- Email: [email protected]
Morphological study of fully and partially isolated early human follicles
FERTILITY AND STERILITY威 VOL. 75, NO. 1, JANUARY 2001 Copyright ©2001 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
Morphological study of fully and partially isolated early human follicles Ronit Abir, Ph.D.,a Benjamin Fisch, M.D., Ph.D.,a Shmuel Nitke, M.D.,a Elimelech Okon, M.D.,b Ahud Raz, M.D.,a,c and Zion Ben Rafael, M.D.a Rabin Medical Center, Beilinson Campus, Petah Tikva, and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
Received March 24, 2000; revised and accepted July 26, 2000. Partially supported by research grants from the Israel Cancer Association (Z.B., R.A.) and Leo Mintz Fund-Tel Aviv University (B.F., R.A.). Presented at the meetings of the European Society for Human Reproduction and Embryology (ESHRE), Goteborg, Sweden, June 21–24, 1998 and Bologna, Italy, June 24 –28, 2000. Address of correspondence: Ronit Abir, Ph.D., IVF Research Laboratory, Department of Obstetrics and Gynecology, Rabin Medical Center, Beilinson Campus, Petah Tikva 49100, Israel (FAX: 972-39240533; E-mail: ronitabir @hotmail.com). a Department of Obstetrics and Gynecology. b Department of Pathology. c Present address: Department of Obstetrics and Gynecology, Soroka Hospital, Beer Sheva, Israel. 0015-0282/01/$20.00 PII S0015-0282(00)01668-X
Objective: To compare the development of fully and partially isolated human follicles by using various culture systems. Design: Human ovarian material was incubated with collagenase and deoxyribonuclease. Fully and partially isolated follicles (30 –50 m) were dissected and studied under light and electron microscopy. The follicles were then cultured on and within various matrices. Fully isolated follicles were also cocultured with stromal cells. Setting: Rabin Medical Center, a major care and referral center. Patient(s): Women undergoing laparoscopy. Intervention(s): None. Main Outcome Measure(s): Microscopy studies, follicular measurements. Result(s): Electron microscopy studies revealed an excess of lipid droplets in the granulosa cells of freshly isolated follicles. An increase in follicular size and granulosa cell number was observed only in the fully isolated follicles cultured within collagen gels for 24 hours. Most of the partially isolated follicles detached from the collagen gels. When cultured on collagen, extracellular matrix, and poly-L-lysine, both the fully and the partially isolated follicles deteriorated within the first 24 hours; coculture with stromal cells had no beneficial effect. Conclusion(s): The excess in lipid droplets in granulosa cells of isolated follicles might suggest that the isolation process does not yield completely healthy follicles. However, despite this finding, our studies show that fully isolated follicles, but not partially isolated follicles, can grow within, but not on, a culture matrix. (Fertil Steril威 2001;75:141– 6. ©2001 by American Society for Reproductive Medicine.) Key Words: In vitro, cryopreservation, full and partial isolation, transmission electron microscopy (TEM), primordial and primary follicles
Most of the follicles in the human ovary are unilaminar primordial or primary follicles (30 –50 m; (1). Their transformation into preantral (multilaminar) follicles presumably occurs through interactions with the surrounding stromal cells, although it is unclear whether these cells transmit inhibitory or stimulatory signals. Stimulation with FSH begins only at the preantral stages. The in vitro maturation of unilaminar follicles derived from cryopreserved-thawed ovaries would be of great benefit to many infertile women, for example, former cancer patients who became infertile after anticancer treatment (1). There are two possible approaches to the culture of unilaminar follicles (1). The first is to culture whole slices of ovarian tissue (organ
culture), such that the structural integrity of the tissue and the interactions between the stromal cells and the follicles are maintained (2). Human follicles have been shown to survive in organ culture, and in some cases, secondary follicles were observed (2). However, despite the simplicity of this culture system, it does not allow direct monitoring at the beginning of culture (0 hour), which makes it unclear whether the follicles indeed originated from unilaminar follicles in culture. Furthermore, the low follicular density in many human ovarian biopsies may lead to culture of empty ovarian slices (1). The second approach is to culture isolated unilaminar follicles. This culture system enables direct monitoring of the follicles during culture (1), but the follicles are taken out of 141
their natural environment, which might eliminate growth induction. We have recently shown that fully isolated follicles can proliferate in a collagen gel culture (3). However, Hovatta et al. (2) found that partially isolated follicles grown on an extracellular matrix (ECM) did not survive well compared with follicles in organ culture. The aim of the present study was to compare the development of fully and partially isolated follicles on and within various matrices. The use of partially isolated follicles maintained the interactions with the stromal cells and allowed us to monitor individual follicles. A secondary aim of our study was to study the morphology of isolated follicles by light and electron microscopy studies.
MATERIALS AND METHODS Ovarian Material Ovarian material was obtained from patients undergoing laparoscopy for various ovarian cysts. These included 20 women aged 17– 40 years (mean ⫾ SD: 30 ⫾ 7). The study was approved by the hospital’s ethical committee.
Cryopreservation and Thawing of Ovarian Tissue In 15 cases, the tissue specimens underwent cryopreservation with a 1, 2 propandiol (PROH; Sigma, St. Louis, MO) and sucrose (Sigma) protocol (3). Before cryopreservation, the samples were transferred through increasing concentration gradients of the freezing solution for several minutes and then frozen slowly in a programmable freezer (Kryo 10; series 10/20, Planer Biomed, Sunbury on Thames, UK). Ovarian pieces were thawed by transferring them in decreasing concentration gradients of PROH and sucrose for several minutes.
Isolation of Follicles The isolation process has been described by us elsewhere (3). In brief, after incubation with a collagenase IX (5,944 U/mL; Sigma) and deoxyribonuclease IV (180 U/mL; Sigma) solution, the follicles were dissected, and fully isolated follicles were released by aspirating them through a fine-bore pipette. Fully isolated follicles ⬍35 m and secondary follicles with at least two full granulosa cell (GC) layers were not cultured because our team has found that they fail to grow in collagen gel culture (3). However, these criteria could not be applied to partially isolated follicles because they were connected to the stromal layer in groups.
Preparation of Stromal Cells After follicular isolation, the digested ovarian fragments were collected in medium and aspirated repeatedly with automatic pipettes (40 –200 m, Finnpipette, Helsinki, Finland). Individual cells were rinsed several times and cocultured with fully isolated follicles. 142
Abir et al.
Partially and fully isolated human follicles
Preparation of Serum Gonadotropin-deficient human blood was obtained from women after pituitary– gonadal suppression by a gonadotropin-releasing hormone agonist (Decapeptyl GmbH, Ferring, Kiel, Germany) for IVF treatment. The hormone was administrated at a dose of 0.1 mg per day for at least 14 days, and ovarian suppression was confirmed by measuring circulating E2 levels. Human serum (HS) was prepared and heat inactivated. The use of this serum enabled us to monitor the levels of human FSH added to the culture medium.
Collagen Gel Solutions Collagen gel solutions were prepared from a rat tail collagen solution according to the manufacturer’s instructions (Collaborative Biomedical Products, Bedford, MA; see ref. 3). No more than three fully isolated follicles or one piece of connective tissue with partially isolated follicles was placed in each well of four-well culture plates (Nunclon, Roskilde, Denmark) and overlaid with a 100-L collagen gel solution. The fully isolated follicles were incubated for 24 hours because longer culture periods have been found to result in disruption of the follicular structure (3). The partially isolated follicles were placed in the incubator for ⱕ4 days.
Collagen, ECM, and Poly-L-Lysine Coatings Rat tail collagen type I solution (Collaborative Biomedical Products) was diluted to a concentration of 50 g/mL, and ECM gel (Sigma) was diluted 1:1. Millicell CM inserts (12-mm diameter, 0.4 m pore size; Millipore, Bedford, MA) were fitted into 24-well plates (Nunclon) and coated with the diluted solutions according to the manufacturer’s instructions (Collaborative Biomedical Products and Sigma, respectively). To allow gelling, ECM-coated inserts were kept for half an hour in a humidified incubator at 37°C. Wells in four-well culture plates (Nunclon) were coated with a poly-L-lysine 0.01% solution (Sigma) and prepared further according to the manufacturer’s instructions. Fully and partially isolated follicles were placed on these culture systems, and stromal cells were added to some of the fully isolated follicles (Table 1).
Culture Medium The culture medium contained Earle’s balanced salt solution (Biological Industries, Beit Haemek, Israel) with 0.47 mM sodium pyruvate (Sigma); 2 mM/L L-glutamine (Biological Industries); 10% FBS (Biological Industries; see Reference (3) or HS, human FSH (Metrodin, Teva Pharmaceutical Industries, Petah Tikva, Israel) at doses of 0, 0.5, and 1 U/mL; and 50 U/mL penicillin G (sodium salt), 50 g/mL streptomycin sulfate, and 0.125 g/mL amphotericin B (Penicillin-Streptomycin-Amphotericin B2 Solution, Biological Industries). The follicles were measured with a calibrated eyepiece micrometer at the beginning of the culture (0 hours) and every 24 hours. Vol. 75, No. 1, January 2001
TABLE 1 Fully and partially isolated human follicles cultured on various coatings.a Group and system Fresh, fully isolated Collagen Collagen ⫹ stromal cells ECM ECM ⫹ stromal cells Poly-L-lysine Poly-L-lysine ⫹ stromal cells Fresh, partially isolated Collagen ECM Poly-L-lysine
0h (follicle m)
n
44.1 ⫾ 14.0 40.0 ⫾ 15.5 39.9 ⫾ 17.3 37.6 ⫾ 15.1 35.8 ⫾ 16.0
50b 55 55 60 60
40.0 ⫾ 18.2
60 Deteriorated within 24 h
45.0 ⫾ 22.0 40.5 ⫾ 22.0 45.0 ⫾ 18.1
60 Deteriorated within 24 h 70 Deteriorated within 24 h 70 Deteriorated within 24 h
After culture
Deteriorated Deteriorated Deteriorated Deteriorated Deteriorated
within within within within within
24 24 24 24 24
h h h h h
Note: Values are means ⫾ SD. All follicles were cultured with HS and 0.5 U/mL. b Number of follicles at 0 h.
morning, with an araldite resin (Sigma) mixture at 60°C for 2 hours while being periodically stirred. After infiltration, the samples were embedded in a fresh mixture of araldite resin (Sigma) and kept at 60°C for 72 hours. Semithin (0.5– 0.75 m) sections for light microscopy (LM) were stained with toluidine blue (BDH Chemicals Ltd, Poole, UK), and ultrathin (1Å) sections for transmission electron microscopy (TEM) were stained with uranil acetate (BDH) and lead citrate (BDH). The follicles were examined with a JEOL (JEM 1010) electron microscope. Signs of atresia in the oocyte and GCs were defined as follows: nuclear changes, number of lipid droplets, vacuoles and changes in shape of follicle or oocyte (4, 5). The index for the number of vacuoles and lipid droplets in the GCs was given a range of 1–3, with 1 indicating very few (up to three per follicle) and 3 indicating many (at least 10 per follicle). Shrunken oocytes with abnormal nuclei and with a large number of vacuoles were considered atretic.
a
Abir. Study of isolated human follicles. Fertil Steril 2001.
Histologic Preparation for Light Microscopy Fresh and frozen-thawed ovarian pieces of 0.5–1 mm (control samples); partially isolated follicles; cultured collagen gels with partially isolated follicles or pieces that were separated from the gels spontaneously, mechanically, or enzymatically at various culture stages; and partially isolated follicles cultured on the various matrices were fixed in Bouin’s solution and prepared for paraffin embedding. Mechanical separation of the follicles from the collagen gels was performed with 21-G needles (Terumo Europe N.V., Leuven, Belgium) attached to 1-mL syringes (Terumo Europe N.V.) and enzymatic separation after treatment with collagenase IX (2,972 U/mL; Sigma) for 30 minutes. Fully isolated follicles equilibrated in the incubator for 1 hour with culture medium before being fixed (control) and collagen gels containing fully isolated follicles after a 24-hour culture period were also prepared for histologic study (3). Follicles and their oocytes were measured with a calibrated eyepiece micrometer. The number of GCs was counted in sections immediately after isolation and after collagen gel culture.
Histologic Preparation for Transmission Electron Microscopy All samples were fixed in 3% glutaraldehyde (Sigma). The following three protocols were used. Control samples. Fresh and frozen-thawed tissue pieces ⱕ0.5–1 mm were fixed and postfixed with 1% osmium tetraoxide (Sigma). These samples were then dehydrated with increasing concentrations of alcohol and finally with acetone (AR grade; Biolab, Jerusalem, Israel). The specimens were infiltrated overnight at room temperature with an acetone-araldite resin (Sigma) solution and, the following FERTILITY & STERILITY威
Partially isolated follicles. The partially isolated follicles were fixed, postfixed, and dehydrated. They were infiltrated at 60°C for only 1 hour with an araldite resin (Sigma) mixture while being periodically stirred, after which they were embedded, processed, and examined. Fully isolated follicles. A proportion of the 24-hour gels were treated with collagenase IX (2,972 U/mL; Sigma) for 30 minutes, and the follicles were recovered and prepared for TEM. They were fixed, postfixed, and dehydrated as described above but without acetone. The follicles were then transferred at room temperature to an araldite resin (Sigma) mixture for a few minutes, after which they were embedded, processed, and examined.
Statistical Analysis Data were statistically analyzed by unpaired Student’s t test, analysis of variance (ANOVA), and 2 test.
RESULTS When examining LM sections of follicles and their respective oocytes from fresh or frozen-thawed tissue before and after isolation, no significant differences in their diameter were found (follicular diameter ⫾ SD: 37.9 ⫾ 11.7 to 45.9 ⫾ 9.0 m; oocyte diameter ⫾ SD: 26.2 ⫾ 10.1 to 30.0 ⫾ 7.5 m; 33– 40 follicles per group). No pyknotic cells or abnormal oocytes were observed under LM. For the TEM studies, 25 control follicles from fresh tissue, 38 control follicles from frozen-thawed tissue, 23 isolated follicles from fresh tissue, and 20 isolated follicles from thawed tissue were examined. There were no identifiable differences in the number and quality of the intracellular organelles between the oocytes and GCs of the partially isolated follicles and the control follicles from both fresh and frozen-thawed tissue, apart from an increase in the number of lipid droplets in the GCs of isolated follicles from both fresh and frozen-thawed tissue. 143
TABLE 2 Fully and partially isolated human follicles cultured within collagen gels.
Group and sera Fresh fully isolated FCS FCS HS HS Thawed fully isolated FCS FCS HS HS Fresh partially isolated HS Thawed partially isolated HS
Human FSH (U/mL)
Follicle size, m ⫾ SD (n) 0h
24 h
0.5 1.0 0.5 1.0
45.0 ⫾ 10.0 (60)a 41.1 ⫾ 14.2 (50) 35.0 ⫾ 6.9 (60) 36.0 ⫾ 11.1 (50)
69 ⫾ 15 (25)b 70.1 ⫾ 11.5 (19) 69.8 ⫾ 13.6 (25) 71.2 ⫾ 13.2 (21)
0.5 1.0 0.5 1.0
35 ⫾ 15 (70) 37.0 ⫾ 12.4 (70) 35.1 ⫾ 9.2 (70) 35.2 ⫾ 16.3 (70)
65.0 ⫾ 15.0 (27) 71.2 ⫾ 12.1 (28) 65.1 ⫾ 15.0 (26) 68.8 ⫾ 14.4 (29)
0.5
33.0 ⫾ 14.2 (88)
follicles detachedc
0.5
37.0 ⫾ 20.0 (60)
follicles detached
a
Number of follicles at 0 hours. b Number of growing follicles after 24 hours in culture. c See more details in Results. Abir. Study of isolated human follicles. Fertil Steril 2001.
These included 0.4 ⫾ 0.6 and 0.4 ⫾ 0.7 lipid droplets for follicles from intact tissue specimens of fresh and frozenthawed ovaries, respectively, as compared with 2.0 ⫾ 0.9 and 2.0 ⫾ 0.8 from isolated follicles of similar groups. There were significantly more atretic changes in oocytes of follicles in intact tissue blocks (control) from frozen-thawed tissue compared with oocytes of isolated follicles from fresh and frozen-thawed tissue (P⬍.01), which included mostly vacuoles (30%), although their number in individual oocytes was low. However, low numbers of vacuoles were also present in oocytes from fresh tissue (20%), and two follicles from this group were very atretic. Moreover, in about half the nonisolated follicles, the oocytes’ nuclei were not centrally located, independent of atretic changes. Tables 1 and 2 represent the results of culturing fully and partially isolated follicles using various culture systems. There were no significant differences in the initial sizes of the follicles among the culture groups. Growth was noted only for fully isolated follicles cultured within collagen gels, and there were no differences between follicles from fresh and thawed tissue (Table 2). GCs numbered on average 14.4 ⫾ 6 in sections of freshly isolated follicles (n ⫽ 66) compared with 45.3 ⫾ 16.2 in sections of follicles that grew in collagen gel culture (n ⫽ 55; P⬍.0001). The two sera types and FSH doses had similar effects on the fully isolated follicles cultured in collagen gels (Table 2), although no growth was observed without FSH addition. Coculture with stromal cells could not be performed in collagen gels because the excess culture medium resulted in lack of gelling. 144
Abir et al.
Partially and fully isolated human follicles
Sixty percent of the slices with the partially isolated follicles detached from the collagen gels (53 follicles from frozen-thawed tissue and 36 follicles from fresh tissue), and the follicles did not survive (Table 2). Of the remainder, 15 of 35 (43%) follicles from frozen-thawed tissue and 10 of 24 (42%) from fresh tissue survived up to 3– 4 days in culture, but no growth was observed. The morphology of the follicles after isolation and culture is shown in Figure 1A to E.
DISCUSSION This study showed that isolated human follicles have more lipid droplets in their GCs than those in intact tissue specimens. Furthermore, only fully isolated but not partially isolated follicles can grow in culture, and only within a supporting collagen gel matrix, not on a matrix. Growth was indicated by an increase in the number of the GCs and in the size of the follicles. These results were similar for follicles from fresh or frozen-thawed ovarian tissue. An increase in lipid droplets in GCs might suggest that our isolation process did not produce very healthy follicles, as in a recent study, radiation induced accumulation of lipid droplets in oocytes of mouse unilaminar follicles (4). Although the lipid droplets may have been caused by the high concentration of collagenase used for isolation, the high percentage (up to 40%) of the fully isolated follicles that grew in collagen gel culture suggests that the increase in lipid droplets did not affect their follicular growth. We have no explanation as to why more atretic changes were observed in oocytes from intact tissue blocks or why, in a large number of these oocytes, the nuclei were not centrally located. The significant number of oocytes from intact frozenthawed tissue with atretic changes perhaps suggests that cryopreservation induces some changes in the follicles. Very recently, Gook et al. (5) conducted TEM studies of follicles after cryopreservation using a similar protocol to ours and found 67% vacuolation in the oocytes of primordial follicles; furthermore, only 56% of the oocytes were normal. Both these rates are higher than those noted in our study. Moreover, in most of our cases, the vacuolized oocytes contained very few vacuoles, whereas Gook et al. (5) did not report the amount of vacuoles found, nor did they examine the quality of the GCs. The differences between the studies may result from slight differences in the freezing and thawing procedures, such as shorter duration of prefreezing and thawing stages or differences in the TEM preparation methods. Because only fully isolated follicles devoid of stromal cells grew in culture, it is possible that the stromal cells induce growth-inhibiting signals rather than promote growth and that the rapid follicular growth was a result of the abolishment of inhibition signals. This would also explain the lack of growth of the partially isolated follicles, even in those remaining in the supporting collagen gels. However, it is also possible that the lack of growth of the partially Vol. 75, No. 1, January 2001
FIGURE 1 Photographs of human follicles. (A) Electron microscope section of a primordial follicle in an intact tissue block (nonisolated). Note the single lipid droplet (L; original magnification, ⫻5,000). (B) Electron microscope section of an isolated follicle. The GC layer is in transition between the primordial stage (flat GC) and the primary stage (cuboidal GC). Note the 11 lipid droplets (L; original magnification, ⫻3,780). (C) Micrograph of a unilaminar follicle that reached multilaminar stages after 24 hours’ culture. Note the multilaminar GC layers surrounding the oocyte and the cellular outgrowths. Size at 0 hours: 45 m; after 24 hours: 90 m (original magnification, ⫻400). (D) Section of a follicle digested from the collagen gel after 24 hours of culture. Size at 0 hours: 45 m; after 24 hours: 75 m. Note the GC layers and the oocyte. (Original magnification, ⫻400; staining with toluidine blue). (E) Section of two partially isolated primordial follicles cultured within a collagen gel for 4 days. Note the partially digested stroma layer (S) and that the follicles remained in their primordial stage after culture (original magnification, ⫻400; staining with hematoxylin and eosin).
Abir. Study of isolated human follicles. Fertil Steril 2001.
FERTILITY & STERILITY威
145
isolated follicles is attributable to stromal damage caused by the enzymatic treatment. Because the number of partially isolated follicles that remained within the collagen gels was low, we cannot draw definitive conclusions. It is noteworthy that our results are similar to the findings reported by Hovatta et al. (2) of low survival of partially isolated follicles cultured on ECM. Our study showed that the small and fragile fully isolated follicles can only grow within a supporting matrix. Despite the increase in GC layers (3) and GC cells, it is important to remember that our studies were purely morphological. Further growth parameters need to be examined, for example, DNA synthesis or bromodeoxyuridine labeling, to completely eliminate the remote possibility of cell swelling because of fluid absorption (3). Unfortunately, the current commercial collagen lots are too diluted to produce firm collagen gels, and such studies cannot be conducted. Nevertheless, it would be worthwhile to try culturing isolated follicles within other matrices, such as agar or agarose. Another noteworthy finding is failure of the follicles to grow beyond 24 hours of culture, possibly because of basement membrane damage caused by the collagenase treatment. If so, it is unclear why this damage did not affect their growth during the first 24 hours of culture. Moreover, because follicular growth in collagen gels was not observed without FSH addition and FSH receptors are not present in primordial follicles (1), it is likely that the follicles cultured were primary– early secondary and not primordial. These findings are consistent with the lack of growth of the smallest unilaminar follicles in collagen gel cultures (3). Because many human ovarian biopsies available for re-
146
Abir et al.
Partially and fully isolated human follicles
search are from older women or from ovaries with cysts, they contain a low number of follicles. Therefore, the ability to culture isolated follicles, and not only follicles in organ culture, is very important for future studies. Although organ culture is both quick and easy to perform (2) compared with cultures of isolated follicles, its use for human biopsy specimens would lead to culture of empty ovarian specimens. Therefore, parallel culture systems for both isolated and nonisolated human follicles need to be developed with optimization of their culture medium.
Acknowledgments: The authors are indebted to Ms. G. Ganzach from the Editorial Board of Rabin Medical Center for the English editing. We are also grateful to J. Zissu and Y. Katz from the Pathology Department at Rabin Medical Center for assistance with the electron microscope.
References 1. Abir R, Fisch B, Raz Ah, Nitke S, Ben Rafael Z. Preservation of fertility in women undergoing chemotherapy. Current approach and future prospects. J Assist Reprod Genet 1998;15:469 –76. 2. Hovatta O, Wright C, Kraustz T, Hardy K, Winston RM. Human primordial, primary and secondary follicles in long-term culture: effect of partial isolation. Hum Reprod 1999;14:2519 –24. 3. Abir R, Roizman P, Fisch B, Nitke S, Okon E, Orvieto R, et al. Pilot study of isolated early human follicles cultured in collagen gels for 24h. Hum Reprod 1999;14:1299 –301. 4. Lee CJ, Park HH, Do BR, Yoon Y, Kim, JK. Natural and radiationinduced degeneration of primordial and primary follicles in mouse ovary. Anim Reprod Sci 2000;59:109 –17. 5. Gook DA, Edgar DH, Stern C. Effect of cooling rate and dehydration regimen on the histologic appearance of human ovarian cortex following cryopreservation in 1, 2 propandiol. Hum Reprod 1999;14:2061– 8.
Vol. 75, No. 1, January 2001
Morphological study of fully and partially isolated early human follicles Ronit Abir, Ph.D.,a Benjamin Fisch, M.D., Ph.D.,a Shmuel Nitke, M.D.,a Elimelech Okon, M.D.,b Ahud Raz, M.D.,a,c and Zion Ben Rafael, M.D.a Rabin Medical Center, Beilinson Campus, Petah Tikva, and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
Received March 24, 2000; revised and accepted July 26, 2000. Partially supported by research grants from the Israel Cancer Association (Z.B., R.A.) and Leo Mintz Fund-Tel Aviv University (B.F., R.A.). Presented at the meetings of the European Society for Human Reproduction and Embryology (ESHRE), Goteborg, Sweden, June 21–24, 1998 and Bologna, Italy, June 24 –28, 2000. Address of correspondence: Ronit Abir, Ph.D., IVF Research Laboratory, Department of Obstetrics and Gynecology, Rabin Medical Center, Beilinson Campus, Petah Tikva 49100, Israel (FAX: 972-39240533; E-mail: ronitabir @hotmail.com). a Department of Obstetrics and Gynecology. b Department of Pathology. c Present address: Department of Obstetrics and Gynecology, Soroka Hospital, Beer Sheva, Israel. 0015-0282/01/$20.00 PII S0015-0282(00)01668-X
Objective: To compare the development of fully and partially isolated human follicles by using various culture systems. Design: Human ovarian material was incubated with collagenase and deoxyribonuclease. Fully and partially isolated follicles (30 –50 m) were dissected and studied under light and electron microscopy. The follicles were then cultured on and within various matrices. Fully isolated follicles were also cocultured with stromal cells. Setting: Rabin Medical Center, a major care and referral center. Patient(s): Women undergoing laparoscopy. Intervention(s): None. Main Outcome Measure(s): Microscopy studies, follicular measurements. Result(s): Electron microscopy studies revealed an excess of lipid droplets in the granulosa cells of freshly isolated follicles. An increase in follicular size and granulosa cell number was observed only in the fully isolated follicles cultured within collagen gels for 24 hours. Most of the partially isolated follicles detached from the collagen gels. When cultured on collagen, extracellular matrix, and poly-L-lysine, both the fully and the partially isolated follicles deteriorated within the first 24 hours; coculture with stromal cells had no beneficial effect. Conclusion(s): The excess in lipid droplets in granulosa cells of isolated follicles might suggest that the isolation process does not yield completely healthy follicles. However, despite this finding, our studies show that fully isolated follicles, but not partially isolated follicles, can grow within, but not on, a culture matrix. (Fertil Steril威 2001;75:141– 6. ©2001 by American Society for Reproductive Medicine.) Key Words: In vitro, cryopreservation, full and partial isolation, transmission electron microscopy (TEM), primordial and primary follicles
Most of the follicles in the human ovary are unilaminar primordial or primary follicles (30 –50 m; (1). Their transformation into preantral (multilaminar) follicles presumably occurs through interactions with the surrounding stromal cells, although it is unclear whether these cells transmit inhibitory or stimulatory signals. Stimulation with FSH begins only at the preantral stages. The in vitro maturation of unilaminar follicles derived from cryopreserved-thawed ovaries would be of great benefit to many infertile women, for example, former cancer patients who became infertile after anticancer treatment (1). There are two possible approaches to the culture of unilaminar follicles (1). The first is to culture whole slices of ovarian tissue (organ
culture), such that the structural integrity of the tissue and the interactions between the stromal cells and the follicles are maintained (2). Human follicles have been shown to survive in organ culture, and in some cases, secondary follicles were observed (2). However, despite the simplicity of this culture system, it does not allow direct monitoring at the beginning of culture (0 hour), which makes it unclear whether the follicles indeed originated from unilaminar follicles in culture. Furthermore, the low follicular density in many human ovarian biopsies may lead to culture of empty ovarian slices (1). The second approach is to culture isolated unilaminar follicles. This culture system enables direct monitoring of the follicles during culture (1), but the follicles are taken out of 141
their natural environment, which might eliminate growth induction. We have recently shown that fully isolated follicles can proliferate in a collagen gel culture (3). However, Hovatta et al. (2) found that partially isolated follicles grown on an extracellular matrix (ECM) did not survive well compared with follicles in organ culture. The aim of the present study was to compare the development of fully and partially isolated follicles on and within various matrices. The use of partially isolated follicles maintained the interactions with the stromal cells and allowed us to monitor individual follicles. A secondary aim of our study was to study the morphology of isolated follicles by light and electron microscopy studies.
MATERIALS AND METHODS Ovarian Material Ovarian material was obtained from patients undergoing laparoscopy for various ovarian cysts. These included 20 women aged 17– 40 years (mean ⫾ SD: 30 ⫾ 7). The study was approved by the hospital’s ethical committee.
Cryopreservation and Thawing of Ovarian Tissue In 15 cases, the tissue specimens underwent cryopreservation with a 1, 2 propandiol (PROH; Sigma, St. Louis, MO) and sucrose (Sigma) protocol (3). Before cryopreservation, the samples were transferred through increasing concentration gradients of the freezing solution for several minutes and then frozen slowly in a programmable freezer (Kryo 10; series 10/20, Planer Biomed, Sunbury on Thames, UK). Ovarian pieces were thawed by transferring them in decreasing concentration gradients of PROH and sucrose for several minutes.
Isolation of Follicles The isolation process has been described by us elsewhere (3). In brief, after incubation with a collagenase IX (5,944 U/mL; Sigma) and deoxyribonuclease IV (180 U/mL; Sigma) solution, the follicles were dissected, and fully isolated follicles were released by aspirating them through a fine-bore pipette. Fully isolated follicles ⬍35 m and secondary follicles with at least two full granulosa cell (GC) layers were not cultured because our team has found that they fail to grow in collagen gel culture (3). However, these criteria could not be applied to partially isolated follicles because they were connected to the stromal layer in groups.
Preparation of Stromal Cells After follicular isolation, the digested ovarian fragments were collected in medium and aspirated repeatedly with automatic pipettes (40 –200 m, Finnpipette, Helsinki, Finland). Individual cells were rinsed several times and cocultured with fully isolated follicles. 142
Abir et al.
Partially and fully isolated human follicles
Preparation of Serum Gonadotropin-deficient human blood was obtained from women after pituitary– gonadal suppression by a gonadotropin-releasing hormone agonist (Decapeptyl GmbH, Ferring, Kiel, Germany) for IVF treatment. The hormone was administrated at a dose of 0.1 mg per day for at least 14 days, and ovarian suppression was confirmed by measuring circulating E2 levels. Human serum (HS) was prepared and heat inactivated. The use of this serum enabled us to monitor the levels of human FSH added to the culture medium.
Collagen Gel Solutions Collagen gel solutions were prepared from a rat tail collagen solution according to the manufacturer’s instructions (Collaborative Biomedical Products, Bedford, MA; see ref. 3). No more than three fully isolated follicles or one piece of connective tissue with partially isolated follicles was placed in each well of four-well culture plates (Nunclon, Roskilde, Denmark) and overlaid with a 100-L collagen gel solution. The fully isolated follicles were incubated for 24 hours because longer culture periods have been found to result in disruption of the follicular structure (3). The partially isolated follicles were placed in the incubator for ⱕ4 days.
Collagen, ECM, and Poly-L-Lysine Coatings Rat tail collagen type I solution (Collaborative Biomedical Products) was diluted to a concentration of 50 g/mL, and ECM gel (Sigma) was diluted 1:1. Millicell CM inserts (12-mm diameter, 0.4 m pore size; Millipore, Bedford, MA) were fitted into 24-well plates (Nunclon) and coated with the diluted solutions according to the manufacturer’s instructions (Collaborative Biomedical Products and Sigma, respectively). To allow gelling, ECM-coated inserts were kept for half an hour in a humidified incubator at 37°C. Wells in four-well culture plates (Nunclon) were coated with a poly-L-lysine 0.01% solution (Sigma) and prepared further according to the manufacturer’s instructions. Fully and partially isolated follicles were placed on these culture systems, and stromal cells were added to some of the fully isolated follicles (Table 1).
Culture Medium The culture medium contained Earle’s balanced salt solution (Biological Industries, Beit Haemek, Israel) with 0.47 mM sodium pyruvate (Sigma); 2 mM/L L-glutamine (Biological Industries); 10% FBS (Biological Industries; see Reference (3) or HS, human FSH (Metrodin, Teva Pharmaceutical Industries, Petah Tikva, Israel) at doses of 0, 0.5, and 1 U/mL; and 50 U/mL penicillin G (sodium salt), 50 g/mL streptomycin sulfate, and 0.125 g/mL amphotericin B (Penicillin-Streptomycin-Amphotericin B2 Solution, Biological Industries). The follicles were measured with a calibrated eyepiece micrometer at the beginning of the culture (0 hours) and every 24 hours. Vol. 75, No. 1, January 2001
TABLE 1 Fully and partially isolated human follicles cultured on various coatings.a Group and system Fresh, fully isolated Collagen Collagen ⫹ stromal cells ECM ECM ⫹ stromal cells Poly-L-lysine Poly-L-lysine ⫹ stromal cells Fresh, partially isolated Collagen ECM Poly-L-lysine
0h (follicle m)
n
44.1 ⫾ 14.0 40.0 ⫾ 15.5 39.9 ⫾ 17.3 37.6 ⫾ 15.1 35.8 ⫾ 16.0
50b 55 55 60 60
40.0 ⫾ 18.2
60 Deteriorated within 24 h
45.0 ⫾ 22.0 40.5 ⫾ 22.0 45.0 ⫾ 18.1
60 Deteriorated within 24 h 70 Deteriorated within 24 h 70 Deteriorated within 24 h
After culture
Deteriorated Deteriorated Deteriorated Deteriorated Deteriorated
within within within within within
24 24 24 24 24
h h h h h
Note: Values are means ⫾ SD. All follicles were cultured with HS and 0.5 U/mL. b Number of follicles at 0 h.
morning, with an araldite resin (Sigma) mixture at 60°C for 2 hours while being periodically stirred. After infiltration, the samples were embedded in a fresh mixture of araldite resin (Sigma) and kept at 60°C for 72 hours. Semithin (0.5– 0.75 m) sections for light microscopy (LM) were stained with toluidine blue (BDH Chemicals Ltd, Poole, UK), and ultrathin (1Å) sections for transmission electron microscopy (TEM) were stained with uranil acetate (BDH) and lead citrate (BDH). The follicles were examined with a JEOL (JEM 1010) electron microscope. Signs of atresia in the oocyte and GCs were defined as follows: nuclear changes, number of lipid droplets, vacuoles and changes in shape of follicle or oocyte (4, 5). The index for the number of vacuoles and lipid droplets in the GCs was given a range of 1–3, with 1 indicating very few (up to three per follicle) and 3 indicating many (at least 10 per follicle). Shrunken oocytes with abnormal nuclei and with a large number of vacuoles were considered atretic.
a
Abir. Study of isolated human follicles. Fertil Steril 2001.
Histologic Preparation for Light Microscopy Fresh and frozen-thawed ovarian pieces of 0.5–1 mm (control samples); partially isolated follicles; cultured collagen gels with partially isolated follicles or pieces that were separated from the gels spontaneously, mechanically, or enzymatically at various culture stages; and partially isolated follicles cultured on the various matrices were fixed in Bouin’s solution and prepared for paraffin embedding. Mechanical separation of the follicles from the collagen gels was performed with 21-G needles (Terumo Europe N.V., Leuven, Belgium) attached to 1-mL syringes (Terumo Europe N.V.) and enzymatic separation after treatment with collagenase IX (2,972 U/mL; Sigma) for 30 minutes. Fully isolated follicles equilibrated in the incubator for 1 hour with culture medium before being fixed (control) and collagen gels containing fully isolated follicles after a 24-hour culture period were also prepared for histologic study (3). Follicles and their oocytes were measured with a calibrated eyepiece micrometer. The number of GCs was counted in sections immediately after isolation and after collagen gel culture.
Histologic Preparation for Transmission Electron Microscopy All samples were fixed in 3% glutaraldehyde (Sigma). The following three protocols were used. Control samples. Fresh and frozen-thawed tissue pieces ⱕ0.5–1 mm were fixed and postfixed with 1% osmium tetraoxide (Sigma). These samples were then dehydrated with increasing concentrations of alcohol and finally with acetone (AR grade; Biolab, Jerusalem, Israel). The specimens were infiltrated overnight at room temperature with an acetone-araldite resin (Sigma) solution and, the following FERTILITY & STERILITY威
Partially isolated follicles. The partially isolated follicles were fixed, postfixed, and dehydrated. They were infiltrated at 60°C for only 1 hour with an araldite resin (Sigma) mixture while being periodically stirred, after which they were embedded, processed, and examined. Fully isolated follicles. A proportion of the 24-hour gels were treated with collagenase IX (2,972 U/mL; Sigma) for 30 minutes, and the follicles were recovered and prepared for TEM. They were fixed, postfixed, and dehydrated as described above but without acetone. The follicles were then transferred at room temperature to an araldite resin (Sigma) mixture for a few minutes, after which they were embedded, processed, and examined.
Statistical Analysis Data were statistically analyzed by unpaired Student’s t test, analysis of variance (ANOVA), and 2 test.
RESULTS When examining LM sections of follicles and their respective oocytes from fresh or frozen-thawed tissue before and after isolation, no significant differences in their diameter were found (follicular diameter ⫾ SD: 37.9 ⫾ 11.7 to 45.9 ⫾ 9.0 m; oocyte diameter ⫾ SD: 26.2 ⫾ 10.1 to 30.0 ⫾ 7.5 m; 33– 40 follicles per group). No pyknotic cells or abnormal oocytes were observed under LM. For the TEM studies, 25 control follicles from fresh tissue, 38 control follicles from frozen-thawed tissue, 23 isolated follicles from fresh tissue, and 20 isolated follicles from thawed tissue were examined. There were no identifiable differences in the number and quality of the intracellular organelles between the oocytes and GCs of the partially isolated follicles and the control follicles from both fresh and frozen-thawed tissue, apart from an increase in the number of lipid droplets in the GCs of isolated follicles from both fresh and frozen-thawed tissue. 143
TABLE 2 Fully and partially isolated human follicles cultured within collagen gels.
Group and sera Fresh fully isolated FCS FCS HS HS Thawed fully isolated FCS FCS HS HS Fresh partially isolated HS Thawed partially isolated HS
Human FSH (U/mL)
Follicle size, m ⫾ SD (n) 0h
24 h
0.5 1.0 0.5 1.0
45.0 ⫾ 10.0 (60)a 41.1 ⫾ 14.2 (50) 35.0 ⫾ 6.9 (60) 36.0 ⫾ 11.1 (50)
69 ⫾ 15 (25)b 70.1 ⫾ 11.5 (19) 69.8 ⫾ 13.6 (25) 71.2 ⫾ 13.2 (21)
0.5 1.0 0.5 1.0
35 ⫾ 15 (70) 37.0 ⫾ 12.4 (70) 35.1 ⫾ 9.2 (70) 35.2 ⫾ 16.3 (70)
65.0 ⫾ 15.0 (27) 71.2 ⫾ 12.1 (28) 65.1 ⫾ 15.0 (26) 68.8 ⫾ 14.4 (29)
0.5
33.0 ⫾ 14.2 (88)
follicles detachedc
0.5
37.0 ⫾ 20.0 (60)
follicles detached
a
Number of follicles at 0 hours. b Number of growing follicles after 24 hours in culture. c See more details in Results. Abir. Study of isolated human follicles. Fertil Steril 2001.
These included 0.4 ⫾ 0.6 and 0.4 ⫾ 0.7 lipid droplets for follicles from intact tissue specimens of fresh and frozenthawed ovaries, respectively, as compared with 2.0 ⫾ 0.9 and 2.0 ⫾ 0.8 from isolated follicles of similar groups. There were significantly more atretic changes in oocytes of follicles in intact tissue blocks (control) from frozen-thawed tissue compared with oocytes of isolated follicles from fresh and frozen-thawed tissue (P⬍.01), which included mostly vacuoles (30%), although their number in individual oocytes was low. However, low numbers of vacuoles were also present in oocytes from fresh tissue (20%), and two follicles from this group were very atretic. Moreover, in about half the nonisolated follicles, the oocytes’ nuclei were not centrally located, independent of atretic changes. Tables 1 and 2 represent the results of culturing fully and partially isolated follicles using various culture systems. There were no significant differences in the initial sizes of the follicles among the culture groups. Growth was noted only for fully isolated follicles cultured within collagen gels, and there were no differences between follicles from fresh and thawed tissue (Table 2). GCs numbered on average 14.4 ⫾ 6 in sections of freshly isolated follicles (n ⫽ 66) compared with 45.3 ⫾ 16.2 in sections of follicles that grew in collagen gel culture (n ⫽ 55; P⬍.0001). The two sera types and FSH doses had similar effects on the fully isolated follicles cultured in collagen gels (Table 2), although no growth was observed without FSH addition. Coculture with stromal cells could not be performed in collagen gels because the excess culture medium resulted in lack of gelling. 144
Abir et al.
Partially and fully isolated human follicles
Sixty percent of the slices with the partially isolated follicles detached from the collagen gels (53 follicles from frozen-thawed tissue and 36 follicles from fresh tissue), and the follicles did not survive (Table 2). Of the remainder, 15 of 35 (43%) follicles from frozen-thawed tissue and 10 of 24 (42%) from fresh tissue survived up to 3– 4 days in culture, but no growth was observed. The morphology of the follicles after isolation and culture is shown in Figure 1A to E.
DISCUSSION This study showed that isolated human follicles have more lipid droplets in their GCs than those in intact tissue specimens. Furthermore, only fully isolated but not partially isolated follicles can grow in culture, and only within a supporting collagen gel matrix, not on a matrix. Growth was indicated by an increase in the number of the GCs and in the size of the follicles. These results were similar for follicles from fresh or frozen-thawed ovarian tissue. An increase in lipid droplets in GCs might suggest that our isolation process did not produce very healthy follicles, as in a recent study, radiation induced accumulation of lipid droplets in oocytes of mouse unilaminar follicles (4). Although the lipid droplets may have been caused by the high concentration of collagenase used for isolation, the high percentage (up to 40%) of the fully isolated follicles that grew in collagen gel culture suggests that the increase in lipid droplets did not affect their follicular growth. We have no explanation as to why more atretic changes were observed in oocytes from intact tissue blocks or why, in a large number of these oocytes, the nuclei were not centrally located. The significant number of oocytes from intact frozenthawed tissue with atretic changes perhaps suggests that cryopreservation induces some changes in the follicles. Very recently, Gook et al. (5) conducted TEM studies of follicles after cryopreservation using a similar protocol to ours and found 67% vacuolation in the oocytes of primordial follicles; furthermore, only 56% of the oocytes were normal. Both these rates are higher than those noted in our study. Moreover, in most of our cases, the vacuolized oocytes contained very few vacuoles, whereas Gook et al. (5) did not report the amount of vacuoles found, nor did they examine the quality of the GCs. The differences between the studies may result from slight differences in the freezing and thawing procedures, such as shorter duration of prefreezing and thawing stages or differences in the TEM preparation methods. Because only fully isolated follicles devoid of stromal cells grew in culture, it is possible that the stromal cells induce growth-inhibiting signals rather than promote growth and that the rapid follicular growth was a result of the abolishment of inhibition signals. This would also explain the lack of growth of the partially isolated follicles, even in those remaining in the supporting collagen gels. However, it is also possible that the lack of growth of the partially Vol. 75, No. 1, January 2001
FIGURE 1 Photographs of human follicles. (A) Electron microscope section of a primordial follicle in an intact tissue block (nonisolated). Note the single lipid droplet (L; original magnification, ⫻5,000). (B) Electron microscope section of an isolated follicle. The GC layer is in transition between the primordial stage (flat GC) and the primary stage (cuboidal GC). Note the 11 lipid droplets (L; original magnification, ⫻3,780). (C) Micrograph of a unilaminar follicle that reached multilaminar stages after 24 hours’ culture. Note the multilaminar GC layers surrounding the oocyte and the cellular outgrowths. Size at 0 hours: 45 m; after 24 hours: 90 m (original magnification, ⫻400). (D) Section of a follicle digested from the collagen gel after 24 hours of culture. Size at 0 hours: 45 m; after 24 hours: 75 m. Note the GC layers and the oocyte. (Original magnification, ⫻400; staining with toluidine blue). (E) Section of two partially isolated primordial follicles cultured within a collagen gel for 4 days. Note the partially digested stroma layer (S) and that the follicles remained in their primordial stage after culture (original magnification, ⫻400; staining with hematoxylin and eosin).
Abir. Study of isolated human follicles. Fertil Steril 2001.
FERTILITY & STERILITY威
145
isolated follicles is attributable to stromal damage caused by the enzymatic treatment. Because the number of partially isolated follicles that remained within the collagen gels was low, we cannot draw definitive conclusions. It is noteworthy that our results are similar to the findings reported by Hovatta et al. (2) of low survival of partially isolated follicles cultured on ECM. Our study showed that the small and fragile fully isolated follicles can only grow within a supporting matrix. Despite the increase in GC layers (3) and GC cells, it is important to remember that our studies were purely morphological. Further growth parameters need to be examined, for example, DNA synthesis or bromodeoxyuridine labeling, to completely eliminate the remote possibility of cell swelling because of fluid absorption (3). Unfortunately, the current commercial collagen lots are too diluted to produce firm collagen gels, and such studies cannot be conducted. Nevertheless, it would be worthwhile to try culturing isolated follicles within other matrices, such as agar or agarose. Another noteworthy finding is failure of the follicles to grow beyond 24 hours of culture, possibly because of basement membrane damage caused by the collagenase treatment. If so, it is unclear why this damage did not affect their growth during the first 24 hours of culture. Moreover, because follicular growth in collagen gels was not observed without FSH addition and FSH receptors are not present in primordial follicles (1), it is likely that the follicles cultured were primary– early secondary and not primordial. These findings are consistent with the lack of growth of the smallest unilaminar follicles in collagen gel cultures (3). Because many human ovarian biopsies available for re-
146
Abir et al.
Partially and fully isolated human follicles
search are from older women or from ovaries with cysts, they contain a low number of follicles. Therefore, the ability to culture isolated follicles, and not only follicles in organ culture, is very important for future studies. Although organ culture is both quick and easy to perform (2) compared with cultures of isolated follicles, its use for human biopsy specimens would lead to culture of empty ovarian specimens. Therefore, parallel culture systems for both isolated and nonisolated human follicles need to be developed with optimization of their culture medium.
Acknowledgments: The authors are indebted to Ms. G. Ganzach from the Editorial Board of Rabin Medical Center for the English editing. We are also grateful to J. Zissu and Y. Katz from the Pathology Department at Rabin Medical Center for assistance with the electron microscope.
References 1. Abir R, Fisch B, Raz Ah, Nitke S, Ben Rafael Z. Preservation of fertility in women undergoing chemotherapy. Current approach and future prospects. J Assist Reprod Genet 1998;15:469 –76. 2. Hovatta O, Wright C, Kraustz T, Hardy K, Winston RM. Human primordial, primary and secondary follicles in long-term culture: effect of partial isolation. Hum Reprod 1999;14:2519 –24. 3. Abir R, Roizman P, Fisch B, Nitke S, Okon E, Orvieto R, et al. Pilot study of isolated early human follicles cultured in collagen gels for 24h. Hum Reprod 1999;14:1299 –301. 4. Lee CJ, Park HH, Do BR, Yoon Y, Kim, JK. Natural and radiationinduced degeneration of primordial and primary follicles in mouse ovary. Anim Reprod Sci 2000;59:109 –17. 5. Gook DA, Edgar DH, Stern C. Effect of cooling rate and dehydration regimen on the histologic appearance of human ovarian cortex following cryopreservation in 1, 2 propandiol. Hum Reprod 1999;14:2061– 8.
Vol. 75, No. 1, January 2001