| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute of Reproductive and Developmental Biology,3 Imperial College London, Hammersmith Hospital, London W12 0NN, United Kingdom
Department of Anatomy and Developmental Biology,4 University College London, WC1E 6BT London, United Kingdom
Department of Mathematics,5 Imperial College London, London SW7 2AZ, United Kingdom
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
20 ng/ml). The effect of FSH on spindle morphology and chromosome alignment during metaphase I was then explored using immunocytochemistry and three-dimensional reconstruction of confocal sections. High FSH had no effect on gross spindle morphology but did alter chromosome congression during prometaphase and metaphase, with the spread of chromosomes across the spindle at this time being significantly greater in oocytes cultured in 2000 ng/ml compared with 2 ng/ml FSH. Analysis of three-dimensional reconstructions of spindles in oocytes matured in 2000 ng/ml FSH shows that chromosomes are more scattered and farther apart than they are following maturation in 2 ng/ml FSH. These results demonstrate that exposure to high levels of FSH during IVM can accelerate nuclear maturation and induce chromosomal abnormalities and highlights the importance of the judicious use of FSH during IVM.
aneuploidy, FSH, gamete biology, IVM, meiosis, oocyte development, spindle
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
During superovulation or IVM, the oocyte is undergoing critical cellular events that enable the egg to resume meiosis and complete cytoplasmic maturation. Following the LH surge in vivo or once removed from the constraints of the antral follicle, the oocyte (which has been arrested since before birth at prophase of the first meiotic division) will resume meiosis. Following germinal vesicle breakdown (GVBD), the first meiotic spindle assembles and paired homologous chromosomes (bivalents) become aligned along its equator before their timely separation during anaphase. Fertilization and preimplantation development depend on the successful completion of cytoplasmic maturation in parallel with nuclear maturation and the accurate segregation of chromosomes at meiosis I [11–13]. It is also recognized that the environment to which the maturing oocyte is exposed has the potential to compromise subsequent embryonic development [14–17], raising concerns about the possible adverse effects of increased exposure to FSH at this time. Furthermore, following resumption of meiosis, the processes of spindle formation and chromosome segregation, with expulsion of the first polar body, are believed to be particularly sensitive to both the physical and chemical environment [18, 19], highlighting the importance of the maturation environment in the genesis of aneuploidy.
In humans, aneuploidy is a major cause of pregnancy loss [13] and increases with age [20]. Chromosomal abnormalities can arise during meiosis through structural or temporal aberrations in the separation of homologous chromosomes during meiosis I or sister chromatids during meiosis II. While anomalies in the structural attachments between chromosomes (chiasmata) or chromatids (cohesins) play a significant role in nondisjunction [21], spindle structure is also thought to be important. Failure of centromere attachment to the spindle can result in the loss of chromosomes. In oocytes, chromosomal displacement from the meiotic spindle does not prevent progression to anaphase I and the completion of the first meiotic division and so the absence of a rigorous check point controlling this stage of the cell cycle establishes a potential mechanism for the genesis of aneuploidy [22, 23]. In the mouse, spindle morphology and chromosome behavior have been examined using antitubulin immunofluorescence coupled with DNA staining [19, 23–29]. This technique allows visualization of both the spindle and chromosomal distribution/attachment to the spindle at specific time points during the meiotic cycle. A number of environmental aneugens have been identified in this way [23, 27, 30, 31]. During IVM, such aneugens disrupt spindle assembly and chromosomal alignment during meiosis I and, if meiosis proceeds, predispose the oocyte to inaccurate segregation of the chromosomes and aneuploidy.
FSH receptors are exclusive to the granulosa cells of maturing follicles [32, 33], which in turn are directly connected to the oocyte via gap junctions [34, 35]. FSH effects on cumulus cells may therefore be directly transmitted to the oocyte via the secondary messenger cAMP and one of its downstream target kinases [36, 37]. One possible manifestation of exposure to abnormally high concentrations of FSH at this time could be alterations in spindle morphology. Indeed, there is accumulating evidence to suggest that the endocrine environment can adversely affect spindle normality. FSH, both in vivo and in vitro, affects the number of oocyte centrosomes, suggesting that the hormonal environment can modulate the dynamics of centrosome-based microtubule assembly, including spindle formation [18]. Animal studies have demonstrated that superovulation with exogenous FSH can induce chromosomal abnormalities [38–40], with repeated exposure to FSH in vivo leading to spindle abnormalities in murine oocytes [41]. Furthermore, oocytes from LHßCTP transgenic mice, which hypersecrete LH, show increased levels of congression failure, where the majority of chromosomes have not aligned on the equator of the meiotic spindle [25]. Finally, in women, FSH levels increase toward the end of reproductive life [42, 43], and it is hypothesized that this may contribute to the increased aneuploidy rates seen in oocytes retrieved from older women [44, 45]. Indeed, spindle morphology and/or chromosome alignment is abnormal in oocytes from older mice [46] and women [47, 48]. Collectively, these studies suggest that exposure to high levels of FSH in vivo may have an adverse effect on follicular function and oocyte health, possibly in relation to spindle assembly and chromosomal alignment.
Recently, we have shown that FSH accelerates oocyte maturation and significantly increases glycolysis in cumulus-oocyte complexes (COCs) undergoing IVM [49]. Our aim in this study was to extend this work and ascertain the potential role of FSH in the genesis of chromosomal abnormalities, more specifically its ability to disrupt chromosomal segregation during meiosis I. The majority of aneuploidies arise during meiosis I [50], and it is during this division that the oocyte is exposed to the abnormal hormonal milieu of the superovulated follicle or the unbalanced environment of IVM. We have examined the incidence of hyperhaploidy in oocytes from COCs that have undergone IVM in increasing concentrations of FSH. Focusing on COCs maturing in low and high concentrations of FSH, we have gone on to examine, in detail, spindle morphology and chromosome alignment in metaphase I oocytes using three-dimensional reconstruction of confocal sections.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Unstimulated 25- to 28-day-old F1 (C57BL/6 x CBA/Ca) female mice (Harlan Olac Ltd., Oxon, UK) were killed by cervical dislocation, the ovaries removed and transferred to M2 medium [51] supplemented with 4 mg/ml bovine serum albumin (BSA, Fraction V; Sigma, Poole, UK). Large antral follicles were punctured with sterile acupuncture needles (AcuMedic, London, UK) to release COCs. Only COCs that consisted of an oocyte surrounded by at least two complete layers of cumulus cells were selected for further culture. Mice were housed and maintained in accordance with the Animals (Scientific Procedures) Act of 1986 and associated codes of practice.
Collection of Control Mature Metaphase II Oocytes
Control oocytes (metaphase II) that had matured in vivo, were collected from unstimulated sexually mature F1 (C57BL/6 x CBA/Ca) female mice on the day following estrus. Oocytes, which had been exposed to FSH in vivo, were collected from superovulated female mice of the same strain. Superovulation was carried out by i.p. injection of 7.5 IU eCG (Folligon; Intervet Cambridge, UK) at 1200 h followed 48 h later by 7.5 IU hCG (Chorulon; Intervet). Females were killed by cervical dislocation on the morning following ovulation, the oviducts removed, and cumulus masses mechanically released from the oviducts. Cumulus cells were removed by brief incubation (30–60 sec) in 150 IU/ml hyaluronidase (Sigma) in M2.
In Vitro Maturation of Oocytes
Minimum essential Eagle medium (MEM; Sigma) supplemented with 3 mg/ml BSA was the basic culture medium used for IVM [49]. The medium was supplemented with human recombinant FSH (Puregon; Organon, Cambridge, UK) as specified below. One nanogram of Puregon contains 0.01 IU FSH.
COCs were taken through four washes of medium before being cultured individually in 5-µl drops of pre-equilibrated medium overlaid with silicone oil (Dow Corning 200/50 cs; BDH, VWR International, Lutterworth, UK) in an atmosphere of 5% CO2 in air at 37°C.
For analysis of the effect of FSH on nondisjunction during anaphase I, COCs were cultured for 15–16 h in MEM supplemented with 0, 2, 20, 200, or 2000 ng/ml FSH. At the end of the incubation, COCs were removed from their incubation drops and cumulus cells were removed by gentle pipetting. Denuded oocytes were scored immediately for stage of nuclear maturation. Oocytes were classified as being at the germinal vesicle stage if the nuclear membrane remained intact and was clearly visible by light microscopy, at metaphase I if there was no visible nuclear membrane, or at metaphase II following extrusion of the first polar body. Denuded oocytes were held until spreading at 37°C in medium containing the same concentration of FSH as that used for their maturation.
For immunohistochemistry of metaphase I spindles, COCs were cultured individually for 6–7 h in MEM (as above) supplemented with 2 or 2000 ng/ml FSH. At the end of the culture period, all COCs were removed from culture drops and denuded by repeated pipetting and examined under an inverted light microscope (Eclipse TE300, Nikon UK, Kingston-upon-Thames, UK). Care was taken to maintain the COCs and oocytes at 37°C throughout. Oocytes that had no visible germinal vesicle and that had not extruded a polar body were deemed to be at the metaphase I stage and were selected for immunocytochemistry (see below).
Spreading of Oocytes for Chromosome Counts
Oocytes that had been isolated from COCs matured for 15–16 h in vitro in increasing concentrations of FSH were spread and stained for counting of chromosomes. The length of time between denuding and spreading ranged from 15 min to 2 h (mean 1 h). During this time, a further 29 oocytes matured from metaphase I to metaphase II.
Oocytes at metaphase stages I or II were anonymized, spread, and fixed as described by Tarkowski [52] with the following modifications. Oocytes were treated with hypotonic solution (1% sodium citrate) for 6–7 min and fixed in 5:1:4 methanol:acetic acid:water for 1 min or until the zona pellucida had disappeared. Oocytes were individually transferred to glass microscope slides and 13 µl of 3:1 methanol:acetic acid fixative was gently dropped onto the oocyte. As the fixative dried, further drops were added; this process was repeated 2–3 times. The oocyte was constantly observed throughout the fixation process using a Leica MZ12 dissecting microscope (Leica Microsystems, Milton Keynes, UK). Slides were dehydrated through an ethanol series (70%, 90%, and 100%; 5 min in each), air dried, and mounted under a coverslip in antifade Vectashield (Vector Laboratories, Peterborough, UK) containing 1.5 µg/ml DAPI (4,6-diamidino-2-phylindole). Slides were stored at 4°C until chromosome counts were performed.
Analysis of Chromosome Spreads
Anonymized slides were analyzed by a single observer (AI) and DAPI-labeled metaphase spreads of the oocytes were visualized using an Axioskop fluorescence microscope (Zeiss, Welwyn Garden City, UK) and images captured with a digital camera (Axiophot 2, Zeiss). Only spreads with clearly identifiable chromosomes were included for analysis; those with clumped, overlapping, or excessively spread chromosomes (outside the field of view) were excluded. For oocytes at metaphase I, the numbers of diploid (20 bivalents, 40 chromosomes in total), hypoploid (<20 bivalents), and hyperploid (>20 bivalents) were counted. For metaphase II oocytes, the numbers of haploid (20 chromosomes, each with two chromatids), hypohaploid (<20 chromosomes), and hyperhaploid (>20 chromosomes) oocytes were counted.
It would be expected that nondisjunction during meiosis would lead to equal numbers of hypohaploid and hyperhaploid oocytes and it is well recognized that chromosome loss during the process of spreading the oocyte can lead to an overestimate of the percentage of hypohaploid oocytes. Many workers have therefore suggested ignoring spreads with missing chromosomes and calculating conservative estimates of the percentage of aneuploid oocytes by doubling the percentage of hyperhaploid oocytes [53, 54]. This is the approach that we have taken.
In addition to numerical abnormalities, other chromosomal abnormalities were noted. In metaphase I oocytes, homologous chromosomes are normally paired and held together by crossovers between the homologues, which only resolve and separate at anaphase I. Occasionally, premature chromosomal separation (PCS), where homologous chromosomes have separated, was seen in metaphase I oocytes, as has been previously observed [27, 55]. In addition, premature separation of sister chromatids (PSSC) was occasionally observed in metaphase II oocytes [23, 27, 54, 56] where the two chromatids of the chromosome have prematurely separated at the centromere before, rather than during, anaphase II. PCS and PSSC [56] were defined as separations greater than the width of a chromosome or chromatid, respectively.
Immunocytochemistry
Following 6–7 h of IVM, individual denuded oocytes that had undergone GVBD were washed briefly in phosphate-buffered saline (PBS; Oxoid Ltd., Basingstoke, Hants, UK) at 37°C to remove the serum, briefly (1 min) fixed in 100% methanol (BDH, analar grade), and washed again in PBS. Fixed oocytes were then incubated for 1 h in 0.1 M lysine (Sigma) in PBS containing 0.1% Triton-X-100 (Sigma). After three washes in PBS, oocytes were transferred into a 10-µl drop of a mouse monoclonal anti-
-tubulin antibody (diluted 1:300 in PBS; Sigma) overlaid with silicone fluid (BDH) in a 60-mm Petri dish and incubated overnight at 4°C. After brief washes in PBS, oocytes were transferred to 10-µl drops of fluorescein isothiocyanata (FITC)-labeled sheep-anti-mouse secondary antibody (Sigma) (diluted 1:50 in PBS) and incubated for 1 h at 37°C. Exposure to light was minimized during and following secondary antibody incubation. Oocytes were washed three times (10 min each) in PBS, then individually mounted in Vectashield containing 1.5 µg/ml diamidino-2-phenylindole (DAPI) (Vector Laboratories, Peterborough, UK). Oocytes were mounted on the slide within a ring of clear nail varnish (three coats thick) to retain the spherical structure of the oocyte and avoid flattening. A coverslip was applied and sealed with nail varnish. For negative controls, primary antibody was omitted.
Three-Dimensional Analysis of Metaphase I Spindle Morphology
Stained oocytes were examined using a Leica SP2 laser scanning confocal microscope (Leica, Milton Keynes, UK) equipped with a 63x, 1.32 numerical aperture objective. The DAPI was stimulated with the 351- and 363-nm laser lines and the emission detected from 414 to 466 nm, then the FITC was separately stimulated with 488-nm laser line and emission detected in the range 511–577 nm to eliminate any bleed through of the emission spectra. The metaphase I spindle, labeled with FITC, and assembled chromosomes, labeled with DAPI, were identified and sequential confocal sections (z-series) at approximately 0.2-µm intervals were taken throughout the spindle. A three-dimensional image of the spindle and chromosomes was then rendered from the collected data using Imaris (Bitplane AG, Zurich, Switzerland), allowing analysis of spindle size and shape, together with chromosomal attachment to the spindle.
First, the spindle was assessed on the basis of gross morphology (Fig. 1). The shape of the spindle was classified as normal if it was barrel shaped (Fig. 1A, i and ii) or abnormal if it was asymmetrical or round (Fig. 1A, iii and iv). The dispersal of chromosomes was defined as normal if they lay on a metaphase plate (Fig. 1Ai) or abnormal if they were scattered (Fig. 1Aii).
|
Measurements of spindle dimensions were made using Imaris (Bitplane AG). Both poles of the spindle were identified and marked while scrolling through the z-series from top to bottom. This allowed the accurate calculation, within Imaris, of the Euclidean distance between the two poles of the spindle, irrespective of the angle at which the spindle was lying (Fig. 1B). The same approach was used to calculate the spindle width at the equator (average of two perpendicular measurements), pole width (average of both poles), and the dispersal of chromosomes (both along and across the spindle) (Fig. 1C).
Three-dimensional measurements of chromosome dispersal were made using isosurfaces produced with Imaris Surpass (Bitplane AG). An isosurface of the DAPI-labeled chromosomes was built (Fig. 2A). Specification of successively lower threshold values then allowed separation of individual isosurfaces, each of which represented either individual chromosomes or small clusters of chromosomes (Fig. 2B). The x, y, and z coordinates of the center of gravity of each isosurface was saved and taken to represent the relative position of the chromosomes or chromosome clusters within three-dimensional space. The distance within three-dimensional space between each and every chromosome cluster (Fig. 2C) within an oocyte was then calculated as follows:
|
Distance between ith and jth point, whose coordinates are, respectively (xi, yi, zi) and (xj, yj, zj) is
 |
Statistical Analysis
Numbers of oocytes cultured in all concentrations of FSH that reached metaphase II, or that were aneuploid, were compared using 
2 analysis for multiple groups. If the difference was significant (P < 0.05), differences between individual pairs of groups were compared using 2 analysis, and the P value was then corrected using the Bonferroni correction (multiplying the P value by the number of comparisons initially made, i.e., n(n – 1)/2, where n is the number of groups). These analyses were performed using GraphPad InStat version 3.0a for Macintosh (GraphPad Software, San Diego, California; www.graphpad.com).
For the three-dimensional analysis of spindle dimensions and chromosome alignment, the spindle dimensions and average distances between chromosomes or chromosome clusters were found to be normally distributed (using the method of Kolmogorov and Smirnov). However, the standard deviations (SDs) for the two groups were significantly different (tested using the method of Bartlett). The two groups were compared either using a two-tailed t-test with a Welch correction or by a Mann-Whitney U-test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Effect of FSH on Nuclear Maturation
As described previously [49], a total of 352 COCs (constituting 18 replicate experiments) from 61 mice were matured in vitro, with between 69–72 COCs being cultured in each concentration of FSH for 15–16 h overnight. At the end of the incubation, oocytes were immediately denuded to determine their maturation status. Incubation with FSH significantly increased the percentage of oocytes that reached metaphase II by the end of the incubation (Table 1) [49].
|
Twenty-nine oocytes classified as metaphase I after denuding went on to extrude their first polar body in the interval, on average 1 h, between denuding and spreading for chromosome analysis. At the time of spreading, a similar proportion of oocytes had reached metaphase II at each dose of FSH (Table 1). The difference between the maturation rate observed at the time of denuding and when the oocytes were spread suggests that high concentrations of FSH could result in faster maturation, with a similar proportion ultimately completing maturation following exposure to all doses of FSH, given sufficient time. There was no significant difference in the length of time between denuding and spreading of oocytes matured at each dose of FSH (P = 0.99, ANOVA).
Effect of FSH on the Chromosomal Constitution of Mouse Oocytes
Table 2 shows the number of metaphase I and metaphase II oocytes that were hypoploid, diploid, or hyperploid (metaphase I); or hypohaploid, haploid, or hyperhaploid (metaphase II), at each concentration of FSH. Analysis of metaphase I oocytes allows identification of intrinsic chromosomal abnormalities of the oocyte that have arisen before meiosis, during mitosis of primordial germ cells in the fetal ovary. Examination of metaphase II oocytes allows analysis of aneuploidies that will have arisen during metaphase/anaphase I of meiosis and hence that would be expected to be susceptible to any adverse factors present in the in vitro environment.
|
Exposure to high levels of FSH increased hyperhaploidy. While the majority of metaphase II oocytes were haploid (Fig. 3A), hyperhaploidy (Fig. 3C) was seen in oocytes that matured in 20, 200, 2000 ng/ml FSH. No hyperhaploidies were found in the control and 2-ng/ml groups (Table 2). The conservative estimates (calculated by doubling the number of hyperhaploid oocytes) of the proportion of aneuploid metaphase II oocytes at 20, 200, 2000 ng/ml FSH are 8.5%, 19.0%, and 19.0%, respectively. On this basis, incubation with increasing concentrations of FSH significantly increased the percentage of aneuploid metaphase II oocytes (P = 0.006, 2 test for trend [57], hyperhaploid oocytes versus oocytes that were not hyperhaploid). In the oocytes that matured in high concentrations of FSH (20 ng/ml), the incidence of hyperhaploid oocytes was significantly higher than in the oocytes matured at low concentrations (0 and 2 ng/ml) (P = 0.015, Fisher exact). As expected, there was no significant relationship between the concentration of FSH and hyperploidy in metaphase I oocytes (Table 2).
|
Premature chromosomal separation was observed in 11 metaphase I spreads (Fig. 3, E and F), and premature separation of sister chromatids in 8 metaphase II oocytes (Fig. 3D and Table 2). There was no significant effect of FSH dose on the incidence of either of these abnormalities.
One oocyte of 55 collected from naturally ovulating mice (1.8%) and 4 of 59 (6.7%) oocytes from superovulated mice were hyperhaploid. In comparison with oocytes matured in vivo, there was no significant effect of superovulation (P = 0.37, Fisher exact test) or IVM (at the lowest dose of FSH, 2 ng/ml; Table 2, P = 1) on inducing aneuploidy. Levels of aneuploidy in superovulated oocytes were not significantly different from those in oocytes matured in vitro (at 2 ng/ml FSH, P = 0.14).
Three-Dimensional Analysis of Metaphase I Spindle Morphology
The presence of hyperhaploidy in the metaphase II oocytes led us to examine the spindles of oocytes exposed to high levels of FSH in vitro, to see whether abnormalities of metaphase I spindle structure and chromosome alignment could have given rise to the aneuploidy seen at metaphase II.
Qualitative analysis Fifty-six COCs (eight replicate experiments) were cultured for 6–7 h in medium containing 2 or 2000 ng/ml FSH, 28 in each condition. The spindles and chromosomes were examined by confocal microscopy. Bipolar metaphase I spindles were anastral and the majority showed a typical barrel shape (Fig. 4, A and B). In a few cases (six), the labeling of the spindle was uneven and did not allow assessment of spindle morphology. In the remaining 50 spindles, spindle morphology and chromosome alignment could be qualitatively categorized according to the criteria described earlier (Fig. 1A and Table 3). The concentration of FSH in the maturation medium had no significant effect on spindle morphology—the proportion of oocytes displaying a classic barrel-shaped spindle was similar in both concentrations of FSH (P = 0.72) (Table 3). Abnormally shaped spindles were less common than symmetrical bipolar spindles in both groups (Table 3) and included asymmetrical bipolar spindles (Fig. 4, C and D) or spherical spindles (Fig. 4, E–H), which were probably still assembling.
|
|
The concentration of FSH did, however, have a significant effect on the alignment of the chromosomes on the metaphase plate. Significantly fewer oocytes cultured in 2000 ng/ml FSH had spindles with well-aligned chromosomes (P = 0.012) (Table 3). Evaluation of the rotated three-dimensional, reconstructed isosurface image of the spindle was important for assessing chromosome alignment, as demonstrated in Figure 4, I—L. At first sight, examination of the confocal projections of two spindles (Fig. 4, I and K) suggests that the chromosomes could be displaced along the spindle. However, rotated isosurfaces of these spindles showed that the chromosomes were, in fact, well aligned on the metaphase plate, confirming that the spindles were lying obliquely to the axis of confocal analysis (Fig. 1B). Evidence of chromosome displacement included what appeared to be premature anaphase at one side of the metaphase plate (Fig. 4, M and N) and displacement of individual chromosomes toward the poles of the spindle (Fig. 4, O–T).
Quantitative analysis Measurement of the dimensions of the spindle and metaphase plate (Fig. 1) supported the qualitative observations above on gross morphology. There was no significant effect of FSH concentration on the length or width of the spindle itself or on the width of the spindle poles (Table 4). However, the spread of chromosomes across the spindle was significantly greater in oocytes that had been cultured in 2000 ng/ml FSH than in 2 ng/ml FSH (P = 0.044; Table 4). This indication that chromosomes were more dispersed was examined in more detail by focusing on the three-dimensional position of chromosomes or chromosome clusters (Fig. 2). There was no significant difference between the number of chromosome clusters that could be resolved from the chromosome isosurface in 2 ng/ ml or 2000 ng/ml FSH (8.6 ± 0.8 and 9.5 ± 0.7, respectively, mean ± SEM). The Euclidean distance between each and every chromosome cluster was calculated (Fig. 2C) and the average distance between chromosome clusters computed for each oocyte. Chromosome clusters were significantly farther apart in oocytes that had been exposed to 2000 ng/ml FSH than 2 ng/ml FSH (P = 0.003, Mann-Whitney test; Fig. 5).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Chromosomal abnormalities are a significant cause of embryonic and fetal loss and arise following the resumption of meiosis, through the unequal segregation of chromosomes during anaphase I or II [13]. The exact mechanism by which FSH induces aneuploidy is not clear. Mitotic cell division is subject to cell-cycle check points that are in place to ensure adequate spindle assembly, followed by the correct alignment of chromosomes along the equatorial plane before the anaphase/telophase transition [59, 60]. Meiotic divisions, particularly in the female, may not be subject to such rigorous control mechanisms, which might explain why these divisions are more prone to error [22, 61, 62]. Several exogenous compounds that induce aneuploidy in oocytes alter microtubular assembly and function, resulting in abnormal spindle morphology and chromosomal alignment [23, 27, 30, 31, 44]. A proportion of mouse oocytes exposed to such chemical aneugens are able to complete meiosis despite having displaced chromosomes and/or chromosome congressional failure [23, 27, 30, 31]. We hypothesized that this may be the mechanism of FSH-induced aneuploidy. Our three-dimensional studies did not show any of the severe spindle abnormalities that have been reported by other groups, such as whirls of unpolarized tubulin or grossly asymmetric spindles, nor did we see extensive detachment of chromosomes from the spindle [23, 27, 30, 31]; hence, FSH does not appear to have a significant spindle-disrupting role. However, our three-dimensional confocal analysis did demonstrate more spindles having chromosomes that were displaced from the metaphase plate and a significant increase in chromosomal dispersal on the spindles of oocytes exposed to high levels of FSH (Tables 3 and 4, Fig. 5). This may reflect defective chromosome congression, which may predispose the oocyte to errors in chromosome division during subsequent anaphase.
Recently, we have found that FSH-stimulated glucose uptake in mouse COCs is mediated by the phosphatidylinositol 3-kinase (PI3-kinase) pathway [49] and there is also accumulating evidence that the PI3-kinase pathway plays a part in the meiotic maturation of mammalian oocytes [63, 64]. Traditionally, FSH was believed to mediate its effects through the activation of protein kinase-A (PKA) via the secondary messenger cyclic AMP [65]. It is now increasingly apparent that other cellular signaling targets, including those within the PI3-kinase pathway, are activated by FSH [36], specifically protein kinase-B (PKB/Akt), which is phosphorylated in response to FSH [37]. One of the downstream targets of phosphorylated PKB/Akt is glycogen synthase kinase-3 (GSK-3) [66], a kinase that has been implicated in the control of microtubule stability during mitotic divisions [67, 68]. PKB/Akt phosphorylates and inactivates GSK-3, and the temporal and spatial association between PKB/Akt, GSK-3, and the mitotic spindle has identified a possible role for both kinases in the attachment of chromosomes to microtubules [69]. Colocalization of active PKB and phosphorylated GSK-3 to the centrosomes and spindle poles led Wakefield et al. [69] to hypothesize that phosphorylated inactive GSK-3 stabilizes microtubules at the centrosomes, facilitating chromosomal attachment to the microtubules. Meanwhile, unphosphorylated, active GSK-3 throughout the main body of the spindle was thought to maintain the dynamic nature of the spindle, allowing mobile capture and movement of chromosomes to the metaphase plate. Their finding that global inactivation of GSK-3 with specific inhibitors led to changes in spindle morphology and abnormal chromosome alignment led to the proposal that inappropriate stabilization of microtubules throughout the length of the spindle would inhibit chromosome congression [69]. Recently, GSK-3 protein has been identified in mouse oocytes, and inhibition of GSK-3 also compromises spindle morphology and function, leading to aberrant segregation of the homologues [70]. We hypothesize that exposure of oocytes to high levels of FSH during meiotic maturation could increase PKB/Akt phosphorylation, resulting in global phosphorylation and inactivation of spindle GSK-3. The resulting increased spindle stability may explain the greater chromosome dispersion in oocytes exposed to the highest concentration of FSH.
There is accumulating evidence that perturbations in the timing of the meiotic cycle may make the oocyte vulnerable to errors in chromosomal division [44, 71]. Oocytes from older CBA females have a higher incidence of aneuploidy than those from younger females and progress more rapidly through the first meiotic prophase during maturation in vitro [44]. The authors hypothesize that the reduction of the "critical period during which attachment and alignment takes place could predispose to nondisjunction, abnormal chromosome orientation, and aneuploidy" [44]. This theory fits well with our findings here, where significantly more oocytes completed nuclear maturation within 16 h when COCs were exposed to 20, 200, and 2000 ng/ml FSH (Table 1) [49], and it was only at these concentrations that hyperploidy was observed (Table 2). Furthermore, three-dimensional analysis showed that the chromosomes are more widely dispersed on the spindles of oocytes exposed to 2000 ng/ml (Fig. 5). It may be that high levels of FSH are overriding cellular checkpoints that ensure that the chromosomes are correctly attached to the spindle, initiating the metaphase/anaphase transition while the chromosomes are still congressing and before metaphase has been fully established.
Alternatively, it is possible that the increased spread of chromosomes in oocytes matured in high concentrations of FSH indicates that these oocytes are at a slightly different stage of meiosis at the time point chosen in this experiment. FSH has been directly shown to alter the tempo of meiosis [72, 73]. Cumulus-enclosed oocytes can be prevented from undergoing spontaneous maturation using inhibitory agents such as hypoxanthine, and FSH overcomes this inhibition. FSH initially delays GVBD in both spontaneously maturing and hypoxanthine-arrested oocytes but later (after 6 h in vitro) stimulates maturation [72, 73]. During establishment of the metaphase spindle, there is pole-to-pole movement of chromosomes as they are captured by microtubules, form bidirectional kinetochore attachments with spindles from both poles, and congress on the metaphase plate [60]. In our study, it is possible that exposure of oocytes to high concentrations of FSH is associated with an initial delay in resumption of meiosis so that, when the oocytes are examined, the chromosomes are at late prometaphase and are still undergoing congression on the metaphase plate and are therefore more dispersed. However, while the spherical spindles with scattered chromosomes (Figs. 4, E–H) are likely to be in early prometaphase, it is harder to be certain whether the spindles with displaced chromosomes are at late prometaphase or are showing signs of congression failure [25]. Three oocytes have spindles with displaced chromosomes that could be at late prometaphase (e.g., Fig. 4, O and P), but the majority have spindles with poor alignment at the equator and many scattered chromosomes (e.g., Fig. 4, Q–S), which bear a strong resemblance to the oocytes with congression failure that have recently been described [25].
Another possibility is that high FSH concentrations facilitate the maturation of oocytes that have preexisting chromosomal errors. It has been suggested that inherent chromosomal abnormalities of the diploid oocyte, before any meiotic divisions, may account for a proportion of aneuploid embryos [74]. This theory postulates that aneuploid oocytes, generated by nondisjunction occurring during the mitotic divisions of the germ line, accumulate in the ovary and are only recruited into the growth phase toward the end of a woman's reproductive life, thus accounting for the increased incidence of aneuploidy with age [74–76]. Alternatively, chromosomally abnormal oocytes may initiate growth at a constant rate throughout reproductive life but may be identified by an, as yet, unknown screening mechanism and removed by atresia. IVM techniques may circumvent such potential protective mechanisms, allowing aneuploid oocytes to reach maturity. Errors arising during germ cell division are not well documented, as analysis of the chromosomal constitution of the oocyte at this stage of the cell cycle is difficult; however, fluorescent in situ hybridization (FISH) analysis of human oocytes and their polar bodies following IVF cycles has clearly demonstrated the existence of trisomic oocytes [77–79]. Furthermore, FISH analysis of immature diploid oocytes isolated from human preantral follicles has shown that a significant proportion of prophase I oocytes have extra signals for chromosomes 13, 21, and X (unpublished results). FSH may facilitate the meiotic resumption of such chromosomally abnormal oocytes, in effect rescuing these oocytes, and increasing the hyperploid rate in oocytes exposed to the highest concentrations of FSH. The observed increase in maturation rate (18%) is similar to the conservative estimate of the hyperploidy rate (19%) at 2000 ng/ml, supporting the hypothesis that FSH may be acting by forcing chromosomally abnormal oocytes to mature. However, the paucity of chromosomally abnormal oocytes that remained at the metaphase I stage at low concentrations of FSH in the current study make this scenario less likely.
Both PSSC and PCS were seen in the chromosome spreads. Because the incidence of both PSSC and PCS appeared independent of dose of FSH and PCS was present in our control group, FSH is unlikely to be responsible for these findings. Furthermore, the separated chromosomes and chromatids remain in the vicinity of each other (Fig. 3), making it difficult to discount the possibility that they have been separated during the spreading process. The absence of individual single chromatids in metaphase II oocytes suggests that precocious chromatid separation is not a significant occurrence during anaphase I.
The IVM of oocytes obtained from medium-sized antral follicles is gaining popularity as a means of generating developmentally competent oocytes for in vitro fertilization (IVF). The production of oocytes in this way has several potential benefits, which include increasing the number of oocytes available for IVF and reducing the need for exogenous gonadotropin treatment, offering an alternative to superovulation. This may be of particular benefit to women who are vulnerable to hyperstimulation. However, the association between the in vitro culture of oocytes and embryos and the genesis of fetal abnormalities in experimental animals [80, 81] has raised concerns about the safety and long-term consequences of IVM and highlighted the importance of clearly establishing the effect that components of the culture environment have on oocyte health. Here we have shown that FSH, which is routinely included in IVM media at doses of up to 1000 ng/ml or more, can increase the proportion of hyperhaploid oocytes, which may in turn give rise to trisomic embryos following fertilization. While exposure to high doses of FSH in vitro does not cause obvious, gross spindle aberrations, it does result in a wider dispersal of chromosomes, reflecting either effects on the spindle itself (and the linked process of chromosomal congression) or differences in the timing of meiosis. Both of these interpretations have been linked to the genesis of aneuploidy. A less plausible explanation is that FSH is rescuing aneuploid primary oocytes that would otherwise have been excluded from the maturing oocyte pool. Whatever the mechanism, this work establishes a direct link between FSH and the genesis of aneuploidy, and in this way supports the long-held view that it is the hormonal imbalance, seen toward the end of a woman's reproductive life, that is responsible for the increased incidence of aneuploidy with age [44, 45]. Finally, these results reinforce the current trend toward more moderate use of FSH in superovulation and IVM, both for clinical and research purposes.
|
FOOTNOTES |
|---|
2 Correspondence. FAX: +44 (0)207 594 2111; k.hardy{at}imperial.ac.uk
Received: 19 May 2004.
First decision: 30 June 2004.
Accepted: 25 August 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
