**3. Perspectives**

224 Current Frontiers in Cryopreservation

tissue subjected to cryopreservation. Ultrastructure of doe rabbit follicles after cryopreservation was well preserved, but stromal cells and fibroblasts were damaged. Such alterations have been observed in human tissues after cryopreservation (Navarro-Costa et al., 2005; Santos et al., 2010). However, fibroblasts can easily be reproduced by cell division, indicating that damage to the stroma can be repaired. Collagen fibers did not seem to be damaged by cryopreservation in this study, but they were sparse in the doe rabbit ovary. This observation may explain the fragility of doe rabbit ovarian tissue during the

Sugar are not systematically associated with permeating CPA, but Marsella et al. (2008), showed the advantageous effect of sucrose. Trehalose has been frequently used in embryo cryopreservation, but not in ovarian tissue cryopreservation. Sucrose and trehalose share the property to stabilise cellular membranes and proteins via the formation of hydrogen bonds with polar residues of phospholipidic membrane. This property allows preserving the membrane integrity under anhydrous conditions. Moreover, it modifies the temperature at which the separation of lipid phase occurs during cooling (Crowe et al., 1984; Crowe et al., 1985; Crowe et al., 2001). As compared to other sugars, trehalose seems to have a higher capacity to preserve biomolecules, cellular membrane and cells in a drying or in a freezing state (Crowe et al., 1996; Storey et al., 1998; Sano et al., 1999; Welsh and Herbert, 1999).

Few comparable studies have been reported in the cryopreservation of different mammalian species. Despite encouraging results in the different studied species, and except in the queen, none of evaluated protocols allows preserving the same proportion of normal follicles than before freezing. Most of authors observed a reduction of normal follicles in frozen/thawed ovarian tissue compared with fresh control when using similar freezing protocols in the mouse (Candy et al., 1997), the goat (Rodrigues et al., 2004), the cow (Lucci et al., 2004), and the ewe (Demirci et al., 2002). As for the queen, no morphological difference was observed in human follicles before and after cryopreservation (Hovatta et al., 1996; Fabbri et al., 2006) Newton observed similar proportions of "viable" follicles after freezing when using DMSO, ethylene glycol or PROH and xenografting (Newton et al.,

Live births in the rabbit doe and follicular growths observed in the bitch after grafting of cryopreserved ovarian tissue shows the efficacy of evaluated freezing protocols. Almodin et al*.* obtained live offspring after grafting of small fragments of cryopreserved rabbit ovarian tissue using 1.5 M DMSO and a very slow post seeding freezing rate (Almodin et al., 2004b). In the bitch, results about in vivo growth obtained after cryopreservation are relatively scarce compared to other species. Ishijima et al. (2006) tried to transplant vitrified ovarian tissue (2M DMSO, 3M PROH, 1M acetamide) into immunodeficient mice during 4 weeks and observed signs of growth in the early follicles (primordial-primary follicles). However, they noticed an important follicular loss occurring just after grafting which is in accordance with our results obtained with slow-frozen tissue. The time necessary for setting up of the neovascularization within the grafted tissue seems to be more deleterious for the cells than the cryopreservation technique itself. Except our results obtained on the bitch ovarian tissue cryopreservation no other results using slow freezing of female germ cells was obtained in this species. Furthermore, live birth has not yet been obtained after ovarian tissue cryopreservation, but the difficulties to mastered *in vitro* maturation and fertilization

steps in canines do not contribute to the development of this technique.

equilibration, freezing and thawing steps (Neto et al., 2005).

1996).

In definitive, the use of factorial fractional experimental design approach allowed us to develop suitable cryopreservation protocol in different species, while reducing the number of experiments and increasing the number of parameters evaluated. However it can be noticed in our model species, but also in the literature, that results can be radically different according to the species. Moreover, among the numerous articles published on ovarian tissue cryopreservation heterogeneous results can be observed in the same species after applying roughly or widely different slow-freezing protocols. One of the candidates to explain such disparity in the obtained results is the amount of ice crystals formed during slow freezing. Indeed, the strategy of slow cooling is to decrease cell temperature enough slowly to allow removal of most of the freezable intracellular water before reaching the ice nucleation temperature. The main objective of this method is to avoid intracellular ice crystal formation which is known to be lethal. However ice crystallization still occurs extracellularly with the risk of tissue shrinkage or disorganization of the tissue components.

As ice formation and melting are exothermic and endothermic phenomena respectively, they can be objectified and studied by thermodynamical measures. Among the various physical methods of analysis, Differential Scanning Calorimetry (DSC) is an interesting tool. In fact, DSC gives the opportunity to measure important parameters of a cryopreservation solution under dynamic conditions. A cryopreservation solution can thus be characterized by its thermal properties such as temperatures of phase transitions and quantity of ice crystallized and melted. Two types of DSC are commonly used: power-compensation DSC and heat-flux DSC. Our team chose the first one in order to study cryopreservation solutions with a more fundamental approach than with biological methods.

The power compensation DSC is based on the "zero balance principle" as explained as follow. The sample and a reference are placed in two microfurnaces which are continuously cooled by liquid nitrogen. The temperature of each microfurnace can be, on the one hand, precisely measured by a temperature sensor and, on the other hand, precisely adjusted by a heating resistor. Each microfurnace contains one sensor and one resistor. The principle of the power compensation DSC is to maintain the two microfurnaces under the same temperature regardless of phase transitions or reactions occurring in the sample. Thus, when a phase transition occurs in the sample, the heat released or absorbed by the sample has to be compensated by the heating resistor which is below the sample. Consequently, the calorimeter measures a difference between the heating powers provided by the two resistors. This difference reveals the phase transition. When this phase transition is crystallization, this difference allows also us to measure the quantity of ice formed.

New Approaches of Ovarian Tissue Cryopreservation from Domestic Animal Species 227

tissue but also in embryo and oocyte vitrification place the vitrification as a gold standard method for germ cells cryopreservation (Saragusty and Arav, 2011). It has been suggested that with time, conventional slow freezing will be replaced entirely by vitrification

As mentioned above, DSC can be a precious aid for the study of vitrification. In fact, DSC allows the detection of phase transitions, both crystallization or glass transition, and can thus determine three thermal properties of vitrification solutions: the vitreous transition temperature, the critical cooling rate and the critical warming rate (Baudot et al., 2007). These three thermal properties are specific to each vitrification solution and could be used for a better utilization of the solutions. The critical cooling rate and the critical warming rate are the cooling and warming speeds above which the crystallization cannot occur between the vitreous transition temperature and the crystallization temperature. The calculation of these critical cooling and warming rates is based on a semi-empiric model developed by Boutron in 1986 according to the "classic" theory of crystallization (Boutron, 1986). This model reproduces well analytically the experimental results, but some approximations are

As described in the introductive part, extracellular crystallization occurs during slow freezing. The penetrating CPAs like DMSO, PROH or ethylene glycol, are mainly used to reduce the volume of freezable ice within the cells, but their respective toxicity for the cells limit the potential concentration usable. However, since the last decades, nonpenetrating CPAs like disaccharides have been widely used in cryoprotective solution to improve cell preservation and reduce the penetrating CPA amount necessary. To our knowledge, the first use of sucrose as cryoprotectant of ovarian tissue was in 1996 with human ovaries (Hovatta et al., 1996). Sugars have been proved to play a role on ice crystal formation and stability (Kuleshova et al., 1999). Previous studies on mouse embryos revealed that sugars increase the homogeny ice transition temperature in freezing solutions. By this way, dehydration of the cells occurs when they are still permeable to water without reaching ice nucleation temperature. Moreover, the trehalose was reported to reduce the size of ice crystals and shorten their elongation during freezing (Sei et al.,

On the basis of the contribution of DSC in other fields of cryopreservation [vitrification of ovary (Baudot et al., 2007), slow-freezing or vitrification of plants (Volk and Walters, 2006; Skyba et al., 2011), slow-freezing of aquatic crustaceans (Issartel et al., 2006)], our team chose a thermodynamic approach to study slow-freezing solutions for ovarian tissue. In fact, DSC allows the measure of the maximal quantity of ice formed in a solution (Qmax). Qmax corresponds to the definition of Boutron : "The heat of solidification are represented by the numbers q of grams of ice whose solidification at 0°C would liberate the same amount of heat as that from 100g of solution on crossing the corresponding peaks. They are close to the real quantities of ice crystallized in % (w/w) of the solution when it is ice which crystallizes. One obtains the heat in calories per 100g of solution by multiplying q by 79.78" (Boutron, 1984). Our team has thus quantified the quantity of ice formed in solutions used for the

cryopreservation of ovarian tissue of different species (table 6).

techniques (Vajta and Kuwayama, 2006).

introduced.

2002).

**3.2 Limit the ice formation** 

Since recent years, several strategies are developed to avoid deleterious effects of the ice crystal formation during cooling, and thermodynamical approaches are more and more associated to these strategies.

Fig. 7. Scheme of power-compensation DSC

The specimen and reference temperatures are controlled independently using separate (identical) ovens. The temperature difference between the sample and reference is maintained to zero by varying the power input to the two furnaces. This energy is then a measure of the enthalpy or hat capacity changes in the test specimen (relative to the reference).
