by identical percentages of PF within the cortex after one week of culture, independent of the cryopreservation procedure. Although we did not have enough material in the three small cats (Black-footed cat, Oncilla and Rusty-spotted cat) to perform the fresh day 7 control, the percentage of viable PF at day 0 fresh and day 7 after freezing was not different. The sole exception was provided by the Geoffroy’s cat and the Amur leopard where we found a significant increase of PF values caused by the huge variation of existent follicles per ovarian cortex piece after seven days post-thaw.
In our study we applied a slow freezing cryopreservation protocol which is successfully used for human ovarian tissue . For feline ovarian cortex, however, only a few reports are available on cryopreservation protocols, both on slow freezing and vitrification. Bosch et al.  and Lima et al.  used almost identical cooling rates and ethylene glycol as cryoprotective agent as in our study, but they abstained using sugar. In both studies only domestic cat ovarian tissue was used and immediately after thawing 58% and after xenotransplantation only 10% of viable follicles were found. Our data indicated no differences in percentage of follicle numbers before and after cryopreservation. In contrast to Bosch et al.  we didn’t perform xenotransplantation of frozen-thawed tissue, but after one week culture still a high amount of follicles were detectable. Other authors vitrified the tissue  and reported the survival of oocytes after thawing again without performing culture examination. Therefore, it is very difficult to compare the results and conclude about the best freezing approach for cat species.
Although the CFN presents only a rough estimation of follicles within the ovarian cortex, we are convinced that this parameter reflects the ovarian oocyte reserve of a female animal very effectively. It was shown that the CFN was influenced by age and size of felid species. Even more, the results after thawing and 7 days culture indicate that still a high amount of viable oocytes can be preserved after the death of an animal. If the methods for realization of the oocyte pool can be developed for ART in the future, the frozen ovarian tissue of rare and valuable felid species can serve a resource bank to increase genetical diversity.