**2.4.1 In the doe rabbit**

214 Current Frontiers in Cryopreservation

plasma membrane. In this way, the authors did evaluate the viability of the frozen-thawed follicles using Calcein-AM and Ethidium homodimer I stains (Live/Dead Viability/Cytotoxicity Kit, Molecular Probes) on enzymatically isolated follicles (Fig. 2).

One of the key components of any preservation protocol must be a functional assay that matches the type of cells or tissue being analyzed. In some cases this determination is quite easy. As for example, the sperm motility is a well-accepted functional assay for this system, whereas the capacity to resume folliculogenesis sounds reasonable for the ovarian cortex. Fresh and cryopreserved rabbit doe ovarian tissues were autografted into young females. Fresh ovarian tissue was grafted on the ipsilateral ovarian pedicle immediately after ovary resection (2 grafts per female), contrary to cryopreserved ovarian tissue grafted on the controlateral ovarian pedicle 24 hours after freezing (one graft per female). Control females were ovariectomised, according to the same surgery resection that used before graft. From height week after graft, grafted rabbit doe were inseminated every 3 weeks in case of negative pregnant diagnosis. Eleven months later, ovarian grafts were removed at necropsy

In the bitch, the growth ability of frozen-thawed ovarian tissue has also been assessed by implantation of small pieces of ovarian tissue into adult female SCID mice. After removing mice ovaries, the canine frozen thawed ovarian tissue was placed intramuscularly in the gluteal superficial muscle. To study the setting up of graft resumption, the graft was harvested at one, eight or 16 weeks and the resumption of the ovarian activity was controlled by vaginal cytology assessment. After harvesting, the grafts were processed for

In the bitch, the alpha smooth muscle actin (alpha-SMA), a marker of the mature blood vessel, was used to assess the vessel density within the ovarian tissue, using a primary antibody directed against alpha-SMA. For each slice, image analysis was performed under direct light microscope to determine the tissue areas. The stained vessels were counted in several fields of the same slice and in several slices per animal to deduce a blood vessel

To illustrate the interest of the use of fractional experimental design, the authors decided to present some of their results obtained in the doe rabbit, in the queen, in the bitch and in the cow during the last 6 years. The doe rabbit was used as a model for the human and the animal applications of the ovarian tissue cryopreservation, because of its biological and breeding characteristics. The cat was considered as a model for the ovarian tissue cryopreservation studies of endangered wild felids; from all the felids, the domestic cat is the only non endangered feline species. The cow was used as model for ruminants, with a special interest for preserving high valuable individuals in combination to embryos and to

**2.3.3 In vivo functionality (autografting model) and in vivo growth potential** 

**(xenografting model)** 

histological assessment.

**2.3.4 Vascularization** 

**2.4 Experimental results** 

density.

to observe follicular structures by histology.

The experimental variability expressed as the repetition of one single combination, showed that both the structural and the estimated models of the experimental design were valid when considering the morphological preservation ratio of the follicles. The concentration of the permeating CPA (*P* = 0.67) and the number of equilibration steps (*P* = 0.19) seemed to have no significant effect on the morphological preservation ratio of ovarian follicles. The nature of the permeating and non-permeating CPA seemed to influence the morphological preservation ratio of the follicles (*P* = 0.08 and *P* = 0.07 respectively) although the nonsignificant difference. DMSO tended to reduce the morphological preservation ratio, as compared with PROH. Morphological preservation ratio was increased in the presence of trehalose compared with sucrose. The freezing rate seemed to be the factor that had the greatest impact on the morphological preservation ratio of the doe rabbit follicles. At a freezing rate of 0.3°C/min we observed a significant increasing of the follicular morphological preservation ratio, as compared with 2°C/min (*P*<0.01). No significant interaction was observed between the nature of the permeating CPA and its concentration.

According to the results of the experimental design, the precise evaluation of the best combination of factors influencing positively the morphological preservation ratio (3 steps equilibration protocol, 1.5M DMSO or 1.5M PROH, medium supplemented with either sucrose or trehalose) was performed. Ovarian pieces were treated according to the results obtained with experimental design. Ovarian cortices were equilibrated (3 steps) in the freezing media based on TCM 199 and 10% FCS, at room temperature. The freezing media was supplemented with 1.5 M DMSO or 1.5M PROH and 0.2 M sucrose or 0.2M trehalose. Freezing of ovarian fragments was slowly performed at 0.3°C/min from the temperature of seeding (-7°C/min) up to -35°C. Thawing, histology, viability tests and electron microscopy evaluation process were performed before and after cryopreservation as described previously.

Fig. 1. Rabbit follicular morphology before (A) and after (B) cryopreservation with PROH and trehalose, with a post-seeding freezing rate at 0.3°C/min

New Approaches of Ovarian Tissue Cryopreservation from Domestic Animal Species 217

in the ovarian stroma and the albuginea. Epithelial cells were often absent as compared to

Ultrastructural analysis of the preantral follicles was performed without preliminary selection on semi-thin sections. TEM analysis showed that most follicles of control ovarian tissue had normal ultrastructure, according to mitochondria, nucleus and nuclear membrane, Golgi apparatus and endoplasmic reticulum cisternae observation. They often had vacuoles in cytoplasm and vesicles. Nevertheless, vacuoles were not characteristic of apoptosis. Cellular membranes of the oocyte and follicular cells were in close connection. In general, ovarian stroma was well organised. Fibroblasts and collagen fibres were

After cryopreservation, oocyte ultrastructure appeared to be similar to the control especially for mitochondria, Golgi apparatus, endoplasmic reticulum, interdigital structure between oocyte and follicular cells (Fig.3). Vesicles and vacuoles were rarely observed. Chromatin of the oocyte was diffused and well preserved. Nevertheless, dark follicular cells or follicular cells without any content were most frequently observed, whereas some follicles showed partial or total disruption of their nuclear membrane whatever the evaluated cryprotective solution. The most important damage observed after cryopreservation was the disorganisation of the ovarian stroma (Fig.3). Fibroblasts showed lack of cytoplasm or important vacuolisation. In general, these damages were less frequently observed after

The experimental variability showed that neither the structural, nor the estimated models of the experimental design were valid when considering the morphological preservation ratio of the follicles or the viability preservation ratio. So, global discrimination of the chemo-physical parameters was not possible. Nevertheless, the influence of the freezing rate after seeding and after -40°C, and the influence of the addition of sucrose in the freezing medium composed of 1.5M CPA were evaluated and analyzed by classical

Before freezing, ovarian tissue presented 72.2 3.6% and 83.8 2.9% of normal follicles (type I) for group 2°C/min and 0.5°C/min post-seeding freezing rate respectively. When freezing with PROH, and whatever the post-seeding freezing rate, proportions of morphologically normal follicles were not significantly reduced after freezing compared to before freezing (69.2 9.1% for 2°C/min group vs. 67.4 2.9°C/min for 0.5°C/min group). After freezing with DMSO, and whatever the post-seeding freezing rate, proportions of type I follicles were significantly reduced (40.8 6.6% after freezing at 2°C/min and 51.6 5.1% after freezing at 0.5°C/min; *P*<0.05). Whatever the post-seeding freezing rate, type III defects were the most frequently observed after freezing. General observation of the ovarian tissue showed a good preservation of the ovarian stroma cells and structure after

Before freezing, ovarian tissue submitted to a free fall into the freezing chamber during the third phase of the freezing process presented 72.2 3.6% of type I follicles without any

the fresh ovarian tissue.

distinguishable (Fig. 3).

cryopreservation using PROH and trehalose.

**2.4.2 Investigations in the queen** 

ANOVA test.

cryopreservation.

Fig. 2. View of rabbit isolated follicles under direct light for selection (A) and under fluorescent light (B) after calcein AM/ethidium homodimer I stains to evaluate viability after cryopreservation with PROH, with a post-seeding freezing rate at 0.3°C/min

In control fragments, we observed 72.6 2.8% and 77.7 3.9% of type I follicles (no significant difference) for sucrose and trehalose control groups respectively. After cryopreservation, no statistical difference of the proportions of type I follicles was found between sucrose and trehalose (50.2 4.1% *vs*. 51.1 1.8% respectively) when using DMSO for cryopreservation. When using PROH as permeating CPA, the proportion of type I follicles was lower after cryopreservation with sucrose as compared to trehalose (55.0 3.8% *vs.* 65.0 3.3% respectively; *P*<0.05). When freezing with trehalose the proportion of type I follicles was higher with PROH as compared to DMSO (65.0 3.3% *vs*. 51.1 1.8% respectively; *P*<0.01). Nevertheless, the proportions of type I follicles were significantly reduced after cryopreservation (from *P*≤0.01 to *P*<0.001), whatever the permeating and the non-permeating CPA. No significant difference was observed between the different groups of frozen ovarian cortices, when considering the morphological preservation ratio.

According to these results, the cryopreservation protocol based on a post-seeding freezing rate at 0.3°C/min and using a freezing medium composed of 1.5M PROH, supplemented with 0.2M trehalose was finally evaluated by orthotopic autografting to observe the potential of the cryopreserved follicles to resume follicular growth and to be fertilized.

Before freezing, type II follicles represented the most important part of follicles with morphological defect (19.1 2.9% and 16.1 3.2% in sucrose and trehalose groups respectively). After cryopreservation, follicular defect of type IV (degenerated follicles) was the most important type of morphological defect: 32.5 4.8% and 24.0 1.9% after freezing using DMSO, with sucrose and trehalose respectively; versus 27.2 5.6% and 18.1 3.0% after freezing using PROH, with sucrose and trehalose respectively. The general aspect of ovarian tissue before and after cryopreservation showed a good preservation of structural architecture (follicular structure and connective tissue). Spaces were observed in some case, in the ovarian stroma and the albuginea. Epithelial cells were often absent as compared to the fresh ovarian tissue.

Ultrastructural analysis of the preantral follicles was performed without preliminary selection on semi-thin sections. TEM analysis showed that most follicles of control ovarian tissue had normal ultrastructure, according to mitochondria, nucleus and nuclear membrane, Golgi apparatus and endoplasmic reticulum cisternae observation. They often had vacuoles in cytoplasm and vesicles. Nevertheless, vacuoles were not characteristic of apoptosis. Cellular membranes of the oocyte and follicular cells were in close connection. In general, ovarian stroma was well organised. Fibroblasts and collagen fibres were distinguishable (Fig. 3).

After cryopreservation, oocyte ultrastructure appeared to be similar to the control especially for mitochondria, Golgi apparatus, endoplasmic reticulum, interdigital structure between oocyte and follicular cells (Fig.3). Vesicles and vacuoles were rarely observed. Chromatin of the oocyte was diffused and well preserved. Nevertheless, dark follicular cells or follicular cells without any content were most frequently observed, whereas some follicles showed partial or total disruption of their nuclear membrane whatever the evaluated cryprotective solution. The most important damage observed after cryopreservation was the disorganisation of the ovarian stroma (Fig.3). Fibroblasts showed lack of cytoplasm or important vacuolisation. In general, these damages were less frequently observed after cryopreservation using PROH and trehalose.

#### **2.4.2 Investigations in the queen**

216 Current Frontiers in Cryopreservation

A B

In control fragments, we observed 72.6 2.8% and 77.7 3.9% of type I follicles (no significant difference) for sucrose and trehalose control groups respectively. After cryopreservation, no statistical difference of the proportions of type I follicles was found between sucrose and trehalose (50.2 4.1% *vs*. 51.1 1.8% respectively) when using DMSO for cryopreservation. When using PROH as permeating CPA, the proportion of type I follicles was lower after cryopreservation with sucrose as compared to trehalose (55.0 3.8% *vs.* 65.0 3.3% respectively; *P*<0.05). When freezing with trehalose the proportion of type I follicles was higher with PROH as compared to DMSO (65.0 3.3% *vs*. 51.1 1.8% respectively; *P*<0.01). Nevertheless, the proportions of type I follicles were significantly reduced after cryopreservation (from *P*≤0.01 to *P*<0.001), whatever the permeating and the non-permeating CPA. No significant difference was observed between the different groups of frozen ovarian cortices, when considering the

According to these results, the cryopreservation protocol based on a post-seeding freezing rate at 0.3°C/min and using a freezing medium composed of 1.5M PROH, supplemented with 0.2M trehalose was finally evaluated by orthotopic autografting to observe the potential of the cryopreserved follicles to resume follicular growth and to be fertilized.

Before freezing, type II follicles represented the most important part of follicles with morphological defect (19.1 2.9% and 16.1 3.2% in sucrose and trehalose groups respectively). After cryopreservation, follicular defect of type IV (degenerated follicles) was the most important type of morphological defect: 32.5 4.8% and 24.0 1.9% after freezing using DMSO, with sucrose and trehalose respectively; versus 27.2 5.6% and 18.1 3.0% after freezing using PROH, with sucrose and trehalose respectively. The general aspect of ovarian tissue before and after cryopreservation showed a good preservation of structural architecture (follicular structure and connective tissue). Spaces were observed in some case,

Fig. 2. View of rabbit isolated follicles under direct light for selection (A) and under fluorescent light (B) after calcein AM/ethidium homodimer I stains to evaluate viability after cryopreservation with PROH, with a post-seeding freezing rate at 0.3°C/min

morphological preservation ratio.

The experimental variability showed that neither the structural, nor the estimated models of the experimental design were valid when considering the morphological preservation ratio of the follicles or the viability preservation ratio. So, global discrimination of the chemo-physical parameters was not possible. Nevertheless, the influence of the freezing rate after seeding and after -40°C, and the influence of the addition of sucrose in the freezing medium composed of 1.5M CPA were evaluated and analyzed by classical ANOVA test.

Before freezing, ovarian tissue presented 72.2 3.6% and 83.8 2.9% of normal follicles (type I) for group 2°C/min and 0.5°C/min post-seeding freezing rate respectively. When freezing with PROH, and whatever the post-seeding freezing rate, proportions of morphologically normal follicles were not significantly reduced after freezing compared to before freezing (69.2 9.1% for 2°C/min group vs. 67.4 2.9°C/min for 0.5°C/min group). After freezing with DMSO, and whatever the post-seeding freezing rate, proportions of type I follicles were significantly reduced (40.8 6.6% after freezing at 2°C/min and 51.6 5.1% after freezing at 0.5°C/min; *P*<0.05). Whatever the post-seeding freezing rate, type III defects were the most frequently observed after freezing. General observation of the ovarian tissue showed a good preservation of the ovarian stroma cells and structure after cryopreservation.

Before freezing, ovarian tissue submitted to a free fall into the freezing chamber during the third phase of the freezing process presented 72.2 3.6% of type I follicles without any

New Approaches of Ovarian Tissue Cryopreservation from Domestic Animal Species 219

control. Morphological defect of type III was the most frequently observed after freezing

A B

Fig. 4. Queen follicular morphology before (A) and after (B) cryopreservation with 1.5M

In conclusion, queen ovarian tissue seems to be well preserved, without any difference compared to the fresh control when freezing with PROH (according to a free fall in the freezing chamber, without influence of neither sucrose nor post-seeding freezing rate)

In the bitch, the estimated model was validated when the viability preservation ratio was considered. The nature of the non-permeating CPA (*P* = 0.37) did not influence the post thawing viability rate of the ovarian follicles. However all the other factors investigated in this experimental design presented a significant effect on the viability rate. The permeating CPA nature (*P*<0.0001) was the factor that influenced more the viability rate of the follicles. Thus, contrarily to the observations in the other species, the DMSO better preserved the evaluated parameter than PROH. The freezing rate had also a major effect on the viability rate (*P*<0.0001) and a slow freezing rate (0.3°C/min) was less deleterious than the rapid freezing rate for the follicles viability. Moreover the equilibration step also significantly affected the follicles viability, with a beneficial effect of the one step equilibration compared to the 3 steps equilibration. However, no interaction was observed in this model. As a result, the fractional experimental design developed in the bitch, suggested that ovarian tissue

PROH and 0.2M sucrose. Post-seeding freezing rate at 2°C/min.

(Fig. 4).

**2.4.3 Investigation in the bitch** 

with PROH. Queen ovarian stroma was well preserved.

A B

difference with samples directly immersed into the liquid nitrogen (86.8 2.5%). Proportion of normal follicles decreased significantly after cryopreservation except after freezing with PROH according to a free fall into the freezing chamber (68.2 9.1% of type I). After freezing using a direct immersion into liquid nitrogen after -40°C, and whatever the CPA, proportions of type I follicles were decreased compared to fresh control.

Fig. 3. Rabbit follicular ultrastructure before (A & B) and after (C & D) cryopreservation with PROH and trehalose, with a post-seeding freezing rate at 0.3°C/min

Before freezing, queen ovarian tissue showed 72.2 3.6% and 74.8 6.3% of type I follicles respectively for group without and with sucrose without any difference between the two control groups. After freezing, addition of sucrose allowed preserving 63.2 13.6% of normal follicles versus 68.2 9.1% without sucrose when associated with PROH, without any significant difference. Contrary to the results after freezing with DMSO, proportion of type I follicles was not significantly different after freezing with PROH compared to fresh control. Morphological defect of type III was the most frequently observed after freezing with PROH. Queen ovarian stroma was well preserved.

Fig. 4. Queen follicular morphology before (A) and after (B) cryopreservation with 1.5M PROH and 0.2M sucrose. Post-seeding freezing rate at 2°C/min.

In conclusion, queen ovarian tissue seems to be well preserved, without any difference compared to the fresh control when freezing with PROH (according to a free fall in the freezing chamber, without influence of neither sucrose nor post-seeding freezing rate) (Fig. 4).

#### **2.4.3 Investigation in the bitch**

218 Current Frontiers in Cryopreservation

difference with samples directly immersed into the liquid nitrogen (86.8 2.5%). Proportion of normal follicles decreased significantly after cryopreservation except after freezing with PROH according to a free fall into the freezing chamber (68.2 9.1% of type I). After freezing using a direct immersion into liquid nitrogen after -40°C, and whatever the CPA,

A B

A B

Before freezing, queen ovarian tissue showed 72.2 3.6% and 74.8 6.3% of type I follicles respectively for group without and with sucrose without any difference between the two control groups. After freezing, addition of sucrose allowed preserving 63.2 13.6% of normal follicles versus 68.2 9.1% without sucrose when associated with PROH, without any significant difference. Contrary to the results after freezing with DMSO, proportion of type I follicles was not significantly different after freezing with PROH compared to fresh

Fig. 3. Rabbit follicular ultrastructure before (A & B) and after (C & D) cryopreservation

with PROH and trehalose, with a post-seeding freezing rate at 0.3°C/min

proportions of type I follicles were decreased compared to fresh control.

A B

C D

In the bitch, the estimated model was validated when the viability preservation ratio was considered. The nature of the non-permeating CPA (*P* = 0.37) did not influence the post thawing viability rate of the ovarian follicles. However all the other factors investigated in this experimental design presented a significant effect on the viability rate. The permeating CPA nature (*P*<0.0001) was the factor that influenced more the viability rate of the follicles. Thus, contrarily to the observations in the other species, the DMSO better preserved the evaluated parameter than PROH. The freezing rate had also a major effect on the viability rate (*P*<0.0001) and a slow freezing rate (0.3°C/min) was less deleterious than the rapid freezing rate for the follicles viability. Moreover the equilibration step also significantly affected the follicles viability, with a beneficial effect of the one step equilibration compared to the 3 steps equilibration. However, no interaction was observed in this model. As a result, the fractional experimental design developed in the bitch, suggested that ovarian tissue

New Approaches of Ovarian Tissue Cryopreservation from Domestic Animal Species 221

compared respectively with sucrose and FCS. No significant interaction was observed

In order to discriminate permeating CPAs, ovarian fragments from 5 cows were frozen using 1.5M DMSO or 1.5M PROH with 4 g/L Albumax® and 0.2M trehalose according to a post-seeding freezing rate at 0.3°C/min. Before freezing, ovarian tissue showed 40.6 12.6% of type I preantral follicles. Proportion of type I follicles was reduced to 20.2 3.9% after cryopreservation using DMSO and to 23.8 3.4% when using PROH. No statistical difference was found between DMSO and PROH. Morphological defects of type II were the most important kind of defect (47.0 11.7%). Proportion of type IV follicular defects was significantly improved compared to control for the both CPAs (47.2 7.8% and 44.8 4.4% for DMSO and PROH respectively). Proportion of type III follicles was constant before and after freezing. Ovarian stroma seems to be affected by cryopreservation and shows spaces

A B Fig. 6. Cow follicular morphology before (A) and after (B) cryopreservation using PROH

As for the other species, the influence of the post-seeding freezing rate was evaluated when freezing with 2M PROH. Proportions of morphologically normal follicles were significantly reduced after freezing compared with fresh tissue, whatever the post-seeding freezing rate 17.6 6.2% after freezing at 2°C/min vs. 57.8 13.0% before freezing and 17.8 6.5% after

In the doe rabbit, nine pups were born from cryopreserved grafted group, suggesting the efficiency of the cryopreservation protocols based on PROH and trehalose and using very slow freezing rate. At necropsy, follicular structures were observed in most of females.

In the bitch, the cryopreservation method optimized with the fractional experimental design and validated by in vitro assessment (morphology, viability, and toxicity) was then evaluated by heterotopic xenografting to determine whether the ovarian tissue integrity and

between the nature of the permeating CPA and its concentration.

and disjoined cells.

and slow freezing rate

freezing at 0.3°C/min vs. 60.0 4.9% before freezing).

**2.4.5 In vivo follicle resumption from cryopreserved ovarian tissue** 

should be cryopreserved in a solution containing 2 M DMSO as permeating CPA supplemented with 0.2 M sucrose or trehalose in a one-step equilibration and frozen at a 0.3°C/min freezing rate (Fig. 5).

Fig. 5. Bitch follicular morphology before (A) and after (B) cryopreservation

Consequently, theses parameters validated by the fractional experimental design for viability assessment were used and applied for morphological assessment of frozen-thawed bitch ovarian tissue and comparison with fresh tissue. So, the morphology of preantral follicles was compared between fresh and cryopreserved tissue. The histological analysis revealed that no significant difference was observed between fresh (89.1 6.1 % type I follicles) and cryopreserved ovarian tissue (82.4 4.4 % type I follicles) when type I follicles were observed. The main abnormality observed on preantral follicle after cryopreservation in the bitch was the oocyte nucleus defect (~8%). In this case, the nucleus often appeared pycnotic, with a reduced size and a densely packed chromatin. Sometimes the nuclear membrane was ruptured. However, ooplasm defect were rarely observed alone, but combined with nuclear defects. In some cases few granulosa cells were absent in the primordial or primary follicles. It is probable that ice crystal formation occurring during cooling be responsible of this partially denuded pattern, by destroying or dislodging some granulosa cell during ice expansion.

#### **2.4.4 Investigations in the cow**

As for the rabbit doe, the experimental design was valid when considering the morphological preservation ratio of the follicles, but not when considering the viability ratio of the follicles. The concentration of the permeating CPA (*P* = 0.59) and the medium (*P* = 0.76) seem to have no significant effect on the morphological preservation ratio of ovarian follicles. The nature of the permeating CPA seemed to influence the morphological preservation ratio of the follicles (*P* = 0.07) although the non-significant difference. PROH tended to improve the morphological preservation ratio, as compared with DMSO. The nature of the non-permeating CPA (*P* = 0.002) and the cells surfactant (*P* = 0.04) had significant influence. Trehalose and Albumax® improved morphological preservation ratio

should be cryopreserved in a solution containing 2 M DMSO as permeating CPA supplemented with 0.2 M sucrose or trehalose in a one-step equilibration and frozen at a

A B

Consequently, theses parameters validated by the fractional experimental design for viability assessment were used and applied for morphological assessment of frozen-thawed bitch ovarian tissue and comparison with fresh tissue. So, the morphology of preantral follicles was compared between fresh and cryopreserved tissue. The histological analysis revealed that no significant difference was observed between fresh (89.1 6.1 % type I follicles) and cryopreserved ovarian tissue (82.4 4.4 % type I follicles) when type I follicles were observed. The main abnormality observed on preantral follicle after cryopreservation in the bitch was the oocyte nucleus defect (~8%). In this case, the nucleus often appeared pycnotic, with a reduced size and a densely packed chromatin. Sometimes the nuclear membrane was ruptured. However, ooplasm defect were rarely observed alone, but combined with nuclear defects. In some cases few granulosa cells were absent in the primordial or primary follicles. It is probable that ice crystal formation occurring during cooling be responsible of this partially denuded pattern, by destroying or dislodging some

As for the rabbit doe, the experimental design was valid when considering the morphological preservation ratio of the follicles, but not when considering the viability ratio of the follicles. The concentration of the permeating CPA (*P* = 0.59) and the medium (*P* = 0.76) seem to have no significant effect on the morphological preservation ratio of ovarian follicles. The nature of the permeating CPA seemed to influence the morphological preservation ratio of the follicles (*P* = 0.07) although the non-significant difference. PROH tended to improve the morphological preservation ratio, as compared with DMSO. The nature of the non-permeating CPA (*P* = 0.002) and the cells surfactant (*P* = 0.04) had significant influence. Trehalose and Albumax® improved morphological preservation ratio

Fig. 5. Bitch follicular morphology before (A) and after (B) cryopreservation

0.3°C/min freezing rate (Fig. 5).

granulosa cell during ice expansion.

**2.4.4 Investigations in the cow** 

compared respectively with sucrose and FCS. No significant interaction was observed between the nature of the permeating CPA and its concentration.

In order to discriminate permeating CPAs, ovarian fragments from 5 cows were frozen using 1.5M DMSO or 1.5M PROH with 4 g/L Albumax® and 0.2M trehalose according to a post-seeding freezing rate at 0.3°C/min. Before freezing, ovarian tissue showed 40.6 12.6% of type I preantral follicles. Proportion of type I follicles was reduced to 20.2 3.9% after cryopreservation using DMSO and to 23.8 3.4% when using PROH. No statistical difference was found between DMSO and PROH. Morphological defects of type II were the most important kind of defect (47.0 11.7%). Proportion of type IV follicular defects was significantly improved compared to control for the both CPAs (47.2 7.8% and 44.8 4.4% for DMSO and PROH respectively). Proportion of type III follicles was constant before and after freezing. Ovarian stroma seems to be affected by cryopreservation and shows spaces and disjoined cells.

A B

Fig. 6. Cow follicular morphology before (A) and after (B) cryopreservation using PROH and slow freezing rate

As for the other species, the influence of the post-seeding freezing rate was evaluated when freezing with 2M PROH. Proportions of morphologically normal follicles were significantly reduced after freezing compared with fresh tissue, whatever the post-seeding freezing rate 17.6 6.2% after freezing at 2°C/min vs. 57.8 13.0% before freezing and 17.8 6.5% after freezing at 0.3°C/min vs. 60.0 4.9% before freezing).

#### **2.4.5 In vivo follicle resumption from cryopreserved ovarian tissue**

In the doe rabbit, nine pups were born from cryopreserved grafted group, suggesting the efficiency of the cryopreservation protocols based on PROH and trehalose and using very slow freezing rate. At necropsy, follicular structures were observed in most of females.

In the bitch, the cryopreservation method optimized with the fractional experimental design and validated by in vitro assessment (morphology, viability, and toxicity) was then evaluated by heterotopic xenografting to determine whether the ovarian tissue integrity and

New Approaches of Ovarian Tissue Cryopreservation from Domestic Animal Species 223

queen and in the cow, Demirci *et al.* observed a high (but not significant) proportion of follicles without any morphological defect after a post-seeding freezing rate of 2°C/min in the ewe (Demirci et al., 2001). Nevertheless, Gook *et al.* also observed a better follicular preservation when using a slow freezing rate (0.3°C/min) with human ovarian tissue (Gook et al., 1999). Whereas Cleary *et al.* observed no difference in terms of follicular growth after grafting, between a conventional embryo freezing protocol (0.3°C/min) and a passive cooling at 1°C/min from 0°C to -84°C on the mouse ovarian tissue (Cleary et al., 2001). Although these two cooling rates (0.3°C/min and 2°C/min) could be considered as slow, these results may be explained by a difference in cell dehydration during the post-seeding step. With rapid cooling rates, we can hypothesise that time required for the exosmose of the cell water is insufficient and consequently promotes the formation of lethal intracellular ice. While at very slow cooling rates, high level of dehydration occurs with concomitant increasing in solute concentration (salting out). Investigations on the freezing rate were extended in the queen with the evaluation of the third freezing phase. When associated with PROH, slow cooling in solid phase seems to be more appropriate. Nevertheless, several authors use cryopreservation protocols with a direct immersion into liquid nitrogen after – 40°C, such as Rodrigues et al. (2004) in the goat, Lucci et al. (2004) in the zebu cow, Santos et al. (2006) in the ewe or Lima et al. (2006) in the cat. Births had been obtained after graft of ovarian fragments frozen with such a protocol in the rabbit doe (Almodin et al., 2004b) and in the ewe (Almodin et al., 2004a), but not in the cat where in vivo follicular growth were observed when using a freezing protocol with controlled third phase (Bosch et al., 2004).

The experimental designs revealed a crucial role of the permeating CPA in the doe rabbit, in the cow and in the bitch, added to a crucial role of the non permeating CPA in the doe rabbit and in the cow. Among the various freezing protocols described in the literature, those using DMSO or PROH as permeating CPA seems to be more efficient, whatever the species. Our results suggest that PROH improve the follicular quality after freezing in the doe rabbit and in the queen. Results of experimental design obtained in the cow suggested a better morphological preservation rate when using PROH. However, in this species, standard comparison between DMSO and PROH doesn't confirm these results. Contrary to that, bitch ovarian tissue seems to be better preserved in freezing medium composed of DMSO. These results were confirmed by the follicular growth observed after xenograft. It can be hypothesized that DMSO penetrate better within the tissue than PROH. Indeed, the bitch ovarian tissue as the goat or the ewe is rich in collagenous fibbers and more fibrous than the doe rabbit ovarian tissue for example. Therefore, a good ability to penetrate within the ovarian tissue is an important characteristic for the chosen CPA Nevertheless, both CPA have sensibly the same molecular weight (PROH: 76.10 g/mol, DMSO: 78.14 g/mol) with a lower weight for PROH. Thus a better penetration of the DMSO cannot be explained by this physical parameter. The explanation may come from the toxicity of both CPAs. Our team also investigated the toxicity of the both CPAs on bitch ovarian tissue after equilibration steps at room temperature without freezing and registered a deleterious effect of PROH

Except in canine and feline models, addition of non-permeating CPA in the freezing medium seems to improve the protective effect of CPAs. The protective effect of PROH seems to be improved when it is associated with trehalose, in the doe rabbit and in the cow. This observation was confirmed by electron microscopy evaluation of the doe rabbit ovarian

compared to DMSO on preantral follicle viability in this species.

the follicular growth potential were maintained. The frozen-thawed tissues were grafted to female SCID mice as previously described. The histological assessments of the follicle population revealed a significant increase in the density and distribution of secondary follicles from eight weeks post grafting compared to the follicle population at 1 week (*P*<0.05). Consequently, the shift from primordial-primary follicles to secondary follicles occurred in a time laps of eight weeks. Moreover morphologically normal follicles were observed until 16 weeks post transplantation and intact secondary follicles with more than 3 layers of granulosa cells and a normal oocyte surrounded by a well-defined zona pellucida were present at this time. Despite a massive follicular loss touching particularly the early follicles and occurring just after grafting, the graft survived long term xenografting. Similarly, after an important loss just after grafting (one week grafts) the stromal cell number increased during the graft period, to reach a density comparable to fresh ovarian tissue at 16 weeks. Otherwise, the vascularization setting-up was assessed by immunohistochemistry as developed previously. The vessel density analysis revealed that already at one week post grafting the vessel density within the graft was comparable to the fresh control ovarian tissue. Moreover, the vessel density tended to increase at 16 weeks post grafting compared to the other groups (fresh control, one and eight weeks, *P*<0.05) even if no significant difference was registered. No antral follicles was present at the end of the graft time, however, persisting vaginal cornification was noticed on the recipients from 5-9 days post grafting, and estrus behavior was observed several times during the graft period in the recipients cages indicating an hormonal activity resulting from the graft. Taken together, these in vivo results confirm the good preservation of the bitch ovarian tissue by applying our cryopreservation method.
