Alterations of the cytoskeleton and polyploidy induced by cryopreservation of metaphase II mouse oocytes

FERTILITY AND STERILITY VOL. 69, NO. 5, MAY 1998 Copyright 1998 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Alterations of the cytoskeleton and polyploidy induced by cryopreservation of metaphase II mouse oocytes A. Eroglu, Ph.D.* T. L. Toth, M.D., and M. Toner, Ph.D.* Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts Received July 28, 1997; revised and accepted November 11, 1997. Supported in part by research grants from The Whitaker Foundation, the American Cancer Society, and TAP Pharmaceuticals. Reprint requests: Mehmet Toner, Ph.D., Massachusetts General Hospital, Center for Engineering in Medicine, 55 Fruit Street, GRB 1401, Boston, Massachusetts 02114 (FAX: 617-374-5665; E-mail: mtoner@sbi.org). * Center for Engineering in Medicine and Surgical Services. Vincent Memorial Obstetrics and Gynecology Service. 0015-0282/98/$19.00 PII S0015-0282(98)00030-2 Objective: To determine cryopreservation-induced alterations in the cytoskeleton of metaphase II mouse oocytes and the implications of these alterations in functionality of the cytoskeleton and polyploidy after fertilization. Design: Comparative study. Setting: Clinical and academic research environment at a medical school teaching hospital. Intervention(s): Oocytes were frozen using a slow-cooling (0.5 C/min) and slow-thawing (8 C/min) protocol in 1.5 M dimethyl sulfoxide and 0.2 M sucrose and were analyzed before and after fertilization. Main Outcome Measure(s): Cytoskeletal alterations, fertilization, and polyploidy rates. Result(s): When analyzed immediately after thawing, the oocytes displayed dramatic cytoskeletal alterations. Only slight recovery was observed upon removal of the cryoprotectants. However, incubation after thawing of 1 hour at 37 C completely reestablished a normal microfilament and microtubule pattern while partially restoring normal spindle morphology and chromosome alignment. Accordingly, insemination immediately after removal of cryoprotectants resulted in a significantly decreased fertilization rate and aberrant dynamics of cytoskeleton-dependent events, whereas oocytes inseminated after the post-thaw incubation displayed fertilization rates and cytoskeletal dynamics comparable to those in controls. Cryopreservation did not increase polyspermy but significantly increased digyny when the oocytes were inseminated after the post-thaw incubation. All digynic eggs displayed an abnormal spindle remnant in comparison with diploid or polyspermic eggs. Conclusion(s): A brief period of incubation after thawing allows recovery and positively affects fertilization and cytoskeletal dynamics. Cryopreservation does not impair the functionality of microfilaments and cytoplasmic microtubules during postfertilization events. Our findings suggest that the increased rate of digyny in cryopreserved oocytes may be related to the spindle disorganization, leading to failure in segregation of the chromosomes, rather than to direct malfunction of the microfilaments in polar body formation. (Fertil Steril 1998;69:944 57. 1998 by American Society for Reproductive Medicine.) Key Words: Cryopreservation, cytoskeleton, digyny, fertilization, oocyte Oocyte cryopreservation is desired for both clinical and legal/ethical reasons in the treatment of human infertility. Although successful cryopreservation of mouse embryos was achieved as early as 1972 (1), oocyte cryopreservation is still far from being a clinically successful procedure. Several clinics around the world have attempted to cryopreserve human oocytes. However, only a few pregnancies with cryopreserved oocytes have been obtained so far (2 4). Studies using metaphase II oocytes showed that the damage associated with cryopreservation is multifactorial and includes oocyte lysis (5 7), premature zona hardening resulting in inhibition of sperm penetration (8), and disruption of microfilaments and spindle microtubules (9 12). Establishment of a clinically feasible cryopreservation technique warrants further fundamental research to clarify the mechanism of cryopreservation-induced damage at the cellular and molecular levels. Among the key concerns associated with oocyte cryopreservation is the disruption of cytoskeletal architecture, which is believed to lead to chromosomal anomalies as well as abnormal cytokinesis. It is well known that appropriate organization of spindle microtubules is essential for correct alignment and segregation of chromosomes. Further, the involvement of microfilaments and microtubules in events after fertilization has been demonstrated by the 944

use of inhibitors. For example, treatment of eggs with microfilament inhibitors (e.g., cytochalasins) prevented spindle rotation, polar body formation, pronuclear migration, and cytokinesis (12 15). In addition, the use of nocodazole, a drug that prevents tubulin polymerization by binding to the monomers, inhibited pronuclear formation and migration (14, 16). Therefore, it is highly likely that any injury causing disorganization of these cytoskeletal elements may ultimately lead to chromosomal anomalies and failure of fertilization and development. Previous studies have shown that cryoprotectants disrupt the cortical microfilament network (11 17). Exposure of metaphase II oocytes to cooling (10, 18, 19), cryoprotectants (9, 17), or a freeze-thaw cycle (12) caused depolymerization and disorganization of the spindle microtubules, eventually resulting in chromosomal scattering. However, the implications of these changes in the subsequent cytoskeleton-dependent events that occur after fertilization (i.e., spindle rotation, formation of the second polar body, pronuclear formation and migration, and cytokinesis) have not yet been studied. In this study, we used a slow freeze-thaw protocol and multiple-label fluorescence microscopy to investigate the effects of cryopreservation on microfilament, microtubule, and DNA patterns, as well as on cytoskeleton-dependent events after fertilization in a period extending from thawing of metaphase II oocytes to the two-cell stage. After the demonstration of cryopreservation-induced cytoskeletal alterations and their reversibility, this study describes the effect of post-thaw recovery on the dynamics of cytoskeleton-dependent events after fertilization. In addition, this study reveals the effects of cryopreservation on polyspermy and digyny (retention of the chromosomes of the polar body in the egg) and provides clues into the underlying mechanism. MATERIALS AND METHODS Media Various media were used for different stages of experiments. (I) Dulbecco s phosphate-buffered saline (PBS; Sigma, St. Louis, MO) supplemented with 4 mg/ml bovine serum albumin (BSA; Fraction V, Sigma) was used for oocyte handling in air and for dilution of cryoprotectants. (II) The freezing medium was PBS supplemented with 4 mg/ml BSA, 1.5 M dimethyl sulfoxide (DMSO; Aldrich, Milwaukee, WI), and 0.2 M sucrose (Sigma). (III) Sperm were dispersed in bicarbonate-buffered human tubal fluid (HTF) medium (Irvine Scientific, Santa Ana, CA) without BSA. (IV) In vitro fertilization and subsequent culture were performed in HTF medium containing 5 mg/ml BSA. Before use, drops of the HTF medium overlaid with silicone oil (Aldrich) were equilibrated overnight under a humidified atmosphere of 5% CO 2 in air. To remove impurities, the silicone oil was washed with HTF medium and then sterilized by autoclaving. Source of Oocytes Oocytes were collected from superovulated, 4- to 6-weekold, virgin F1 hybrid mice (C57B1/6 DBA; Charles River Laboratories, Wilmington, MA) maintained on a constant light-dark cycle (7 A.M. to7p.m. light; 7 P.M. to7a.m. dark). Before initiation of the study, approval was obtained from the Subcommittee on Research Animal Care at Massachusetts General Hospital. Superovulation was induced by intraperitoneal injections of 7.5 IU of pregnant mare serum gonadotropin (PMSG; Sigma) and 7.5 IU of hcg (Sigma) administered 49 50 hours apart. Thirteen to 14 hours after hcg injection, the oviducts were excised from euthanized females and were immediately placed in PBS. Freshly ovulated oocytes were released into PBS by puncturing the ampulla of each oviduct under a stereomicroscope (Nikon SMZ-2B, Melville, NY). Cumulus cells were then removed by a brief (3 4 minutes) exposure to 120 IU/mL bovine testis hyaluronidase (Type VI-S; Sigma) in PBS at ambient temperature. Immediately after dispersion of the cumulus, the oocytes were washed three times in PBS, and those with normal morphology were selected for experimentation. Cryopreservation of Oocytes We used the cryopreservation protocol described previously (20), with two modifications: [1] 0.2 M sucrose was added to the freezing medium as a nonpermeating cryoprotectant in addition to 1.5 M DMSO, and [2] upon thawing, DMSO was directly removed from the oocytes by placing them in a PBS solution containing 0.2 M sucrose at ambient temperature. For the addition of cryoprotectants, morphologically intact oocytes were transferred directly to precooled (4 C) 1.5 M DMSO 0.2 M sucrose in PBS for 30 minutes at 4 C. Ten to 20 oocytes were then aspirated into precooled 0.25-mL plastic straws (TS Scientific, Perkasie, PA) containing the same freezing medium. All of these manipulations were performed in a cold room at 4 C. After addition of the cryoprotectants, the straws were quickly transferred to a programmable freezer (Planer Kryo 10-II; TS Scientific) at 0 C and were cooled to 7 C at 2 C/min. Extracellular ice was seeded at 7 C by touching the straws with metal forceps prechilled in liquid nitrogen. After 15 minutes at this temperature, the straws were cooled to 80 C at a rate of 0.5 C/min and then plunged into liquid nitrogen. The oocytes were stored in liquid nitrogen for 1 15 days before thawing. For thawing, the straws were reintroduced into the programmable freezer at 80 C and were slowly warmed to 4 C at a rate of 8 C/min. Upon reaching 4 C, the straws were removed from the freezer and their contents were released directly into 0.2 M sucrose in PBS at ambient temperature for the removal of DMSO. After being held in FERTILITY & STERILITY 945

the 0.2-M sucrose solution for 10 minutes, the oocytes were transferred to PBS at ambient temperature for an additional 10 minutes. Before use, the oocytes were washed first in PBS and then in HTF at 37 C. Survival of oocytes after a freezethaw cycle was assessed by morphologic criteria. The oocytes were classified as morphologically viable if they displayed homogeneous cytoplasm, an intact plasma membrane, and the zona pellucida (ZP). In Vitro Fertilization and Culture In vitro fertilization and culture of inseminated oocytes were performed at 37 C under a humidified atmosphere of 5% CO 2 in air. For IVF, sperm were obtained from the cauda epididymides of a mature (4 6-month-old) BDF1 male (Charles River Laboratories). The cauda epididymides were dissected and placed in a large drop (0.5 ml) of preequilibrated HTF medium without BSA. Sperm were released into the medium by gently puncturing the epididymides (5 7 times) with a hypodermic needle and were allowed to disperse for 15 minutes at 37 C. After dispersion, the sperm concentration was determined using a hemocytometer. To obtain a final concentration of 1 2 10 6 sperm/ml, we added an appropriate volume of the sperm suspension to each insemination dish containing HTF with BSA supplementation. The insemination dishes were then incubated for 1 2 hours to capacitate sperm before the addition of oocytes. After 6 hours of incubation with sperm, the oocytes were washed in PBS and HTF and then were cultured in HTF. Immunocytochemical Labeling All chemicals used for fixation and labeling were bought from Sigma unless stated otherwise. Simultaneous fixation and extraction of oocytes were performed in a fixative (2% formaldehyde, 0.1% Triton X-100, 0.1 M Pipes (1,4-piperazine bis[ethanesulfonic acid]), ph 6.9, 5 mm MgCl 2 6H 2 O, 2.5 mm ethyleneglycol-bis-( -aminoethylether)-n, N, N, N -tetraacetic acid) containing 1 M taxol, 0.01% aprotinin, and 50% deuterium oxide to stabilize the microtubules (21). Oocytes from each experimental condition were transferred directly to the fixative without any manipulations before fixation (e.g., the ZP was not removed by acid or by enzyme treatment). Fluctuations in temperature and ph before fixation that might result in rapid depolymerization or polymerization of microtubules were avoided. After treatment with the fixative for 30 40 minutes at 37 C, the oocytes were washed in a blocking solution consisting of 0.02% sodium azide, 2% normal goat serum, 2% powdered milk, 2% BSA, 0.1 M glycine, and 0.01% Triton X-100 in PBS and were stored in the same blocking solution at 4 C until labeling within 1 30 days. Sperm attached to the ZP of inseminated oocytes were removed by repeated gentle micropipeting of the oocytes after fixation. To evaluate the microtubule, filamentous actin (f-actin) (microfilament), and DNA patterns of the cryopreserved oocytes, we performed triple fluorescence labeling. For tubulin labeling, fixed oocytes were incubated for 60 minutes at 37 C with a primary mouse anti -tubulin monoclonal antibody (Amersham, Arlington Heights, IL), diluted 1:100 in the aforementioned blocking solution, and subsequently were washed three times in the same blocking solution. These steps were followed by simultaneous exposure of the oocytes to a secondary fluorescein-conjugated goat antimouse immunoglobulin (Boehringer-Mannheim, Indianapolis, IN) and rhodamine-phalloidin (to visualize f-actin) (Molecular Probes, Eugene, OR), diluted 1:100 and 1:50, respectively, in the blocking solution, for 60 minutes at 37 C. After three washes in the blocking solution, the oocytes were mounted in an 80% glycerol-pbs mixture containing 1% n-propyl gallate as an antifading agent and 5 g/ml Hoechst 33342 (Molecular Probes) to label the DNA. Microscopy and Image Processing Differential interference contrast and fluorescence microscopy of specimens were performed using a Nikon Diaphot 300 microscope with the following objectives: Nikon Fluor 40 differential interference contrast (N.A. 1.3) and Nikon Plan Apo 60 differential interference contrast (N.A. 1.4). The microscope was equipped with fluorescein, rhodamine, and Hoechst selective filter sets and was interfaced with a Hamamatsu SIT camera (C2400 Photonic Microscopy, Oak Brook, IL). Images were acquired on a Macintosh Quadra 950 (Apple Computer, Cupertino, CA) with IPLab Spectrum software (Signal Analytics, Vienna, VA). Acquired images were processed using Adobe Photoshop 3.0 and Aldus Persuasion 2.1 software packages (Adobe Systems, Mountain View, CA), and results were printed on a Tektronix Phaser IISDX printer (Tektronix, Wilsonville, OR). Experimental Groups The effects of cryopreservation on the oocyte cytoskeleton were studied at various times both before and after fertilization, as follows. Prefertilization Experiments In this series of experiments, cryopreserved metaphase II oocytes were randomly divided into three experimental groups after thawing to evaluate microfilament and microtubule patterns, spindle organization, and chromosome alignment with respect to control oocytes. Only surviving oocytes as determined by morphologic criteria were fixed for examination. Group I comprised cryopreserved oocytes that were released into 0.2 M sucrose in PBS immediately after thawing and were fixed after 2 3 minutes. Group II oocytes were fixed after thawing and removal of cryoprotectants, as described earlier. Group III oocytes were fixed after thawing, removal of cryoprotectants, and incubation for 1 hour at 37 C. The control group included freshly ovulated oocytes, which were denuded, washed three times, and incubated in HTF for 1 hour at 37 C before fixation. Cytoskeletal patterns of the control oocytes were considered normal, and deviations from these were scored as abnormal. Spindle morphology was evalu- 946 Eroglu et al. Cytoskeleton and polyploidy in oocytes Vol. 69, No. 5, May 1998

ated separately from other microtubular changes in the cytoplasm, such as aster or network formation. Postfertilization Experiments Many events that occur after fertilization (i.e., spindle rotation, polar body formation, pronuclear formation and migration, and cytokinesis) are dependent on proper function of the cytoskeleton. Effects of cryopreservation on such events, as well as on polyspermy and digyny, were studied by inseminating the oocytes both immediately after (group I) and 1 hour after (group II) the removal of cryoprotectants. Groups I and II in the postfertilization experiments correspond to groups II and III in the prefertilization experiments, respectively, with the exception that oocytes were inseminated after analogous treatment instead of fixation. Oocytes were fixed at three time points (3 hours, 8 hours, and 24 hours) after insemination. The introduction of oocytes to the sperm suspension was considered time zero. These points were chosen according to data obtained in preliminary experiments of the dynamics of events after fertilization. Oocytes were considered fertilized when they were activated after sperm penetration, as assessed by differential interference contrast and fluorescence microscopy. Specifically, oocytes fixed 3 hours after insemination were considered fertilized when they resumed the second meiotic division and began to form the second polar body after sperm penetration, which induced local changes in f-actin organization. At 8 hours after insemination, in addition to the 3-hour criteria, the presence of at least two pronuclei (2PN) with an incorporated sperm tail was considered evidence of fertilization. At 24 hours after insemination, fertilization was determined by the number of two-cell embryos. In the case of developmental arrest at the pronuclear stage, an egg was considered fertilized if it displayed 2PN and the second polar body exhibited DNA staining. Evaluation of each event after fertilization was performed at different times. At 3 hours after insemination, we evaluated spindle rotation, second polar body formation, and pronuclear formation in conjunction with cytoskeletal reorganization. At 8 hours after insemination, pronuclear migration was examined in addition to spindle rotation and the formation of the second polar body and pronuclei, whereas cytokinesis (cleavage to the two-cell stage) was studied at 24 hours after insemination. The spindle was considered rotated when its long axis became perpendicular to the cell surface. Formation of the second polar body was scored as normal when the border of the second polar body was completely distinguishable by f-actin staining. Pronuclei were considered migrated when male and female pronuclei met in the center of the egg. Polyspermy was confirmed by the observation of more than one penetrated sperm head and sperm tail (3 hours after insemination) or by the observation of 2PN with corresponding sperm tails in the presence of a second polar body containing DNA staining (8 hours after insemination). Eggs were considered digynic when they displayed 2PN with one missing sperm tail and either no second polar body or a second polar body exhibiting no DNA staining. Statistical Analysis All results were analyzed by 2 test. However, Yates corrected 2 test or Fisher s exact test was used when one of the expected values in a contingency table was 10 or 5, respectively. Differences between groups were considered statistically significant at P 0.05. Experiments were performed at least three times for each group except for group I in the prefertilization experiments, which were performed twice. RESULTS Of 515 frozen oocytes, 488 (95%) survived the cryopreservation protocol, and 463 (95%) of those surviving were analyzed subsequently for cytoskeletal alterations. The difference between the number of oocytes that survived and the number of those analyzed is due to loss during various manipulations and the exclusion of seven parthenogenetically activated oocytes. Prefertilization Experiments To quantify alterations induced by cryopreservation in the cytoskeleton and in chromosome alignment, we analyzed the cryopreserved oocytes with respect to control oocytes that did not undergo cryopreservation. Control oocytes normally displayed uniform f-actin staining around the cortex, except in the area overlying the second meiotic spindle. This area was characterized by three distinctive zones: [1] a central area of intense f-actin staining just over the spindle, [2] an intermediate zone of relatively weak f-actin staining, and [3] an f-actin rich outside ring (Fig. 1A). Microtubules in control oocytes were observed exclusively in the second meiotic spindle, which appeared anastral and barrel-shaped with chromosomes aligned at the equatorial plate. The spindle was localized cortically and parallel to the cell surface. Approximately 97% of control oocytes displayed a normal f-actin pattern, spindle organization, and chromosome alignment, and all control oocytes showed a normal microtubule pattern. We next examined the effect of cryopreservation on f- actin organization. When fixed immediately after thawing (group I), the oocytes displayed nonuniform f-actin staining in the cortex, and the microfilament-rich area overlying the chromosomes on the meiotic spindle was no longer apparent (Fig. 1B). In addition, the submembranous microfilament network was disrupted, especially in the microfilament-rich area overlying the chromosomes (Fig. 1B). As summarized in Table 1, all oocytes in group I showed an abnormal f-actin pattern (P 0.0001 compared with control oocytes). Removal of cryoprotectants before fixation (group II) resulted FERTILITY & STERILITY 947

FIGURE 1 Cytoskeletal organization and chromosome alignment of metaphase II mouse oocytes during the period after thawing. Overlaid fluorescence images depict microfilaments (f-actin) (red), microtubules (green), and DNA (blue). (A), An unfrozen control oocyte fixed after 1 hour of incubation at 37 C. Arrowhead indicates the f-actin rich ring outside the relatively f-actin depleted intermediate zone. (B), A cryopreserved oocyte fixed immediately upon thawing (without removal of cryoprotectants). Arrowhead shows one of several microtubular asters; arrow indicates the first polar body. (C), A cryopreserved oocyte fixed immediately after the removal of cryoprotectants. (D), A cryopreserved oocyte fixed after the removal of cryoprotectants and subsequent incubation for 1 hour at 37 C. Arrowhead indicates f-actin rich ring; arrow shows the first polar body. Bar, 20 m. in restoration of a normal f-actin pattern in 14% of the oocytes. This percentage was not statistically different from that of group I (P 0.0908) and was significantly lower than that of the control group (P 0.0001). Although the disrupted microfilament network was repaired in the remaining group II oocytes (86%), as indicated by relatively uniform f-actin staining in the cortex, the characteristic microfilament-rich area overlying the chromosomes was not reestablished (Fig. 1C). However, when cryopreserved oocytes were incubated for 1 hour at 37 C after the removal of cryoprotectants (group III), 94% of the oocytes displayed normal microfilament organization, indicating full recovery from cryopreservation-induced alterations (Fig. 1D). We also evaluated the effect of cryopreservation on microtubule organization (Table 1). Tubulin staining demonstrated that all group I oocytes underwent dramatic changes, such as the appearance of numerous microtubular asters and a microtubular network formed by radiating arrays of these asters, as shown in Figure 1B. The number of group I oocytes showing these changes was statistically significant compared with control oocytes (P 0.0001). On removal of 948 Eroglu et al. Cytoskeleton and polyploidy in oocytes Vol. 69, No. 5, May 1998

TABLE 1 Effects of cryopreservation on the cytoskeleton and chromosome alignment of metaphase II mouse oocytes. Group No. of repeats Morphologic survival* Normal f-actin pattern Normal microtubule pattern Normal barrelshaped spindle Normal chromosome alignment Control 3 97% (29/30) 100% (30/30) 97% (29/30) 97% (29/30) Group I 2 92% (22/24) 0 (0/22) 0 (0/22) 9% (2/22) 41% (9/22) (Fixed upon thawing) Group II 4 96% (53/55) 14% (7/49) 69% (34/49) 18% (9/49) 78% (38/49) (Fixed after thawing and removal of cryoprotectants) Group III 4 91% (64/70) 94% (51/54) 100% (54/54) 70% (38/54) 80% (43/54) (Fixed after thawing, removal of cryoprotectants, and incubation for 1 h at 37 C) * Percentage of surviving oocytes assessed by homogeneous ooplasm, intact plasma membrane, and ZP (no. of surviving oocytes/no. of frozen oocytes). Percentage of oocytes showing normal F-actin and microtubule pattern, spindle shape, or chromosome alignment (no. of oocytes with normal characteristic/no. of oocytes analyzed). Differences between the numbers of oocytes that survived and those analyzed reflect the small number of oocytes lost during various manipulations. Patterns displayed by nonfrozen metaphase II control oocytes were considered normal.,, Within each column, values with different symbols are significantly different (see text for details). the cryoprotectants (group II), however, the oocytes recovered from cryopreservation-induced changes in the microtubule pattern. Microtubular asters and the network disappeared in 69% of group II oocytes, as illustrated in Figure 1C. Although this recovery was significant when compared with group I (P 0.001), the number of oocytes with a normal microtubule pattern in group II was still significantly less than that of control oocytes (P 0.0021). A post-thaw incubation for 1 hour at 37 C after removal of cryoprotectants (group III) completely reversed the cryopreservation-induced changes in microtubule pattern in all oocytes (Fig. 1D), representing a statistically significant recovery compared with group I (P 0.0001) and group II (P 0.0001) oocytes. To explore further the effects of cryopreservation on the oocyte cytoskeleton, we evaluated spindle morphology in each group. Only 9% of the cryopreserved oocytes fixed upon thawing (group I) displayed a barrel-shaped spindle (Table 1; P 0.0001 compared with control oocytes). The remaining oocytes (91%) in group I exhibited one of the following abnormal spindle characteristics: astral fusiform, reduction to poles, or a partial spindle. Removal of cryoprotectants (group II) did not significantly increase the proportion of cryopreserved oocytes having a barrel-shaped spindle compared with those in group I. In fact, normal spindle morphology was significantly lower in group II oocytes than in control oocytes (P 0.0001). Most oocytes in group II displayed an elongated and attenuated spindle (Fig. 1C), while a few had a fusiform spindle. However, after the incubation of oocytes for 1 hour at 37 C after thawing (group III), the barrel-shaped spindle appeared in 70% of the cryopreserved oocytes, as illustrated in Figure 1D. Although restoration of a barrel-shaped spindle was statistically significant when compared with group I (P 0.0001) and group II (P 0.0001), the percentage of group III oocytes with a normal spindle was still significantly less than that of control oocytes (P 0.0096). Finally, chromosome alignment was examined in all groups of oocytes (Table 1). Upon thawing, 41% of the cryopreserved oocytes (group I) had chromosomes that were aligned normally at the equatorial plate regardless of the spindle shape. This percentage was significantly lower than that of the control oocytes exhibiting normal chromosome alignment (P 0.0001). However, after the removal of cryoprotectants (group II), 78% of the cryopreserved oocytes showed normal chromosome alignment. Although this percentage was significantly greater than that observed in group I (P 0.006), it was lower than that in the controls (P 0.0483). When incubation was used after thawing (group III), 80% of the oocytes displayed normal chromosome alignment. This was still significantly lower than that of the controls (P 0.0484); thus, 1 hour of incubation could only partially reverse the cryopreservation-induced damage to chromosome alignment. Postfertilization Experiments To elucidate the effect of post-thaw handling on the cytoskeleton-dependent events after fertilization and polyploidy, we divided the cryopreserved oocytes into two groups. In group I, cryopreserved oocytes were inseminated immediately after removal of cryoprotectants; in group II, the oocytes were incubated for 1 hour at 37 C before insemination. Fertilization of Cryopreserved Oocytes Fertilizability of cryopreserved oocytes was determined in each group at 3 hours, 8 hours, and 24 hours after insemination by assessing the activation of oocytes after FERTILITY & STERILITY 949

TABLE 2 Postfertilization events in mouse eggs cryopreserved at the metaphase II stage. Group No. of repeats No. of hours after insemination Fertilization rate* Spindle rotation Second polar body formation Pronuclear formation Pronuclear migration Completed cytokinesis Control 6 3 92% (55/60) 95% (52/55) 95% (52/55) 27% (15/55) 8 95% (62/65) 100% (62/62) 100% (62/62) 100% (62/62) 79% (49/62) 24 97% (38/40) 95% (61/64) Group I 5 3 62% (28/45) 32% (9/28) 29% (8/28) 0 (0/28) (Inseminated after thawing 8 73% (38/52) 95% (36/38) 95% (36/38) 97% (37/38) 43% (16/37) and removal of 24 79% (45/57) 87% (39/45) cryoprotectants) Group II 6 3 86% (48/56) 88% (42/48) 88% (42/48) 19% (9/48) (Inseminated after thawing, 8 92% (54/59) 93% (50/54) 93% (50/54) 98% (53/54) 66% (35/53) removal of cryoprotectants, and incubation for 1hat 37 C) 24 93% (64/69) 92% (59/64) * Percentage of eggs fertilized (no. of eggs fertilized/no. of eggs analyzed). Eggs were considered fertilized when they were activated after sperm penetration. Percentage of eggs showing the respective event (no. of eggs with the respective event/no. of eggs fertilized). Percentage of the two-cell embryos after culture for 24 h (no. of eggs cleaved/no. of eggs fertilized)., Within each column, values with different symbols for corresponding rows (3 h, 8 h, and 24 h after insemination) are significantly different (see text for details). sperm penetration. These results are summarized in Table 2. When examined at 3 hours after insemination, fertilization rates in group II (86%) and in the control group (92%) were statistically similar, and both were significantly higher than those of group I (62%), with P values of 0.0129 and 0.0006, respectively. At 8 hours after insemination, statistically similar fertilization rates were observed in group II (92%) and in control oocytes (95%). Although some improvement in the fertilization rate of group I oocytes (73%) was evident after extended incubation until 8 hours after insemination, it was still significantly lower than that of group II (P 0.0202) and of controls (P 0.0017). At 24 hours after insemination, the fertilization rate of group I (79%) did not change significantly and remained low when compared with that of group II (93%, P 0.046) and controls (97%, P 0.0043). Postfertilization Cytoskeletal Organization By 3 hours after insemination, eggs from different groups were found at various stages of spindle rotation and second polar body formation. Accordingly, the eggs displayed different microfilament and microtubule organizations (Fig. 2A C). A number of eggs (typically from group I) exhibited a nonrotated spindle. These eggs formed a shoulder (domain) rich in f-actin over each set of maternal chromosomes as they had migrated to the poles of the spindle (Fig. 2B). The equator of the spindle appeared to be attached to the membrane where a furrow was formed between the shoulders. Intense f-actin staining at the furrow indicated the formation of a microfilament ring around the equator. In most cases, the shoulders over each set of chromosomes displayed f-actin organization similar to that observed over the chromosomes on the metaphase II plate (inset in Fig. 2B). A similar domain rich in f-actin was also observed over the swollen, decondensed sperm head (Fig. 2C). Otherwise, f-actin staining was uniform in the cortex. It is noteworthy that chromosome segregation appeared to be incomplete in a few eggs with nonrotated spindles (14% in group I and 10% in group II). These eggs displayed f-actin organization between the metaphase II pattern and the pattern described above. With respect to the microtubule pattern, fine cortical asters appeared in 36% of the eggs with a nonrotated spindle; in the remaining 64%, microtubules were observed exclusively in the spindle. Group II and control eggs fixed 3 hours after insemination typically showed a rotated spindle and a further arrangement of microfilaments for second polar body formation (i.e., the border of the second polar body was becoming or had become clear and intensely stained for f-actin) (Figs. 2A, 2C). The majority of these eggs (95%) had microtubular asters. By 8 hours after insemination, spindle rotation and formation of the second polar body and pronuclei were complete in all groups (Figs. 2D F), and arrangement of the microfilaments and microtubules appeared to be changing according to the location of the pronuclei. Most of the group I eggs displayed nonmigrated, eccentric pronuclei (Fig. 2E). The microtubule pattern was characterized by a network formed by several cortical asters. In contrast to group I, the majority of control and group II eggs displayed pronuclei that had migrated to the center of the cell (Figs. 2D, 2F). In these eggs, the microtubular network became denser and the f-actin staining around the pronuclei increased as they migrated toward the cell center. Eggs examined 24 hours after insemination generally 950 Eroglu et al. Cytoskeleton and polyploidy in oocytes Vol. 69, No. 5, May 1998

FIGURE 2 Cytoskeletal organization of fertilized mouse eggs at 3 hours, 8 hours, and 24 hours after insemination. Overlaid fluorescence images depict microfilaments (f-actin) (red), microtubules (green), and DNA (blue). (A), An unfrozen control egg fixed at 3 hours after insemination displaying rotated spindle and formed second polar body. Arrow indicates recondensing sperm head and overlying f-actin organization, which are slightly out of focus and not clearly visible. (B), A typical cryopreserved egg inseminated immediately after removal of cryoprotectants and fixed at 3 hours after insemination (group I). Arrow indicates recondensing sperm head. Inset shows the two f-actin rich shoulders from the top view in another group I egg having a nonrotated spindle. (C), A typical cryopreserved egg inseminated after a 1-hour incubation at 37 C and fixed 3 hours after insemination (group II). Arrow shows the recondensing sperm head, which induced typical f-actin organization in the cortex. (D), A fertilized control egg fixed at 8 hours after insemination. Arrowheads show migrated male and female pronuclei. Arrow indicates the nucleus of the polar body, which is partially obscured by intense staining for f-actin around the second polar body. (E), A typical cryopreserved and subsequently fertilized egg inseminated immediately after the removal of cryoprotectants and fixed at 8 hours after insemination (group I). Arrowheads show nonmigrated male and female pronuclei; arrows indicate two second polar bodies (each with a small nucleus) formed around the two poles of the tripolar spindle. Small arrow shows focal microfilament enrichment over the female pronucleus. (F), A typical cryopreserved and subsequently fertilized egg inseminated after a 1-hour incubation at 37 C and fixed 8 hours after insemination (group II). Arrowhead shows migrated and overlaid male and female pronuclei. Arrow indicates the nucleus of the polar body. (G), An embryo at the two-cell stage derived from an unfrozen control oocyte. Arrow indicates the second polar body and its nucleus. (H), An uncleaved pronuclear egg derived from a cryopreserved oocyte inseminated immediately after the removal of cryoprotectants and fixed at 24 hours after insemination (group I). Arrow shows the second polar body, which is intensely stained for f-actin and contains a small nucleus. Note that this polyspermic egg contains four pronuclei. (I), An embryo at the two-cell stage derived from a cryopreserved oocyte inseminated after a 1-hour incubation at 37 C and fixed 24 hours after insemination (group II). Arrow indicates a polar body containing a small nucleus. Bar, 20 m. FERTILITY & STERILITY 951

displayed a normal pattern of cytokinesis and cytoskeletal organization in all groups. In two-cell embryos, the DNA pattern was characterized by a large central nucleus with a varying number of prominent nucleoli in each blastomere and a small nucleus in the polar body (Figs. 2G, 2I). Microtubules appeared in both blastomeres in the form of a dense network surrounding the nuclei in the blastomeres as well as in the second polar body. Microfilaments were also found enriched around nuclei, in the cleavage furrow, and around the second polar body. In all three groups, a small number of eggs remained uncleaved at 24 hours after insemination. These eggs displayed no apparent abnormalities in the cytoskeleton. However, most of them were polyploidic (Fig. 2H). Quantification of Postfertilization Events Related to the Cytoskeleton To evaluate the implications of our findings after fertilization, we analyzed large numbers of oocytes in all groups and compared the differences using statistical approaches (Table 2). One of the cytoskeleton-related events evaluated after fertilization was spindle rotation. The percentages of fertilized eggs having a rotated spindle at 3 hours after insemination (Fig. 2A) were 95%, 32%, and 88% for controls, group I, and group II, respectively. These results represent a significant difference between group I and the other two groups (P 0.0001 for group I vs. group II; P 0.0001 for group I vs. control) (Table 2). However, at 8 hours after insemination, a similar percentage of eggs in group I (95%), group II (93%), and the control group (100%) displayed a rotated spindle. These findings clearly indicate that the low rate of spindle rotation in group I at 3 hours after insemination was due to a delay rather than to complete inhibition of the cytoskeleton-related process. Formation of the second polar body was then investigated. The results were similar to those found for spindle rotation. Significantly fewer group I eggs (29%) formed the second polar body at 3 hours after insemination than in group II (88%, P 0.0001) or in controls (95%, P 0.0001). At 8 hours after insemination, however, the percentage of eggs exhibiting second polar body formation in group I (95%) reached the same level as in group II (93%) and controls (100%) (Table 2). There were no significant differences between group II and the control group regarding second polar body formation at 3 hours and 8 hours after insemination, indicating a complete recovery from cryopreservation with respect to second polar body formation. The two other events examined after fertilization involved pronuclear formation and migration. At 3 hours after insemination, small pronuclei developing toward normal size were already found in 19% of group II and in 27% of the control eggs, whereas no group I eggs fixed 3 hours after insemination displayed such pronuclei (Table 2). These findings represent a statistically significant difference between group I and the control group (P 0.0059) and between group I and group II (P 0.0382). At 8 hours after insemination, pronuclear formation in group I (97%) was nearly complete and at a percentage not significantly different from that in group II (98%) or the control group (100%). With respect to pronuclear migration at 8 hours after insemination, 43% of group I, 66% of group II, and 79% of the control eggs had completed pronuclear migration, representing a statistically significant difference between group I and the control group (P 0.0007) and between group I and group II (P 0.0318) (Table 2; Figs. 2D F). Finally, cytokinesis (cleavage to the two-cell stage) was determined at 24 hours after insemination in each group (Table 2). Among the fertilized eggs, cleavage rates to the two-cell stage were comparable in group I (87%), group II (92%), and controls (95%), indicating that cryopreservation did not impair cytokinesis. Polyploidy in Cryopreserved Oocytes To determine the effect of cryopreservation on polyploidy, we analyzed frozen-thawed and nonfrozen control oocytes for polyspermy at 3 hours and 8 hours after insemination and for digyny only at 8 hours after insemination. This time point was used for the evaluation of digyny because of the need to wait for the completion of second polar body formation in all groups. The incidence of digyny in the eggs fixed at 8 hours after insemination was significantly higher in group II (19%) than in the control group (0) (P 0.0013) but was not significantly different between group I (5%) and group II or between group I and the control group (Fig. 3A). Among the total of 12 digynic eggs (2 in group I and 10 in group II), two types were distinguished depending on the presence of the second polar body. The first type of digynic eggs (2 in group I, 5 in group II) failed to form a normal second polar body, resulting in the retention of both sets of maternal chromosomes in the egg and the formation of a female pronucleus or two female pronuclei (Figs. 4A, 4B). These eggs exhibited a total of three regular pronuclei in most cases, one or two small round bodies that were stained for f-actin but displayed no DNA staining, and a weakly stained first polar body with degenerated chromosomes. Instead of a prominent spindle remnant, these eggs contained a tiny remnant of the spindle consisting of a few microtubular fibers with the small round body or bodies located on both sides of this spindle remnant (Fig. 4A). The presence of only one incorporated sperm tail as observed by differential interference contrast microscopy (Fig. 4B) confirmed that although one pronucleus was of paternal origin, the two others were of maternal origin. The second type of eggs considered to be digynic (5 in group II) formed a relatively normal second polar body which, however, exhibited no DNA staining (Figs. 4C, 4D). 952 Eroglu et al. Cytoskeleton and polyploidy in oocytes Vol. 69, No. 5, May 1998

FIGURE 3 Digyny (A) and polyspermy (B) in mouse eggs cryopreserved at the metaphase II stage. Data were pooled from six, five, and six experiments for controls, group I, and group II, respectively, and were expressed as percentage of eggs fertilized. The number at the bottom of each column represents the total number of eggs analyzed in each respective trial. Differences between the control group and group II with respect to digyny were significant at P 0.0013. These eggs displayed three regular pronuclei in all cases. They also exhibited a weakly stained or unstained first polar body with degenerated chromosomes, a sperm tail, and a tiny spindle remnant that seemed to be rotated, with one pole in the egg and the other one either inside or outside the second polar body. Although the spindle remnant in the second type of digynic eggs was more normal than that in the first type of digynic eggs, it was still considerably smaller than the regular spindle remnants observed in diploid or polyspermic eggs (Figs. 4A, 4C). Other cytoskeletal characteristics of both types of digynic eggs were similar to those of diploid eggs described earlier. Rates of polyspermy in group I (7%), group II (13%), and in the control group (18%) were not statistically different when analyzed at 3 hours after insemination (Fig. 3B). Similarly, at 8 hours after insemination, differences in the percentages of polyspermic eggs in group I (8%), group II (9%), and in the control group (16%) were not statistically significant (Fig. 3B). The cytoskeletal organization of the polyspermic eggs fixed 3 hours and 8 hours after insemination resembled that of normal fertilized eggs, except that the FERTILITY & STERILITY 953

former exhibited an enrichment of microfilaments over each sperm head (Fig. 4E) and the latter displayed an extra pronucleus or pronuclei (Figs. 4F, 4G). DISCUSSION This study explores the effects of cryopreservation on cytoskeletal organization and dynamics in metaphase II mouse oocytes after a slow freeze-thaw cycle, through fertilization, and up to the two-cell stage. Using multiple-label fluorescence microscopy targeting microfilaments, microtubules, and DNA, we accurately defined cryopreservationinduced alterations in the cytoskeleton, the meiotic spindle, and chromosome alignment. In addition, we monitored the degree of recovery from these alterations and the subsequent arrangement of the cytoskeletal elements after fertilization. This study shows that when an appropriate protocol is applied for cryopreservation and the oocytes are handled properly, fertilization and cleavage rates comparable to control values can be achieved after cryopreservation of mouse metaphase II oocytes. Cytoskeletal Organization and Dynamics Before and After Fertilization Oocytes examined immediately after thawing displayed several profound cytoskeletal alterations: disruption of the submembranous microfilament network, extensive appearance of microtubular asters and arrays, and disorganization of the spindle leading to chromosomal scattering. We concluded that these alterations were primarily due to the freeze-thaw process because exposure of oocytes to the cryoprotectants at 4 C for 30 minutes did not induce any significant cytoskeletal alterations except for some fine microtubular asters (data not shown). A brief incubation of 1 hour at 37 C after thawing and subsequent removal of the cryoprotectants resulted in a complete recovery from cryopreservation-induced alterations in both microfilament and microtubule patterns. The restoration of a normal barrel-shaped spindle and normal chromosome alignment following incubation after thawing was found in most, but not in all, of the cryopreserved oocytes. These findings are consistent with earlier observations in which spindle organization and chromosomal alignment partially recovered after exposure of mouse oocytes to cooling (10) or cryoprotectants (9). Similar results have been reported by Aigner et al. (12), who observed severe spindle damage and the formation of microtubular asters and networks shortly after thawing. They also reported that after recuperation at 37 C, the number of metaphase II mouse oocytes with normal spindle organization increased and the asters and networks disappeared. By contrast, George and Johnson (22) reported no statistically significant difference between frozen and control metaphase II mouse oocytes in microfilament organization, spindle morphology, or chromosome alignment after thawing. This discrepancy may be due to the fact that the fixation procedure of George and Johnson (22) took 30 minutes, during which the oocytes were held at 37 C and had time to recover. In our study, fixation immediately after thawing was performed within 3 minutes at ambient temperature. In addition, the percentage of control oocytes that showed normal spindle morphology and chromosome alignment in that study was low (78%) relative to our study (97%) and to the study of Aigner et al. (12) (91%). The lower incidence of normal spindle organization and chromosome alignment in the control group observed by George and Johnson (22) might be due to the exposure of control oocytes to suboptimal conditions. In fact, they also reported that after a post-thaw incubation period (3 hours at 37 C), the proportion of control oocytes that displayed normal spindle organization and chromosome alignment increased to 99%, while that for cryopreserved oocytes remained unchanged (70%). These results after incubation for control and frozen-thawed oocytes are similar to those (97% and 70%, respectively) observed in this study after incubation for 1 hour at 37 C. Studies in other species have reported various cytoskeletal responses to cooling, to cryoprotectants, and to the complete process of cryopreservation. With respect to the human oocyte, somewhat opposing results have been reported. Pickering et al. (18) observed that exposure to room temperature for 10 minutes disrupted the spindle in half of the oocytes, whereas 30 minutes of exposure to cold resulted in total spindle disruption in all cases. In their study, approximately 30% of the oocytes exposed to room temperature were able to restore a normal spindle after 1 4 hours of recovery at 37 C. They explained this low level of recovery as compared with murine oocytes by the relative absence of pericentriolar material in the human oocyte. On the other hand, Gook et al. (7) reported that 60% of human oocytes surviving cryopreservation displayed a normal spindle and chromosome alignment, compared with 81% for controls. This observation was related to the protective effect of propanediol, which was used as a cryoprotectant. Vincent et al. (17) observed that exposure of rabbit oocytes to DMSO did not change microfilament organization, whereas treatment with propanediol eradicated the cortical microfilaments. However, these changes were fully reversible after removal of the cryoprotectant. They also reported that exposure of oocytes to room temperature did not affect spindle organization, but freezing and thawing resulted in abnormal spindle morphology, which was partly reversible after post-thaw incubation for 3 hours at 37 C. In contrast, bovine oocytes appear to be more sensitive to cold. Exposing bovine oocytes to room temperature for 30 minutes depolymerized the spindle microtubules in 93% of the oocytes (19). After a 1-hour recovery at 37 C, only 8% of the oocytes displayed a normal spindle. Together, these data demonstrate a variable, species-dependent response to cooling, cryoprotectants, and cryopreser- 954 Eroglu et al. Cytoskeleton and polyploidy in oocytes Vol. 69, No. 5, May 1998

FIGURE 4 Cytoskeletal organization of digynic and polyspermic mouse eggs cryopreserved at the metaphase II stage. Overlaid fluorescence images depict microfilaments (f-actin) (red), microtubules (green), and DNA (blue). (A), A digynic egg inseminated after thawing and after removal of cryoprotectants (group I). Arrowheads show one relatively eccentric and two overlapped centric pronuclei. Arrow indicates a small round body, which is formed next to the tiny spindle remnant and is stained for f-actin but exhibits no DNA staining. Inset depicts the spindle remnant of a control egg at the same magnification and under the same conditions for comparison purposes. (B), A differential interference contrast image of the same egg shown in A. Arrowhead shows the single incorporated sperm tail in the cytoplasm. Arrow indicates the remnant of the first polar body (PB1). (C), A digynic egg inseminated after post-thaw incubation for 1 hour at 37 C (group II). Arrowheads show three centric pronuclei. Arrow indicates the second polar body, which is intensely stained for f-actin but exhibits no DNA staining, indicating that both sets of maternal chromosomes have remained in the egg and have formed two female pronuclei. (D), Differential interference contrast image of the same egg shown in C. Arrowhead shows part of the single incorporated sperm tail. Arrows indicate the second polar body and the remnant of the first polar body. (E), A trispermic egg inseminated after post-thaw incubation for 1 hour at 37 C and fixed at 3 hours after insemination. Arrows show swollen, decondensed sperm heads. Arrowhead indicates one of the microtubular asters. (F), A dispermic egg inseminated after the removal of cryoprotectants and fixed at 8 hours after insemination. Arrows show male pronuclei. Arrowhead indicates focal enrichment of microfilaments over a male pronucleus, which is close to the cell surface. (G), Differential interference contrast image of the dispermic egg shown in F. Arrowheads indicate two incorporated sperm tails adjacent to each male pronucleus. Bars, 20 m. FERTILITY & STERILITY 955