Article Comparative study between intact and non-intact intramuscular auto-grafted mouse ovaries

Transcription

1 RBMOnline - Vol 18 No Reproductive BioMedicine Online; on web 21 November 2008 Article Comparative study between intact and non-intact intramuscular auto-grafted mouse ovaries Hussein Eimani received his PhD from Tarbiat Modarres University of Tehran, Iran. His research area focused on oocyte in-vitro maturation and in-vitro folliculogenesis and vitrification. Further training and specialization in infertility and assisted reproductive technology was undertaken at the Royan Institute, Iran. He is now associate professor in embryology. His research focusing on maturation of immature oocytes and animal cloning has led to numerous publications in national and international journals. Dr Hussein Eimani Hussein Eimani 1,2,5, Saiedeh Fatemeh Siadat 3,4, Poopak Eftekhari-Yazdi 1, K Parivar 3, Mojtaba Rezazadeh Valojerdi 1, Abdolhossein Shahverdi 1 1 Embryology Department, Reproductive Medicine Research Center, Royan Institute, ACECR, Tehran, Iran; 2 Department of Anatomy, Faculty of Medicine, Baqyiatallah University of Medical Sciences; 3 Science and Research Branch, Islamic Azad University, Tehran, Iran; 4 Department of Biology, Faculty of Science, Science and Research Branch, Islamic Azad University, Tehran, Iran 5 Correspondence: Department of Embryology, Royan Institute, 12 Hafez St, Banihashem St, Resalat St, PO Box , Tehran, Iran; eimani@yahoo.com Abstract Anticancer treatments often lead to ovarian failure and infertility. Cryopreservation and subsequent transplantation of the ovaries is one of the solutions that has been adopted as a means of preserving fertility, but primitive ischaemia in the grafted ovary that damages the oocyte pool is considered to be a possible problem. In order to improve blood supply and follicle preservation, two incisions were made in the ovaries before an intramuscular auto-grafting procedure and these non-intact ovaries were compared with the intact ovaries that were also auto-grafted intramuscularly. Follicle numbers and apoptosis were examined in intact and non-intact groups after 1, 2 and 3 weeks post-grafting. The results were compared with the control ovaries, which were not incised and grafted. Although follicle survival in both grafted groups was lower than in the controls (P 0.05), survival of follicles in the grafted ovaries (n = 19) was improved by incision prior to grafting. In addition, the estimated number of follicles decreased in grafted ovaries compared with non-grafted ovaries. Generally, it seems that this procedure is a promising method to preserve ovarian function; further studies are required to improve the success of maintaining ovaries after transplant. Keywords: apoptosis, autologous transplantation, follicle, ischaemia, mouse ovary Introduction Cancer patients treated by either chemo- or radiotherapy frequently suffer from ovarian failure and infertility. One of the new emerging techniques to preserve the reproductive potential of such patients is cryopreservation of ovarian cortical tissue prior to treatment, and its re-transplantation after healing (Newton, 1998; Sonmezer et al., 2004; Isachenko et al., 2006; Maltaris et al., 2007). Ovarian transplantation is a developing technology, and a variety of methodologies and grafting sites have been investigated in mice (Gunasena et al., 1997; Sztein et al., 1998; Liu et al., 2000, 2001, 2002; Migishima et al., 2003; Waterhouse et al., 2004) sheep (Aubard et al., 1999; Arav et al., 2005; Bordes et al., 2005; Bedaiway et al., 2007) and humans (Gook et al., 2001; Oktay et al., 2001; Donnez et al., 2004, 2006). One type of ovarian transplantation strategy is to divide the tissue into groups for either xenotransplantation or auto-transplantation. Ovarian tissue xenotransplantation (transplantation between different species) could eliminate the risk of cancer transmission in cancer patients, but this has been used purely for experimental purposes. It will not be suitable for clinical applications unless the safety and ethical issues can be resolved (Kim et al., 2005). Auto-transplantation (transplanting the ovarian tissue back into the patient), can be orthotopic or heterotopic. Orthotopic transplantation seems to be the natural choice for preservation of fertility (Donnez et al., 2004), but the potential risk of retransferring cancer cells along with the grafts (Shaw et al., 1996) is a major drawback for this option, although this risk appears to be low (Revel et al., 2004; Rogerio et al., 2005). On the other hand, the invasive surgical technique, adhesions Published by Reproductive Healthcare Ltd, Duck End Farm, Dry Drayton, Cambridge CB23 8DB, UK

2 54 and difficulties in directly monitoring follicle growth at an orthoptic site have prompted investigation of alternative graft sites. An alternative to orthotopic grafting is the use of other sites (heterotopic grafting). This may be valuable if these sites provide a better survival rate of the ovarian follicles (Aubard et al., 1999). Heterotopic grafting to sites such as subcutaneous (Aubard et al., 1999; Weissman et al., 1999; Nisolle et al., 2000; Van den Broecke et al., 2001), kidney capsule (Gook et al., 2001; Liu et al., 2001; Kaneko et al., 2003; Kagawa et al., 2007) or intramuscular (Israely et al., 2003, 2006) tissues allow the endocrine system to continue its function (Waterhouse et al., 2004), as well as offering the possibility of easier access and collection of oocytes from these heterotopic grafts for in-vitro maturation and IVF (Liu et al., 2001; Waterhouse et al., 2004). Kidney capsule transplantation provides a rich angiogenic environment, but recovery of follicles and oocytes requires a more invasive procedure. The number of damaged follicles in the intramuscular graft site is significantly lower than those in the subcutaneous site (Israely et al., 2003), and improved follicular survival was reported when the xenograft was transplanted entirely into a skeletal muscle (Israely et al., 2004). Moreover, temperature and pressure changes in the subcutaneous space could damage the oocytes (Donnez et al., 2004). For the above-mentioned reasons and to facilitate the potential recovery of oocytes for IVM and IVF, muscle was chosen as the site for transplantation in this study. In order to optimize the graft size, various sizes of rat ovarian grafts have been studied for autotransplantation, ranging from intact ovaries to one-eighth of the ovary; no significant differences were observed between the average grades of the different graft sizes. The follicles developed to the antral stage only in the half-ovarian grafts (Israely et al., 2003). Also, in many studies researchers used ovarian cortical strips (Oktay et al., 2001; Kiran et al., 2004); hemi ovaries (Waterhouse et al., 2004; Bordes et al., 2005), or in particular whole ovaries with their vascular pedicle, in order to shorten the ischaemic period by experimental intervention or rapid perfusion after grafting (Jeremias et al., 2002; Israely et al., 2006). Recently, Imhof et al. (2006) demonstrated the grafting of whole frozen stored ovine ovaries by microvascular anastomosis, but they concluded that there was no advantage compared with ovarian tissue transplantation, due to the existing poor follicular survival and higher surgical risks. In ovarian tissue transplantation, the graft (intramuscular or any other site) should be large enough to contain the maximum pool of oocytes, and also should be too small to minimize ischaemic injury (Israely et al., 2003). In order to maintain both these criteria, two incisions were made in some ovaries without separating the incised parts. Follicle preservation was then assessed in these ovaries compared with the intact ovaries, which were also autografted intramuscularly. Materials and methods Animals Animal experiments were carried out according to the Declaration Helsinki and the Guiding Principles in the Care and Use of animals (DHEW publication, NIH, 80 23). Ovarian transplantation Female NMRI mice, 4 6 weeks old, were anaesthetized with intraperitoneal ketamine (50 mg/ml; Rotexmedica GMBH, Trittau, Germany) and xylazine (20 mg, 2%; Alfasan, Woerden, The Netherlands). The left ovary was exposed through a small dorsolateral incision and separated from the oviduct, and the fat was dissected away at room temperature. The graft was inserted in the same mouse through a 0.5 cm full thickness dermal incision, through a cut along the gluteus superficialis muscle fibres of the hind limb, and superficially attached to the muscle beneath, facing the skin (Israely et al., 2006). The muscle fibre was sutured with 4 0 surgical thread (non-absorbable silk; Ethicon, New Jersey, USA) in order to allow detection of the transplantation site after killing the animal (cervical dislocation). The peritoneum and the skin were sutured with 5 0 absorbable threads (Ethicon) and 4 0 non-absorbable threads (silk; Ethicon) respectively. After the grafting procedure, the animals were allowed to recover. Two grafting procedures were used; in one method, two incisions were made in the ovaries (non-intact ovary) without separation and the incised ovary was then transplanted into muscle (n = 10); in the second method, after cleaning fat from the ovary, the intact ovaries were transplanted into muscle without making an incision (n = 9). Recovery of grafts In both experimental groups (intact and non-intact ovary methods), the mice were killed by cervical dislocation after a period ranging from 1 to 3 weeks. All grafted ovaries were removed and fixed with Bouin s solution, while the right ovaries in the same mice were randomly removed and fixed in the same manner as the control group. Histology Grafts from the intramuscular transplants were retrieved with the surrounding muscle as marked by the suture, placed overnight in Bouin s fixative, then transferred into 70% ethanol and stored at room temperature until processing. Fixed tissues were embedded in paraffin blocks and the whole ovary was sectioned serially at 6 μm thickness. Three sequential sections were placed on each slide. Alternate slides were stained with haematoxylin and eosin (H & E), and the remaining slides were stained with a TUNEL kit. Apoptosis was examined in the grafts using an in-situ cell death detection kit [horseradish peroxidase (POD); Roche, Mannheim, Germany] according to the manufacturer s instructions. Briefly, after deparaffinizatin, hydrated sections were incubated in a humidified chamber with 0.15% Triton X-100 (Sigma-Aldrich, St Louis, MO, USA) at room temperature for min, and the slides were then rinsed twice with PBS. The TUNEL reaction mixture were added, and the slides were incubated in a humidified atmosphere in the dark; slides were again rinsed three times with PBS, then converter POD was added and the slides were incubated in a humidified chamber for 30 min at 37 C. The slides were again rinsed three times in PBS. Diaminobenzidine (DAB) substrate or alternative POD substrates were added. Slides were incubated three times with PBS and were then counterstained with haematoxylin.

3 Morphometric analysis of grafts All of the haematoxylin eosin slides (about slides for each ovary) were used for to evaluate follicle counts and survival in the grafts. Follicles were counted and classified into four groups: (i) primordial: follicles with one oocyte surrounded by one layer of flattened cells; (ii) primary: follicles with one oocyte surrounded by one layer of cuboidal cells; (iii) preantral: follicles with more than one layer of cuboidal cells without an antrum; and (iv) antral: follicles with more than one layer of cuboidal cells with an antral cavity in granulose cells. Healthy follicles were counted in one of the sections on each H & E slide. To avoid double counting of the same follicle and to ensure inclusion of the largest cross-sections of follicles, counts were performed in the sections in which the oocyte nucleolus was seen within the germinal vesicle (nucleus). Primordial follicles were counted only when the oocyte had a definite nuclear membrane. Statistical analysis All data were analysed using Microsoft Office Excel. One-way analysis of variance (ANOVA) was used to compare the mean number of follicles present in grafted and non-grafted ovaries (control), at different intervals after transplantation. Chi-squared test was used to compare the number of intact and non-intact ovaries in which healthy antral follicles were found. One-sample Kolmogorov Smirnov test was used before ANOVA to see if group distribution was normal or not. Two-by-two comparisons between means were performed using Fisher s least significant differences (LSD). All data are expressed as mean ± SEM. A P-value of <0.05 was considered statistically significant. Results A total of 19 ovaries was collected from 19 mice and grafted into muscle tissue (gluteus superficialis) of the same mice (autologous graft). Nine of the 19 ovaries were intact ovaries that were transplanted into muscle tissue. Three of these transplanted ovaries were removed after 1 week (IW.1N); three of them after 2 weeks (2W.1N) and the last three were removed after 3 weeks (3W.1N); the total healthy follicle number was then counted (Table 1). Two incisions were made in the remaining 10 ovaries before grafting into muscle tissue. The same process described above was also performed on these ovaries (Table 1). Three of the above mice were randomly selected as the control group, and their right ovaries were fixed. The number of primordial, primary, prenatal and antral follicles in each grafted ovary and in the control group was counted and shown as mean ± SEM. The number of follicles in the control group was significantly higher than that in the experimental groups (P 0.05, Table 1). Intramuscular ovary transplantation: intact ovary Antral follicles were present in low numbers in the intact ovary groups: two of the nine ovaries had antral follicles (Table 1). Among the follicles assessed, primordial follicles were the most common, followed by primary follicles present in the highest numbers, then by preantral follicles and antral follicles, i.e. in terms of prevalence, primordial > primary > preantral > antral. This ratio was similar in the grafts assessed after 1-, 2- and 3-week intervals in the intact groups. Damaged areas were limited to central regions of the graft and were confirmed by TUNEL staining (Figure 1). With increasing time after grafting, the damage gradually extended and only follicles at the graft periphery, located away from the necrotic centre, survived. In these groups, it appeared that the number of follicles increased with the passing of time, but the difference was not significant. Dimensions of intact ovaries were about mm before grafting. Intramuscular ovary transplantation: nonintact ovary While the graft (in muscle as well as in other sites) should be large enough to contain the maximum pool of oocytes, it should also be small enough to minimize ischaemia Table 1. Number of healthy follicles in intact grafted, non-intact grafted and control ovaries at 1, 2 or 3 weeks after transplantation. Type of Number No. of follicles (mean ± SEM) ovary of ovaries a Primordial Primary Preantral Antral 1W.IN 3 (1) 23.7 ± ± ± ± 0.3 2W.IN 3 (0) 33.0 ± ± ± ± 0.0 3W.IN 3 (1) 39.3 ± ± ± ± 1.0 1W.N.IN 3 (2) 49.0 ± ± ± ± 0.3 2W.N.IN 4 (3) 72.7 ± ± ± ± 1.0 3W.N.IN 3 (3) 45.3 ± ± ± ± 0.9 Control 3 (3) ± ± ± ± 10.6 IN = intact ovary; N.IN = non-intact ovary (incised ovary); W = weeks. a The value in brackets is the number of ovaries in which healthy antral follicles were found. The difference between control and experimental groups was significant (P < 0.05), but the difference between experimental groups was not significant. 55

4 Figure 1. Apoptosis in transplanted and non-transplanted ovaries. (a c) Transplanted intact ovary stained with H & E. (d f) Transplanted intact ovary stained with TUNEL to identify apoptosis. (g i) Control ovary stained with H & E. (j l) Control ovary stained with TUNEL to identify apoptosis. m = muscle; O = oocyte; p = preantral follicle. Arrows in (c) indicate primordial follicles. Arrows in (e f) indicate TUNEL-positive follicle. *Indicates TUNEL-negative primordial follicle. (a, d) Bar = 200 μm; (b, e) bar = 100 μm; (c, f) bar = 20 μm; (g, j) bar = 1000 μm; (h, k) bar 200 μm; (i, l) bar = 100 μm. 56 reperfusion injury. In other words, in order to improve blood supply to the centre of the graft, two incisions were performed in each ovary (non-intact ovaries,) and these were then transplanted into the muscle. The number of surviving follicles in these ovaries was compared with the previous group (intact ovary). The number of healthy antral follicles in the non-intact group (non-intact ovaries, Figure 2e h) was greater than the number found in the intact ovaries. Eight out of 10 ovaries had antral follicles. The number of follicles in the non-intact ovaries was greater than the number found in the intact ovaries in all follicle stages, but the difference was not significant. For example, the number of primordial follicles in group 4 (week 1, non-intact) was 49 ± 22.6, and in group 1 (week 1, intact) was 23.7± 20.7 (Table 1); in both groups the follicles in the grafts were counted 1 week after grafting, and the only difference was the presence of incisions in group 4 (1W.N.IN) (Table 1). Dimensions of non-intact ovaries were also about mm before grafting, but the exact dimension of the whole ovary after grafting was not measurable due to the unclear margin between the muscular and the ovarian tissue. In the control group nearly all of the follicles were healthy (Table 1), whereas in the grafts that were transplanted for 1 3 weeks, most of the developing follicles were degenerate. Damaged areas were limited to the central

5 region of the graft, and within the grafts transplanted into the intact ovary group; this was confirmed by TUNEL staining (Figure 1). The number of follicles was higher in group 5 (N.IN. 2W) in comparison with other groups, but the difference was not significant. Improved protection of antral follicles in non-intact ovaries In addition to the primordial follicles, developing antral follicles were detected within the grafts (Figure 2e h). In the non-transplanted ovaries of the control group, nearly all of the follicles were healthy, whereas in the intact grafts most of the developing follicles were degenerate, with healthy antral follicles rarely detected. Although, in the non-intact groups most of the developing follicles were also degenerate, the mean of healthy antral follicles in this group was higher than in the intact group. The mean number of healthy antral follicles after 1 week was 0.3 ± 0.3 versus 0.7 ± 0.3, after 2 weeks was zero versus 2.5 ± 1, and after 3 weeks was 1 ± 1 versus 2.7 ± 0.9 in the intact and the non-intact ovaries respectively. Therefore, the number of ovaries with healthy antral follicles were also counted, revealing that in the intact ovaries, 2/9 (22.22%) had healthy antral follicles, whereas in the non-intact ovaries 8/10 (80%) had healthy antral follicles. This difference was significant (P 0.012) (Figure 3). DNA fragmentation in grafted ovaries Apoptotic cells were detected in atretic follicles (Figure 1d f). Most of the follicles in the cortical region of the ovary did not display staining for apoptotic DNA fragmentation (Figure 1). A few scattered TUNEL positive primordial oocytes or surrounding granulosa cells were present throughout the cortical region of grafted ovaries (Figure 1f) but these ovaries showed slightly positive staining for DNA fragmentation in medullary regions during the process of grafting (Figure 1e). Damaged areas within the intact grafted ovaries were shown by TUNEL staining (Figure 1d f). Positive cells were seen as dark brown in colour. Discussion This work compares the intramuscular auto-grafting of intact versus incised mouse ovaries, and examines follicle numbers and apoptosis at 1, 2 and 3 weeks post-grafting. The results are compared with control ovaries that were not incised and grafted. Figure 2. Antral follicle in a control ovary (a d) and grafted non-intact ovaries 2 weeks post-transplantation (e h). Sections have been stained with H & E. h = hole. (a) Bar = 1000 μm; (e) bar = 100 μm; (b, f) bar = 200 μm; (c, g) bar = 100 μm; (d, h) bar = 20 μm Percentage (%) Non-intact ovaries Intact ovaries Figure 3. Percentage of ovaries in which healthy antral follicles were found. The difference between groups is significant (P = 0.012). 57

6 58 Although follicle survival in both graft groups was found to be lower than in controls, survival of follicles in the grafted ovaries was improved by incising the ovary prior to grafting, compared with follicle survival in intact grafted ovaries. The major objective of ovarian transplantation is to preserve fertility by minimizing loss of primordial follicles and oocytes and optimizing follicular development (Israely et al., 2006). So far, a variety of methodologies have been investigated in order to preserve follicles from ischaemic damage. Imhof et al. (2006) demonstrated the grafting of whole frozen stored ovine ovaries by microvascular anastomosis, but they concluded that there was no advantage in whole ovary transplantation by vascular anastomosis compared with ovarian tissue transplantation. This is due to poor follicular survival and higher surgical risks. Another approach to reduce ischaemic damage is transplantation of the ovary to a site that has a good blood supply. A subcutaneous location has been used for ovary transplantation, and although this provides a convenient location, it has a relatively poor blood supply and poor angiogenic capacity. The kidney capsule is another transplantation site that provides a rich angiogenic environment, but recovery of follicles and oocytes requires a more invasive procedure (Israely et al., 2004). Israely et al. (2003) compared two transplantation sites, subcutaneous and intramuscular, and suggested that ovarian maintenance was markedly better in intramuscular transplants. Therefore, in this study muscle was chosen as a transplantation site, as it is more homogeneous and rich in vasculature. A subcutaneous site was not used in the present study, as this is heterogeneous and has relatively poor blood vessel support. In the present study, after counting all follicle stages, the number of primordial follicle was found to be the most frequent in all groups. This may be due to overcounting, or to resistance of these follicles to ischaemic damage. The diameter of the nucleolus of oocytes in preantral and antral follicles is 7 10 μm, and thus there is a risk of over-counting. Great care was taken to avoid double counting in adjacent sections when dealing with the preantral or antral follicles. The diameter of the nucleolus as a marker for primordial follicles is estimated to be 2 μm, so the risk of over-counting is reduced in sections of 6 μm thickness (Liu et al., 2002). Primordial follicles are more resistant to the effect of ischaemia than are growing follicles, presumably because they are dormant and have a low metabolic rate (Liu et al., 2000). Moreover, undifferentiated primordial oocytes are relatively quiescent, small, and do not have a zona pellucida and cortical granules (Liu et al., 2001). In one study by Liu et al. (2001), approximately 50% of the primordial follicle population survives in isologous grafts in mice (Liu et al., 2001). Therefore, it seems that the high primordial follicle number compared with other follicle stages in the grafts in the present study is due to resistance of these follicles to ischaemic damage. Due to these characteristics, in human studies cortical strips are usually used for transplantation.(oktay et al., 2001; Van den Broecke et al., 2001; Kiran et al., 2004), although in animal studies immature ovaries, which are rich in primordial follicles, are usually used for grafting. In this study, adult mouse ovaries, which had more antral and preantral follicles compared with immature mouse ovaries, were grafted. These large follicles are also more sensitive to ischaemic damage. Therefore, the total number of large follicles was lower than the number of small follicles in all grafts. This was in accordance with previous studies (Bland et al., 1968; Felicio et al., 1983). Bland et al. believed that there was a direct correlation between the size of the follicles and their susceptibility to insufficient blood supply; i.e. larger antral follicles invariably underwent degeneration while smaller ones survived. In species with large ovaries, ovary fragments (Oktay et al., 2001) or hemi-ovaries (Waterhouse et al., 2004; Bordes et al., 2005) are usually used instead of intact ovaries, in order to achieve a better blood supply. Therefore, in this study, two incisions were made in the ovary without separating the portions from each other, and the ovary was then transplanted to the muscle tissue. In the present investigation, incised ovaries (non-intact ovaries) transplanted intramuscularly showed an improvement in follicle maintenance. Many follicles, from the primordial through the antral stages of development, could be found in the grafts, especially within the incised ovaries, suggesting a better blood supply in these groups. In addition, the present study indicated that apparent tissue damage in the intact grafted ovary was seen within the 1st week (Figure 1a d). Primordial and primary follicles were fewer in number after 1 week compared with 2 and 3 weeks posttransplantation. However, this difference was not significant; the 1st week is probably the period during which the substantial damage, mainly due to ischaemia reperfusion injuries occurs. Israely et al. (2004) have previously reported that ovarian grafts transplanted into intact tissue were vascularized by day 6 after transplantation, while on earlier days, the grafts remained mostly avascular. In the present study, grafted tissue was probably vascularized after 1 week (Table 1). The distribution of apoptotic cells showed that granulosa cells and stromal cells were more prone to be affected by the procedure than the primordial oocytes. This was confirmed by Kim et al. (2004). They found that ovarian cortex could tolerate ischaemia at least for 3 h, whereas stromal cells appeared to be more vulnerable to ischaemia compared with primordial follicles. Furthermore, the edges of the grafted tissue showed better follicular maintenance, especially in the intact grafted ovaries, and necrotic regions in the medulla of the graft were detected in this group (see Figure 1). This was in accordance with previous studies (Liu et al., 2002). Israely et al. (2003) also reported that the ovarian cortex showed better follicular maintenance, probably due to a better blood supply. They also showed that necrotic damage was apparent in the ovarian medulla. It was also revealed that the number of healthy follicles increased in the non-intact ovary group after 2 weeks; however, this increase was not significant and these findings suggest that early transitory damage during the first week may be overcome at later stages (after 2 weeks). A similar recovery of initial damage was also observed in the intramuscular grafts within 6 7 days post-implantation in the study by Israely et al. (2003). Retarded growth of ovarian grafts has been reported after transplantation of ovarian tissue to the kidney capsule of intact (non-ovariectomized) recipients (Liu et al., 2002).

7 The removal of the recipient s ovaries gives rise to two physiological changes: circulating gonadotrophins are elevated, and factors secreted by the intact recipient ovaries are eliminated. FSH stimulates granulosa cell mitosis, follicle maturation and inhibits granulosa cell apoptosis. Gonadotrophins also play a positive role in the vascular response after transplantation (Dissen et al., 1994). Therefore, high concentrations of FSH in ovariectomized recipients may facilitate follicular survival and development in transplanted ovaries. Liu et al. (2002) also noted that the proportion of growing follicles in grafted ovaries in ovariectomized recipients was higher than in-vivo 2-week-old mouse ovaries. Mice that were used in this study as recipients were the same as donor mice in which the right ovary remained in its place and the left ovary was grafted into muscle on the same side. After 1, 2 and 3 weeks, both grafted and non-grafted ovaries were removed and the follicles examined; the contralateral ovary was present during this time. Consequently, being under the effect of the remaining ovary, there was probably no significant increase in gonadotrophins, and therefore no positive effect of gonadotrophins on granulosa cell mitosis and follicle maturation in these mice. The present study also revealed that some large follicles were damaged in the incised ovaries while creating the incision. These incisions caused some follicles to lose their granulosa content as well as the oocyte. Some spaces were seen in place of those follicles (Figure 2e). Because of a better blood supply and the presence of mouse gonadotrophins, some primordial or primary follicles may be able to grow in these grafts, so it appeared that damage was lower in these groups (Figure 3). Finally, the present study showed that although follicle survival in both graft groups was lower than that in the control group, survival of follicles in the grafted ovaries improved by incising the ovary prior to grafting, compared with follicle survival in intact grafted ovaries. In this study, superficial muscle grafting (Israely et al., 2006) rather than subcutaneous grafting was used, because it was simple and had rapid graft transplantation, better blood supply, convenient in-vivo tracking and easy access for oocyte retrieval. Therefore, the future use of this method is recommended in order to further IVM and IVF. Acknowledgements This work was supported by the Royan Institute. The authors are grateful to Dr Shadab Jabbarpoor, Mr Hiva Alipour and especially to Dr Naghii for his thorough editing of the manuscript and revision of the English version. References Arav A, Revel A, Nathan Y et al Oocyte recovery, embryo development and ovarian function after cryopreservation and transplantation of whole sheep ovary. Human Reproduction 20, Aubard Y, Piver P, Cogni Y et al Orthotopic and heterotopic autografts of frozen thawed ovarian cortex in sheep. Human Reproduction 14, Bedaiwy M A, Falcone T 2007 Harvesting and auto-transplantation of vascularized ovarian grafts: approaches and techniques. Reproductive BioMedicine Online 14, Bland K P, Donovan B T 1968 Auto transplantation of the ovaries to the kidneys or uterus on the oestrous cycle of the guinea-pig. Journal of Endocrinology 41, Bordes A, Lornage J, Demirci B et al Normal gestations and live births after orthotopic autograft of vitrified warmed hemi-ovaries into ewes. Human Reproduction 20, Dissen GA, Lara HE, Fahrenbach WH et al Immature rat ovaries become revascularized rapidly after autotransplantation and show gonadotrophin-dependent increase in angiogenic factor gene expression. Endocrinology 134, Donnez J, Dolmans MM, Demylle D et al Restoration of ovarian function after orthotopic (intraovarian and periovarian) transplantation of cryopreserved ovarian tissue in a woman treated by bone marrow transplantation for sickle cell anemia: case report. Human Reproduction 21, Donnez J, Dolmans MM, Demylle D et al Livebirth after orthotopic transplantation of cryopreserved ovarian tissue. Lancet 364, Felicio LS, Nelson JF, Gosden RG et al Restoration of ovulatory cycles by young ovarian grafts in ageing mice: potentiation by long-term ovariectomy decreases with age. Proceedings of the National Academy of Sciences of the USA 80, Gook D A, McCully B A, Edgar DH et al Development of antral follicles in human cryopreserved ovarian tissue following xenografting. Human Reproduction 16, Gunasena K T, Villines P M, Critser E S et al Live births after autologous transplant of cryopreserved mouse ovaries. Human Reproduction 12, Imhof M, Bergmeister H, Lipovac M et al Orthotopic microvascular reanastomosis of whole cryopreserved ovine ovaries resulting in pregnancy and live birth. Fertility and Sterility 85, Isachenko V, Montag M, Isachenko E et al Effective method for in-vitro culture of cryopreserved human ovarian tissue. Reproductive BioMedicine Online 13, Israely T, Nevo N, Harmelin A et al Reducing ischemic damage in rodent ovarian xenografts transplanted into granulation tissue. Human Reproduction 21, Israely T, Dafni H, Nevo N et al Angiogenesis in ectopic ovarian xenotransplantation: multiparameter characterization of the neovasculature by dynamic contrast-enhanced MRI. Magnetic Resonance in Medicine 52, Israely T, Dafni H, Granot D et al Vascular remodeling and angiogenesis in ectopic ovarian transplants: a crucial role of pericytes and vascular smooth muscle cells in maintenance of ovarian grafts. Biology of Reproduction 68, Jeremias E, Bedaiwy MA, Gurunluoglu R et al Heterotopic autotransplantation of the ovary with microvascular anastomosis: a novel surgical technique. Fertility and Sterility 77, Kagawa N, Kuwayama M, Nakata K et al Production of the first offspring from oocytes derived from fresh and cryopreserved preantral follicles of adult mice. Reproductive BioMedicine Online 14, Kaneko H, Kikuchi K, Noguchi J et al Maturation and fertilization of porcine oocytes from primordial follicles by a combination of xenografting and in vitro culture. Biology of Reproduction 69, Kim SS, Kang1 H G, Kim N H et al., 2005 Assessment of the integrity of human oocytes retrieve from cryopreserved ovarian tissue after xenotransplantation. Human Reproduction 20, Kim SS, Yang HW, Kang HG et al Quantitative assessment of ischemic tissue damage in ovarian cortical tissue with or without antioxidant (ascorbic acid) treatment. Fertility and Sterility 82, Kiran G, Kiran H, Coban YK et al Fresh autologous transplantation of ovarian cortical strips to the anterior abdominal wall at the Pfannenstiel incision site. Fertility and Sterility 82, Liu J, Van der Elst J, Van den Broecke R et al Early massive follicle loss and apoptosis in heterotopically grafted newborn mouse ovaries. Human Reproduction 17, Liu J, Van der Elst J, Van den Broecke R et al Live offspring by in vitro fertilization of oocytes from cryopreserved primordial mouse follicles after sequential in vivo transplantation and in vitro 59

8 maturation. Biology of Reproduction 64, Liu J, Van der Elst J, Van den Broecke R et al Maturation of mouse primordial follicles by combination of grafting and in vitro culture. Biology of Reproduction 62, Maltaris T, Seufert R, Fischl F et al. 2007The effect of cancer treatment on female fertility and strategies for preserving fertility. European Journal of Obstetrics and Gynecology and Reproductive Biology 130, Migishima F, Suzuki-Migishima R, Song SY et al Successful cryopreservation of mouse ovaries by vitrification. Biology of Reproduction 68, Newton H, 1998 The cryopreservation of ovarian tissue as a strategy for preserving the fertility of cancer patients. Human Reproduction Update 4, Nisolle M, Casanas-Roux F, Qu J et al Histologic and ultrastructural evaluation of fresh and frozen thawed human ovarian xenografts in nude mice. Fertility and Sterility 74, Oktay K, Economos K, Kan M et al Endocrine function and oocyte retrieval after autologous transplantation of ovarian cortical strips to the forearm. Journal of the American Medical Association 286, Revel A, Schenker J et al Ovarian tissue banking for cancer patients: is ovarian cortex cryopreservation presently justified? Human Reproduction 19, Rogerio A, Lobo M D 2005 Potential options for preservation of fertility in women. New England Journal of Medicine 353, Shaw JM, Bowles J, Koopman P, et al Fresh and cryopreserved ovarian tissue samples from donors with lymphoma transmit the cancer to graft recipients. Human Reproduction 11, Sonmezer M, Oktay K 2004 Fertility preservation in female patients. Human Reproduction Update 10, Sztein J, Sweet H, Farley J et al Cryopreservation and orthotopic transplantation of mouse ovaries: new approach in gamete banking. Biology of Reproduction 58, Van den Broecke R, Liu J, Handyside A et al Follicular growth in fresh and cryopreserved human ovarian cortical grafts transplanted to immunodeficient mice. European Journal of Obstetrics and Gynecology and Reproductive Biology 97, Waterhouse T, Cox S L, Snow M et al Offspring produced from heterotopic ovarian allograft in male and female recipient mice. Reproduction 127, Weissman A, Gotlieb L, Colgan T et al Preliminary experience with subcutaneous human ovarian cortex transplantation in the NOD SCID mouse. Biology of Reproduction 60, Declaration: The authors report no financial or commercial conflicts of interest. Received 4 February 2008; refereed 12 February 2008; accepted 15 August