Successful fertility following optimized perfusion and cryopreservation of whole ovary and allotransplantation in a premature ovarian insufficiency rat model

Histological evaluation and comparison of fresh ovaries versus cryopreserved/thawed ovaries demonstrated a well-preserved general morphology of the follicles in the optimized group, although there were areas of decreased stromal density or oocyte cytoplasm vacuolations in some samples (Fig. 1). However, more abnormal follicles could be seen in the conventional group. Data for the follicles at all developmental stages are reported in Table 1. The number of follicles in the primordial and primary follicle classes significantly decreased in the optimized group compared with the fresh group (P < 0.05). There was no significant difference in the number of secondary follicles and antral follicles between the two groups. In addition, 77.6% of the total follicles were counted in the optimized group, which was significantly higher than that of the conventional group (P = 0.002), although this value was still significantly reduced compared with the fresh group. The quantitative analysis of the stromal cells showed that there was no significant difference between the optimized group and the control group (P > 0.05) (Fig. 2).

TUNEL-positive cells and Bax-positive cells were detected in the three groups, and representative images are shown in Fig. 3. The results were similar in the TUNEL and Bax analyses. Very few positive staining signals were detected in the fresh group. Positive staining signals in the conventional group were found in every stage of the follicles and the stromal cells, which were significantly increased in number compared with the fresh group (P < 0.01). While positive staining signals in the optimized group were found mostly in the antral follicles, they were significantly reduced in number compared with the conventional group (P < 0.01).

All 25 rats treated with CTX survived. The weight of the rats slightly decreased during CTX treatment but recovered after 2 weeks of CTX treatment. The rats showed abnormal estrus cycles 7-10 days after CTX injection, with prolonged and irregular diestrum or absence of preoestrus and estrus. The estrus cycles of treated rats showed an average duration of 8.00 ± 2.83 days, and were still disordered after 2 weeks of CTX treatment; this duration was significantly longer than that of the control group (3.98 ± 0.85 days; P < 0.05). After 2 weeks of CTX treatment, serum FSH levels in the POI group were significantly higher than the control group (P < 0.01). Serum estradiol-17β and progesterone levels in the POI group were significantly lower than the control group after 2 weeks of CTX treatment (P < 0.01) (Table 2). Contracted ovaries and uteruses were clearly present in the POI group when the abdomen was surgically opened for transplantation.

The median duration of transplantation surgery from incision to closure was 54.00 ± 9.64 min, and the estimated blood loss was < 1 ml in all animals. In all 25 allotransplanted rats, recirculation of the graft was satisfactory based on a color change from white to reddish, with pulsation through the anastomosed vessels. The weight of the rats began to increase after 4 days of significant reduction. Five rats died within 3 days after transplantation due to infection or anastomotic stenosis with thrombosis, while the remaining 20 rats survived after transplantation.

Of the 20 live rats, 14 (70%) began to recover estrous cycles an average of 14.43 ± 2.98 days after WOCP&TP according to vaginal smears. Normal estrous cycles (4.65 ± 1.63 days) were seen 4 weeks after transplantation (P > 0.05). Similarly, serum levels of FSH significantly declined 2 weeks after transplantation (P < 0.05), but were still higher than levels in the control group (P < 0.01). Four weeks after transplantation, the serum levels of FSH continued to decrease and returned to normal levels (P > 0.05). Serum levels of progesterone and estradiol-17β significantly increased to normal levels 2 weeks after transplantation (P > 0.05). Serum levels of AMH significantly increased 2 weeks after transplantation (P < 0.05), but were still lower than levels in the control group (P < 0.05). Four weeks after transplantation, serum levels of AMH continued to increase and returned to normal levels (P > 0.05) (Fig. 4). The normal hormone levels of rats were maintained until they were sacrificed.

The ovaries of the four rats that became pregnant were of normal appearance and vascularity and contained morphologically normal, well vascularized corpora lutea and all developmental stages of follicles. As seen in Fig. 5, the transplanted right ovary appeared normal 8 months after WOCP&TP, with the left ovary absent in the rat that developed a second pregnancy. There were follicles from primordial to antral stages and corpora lutea in the right ovary. The right transplanted uterine segment was the same size as the contralateral uterine segment. The average weight of the ovaries in the four rats that became pregnant was 25.84 mg, comparable with that of the control group (25.79 mg; P > 0.05). The ovaries of the 10 rats that did not become pregnant were of normal appearance in three rats and showed adhesions in seven rats. The average weight of the ovaries in the 10 rats that did not become pregnant was 22.84 mg, which was not different from that of the control group (P > 0.05). However, the number of primordial and pre-antral follicles in the 14 rats that resumed estrous cycles was significantly reduced compared with that of the control group (P < 0.05).

The remaining six rats showed a total absence of all cells on the vaginal smear, revealing no evidence of ovarian function restoration following WOCP&TP. These rats had significantly higher levels of FSH and lower levels of progesterone, estradiol-17β, and AMH at the first month; after 1 month, levels of progesterone, estradiol-17β, and AMH were maintained or continued to decline to being undetectable (P < 0.01) (Fig. 4). The transplanted ovaries and vascular pedicle in these non-restored rats were atrophic or absent with severe adhesions at the time of postmortem examination, and the weight of the ovaries could not be calculated.

Fourteen rats in the POI group that showed recovery of hormone levels were mated after 1 month of WOCP&TP. Four rats (20%) were pregnant within 2 months after mating and all delivered healthy live births. The total number of offspring was 22, and the number of offspring per litter was ranged from 4 to 6, which was statistically less than the normal yield of 8-13 (P < 0.05). A summary of the time of delivery and gender of offspring is presented in Fig. 6a. One month after delivery, one of the four rats underwent a second mating, became pregnant, and delivered a second litter of four live offspring. The first filial generation of rats was mated at 8-10 weeks age. They gave live birth to the second generation of offspring. In the same way, two rats of the second generation were mated when they were 8-10 weeks old, and the third generation of offspring was delivered. A family tree of the mated rats that produced three generations of offspring is shown in Fig. 6b. The number of live births in the filial generations ranged from 6 to 9, with no statistical difference compared with the normal yield (P > 0.05). All offspring had no abnormalities in appearance (Fig. 6c).

Healthy adult female Lewis rats aged 8-10 weeks and weighing 180-200 g were obtained from Bikai Corporation (Shanghai, China). The animals were fed at the Animal Experiment Efficacy Evaluation Center, School of Pharmacy, Fudan University. The animals were housed in groups of 5 per cage and were maintained under standard laboratory conditions (12 h of light, 12 h of dark; 25 °C) for 7-10 days to acclimatize. During acclimatization, standard mouse chow and water were available ad libitum. Vaginal smears were obtained daily. Only rats showing at least two consecutive normal 5-day vaginal estrus cycles were included in the experiments.

Fifty Lewis rats were divided into two groups: the POI group and the control group. To establish animal models of chemotherapy-induced ovarian failure, 25 rats received a loading dose of intraperitoneal CTX 50 mg/kg (Baxter Oncology GmbH, German) followed by a daily intraperitoneal CTX injection of 8 mg/kg for 14 consecutive days. At the same time, 25 rats in the control group received 1 ml of 0.9% sodium chloride. All animals were monitored and weighed daily to observe their condition. Vaginal smears were obtained daily. Blood samples were collected to determine serum levels follicle stimulating hormone (FSH), progesterone, and estradiol-17β levels by enzyme-linked immunosorbent assay kits (Shanghai, China). Rats diagnosed with POI were transplanted after 2 weeks of treatment.

Forty Lewis rats were anaesthetized with sodium pentobarbital (40 mg/kg). With the aid of an operating microscope (GX.SS.22-3, Shanghai Medical Optical Instruments Co., China), the right ovary, fallopian tubes, and the upper third of the right uterus were dissected and branches of their vessels were divided. The ovaries were removed en bloc with their arteries and veins attached, with the ovarian vessels dissected to create short cuffs of the aorta and vena cava. The aorta was cannulated using a 2.5 F (0.75 mm OD) intravenous cannula that was tied securely in place, using 5-0 mersilk ligatures. The ovaries were perfused with heparinized Ringer’s solution via aorta cannulation in vitro. The whole ovary was cut down and subsequently immersed at 4 °C until required. Ovaries were randomly assigned to three groups. In the fresh group (n = 5), the ovary was cut from the pedicle and immediately fixed in 4% paraformaldehyde for subsequent analysis. In the conventional group (n = 5), the ovary was perfused in a conventional mode at a rate of 0.35 ml/min with M2 medium containing sucrose as well as an increasing concentrations of 0 M, 1 M, and 1.5 M DMSO for 10 min, respectively, cryopreserved and thawed, then cut from the pedicle and immediately fixed in 4% paraformaldehyde for subsequent analysis. In the optimized group (n = 5), the ovary was perfused in the new mode (described below), cryopreserved and thawed, then cut from the pedicle and immediately fixed in 4% paraformaldehyde for subsequent analysis. In the POI group, as donors (n = 25), the ovary was perfused in the new mode, cryopreserved and thawed, then transplanted into the POI rats.

The apparatus used for perfusion of the whole ovary is newly designed by our research team (Fig. 7). The apparatus mainly consists of two pulse control pumps (SP-12, Ever Seiko Corporation, Japan) with outlet and inlet pipes and three containers. The perfusion speed is controlled accurately by the peristaltic pump of the apparatus. The solution in the first container, which held an M2 medium containing 0.2 M trehalose and 3 M DMSO, was pumped into the second container at a rate of 0.15 ml/min. The solution in the second container, an M2 medium containing 10% fetal bovine serum, was pumped into the third container at a rate of 0.35 ml/min. A vein detained needle (24 G, 0.7 mm × 19 mm, BD, China) at the end of the output pipe was connected to the aortic cannula. The solution in the third container was an M2 medium containing 0.1 M trehalose, 1.5 M DMSO, and 10% fetal bovine serum. The whole ovary with the pedicle in the third container was perfused for 20 min, so that the concentration of DMSO increased from 0 M to 1.5 M in the first 15 min and then maintained at 1.5 M for the last 5 min. The whole ovaries were perfused in a mode of linearly increasing DMSO concentration at a constant speed of 0.35 ml/min for 20 min before cryopreservation. The media was adjusted to a pH of 7.3-7.4 and sterilized using a 0.22 μm syringe filter (Merck Millipore, China) prior to use. All solutions were refrigerated before use and all perfusions were carried out on ice.

After perfusion, the grafts were immersed in cryoprotective solution and transferred to 25-ml Nalgene cryovials. With the aid of a Planer controlled-rate freezer (Kryo 550-16; Planer plc, Middlesex, UK) a slow-freezing cryopreservation protocol was utilized, which involved cooling to − 7 °C at 2 °C/min before manual seeding was performed. Cooling continued at 0.3 °C/min to − 40 °C and subsequently at 10 °C/min to a final temperature of − 140 °C. The cryogenic vials were then plunged into and stored in liquid nitrogen for at least 2 weeks prior to evaluation or transplantation.

Thawing was carried out rapidly by swirling the ovaries in air for 1 min and then in a water bath at 40 °C for 5 min. After being completely thawed, the grafts were transferred to a perfusion tray and immediately reperfused. The CPA was washed out by reversing the concentration gradient at the same flow rate as before. The cannula was then eased from the artery that was then ready for reanastomosis or examination. The grafts were held in preservation solution on ice until transplantation within 15 min.

Recipient POI rats were anaesthetized with sodium pentobarbital (40 mg/kg). The transplantation operation was performed according to the methodology described by Wang et al. The aorta and vena cava below the left renal artery were dissected free and prepared for end-to-side anastomosis. Longitudinal incisions of 2-3 mm were cut in the great vessels and flushed with heparinized Ringer’s solution prior to aortic-aorta and veno-venous anastomoses using continuous suturing with 10/0 nylon (Ethicon Inc., USA). Both ovaries, fallopian tubes, and the upper third of the uterus were removed. The uterine segment of the cryopreserved organ was sutured to the corresponding end of the host’s organ with 7/0 synthetic absorbable sterile surgical suture (Ethicon Inc., USA). The average duration of the operation was 1 h and in every case blood was observed to surge back into the organs after reanastomosis. The rats in the control group underwent a sham operation.

Vaginal smears were prepared daily to monitor ovarian activity for more than 2 months after transplantation. Whole blood samples were collected every 2 weeks from the orbit after transplantation. After resting at 25 °C, the blood samples were centrifuged 3000 xg for 15 min to obtain plasma, which was stored at − 20 °C until assay performance. Levels of FSH, estradiol-17β, progesterone, and anti-Müllerian hormone (AMH) were determined by enzyme-linked immunosorbent assay kits (Shanghai, China).

Healthy adult male Lewis rats aged 10 weeks were obtained from Bikai Corporation (Shanghai, China). The female rats that recovered ovarian function based on examination of ovarian cyclicity were mated. One month after delivery, one rat underwent a second mating. The first filial generation of offspring was mated when they were 8-10 weeks old. In the same way, the second generation of offspring was mated when they were 8-10 weeks old. The number of live births and the physical appearance of the offspring were observed.

Rats were sacrificed 8 months after transplantation. The transplanted ovary was dissected and fixed in 4% paraformaldehyde for later histological analysis. After routine paraffin embedding, thin sections (4-5 μm thick) of all ovarian tissue were prepared and stained with hematoxylin and eosin. Within these sections, the number of follicles present were recorded and classified as primordial (oocyte surrounded by a single flat layer of follicle epithelial cells/pregranulosa cells), primary (single layer of cuboidal granulosa cells), secondary (two or more layers of granulose cells, no antrum), or antral (presence of an antrum), similar to previously described methods. Evaluation of the normality of the follicles was based on examination of the integrity of the basement membrane, cellular density, presence or absence of pyknotic bodies, and integrity of the oocyte. Based on these criteria, follicles were classified as morphologically normal or abnormal. In addition, the density of the stromal cells in the ovarian medulla was analyzed under high magnification (× 400).

Apoptotic cells were identified with a terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end-labeling (TUNEL) kit (Roche Diagnostics, America) and a Bax kit (Abcam Inc., England) according to the manufacturer’s manual. Positive cells appeared as brown granules. Blinded samples were observed under microscopy (Olympus DP73, Japan). Five different regions in each slide were randomly selected and pictured. The numbers of TUNEL-positive cells and Bax-positive cells were calculated as the average of the numerical values using image analysis software Image-Pro Plus Vision 6.0 (Media Cybernetics, America).

Statistical analyses were performed using SPSS 14. All data were expressed as means ± SEM. The Student t test was used to determine significant differences between two groups. A one-way analysis of variance with least significant difference tests was used to determine significant differences among the four groups. P < 0.05 was considered to be statistically significant.