| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
University of Bielefeld,3 Faculty of Biology, Institute of Gene Technology/Microbiology, D-33501 Bielefeld, Germany
Department of Gynaecology and Obstetrics,4 Women's Hospital, Justus-Liebig-University Giessen, D-35392 Giessen, Germany
Department of Biology,5 Gettysburg College, Gettysburg, Pennsylvania 17325
ABSTRACT
2-Methoxyestradiol (2-ME) is a metabolite of 17beta-estradiol and a natural component of follicular fluid. Local concentrations of 2-ME may be increased by exposure to environmental pollutants that activate the expression of enzymes in the metabolic pathway from 17beta-estradiol to 2-ME. It has been suspected that this may have adverse effects on spindle formation in maturing oocytes, which would affect embryo quality. To study the dose-response patterns, we exposed denuded mouse oocytes to 2-ME during in vitro maturation. Meiotic progression, spindle morphology, centrosome integrity, and chromosome congression were examined by immunofluorescence and noninvasive polarizing microscopy (PolScope). Chromosomal constituents were assessed after spreading and C-banding. 2-ME sustained MAD2L1 expression at the centromeres and increased the number of meiosis I-blocked oocytes in a dose-dependent manner. 2-ME also caused dramatic dose-dependent increases in the hyperploidy of metaphase II oocytes. Some of these meiosis II oocytes contained anaphase I-like chromosomes, which suggests that high concentrations of the catecholestradiol interfere with the physical separation of chromosomes. Noninvasive PolScope analysis and tubulin immunofluorescence revealed that perturbations in spindle organization, which resulted in severe disturbances of the chromosome alignment at the spindle equator (congression failure), were caused by 2-ME at meiosis I and II. Pericentrin-positive centrosomes failed to align at the spindle poles, and multipolar spindles and prominent arrays of cytoplasmic microtubule asters were induced in 2-ME-exposed metaphase II oocytes. In conclusion, a micromolar level of 2-ME is aneugenic for mammalian oocytes. Therefore, exposure to 2-ME and conditions that increase the intrinsic local concentration of 2-ME in the ovary may affect fertility and increase risks for chromosomal aberrations in the oocyte and embryo.
environment, estradiol, meiosis, oocyte development, toxicology
2-Methoxyestradiol (2-ME) is an endogenous metabolite of 17β-estradiol (E2) [1–4]. 2-ME is a potent inhibitor of angiogenesis [5], modulates steroidogenesis and luteal cell function, and affects cell proliferation and differentiation in the ovary [6–8]. 2-ME is normally produced in the granulosa cells of the ovary [4, 9, 10], depending upon the stage of the estrus cycle [11–14]. The hydroxylation of E2 is catalyzed by the cytochrome 450 CYP1A1 and CYP1B1 enzymes [1, 9], which are expressed constitutively in the granulosa and luteal cells of the ovary. Conversion of 2-hydoxyestradiol to 2-ME is catalyzed by catechol-O-methyltransferase (COMT), an enzyme that is also expressed in human ovarian tissue [15].
Exposure to environmental pollutants may cause a critical increase in the local concentration of 2-ME in the ovary through the activation of the arylhydrocarbon receptor (AHR) [16], which is a basic helix-loop-helix transcription factor that is required for normal germ cell dynamics in the ovary [17, 18]. AHR induces the expression of cytochrome P450-metabolizing enzymes [19–21]. Persistent organic pollutants, such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) [15, 21–24] may generate acute or chronic local increases in 2-ME in the oocyte microenvironment, especially when the level of estrogen in the follicular fluid is high, since these pollutants activate the AHR as well as both enzymes [13, 16, 25, 26].
2-ME binds to tubulin at the same site as colchicine and nocodazole [26] and can alter the kinetics of polymerization [27, 28]. The oocyte spindle is highly dynamic, especially during the meiotic prometaphase and metaphase stages [29–32]. Therefore, exposure to 2-ME may affect spindle integrity and function, as is the case for other aneugenic drugs, such as taxol [33, 34]. Exposure to 3 µM 2-ME has been shown to cause spindle abnormalities in bovine oocytes and to inhibit embryo development [16]. Therefore, we studied the dose-response of in vitro maturing mouse oocytes to 2-ME with respect to oocyte degeneration, spindle aberrations, disturbances in chromosomes alignment at the spindle equator (congression failure), induction of meiotic delay or arrest, and persistence of the MAD2L1 protein at centromeres, indicating prolonged activity of the spindle assembly/attachment checkpoint (SAC).
At anaphase I of meiosis, the anaphase-promoting complex/cyclosome (APC/C), which is an ubiquitin-ligase complex, is activated, thereby targeting unique proteins for proteolysis by proteasomes [35]. Proteolysis of securin releases and activates separin, a protease that cleaves meiotic cohesin proteins and induces loss of chromatid cohesion and chiasma resolution at meiosis I [36, 37]. The SAC is a mitotic/meiotic surveillance mechanism that prevents APC/C activation and precocious progression into anaphase in the absence of a spindle or of unattached chromosomes that are not fully under tension from spindle fibers [38, 39, 40]. The SAC appears to be constitutively expressed at prometaphase I of mammalian oogenesis [40, 41]. The SAC is also induced or prolonged in response to microtubule-depolymerizing drugs, and this is associated with the persistent expression of the checkpoint protein Mitotic Arrest Deficient 2 (MAD2, now defined as MAD2L1) at the kinetochores of metaphase chromosomes [40, 42, 43]. MAD2L1 binds to and inhibits the specific activator CDC20 of the APC/C when the SAC is induced, thereby preventing progression to anaphase [41], as is the case when oocytes are exposed to nocodazole [40, 44, 45]. MAD2L1 is present at the kinetochores of chromosomes during meiotic prometaphase I in mouse and rat oocytes but disappears when the chromosomes become aligned at the metaphase I plate [40, 42, 43]. MAD2L1 is not found on kinetochores during anaphase I [40]. In addition, MAD2L1 is expressed at kinetochores of well-aligned metaphase II chromosomes during the constitutive second meiotic arrest in mammalian oocytes under the influence of the cytostatic factor (CSF) [40, 42], although it is not an essential component of CSF [44, 45]. Expression at kinetochores was analyzed in 2-ME-exposed oocytes at meiosis I arrest, after escape and progression to anaphase I, and in control and 2-ME-exposed metaphase II oocytes.
Orientation-independent enhanced polarization microscopy using electronically controlled liquid crystal polarizing optics (PolScope) has been employed to analyze noninvasively spindle formation in living control or drug-exposed mammalian oocytes [34, 46–48]. Pericentrin is one of the constitutive functional components of the oocyte microtubule organizing centers (MTOCs) or centrosomes. Percentrin immunofluorescence has been used to assess centrosomal integrity, recruitment of centrosomes by chromosomes, and centrosomal alignment at the spindle poles during normal meiotic division [49, 50], and to detect aberrations in oocytes exposed to pollutants, such as bisphenol A (BPA) [51], which are suspected to possess aneugenic properties in female meiosis [52].
Finally, we studied the dose-response of oocytes to 2-ME in in vitro-maturing mouse oocytes with respect to errors of chromosome segregation (hyperploidy and polyploidy) at anaphase I, and to failures of chromosome congression at meiosis II, which possibly predispose to meiosis II errors.
The goals of the present study were to determine the effects of 2-ME on spindle dynamics, chromosome congression, and errors in chromosome segregation at meiosis I, and to assess the metaphase II spindle structures and chromosome alignments, as well as the effects of 2-ME on the persistence of the putative spindle checkpoint protein MAD2L1 at the centromeres of meiosis I-blocked oocytes.
All reagents were purchased from Sigma (Deisenhofen) unless stated otherwise. Outbred MF1 mice were originally obtained from Harlan Winkelmann (Borchen, Germany). They were housed under conditions of constant temperature (21 ± 0.5°C) and humidity (55 ± 10% relative humidity) with a photoperiod of 12 h, and were fed ad libitum. Animal care and experimental procedures were in accordance with the National Guidelines. Ovaries were isolated from sexually mature MF1 mice (7–20 weeks of age) on the day of diestrus of the natural cycle. They were placed in prewarmed M2 medium [53] that contained 14 mg/ml BSA. In vitro maturation of denuded mouse oocytes was performed as described previously [48, 54]. Fully grown oocytes with germinal vesicles (GVs) were collected from the ovaries within 30 min, to maintain synchronous maturation. At least three replicate experiments were performed for each set, including controls. For the maturation experiments, isolated and denuded oocytes were placed in 1 ml of M2 medium that contained 4 mg/ml BSA with or without 2-ME or solvent (0.1% dimethyl sulfoxide, DMSO) in Nunclon 4-well dishes and cultured in an ungassed incubator under saturated humidity at 37°C for 9 h and 10 h (maturation to metaphase I) or for 13 h and 16 h (maturation to metaphase II).
Collection and Maturation of Oocytes
2-ME was dissolved in DMSO to give a stock solution of 10 mM 2-ME. The stock was diluted to the final concentration immediately before use. Meiotic progression was assessed after 16 h of culture for all oocytes that were later spread for chromosome analysis or examination of spindles and chromosomes. The data were pooled from all the experiments. The numbers of oocytes with GVs, germinal vesicle breakdown (GVBD), and polar bodies (PBs) were analyzed at the end of the culture period. For meiosis I analysis, oocytes were processed for immunofluorescence after 9 h and 10 h of in vitro maturation without or with 3.75 µM 2-ME. Meiosis I analysis showed that it was impossible to distinguish between the normal prometaphase I stage with chromosomes that were still in the process of congression and the disturbed metaphase I, in which the chromosomes failed to assemble at the equator. At 13 h and 16 h, control oocytes are in the constitutive metaphase II arrest and should have normal meiosis II spindles and well-aligned chromosomes. Therefore, the presence and morphology of the spindle and the chromosomal constitution were determined for the control and 2-ME-exposed mouse oocytes that were matured in vitro for 13 h (PolScope) or 16 h (immunofluorescence), using noninvasive PolScope analysis [34, 48] and fixation and indirect tubulin-immunofluorescence, respectively. For the chromosome analyses, oocytes were spread after 16 h of maturation.
Non-Invasive PolScope Microscopy
For PolScope-microscopy, a Nikon DIAPHOT 300 inverted microscope equipped with Hoffman interference optics, 10x and 40x strain-free objective lenses, a heated stage and the SpindleView imaging system (CRI Inc., Woburn, MA) was employed, as previously described [34, 48]. Mouse oocytes with PBs were placed individually in a micro-droplet of 3 µl prewarmed M2 medium covered with prewarmed mineral oil (M8410) in a WillCo Wells dish (WillCo Wells BV, Amsterdam, The Netherlands). The dish was maintained on a 37°C heating stage (HT-300, MTG Germany) to prevent transient depolymerization of the spindle microtubules by cooling [34, 48]. Metaphase II oocytes were imaged with a 40x strain-free objective lens [48]. The number of oocytes that expressed a birefringent spindle and exhibited severe spindle aberrations was determined for each experimental group.
Indirect anti-tubulin immunofluorescence was used to analyze the spindle morphology and chromosome behavior in fixed GVBD and PB-stage oocytes. For fluorescence microscopy, extraction, fixation, and antibody reactions were performed as described previously [48, 54]. Briefly, for quantitative tubulin and pericentrin immunofluorescence analysis, metaphase II oocytes were extracted for 45–60 min at 37°C in a microtubule-stabilizing solution (25% [v/v] glycerol, 2% Triton X-100, 50 mM KCl, 0.5 mM MgCl2, 25 mM Hepes [pH 6.75], 20 µM phenylmethylsulfonyl fluoride [PMSF], 5 mM EGTA). After attachment to a poly-L-lysine- coated slide, the oocytes were fixed briefly in cold methanol (–20°C). For double-staining of tubulin and pericentrin, extracted, methanol-fixed oocytes were incubated with anti-
-tubulin antibody (Sigma), followed by anti-pericentrin antibody (Covance, Berkeley, CA). The oocytes were then treated with polyclonal anti-mouse fluorescein isothiocyanate (FITC)-conjugated antibody and polyclonal tetramethyl rhodamine isothiocyanate (TRITC)-labeled anti-rat antibody. Chromosomes were stained with 10 µg/ml 4,6-diaminidino-2-phenylindole (DAPI) or 1 µg/ml propidium iodide, as described previously [48, 54]. For routine analysis, spindle images were recorded with a Zeiss Axiophot fluorescence microscope equipped with a sensitive coupled charge device (CCD) camera (SensiCam; PCO CCD Imaging, Kelheim, Germany) [48].
Depending on the stage of spindle formation and chromosome distribution, meiosis I oocytes were assigned to groups of oocytes with GVBD, which had no clearly recognizable bipolar spindle (corresponding to the circular bivalent and early prometaphase I stages), those with a bipolar spindle were assigned to late prometaphase I and metaphase I, or those in anaphase I or oocytes with a first polar body that had progressed to meiosis II. Chromosome congression was analyzed in all the oocytes that had a bipolar spindle. Meiosis II oocytes were analyzed after culture for 16 h. Oocytes that failed to align their chromosomes at the spindle equator were assigned to two groups: 1) a group that had congression failure with displacement of only some chromosomes, while the majority of the chromosomes appeared to be at the spindle equator (some unaligned chromosomes); and 2) a group that had totally unordered chromosomes, i.e., oocytes with major congression failure, resulting in unordered and scattered chromosomes.
An alternative fixation protocol was also employed, in which oocytes were introduced into a fibrin clot, fixed with formaldehyde, reacted with antibody, and then stained for spindle analysis by confocal microscopy as described previously [48, 54]. Serial images of the spindle were combined to assess spindle morphology, centrosomal distribution, and chromosome congression.
Oocytes matured for 16 h and arrested in meiosis I or metaphase II were also spread according to the procedure of Hodges and Hunt [55], to assess the expression of the putative protein of the spindle assembly checkpoint, MAD2L1, at the centromeres of homologous chromosomes and of dyads using a polyclonal mouse antibody to MAD2L1 (kindly donated by Beth Weaver, Ludwig Institute for Cancer Research and Department of Cellular and Molecular Medicine, UCSD, La Jolla, CA) [56]. For sensitive MAD2L1 staining with low background, spread oocytes were first incubated with a polyclonal rabbit anti-MAD2L1 antibody, followed by sequential labeling with a goat polyclonal anti-rabbit FITC-conjugated antibody (Sigma) and amplification by labeling with a rabbit anti-fluorescein antibody and a goat anti-rabbit antibody conjugated to Alexa Fluor 488 (Molecular Probes, Eugene, OR). Finally, the chromosomes were stained with DAPI.
For chromosome spreading, a protocol based on a modification of the Tarkovski method was employed [48, 54, 57]. Briefly, oocytes were placed into 1% sodium citrate, and then individually fixed and spread in ice-cold methanol/acetic acid (3:1). Chromosomes were stained by C-banding according to standard methods [54, 57].
Based on the chromosomal constitution, oocytes were scored as a germinal vesicle (GV) with intact nucleus, as oocytes that contained bivalents at meiosis I (Fig. 1A) or as oocytes that possessed dyads (Fig. 1, B–D). Oocytes without PB that contained twice the haploid number of dyads (Fig. 1D) were included in the group that progressed to metaphase II (Table 3). Those with near diploid numbers of dyads with at least 36 individually recognizable chromosomes were counted for the assessment of polyploidy among all the oocytes with dyads (including GVBD and PB oocytes). The number of hyperploid meiosis II oocytes with first polar body with more than 20 dyads (Fig. 1C) or the respective number of chromatids (monads) and dyads was analyzed. The presence of a single chromatid or pairs of chromatids in oocytes was assessed independent of ploidy in all analyzable GVBD and PB meiosis II oocytes. This was termed ‘predivision'. Finally, the number of oocytes with polar bodies in meiosis II, in which chromosomes were in an anaphase I-like stage of chromatid separation and were not spatially separated from each other, was assessed (nondisjunction; Fig. 1, E and E').
|
|
Isolated oocytes from more than three mice were randomly assigned to control, solvent control, and 2-ME treatment groups for each experiment. Each experiment was performed at least three times and the results were subsequently pooled. A chi-square test with Yates correction was used for statistical analysis to compare meiotic progression, nuclear maturation, chromosomal constitution, and spindle aberrations and chromosome displacement between each of the treatment groups and the control and solvent control groups. For the statistical analysis, P < 0.001 was considered significant.
Maturation of Oocytes in the Presence of 2-ME
2-ME did not affect the proportion of degenerate oocytes (Table 1). Whereas most of the control oocytes, oocytes exposed to solvent, and oocytes treated with 1.25 µM 2-ME spontaneously resumed maturation and developed to metaphase II, 2-ME at doses of 2.5 µM and 3.75 µM induced a small increase in the numbers of oocytes arrested at the GV stage. The increase reached significance in oocytes that were exposed to 5 µM and 7.5 µM 2-ME (P < 0.01 and P < 0.001, respectively) (Table 1).
|
Most of the oocytes derived from unstimulated cycles of the control, solvent control, and 1.25 µM 2-ME-treated groups that were able to resume meiosis emitted a first polar body (>80%). However, higher concentrations of 2-ME induced meiotic delay or arrest, since the numbers of GVBD oocytes dose-dependently and significantly increased in the groups of oocytes exposed to 2.5 µM, 3.75 µM, 5 µM, and 7.5 µM 2-ME (P < 0.001) (Table 1), and this was associated with a dramatic reduction in the numbers of oocytes with PB.
Chromosomal Constitution of Metaphase II Oocytes Matured in the Presence of 2-ME
There was no evidence for activation and pronuclear formation by 2-ME, as only 0.3% of the oocytes in the control, the solvent control, and the 2.5 µM 2-ME groups and 0% of the oocytes in the other groups were activated. As expected from the increased number of oocytes with GVBD, the number of oocytes with bivalents (Fig. 1A) increased significantly in the presence of 2-ME from 13.7% and 7.3% in the control and solvent control group, respectively, to 43.3% in the group matured in 2.5 µM 2-ME and 66.7% in the group exposed to 7.5 µM 2-ME (P < 0.001) (Table 2). Concomitantly, the numbers of oocytes with dyads decreased dose-dependently.
|
Chromosomal analysis revealed that the majority of the control oocytes had a normal chromosomal constitution of 20 dyads (Fig. 1B). The number of diploid metaphase II oocytes (polyploids; Table 3), i.e., those that failed to emit a first polar body and contained 40 instead of 20 dyads (Fig. 1D), was not increased by chronic exposure to 2-ME. There was no evidence that 2-ME induced the precocious detachment of sister chromatids (predivision) to increase the numbers of oocytes with unpaired chromatids at metaphase II. However, hyperploidy increased dose-dependently (Fig. 1C) in groups exposed to 2-ME at concentrations equal to or greater than 3.75 mM compared to the control and solvent control. A significant increase was already observed at 3.75 µM 2-ME (P < 0.05). At 5 µM and 7.5 µM 2-ME, hyperploidy reached 23.5% and 100%, respectively (P < 0.001; Table 3).
Only those oocytes that were exposed to more than 5 µM 2-ME possessed anaphase I-like chromosomes, with distally associated telomeres of chromatids, although they had apparently progressed to meiosis II and possessed a first polar body (Table 3; Fig. 1, E and E').
Spindle Morphologies of Living Meiosis II Oocytes examined by PolScope
Only a few metaphase II oocytes in the control and 2-ME-exposed groups did not possess a birefringent spindle (Fig. 2B). Over 90% of all oocytes matured without or in the presence of 2-ME expressed a spindle under PolScope microscopy (Table 4). The spindles in the controls were barrel-shaped with more or less flat spindle poles (Fig. 2A). Although spindle abnormalities, such as small, elongated spindles, asymmetric spindles, and spindles with tattered poles, were occasionally observed in control oocytes, they were much more prevalent in oocytes that were cultured in the presence of the high concentrations of 2-ME. From the limited number of metaphase II oocytes, it appeared that the frequencies of severe spindle abnormalities, such as very short pole-to-pole distance and pointed spindle poles with low birefringence (Fig. 2, C and D), increased in the presence of 2-ME (Table 4).
|
|
Disturbances in Bipolar Spindle Formation and in Chromosome Congression Induced by 2-ME in Meiosis I Oocytes
Observations of meiosis II oocytes using PolScope suggested that 2-ME interfered with spindle formation, which may have predisposed the oocytes to meiotic errors at anaphase I. To test this hypothesis, control oocytes and oocytes exposed to 3.75 µm 2-ME, a concentration that induces increases in hyperploidy, were fixed after 9 h and 10 h of culture. Fifty-eight percent of the control oocytes had already established a bipolar spindle after 9 h of culture and were in late prometaphase I or metaphase I. Six percent had entered anaphase I or emitted a polar body (Table 5). In contrast, bipolar spindle formation was severely disturbed in 2-ME-exposed oocytes, and nearly two thirds of all the oocytes had unordered chromosomes and a multipolar, unorganized spindle. None of the 2-ME-exposed oocytes had progressed to meiosis II. After 10 h of culture, nearly two thirds of the control oocytes possessed bipolar spindles, and nearly 30% had a polar body or had entered anaphase I, in contrast to the 2-ME exposed oocytes, which had predominantly disorganized spindles and lacked oocytes with polar bodies (Table 5). After 9 h of culture, only 24.1% of the control oocytes had aligned chromosomes, while the percentage increased to 60.9% at 10 h, presumably because the oocytes had reached late metaphase I and were close to anaphase I progression. In contrast, nearly 90% of all the oocytes with bipolarly organized spindles in the 2-ME group had unaligned chromosomes after 9 h, and 100% of these oocytes with bipolar spindle exhibited chromosome congression failure after 10 h of culture, which was significantly different from the controls (Table 5).
|
2-ME-Induced Spindle Aberrations and Congression Failure of Chromosomes Detected by Immunofluorescence at Meiosis II
Since it is impossible to distinguish between a normal prometaphase I stage and a disturbed metaphase I with chromosome congression failure, we analyzed meiosis II oocytes for spindle morphology and chromosome congression after 16 h of culture, by which stage they should have developed to metaphase II. Most of the metaphase II oocytes of the control, solvent control, and the 2.5 µM 2-ME-exposed groups contained typical barrel-shaped, compact metaphase II spindles and well-aligned chromosomes (Fig. 2, E, F, I, and M). During meiotic progression of mouse oocytes, microtubules associated with cytoplasmic microtubule-organizing centers in the prominent asters decrease in size and number, and most of the microtubules become organized within the bipolar spindle. Oocyte spindles are defined as anastral. Accordingly, in the present study, oocytes with prominent asters in the cytoplasm were uncommon in the controls matured for 16 h (Fig. 2, E, F, I, and M). Oocytes that had bipolar spindles with microtubule asters located away from the spindle poles at the periphery of the spindle body or oocytes that had asymmetric spindles were scored as bipolar spindles with mild aberrations (Fig. 2K; Table 6). Severe spindle aberrations, characterized by prominent extra asters and microtubule bundles extending outward from the spindle poles (Fig. 2, G, L, and N), multipolar spindles (Figure 2 G, L, N), and unusually shaped, elongated, wavy, and asymmetric spindles (Fig. 2, h and N), were frequently observed in oocytes exposed to 2-ME at doses equal to or greater than 3.75 µM (Table 6). In particular, increased numbers of small or large cytoplasmic asters were characteristic of oocytes exposed to higher concentrations of 2-ME (Fig. 2, K, L, and N).
|
While the majority of the pericentrin-positive centrosomes assembled at the spindle poles of the control and solvent control oocytes (Fig. 2M), pericentrin-positive and pericentrin-negative asters were highly enriched in the metaphase II oocytes exposed to 2-ME (Fig. 2N). Concomitantly, chromosomes were unaligned and failed to congress at the spindle equator in a large proportion of the meiosis II oocytes exposed to doses of 2-ME equal to or greater than 3.75 µM (Fig. 2, G, H, K, and L). Congression failure was frequently associated with displacement of one or several chromosomes from the equator, resulting in localization of several chromosomes close to the spindle poles (Fig. 2K), while other chromosomes still aligned at the equator (Fig. 2K). A high percentage of the oocytes exposed to 5.0 µM or 7.5 µM 2-ME had totally disordered and scattered chromosomes, although most of the chromosomes appeared to be still attached to spindle fibers (Fig. 2, G, H, and L; Table 6). Significant increases in aberrant spindles and chromosome congression failure at meiosis II were observed for the groups of oocytes exposed to 3.75 µM 2-ME (P < 0.001). Of the few oocytes that escaped meiotic arrest upon exposure to 7.5 µM 2-ME and progressed to meiosis II, all had abnormal spindles and unaligned chromosomes (Table 6).
Persistence of Centromeric Expression of MAD2L1 after 2-ME Exposure
The failure to align chromosomes at the spindle equator after 9 h and 10 h of culture, the dose-dependent increase in the number of oocytes arrested at meiosis I after 16 h of culture, and the high number of meiosis II oocytes with apparently mono-attached and unaligned chromosomes suggest that 2-ME prolongs the SAC in the presence of improperly attached chromosomes. Therefore, we analyzed the expression of a putative component of the SAC, the MAD2L1 protein, at the centromeres of the meiosis I arrested oocytes that were cultured for 16 h in the presence of 2-ME (Fig. 2P). MAD2L1-positive fluorescent foci were observed at most centromeres of the 2-ME-treated, meiosis I-blocked oocytes, similar to the expression pattern of MAD2L1 at the centromeres of metaphase I-arrested oocytes exposed to nocodazole (Fig. 2R). The kinetochores of 2-ME-treated oocytes that had progressed to metaphase II were also positive for MAD2L1, comparable to the controls (Fig. 2, O and O'). In contrast, no MAD2L1 foci were found at the centromeres of anaphase I oocytes that had recently escaped the 2-ME block (Fig. 2, S and S').
Effects of 2-ME on Cell Cycle Progression
In contrast to most somatic cell types that become arrested at the late G2-phase by 2-ME [58–61], there was only a minor increase in GV-arrested dictyate-stage oocytes upon exposure to very high concentrations of 2-ME. The response to 2-ME is similar to those to other antimitotic chemicals, such as nocodazole [48], in which alterations to the cytoskeleton cause cell cycle delay and M-phase arrest but do not interfere much with the transition from the G2 phase to the M phase. 2-ME blocked progression from metaphase I to meiosis II in a high percentage of mouse oocytes exposed to the higher concentrations of 2-ME. 2-ME also caused delays in the formation of a typical bipolar spindle at meiosis I and induced severe spindle abnormalities and aberrant spindle pole organization at metaphase II. The scattered chromosomes at meiosis I, chromosome displacement, and the detachment of chromosomes from the metaphase plate seen in metaphase II oocytes by tubulin immunofluorescence imply that 2-ME perturbs congression of chromosomes, thereby prolonging meiotic maturation by maintaining persistent activity of the spindle attachment checkpoint (SAC) [36, 39, 62]. In fact, the expression of MAD2L1 at the centromeres of bivalent chromosomes of meiosis I-blocked oocytes provides evidence for the first time that the spindle checkpoint may be prolonged in 2-ME-induced meiotic arrest, which is comparable to a nocodazole block [40, 56, 63]. As expected, those oocytes that escaped the checkpoint and progressed to anaphase I did not exhibit MAD2L1 at the centromeres, while constitutive meiosis II arrest was characterized by the presence of MAD2L1 at the centromeres of the chromatids of both 2-ME-treated and control oocytes.
In general, the loss of attachment and/or loss of tension on chromosomes may induce the SAC in mitosis and in oocyte meiosis (reviewed in [62–66]). Currently, it appears that the basic scheme for mitotic SAC signaling is conserved during meiosis I in mammalian oocytes [62]. However, oocytes may be especially susceptible to disturbances because the absence of tension alone may be insufficient to activate a robust SAC [41, 67, 68]. Alternatively, signals induced by low-dose aneugens with subtle effects on the tension/attachment of chromosomes in oocytes may be too weak to induce persistent expression of the putative spindle checkpoint protein MAD2L1 at centromeres and SAC activation [62], as may be the case with 2-ME. As a consequence, 2-ME doses of 3.75 µM and 5 µM, which did not entirely arrest meiosis, increased aneuploidy substantially.
Spindle Aberrations, Merotelic Attachments, and Meiotic Arrest Caused by 2-ME
Physical or nocodozole-induced depolymerization of microtubules induces shrinking of the spindles in mammalian oocytes [48]. In contrast, exposure of mouse metaphase II oocytes to taxol caused a pronounced increase in spindle pole-to-pole distance [34]. There was no consistent change in spindle length at meiosis II when 2-ME was present throughout maturation. Instead, loss of spindle integrity, alterations in shape, and tapered spindle poles with low birefringence were observed. In addition, increases in the numbers of large and small cytoplasmic asters of microtubules and the formation of multipolar spindles were characteristic of 2-ME-exposed oocytes, which indicate that the spatio-temporal regulation of microtubule kinetics and organization was disturbed. The analysis of meiosis I oocytes supports this notion, i.e., the formation of a bipolar spindle is hampered and chromosomes fail to congress at the spindle equator at a time when the control oocytes have already entered anaphase I.
MAD2L1 protein was uniformly found at all centromeres of the 2-ME-blocked meiosis I oocytes arrested for more than 8 h at meiosis I (after 16 h of culture). Loss of MAD2L1 from the centromeres of oocytes that progressed to meiosis II may be a consequence of merotelic attachment (attachment of one kinetochore with microtubules to opposite spindle poles) or multipolar attachments, such that saturated occupation/tension of microtubules on kinetochores results in anaphase I progression. Merotelic attachment cannot be detected by the SAC during mitosis [69, 70]. In addition, the attachment to nonpolar, cytoplasmic asters may generate tension and release from the SAC. Furthermore, oocytes may be incapable of delaying anaphase I in the presence of one or a few unaligned chromosomes that contribute to high susceptibility to meiotic error [67].
Indeed, the few meiosis II oocytes that progressed to metaphase II under the influence of a high 2-ME concentration (7.5 µM) had aberrant spindles and were aneuploid. While increases in hyperploidy were also common in taxol-exposed mouse oocytes [33], comparable to our observations of 2-ME-exposed oocytes, taxol additionally caused increases in parthenogenetic activation, precocious sister chromatid separation, and polyploidy [33]. The latter aberrations were not detected in 2-ME-exposed mouse oocytes. Therefore, it appears that 2-ME, unlike taxol and nocodazole, does not interfere with the synchrony of nuclear progression and cytokinesis [33, 48].
Disruption of Centrosomal Integrity by 2-ME
It is well known that the sorting and redistribution of centrosomal components, such as gamma-tubulin, NuMA, and pericentrin, from the cytoplasm and cortical MTOCs to spindle sites is required for normal oocyte maturation [30, 48, 71, 72]. We show here for the first time that 2-ME disrupts the distribution of pericentrin, in similarity to environmental chemicals with weak estrogenic activity, such as bisphenol A (BPA) [51]. Oocytes express estrogen receptor-beta [73], while 2-ME has only low affinity for estrogen receptors [25, 74]. Furthermore, oocytes are transcriptionally repressed when resuming meiosis [75, 76], which suggests that the adverse effects of 2-ME may be mainly nongenomic. Therefore, the disruption of spindle formation and centrosome integrity by 2-ME is probably important in the genesis of aneuploidy [77, 78]. Notably, exposures of oocytes to diethylstilbestrol (DES), BPA, E2, and 2-ME causes similar disturbances in spindle formation in mammalian oocytes, i.e., disorganization of spindle shape, loss of bipolarity, and aster formation [16, 50, 79]. Apart from altering microtubule polymerization kinetics, common targets, which are responsible for some of the effects of this class of chemicals, may exist in oocytes [77]. 2-ME is known to be the most potent inducer of centrosomal disintegration in V79 cells, as compared to dihydroequilin 3-methyl ether, equilin 3-methyl ether, 17-estradiol, 17β-estradiol 3-methyl ether, E2, dihydroequilin, and estrone 3-methyl ether, and has been shown to cause the induction of multiple signals of gamma-tubulin at centrosomes characteristic of centrosome splitting [77]. This mirrors the effects of 2-ME on pericentrin distribution in mouse oocytes.
2-ME-Induced Nondisjunction and Aneuploidy
Exposure of oocytes to micromolar levels of 2-ME caused errors in chromosome segregation and nondisjunction. The threshold for significant increases in hyperploidy was close to 3.75 µM 2-ME, which is a nonphysiologically high concentration of the sterol. In addition, 2-ME-exposed oocytes occasionally possessed chromosomes in which the distal arms of the chromatids were still facing each other or attached to each other in an anaphase I-like fashion, although the oocytes had apparently progressed to meiosis II and possessed a polar body. It is unlikely that 2-ME causes irreversible inhibition of detachment associated with disturbed proteolysis of cohesin proteins, since we did not observe normal bivalents at meiosis II that were still connected by chiasmata [63].
Aneuploidy in mouse oocytes was increased by 3.75 µM 2-ME, which is similar to the concentrations that induce chromosomal aberrations in the blastomeres of bovine embryos from maturation-exposed oocytes [16]. Therefore, these observations support the notion that environmental exposure that causes local increases of 2-ME to micromolar levels may compromise oocyte maturation and quality [13, 16]. 2-ME is a major metabolite of estradiol in human serum; its concentration peaks at midcycle, like that of estradiol [2, 3]. Although analysis by gas chromatography-mass spectroscopy have detected maximal concentrations of 0.03 and 0.82 µM in human and mare follicular fluid, respectively [2, 4], which correspond to about one to two orders of magnitude below the range that induces spindle aberrations and aneuploidy in mouse oocytes, local 2-ME concentrations may be higher. AHR may be significantly increased by organic pollutants [13], and the relative abundances of CYP enzymes and estrogen may be of sufficient physiological significance [80, 81] to increase the site-specific, local 2-ME concentration. Up to 1 pg/ml polychlorinated dibenzodioxins and dibenzofurans have been found in human follicular fluid [25], which might activate the ARH [13]. Since estradiol concentrations are high in the follicular fluids of healthy human follicles (about 500–600 ng/ml at midcycle [26]), and granulosa cells efficiently produce 2-ME, 2-ME may accumulate in cell membranes or cellular compartments due to its lipophilic nature. It is noteworthy that 2-ME is currently used as a potent chemotherapeutic agent that not only changes microtubular stability but also alters the levels of detyrosinated and acetylated tubulin in human cancer cells [82]. Thus, further studies are required to analyze the complex activity of persistent environmental pollutants in the ovary, and to understand the mechanisms by which 2-ME levels may rise, modulate folliculogenesis, and affect oocyte quality directly and indirectly. The extremely high rates of hyperploidy in oocytes exposed to 5 µM 2-ME show that this catecholestradiol has to be regarded as a potent meiotic aneugen in oocytes, while higher concentrations primarily caused meiotic arrest.
The current observations support the notion that young mammalian oocytes may respond to disturbed spindle formation [44, 48, 63]. However, the spindle checkpoint in oocytes appears to be comparatively leaky and less stringent in female meiosis as compared to male meiosis [62, 64, 83–87], especially in aged oocytes. Aged oocytes contain reduced levels of mRNA transcribed from genes in the checkpoint signaling pathway, such as mad2l1, as compared to young oocytes [83, 84]. Therefore, decreased defense against disruptions of spindle function and chromosome attachment by aneugens (such as 2-ME) may contribute to the high susceptibility of aged oocytes to errors in chromosome segregation [41, 62], even at comparatively low doses of the aneugen.
ACKNOWLEDGMENTS
We thank Rudolf Eichenlaub for critical reading of the manuscript. The expert technical assistance of Ilse Betzendahl is gratefully acknowledged.
FOOTNOTES
1Supported by an EU grant (QCRT-2000-00058). 
Correspondence: 2Ursula Eichenlaub-Ritter, Univ. Bielefeld, Fak. Biol. IX, D-33501 Bielefeld, Germany. FAX: 49 521 1066015; e-mail: EiRi{at}uni-bielefeld.de
Received: 27 June 2006.
First decision: 23 July 2006.
Accepted: 29 December 2006.
REFERENCES
