Hum. Reprod. Advance Access originally published online on November 17, 2006
Human Reproduction 2007 22(3):815-828; doi:10.1093/humrep/del442
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Influence of follicular fluid meiosis-activating sterol on aneuploidy rate and precocious chromatid segregation in aged mouse oocytes
1 Research Laboratories of Schering AG, Berlin and 2 Fakultät für Biologie, Gentechnologie/Biotechnologie, University Bielefeld, Bielefeld, Germany
3 Present address: Charite, Experimental Medicine, Institute of Cell Biology and Neurobiology, 12207 Berlin, Germany
4 To whom correspondence should be addressed at: University of Bielefeld, Faculty of Biology IX, Universitätsstrasse 25, D-33501 Bielefeld, Germany. E-mail: eiri{at}uni-bielefeld.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
BACKGROUND: Follicular fluid meiosis-activating sterol (FF-MAS) protects young oocytes from precocious chromatid separation (predivision). Reduced expression of cohesion and checkpoint proteins and predivision has been hypothesized to occur in age-related aneuploidy in oocytes. METHODS: To know whether FF-MAS also protects aged oocytes from predivision and from age-related non-disjunction, we analysed chromosome constitution in mouse oocytes matured spontaneously with or without 10 µM FF-MAS and in hypoxanthine (HX)-arrested young and aged oocytes induced to resume maturation by FF-MAS. Messenger RNA for checkpoint protein MAD2 and cohesion protein SMC1
was compared between oocytes matured with or without FF-MAS. RESULTS: Aged oocytes possessed many bivalents with single distal chiasma at meiosis I. Predivision was especially high in aged oocytes cultured sub-optimally to metaphase II in 
-minimum essential medium (-MEM). FF-MAS reduced predivision significantly (P < 0.001) but neither reduced non-disjunction nor induced aneuploidy in aged oocytes. Polyploidy was high in FF-MAS-stimulated maturation, in particular in the aged oocytes (P > 0.001). Relative levels of Smc1 mRNA appeared increased by maturation in FF-MAS, and mitochondrial clustering was restored. CONCLUSIONS: Sister chromatids of aged oocytes appear to be highly susceptible to precocious chromatid separation, especially when maturation is under sub-optimal conditions, e.g. in the absence of cumulus and FF-MAS. This may relate to some loss of chromatid cohesion during ageing. FF-MAS protects aged oocytes from predivision during maturation, possibly by supporting Smc1 expression, thus reducing risks of meiotic errors, but it cannot prevent age-related non-disjunction. Aged oocytes appear prone to loss of co-ordination between nuclear maturation and cytokinesis suggesting age-related relaxed cell cycle control.
Key words: oocyte/maturation/meiosis/non-disjunction/age
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
High fidelity of chromosome segregation is essential to maintain genomic stability. Chromosomal abnormalities in gametes predispose to aneuploidy in the embryo that can cause implantation failure, congenital abnormalities, spontaneous abortion, or mental retardation and genetic disease in the offspring. Advanced maternal age and depletion of the follicle pool associated with high risks of errors in chromosome segregation at oogenesis are the most important aetiological factors in human trisomy (Eichenlaub-Ritter, 1998
; Hassold and Hunt, 2001; te Velde and Pearson, 2002; Lynn et al., 2004) and in sub-fertility in perimenopausal women (Sandalinas et al., 2002; Pellestor et al., 2003; Gutierrez-Mateo et al., 2004; Kuliev and Verlinsky, 2004; Baird et al., 2005). It is unknown which factors are responsible for subtle and distinct changes in expression patterns (Steuerwald et al., 2001; Hamatani et al., 2004) and in cell cycle progression of aged oocytes (Eichenlaub-Ritter and Boll, 1989), which can result in reduced oocyte quality and high susceptibility to meiotic errors (for discussion, see Eichenlaub-Ritter et al., 2004). Inefficient bi-directional signalling between oocytes and the somatic compartment (Eppig et al., 2002), increase in mitochondrial dysfunction (Van Blerkom, 2000; Eichenlaub-Ritter et al., 2004; Thouas et al., 2005) as well as reduced expression of proteins in checkpoint control (Steuerwald et al., 2001), and the precocious loss of chromosome cohesion (Angell, 1991; Hodges et al., 2005) have been suspected to contribute to high risk of meiotic errors.
Most maternal age-related errors in oogenesis occur at meiosis I, including aneuploidies derived by errors in segregation of sister chromatids (for review, see Hassold and Hunt, 2001; Eichenlaub-Ritter et al., 2004; Lynn et al., 2004). In fact, chromosomal analyses of spare or donated human oocytes and of first and second polar bodies suggest that two mechanisms contribute to age-related chromosomal imbalance (Pellestor et al., 2002; Kuliev and Verlinsky, 2004). Errors in segregation of homologous chromosomes (commonly referred to as ‘non-disjunction’) lead to meiosis II oocytes with extra or missing dyads. Precocious separation of sister chromatids before anaphase II (predivision) can be recognized by the presence of chromatids in the meiosis II oocyte (Angell, 1991, 1995; Clyde et al., 2001; Pellestor et al., 2002, 2003). Both mechanisms appear to be responsible for chromosomal imbalance in oocytes and embryos of aged women. A recently established mouse model for age-related errors in chromosome segregation at oogenesis suggests that precocious loss of chromatid cohesion along chromatid arms may result in chiasma terminalization and formation of ‘susceptible meiotic configurations’, e.g. bivalents with only one distal attachment (Hodges et al., 2005). Precocious loss of cohesion apparently also contributes to predivision in young oocytes maturing under sub-optimal conditions (Cukurcam et al., 2003).
Thus, denuded mouse oocytes cultured in -minimum essential medium (-MEM) frequently contained one or several pairs of chromatids at metaphase II (Cukurcam et al., 2003). We now analysed the rate of predivision in aged oocytes of the same mouse strain when they were cultured in -MEM. If there is a constitutional loss of cohesion during oocyte ageing, one would expect that the aged oocytes are especially susceptible to spontaneous as well as culture-induced predivision.
The culture-induced predivision of chromatids in young denuded oocytes was significantly reduced when they matured in the presence of follicular fluid meiosis-activating sterol (FF-MAS) (Cukurcam et al., 2003). FF-MAS is a lipophilic sterol (4,4-dimethyl-5-cholest-8,14,24-trien-3-ol) present in human follicular fluid (Byskov et al., 1995) and a natural intermediate in the cholesterol biosynthetic pathway between lanosterol and testis-meiosis-activating sterol (T-MAS) (Byskov et al., 2002; Grondahl et al., 2003). FF-MAS induces resumption of maturation in fully grown, meiotically-arrested oocytes of several species (Byskov et al., 1995; Hegele-Hartung et al., 1999, 2001; Grondahl et al., 2000a; Cavilla et al., 2001; Faerge et al., 2001a; Donnay et al., 2004). Lanosterol was inactive in the stimulation of maturation (Downs et al., 2001) as well as protection from predivision in young oocytes, suggesting a specific activity of FF-MAS (Cukurcam et al., 2003). Resumption of maturation stimulated by FF-MAS to overcome meiotic inhibition by hypoxanthine (HX) requires long periods, and germinal vesicle breakdown (GVBD) occurs with a delay compared with spontaneous maturation or maturation in vivo. FF-MAS does not appear to be the major obligatory physiological factor in the induction of oocyte maturation in vivo (Hegele-Hartung et al., 1999; Baltsen, 2001; Downs et al., 2001; Faerge et al., 2001a; Byskov et al., 2002; Cao et al., 2004; Coticchio et al., 2004). However, FF-MAS appears to improve oocyte quality by promoting nuclear progression to meiosis II (Cukurcam et al., 2003; Marin Bivens et al., 2004a) and supporting spindle formation (Hegele-Hartung et al., 1999). Therefore, we wanted to obtain more information on the possibly beneficial activities of FF-MAS during culture in -MEM, now focusing on the effects of FF-MAS on chromosome behaviour in aged oocytes. We tested whether FF-MAS reduces rates of predivision as well as age-related non-disjunction.
Initially, chromosomal constitution of young and aged oocytes of the CBA/Ca mouse matured in vitro for only 3 or 6 h was analysed in a medium that does not induce extensive predivision (a modified M2 medium; Eichenlaub-Ritter and Boll, 1989). The number of bivalents with single distal chiasma and univalents at meiosis I was determined. According to the prediction by Hodges et al. (2005), one would expect more bivalents with only one distal chiasma in aged compared with young oocytes. After 3 h of culture, oocytes had undergone GVBD initiating spindle formation. After 6 h, most oocytes had progressed to late prometaphase I. Chromosomes were therefore presumably under increasing tension from spindle attachments. Differences in the numbers of univalents between oocytes cultured for 3 and 6 h should therefore indicate loss of chromatid cohesion due to pulling forces on chromosomes during spindle formation at meiotic progression.
Furthermore, chromosomal constitution was analysed after 16 h of culture in M2 medium without FF-MAS, when most oocytes had reached metaphase II to determine the rate of non-disjunction and of constitutional predivision (Eichenlaub-Ritter and Boll, 1989). In addition, the presence of precociously separated homologous chromosomes was assessed in partially immature mouse oocytes cultured in M2 medium, which initially resumed maturation but then became arrested at meiosis I for prolonged periods (until 16 h of culture). This should reveal susceptibility to precocious loss of cohesion, induced by extended metaphase I arrest with respect to oocyte age.
In the second set of experiments in this study, oocytes were chromosomally analysed when they matured spontaneously to metaphase II under sub-optimal conditions in -MEM without and with FF-MAS to assess FF-MAS protective effect on predivision. Because we had preliminary evidence that mitochondrial distribution was disturbed in naked oocytes maturing in -MEM without FF-MAS (Eichenlaub-Ritter et al., 2004), oocytes were also stained by Mitotracker to compare the association of mitochondria with the spindle at meiosis I (clustering) between maturation without and with FF-MAS.
Post-ovulatory ageing was shown to predispose mouse oocytes to precocious loss of chromatid cohesion at meiosis II accompanied by a decrease in the concentration of mad2 mRNA (Steuerwald et al., 2005). We therefore employed real-time RT–PCR to compare Mad2 mRNA concentrations in oocytes matured without and with FF-MAS. In addition, reduction in the concentration of mRNA for Smc1 encoding a meiotic cohesion protein has been implicated in reduced cohesion of sister chromatids in aged oocytes (Hodges et al., 2005). Smc1 mRNA was therefore analysed in groups of oocytes cultured without or with FF-MAS by real-time RT–PCR with -actin as standard.
Finally, we performed experiments to decide whether FF-MAS is capable of stimulating nuclear maturation in aged oocytes, comparable with young oocytes, when they are cultured in the presence of HX, a meiotic inhibitor. For this, maturation rate and chromosomal constitution were assessed in the oocytes of the two age groups when the meiotic block was overcome by FF-MAS.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Chemicals
FF-MAS, 1 mg/ml, (Schering AG, Berlin, Germany) was dissolved in 100% ethanol and diluted to a final concentration of 10 µM/l in
-MEM shortly before use (Hegele-Hartung et al., 1999, 2001; Cukurcam et al., 2003). -MEM without ribonucleosides (Gibco, Karlsruhe, Germany) was supplemented with 8 mg/ml human serum albumin (Sigma, Dreisenhofen, Germany), 0.23 mM sodium pyruvate (Sigma), 2 mM glutamine (ICN, Eschwege, Germany), 100 IU/ml penicillin (ICN) and 100 µg/ml streptomycin (ICN) shortly before use. HX (Sigma), 1 mg/ml, was dissolved in 1 M NaOH and used at a final concentration of 3 mM in medium.
A modified M2 medium containing calcium lactate (2 mM) in addition to calcium chloride was employed for culture of oocytes in the control as there was a difference in chromosome constitution between young and aged oocytes of the CBA/Ca mouse matured in this medium (Eichenlaub-Ritter and Boll, 1989). Modified M2 medium was essentially similar to the normal M2 medium (Hogan et al., 1994) and contained 99.2 mM sodium chloride, 2.6 mM potassium chloride, 1.8 mM calcium chloride, 0.4 mM sodium dihydrogenphosphate, 0.5 mM magnesium chloride, 4.2 mM sodium carbonate, 20 mM HEPES buffer, 22 mM sodium lactate, 0.3 mM sodium pyruvate, 5.6 mM glucose and 2 mM calcium lactate plus antibiotics (penicillin and streptomycin), phenol red and 4 mg/ml bovine serum albumin (BSA, Fraction V, Sigma).
Animals and treatments
Inbred female mice of the CBA/Ca strain were used for the experiments as the follicle pool becomes depleted in females of this strain by the end of the reproductive period, similar to the human (Gosden, 1975; Fabricant and Schneider, 1978; Brook et al., 1984; Eichenlaub-Ritter et al., 1988). CBA/Ca mice were originally obtained by Harlan-Winkelmann (Borchen, Germany) and were kept in the institutional facilities under a 12-h light–dark cycle (0700–1900 hours) and fed ad libitum. Oocytes were obtained from the ovaries of hormonally untreated mice at diestrus of the natural cycle. In the young and aged groups, oocytes were isolated from 6- to 16-week-old CBA/Ca mice or from females approaching the end of their reproductive period (
9 months of age), respectively. Oocytes used for maturation in modified M2 medium for cytogenetic analysis at 3 or 6 h of culture were from 10 and 12 young mice and from 16 and 18 aged mice, respectively. Those used for spreading after 16 h of culture in modified M2 medium were from 16 young and 37 aged mice. For isolation of oocytes matured in -MEM, a total of 171 young mice and 276 aged mice were used. The young group included oocytes from a previous study (772 in the -MEM, 441 in the -MEM plus FF-MAS, 468 in the -MEM plus HX and 754 in the -MEM plus HX plus FF-MAS group) (Cukurcam et al., 2003). Aged oocytes from each experimental group had already been collected, cultured and analysed in parallel with the young ones during the initial study. However, the numbers of young and aged oocytes were insufficient for statistical analysis and therefore increased for the current report to obtain data for comparisons between the age groups. Isolation of oocytes was performed as previously described (e.g. Eichenlaub-Ritter and Boll, 1989; Soewarto et al., 1995; Cukurcam et al., 2003).
A previous study revealed differences in meiotic timing and aneuploidy rate between young and aged oocytes when oocytes were cultured in modified M2 medium (Eichenlaub-Ritter and Boll, 1989). This medium was therefore chosen for the control in the present study. Oocytes from large antral follicles were cultured in 1 ml of modified M2 medium or supplemented -MEM (Hegele-Hartung et al., 1999; Cukurcam et al., 2003) in Nuclon 4-well dishes. Resumption of spontaneous maturation (GVBD) and progression to metaphase II [polar body (PB) formation] were analysed after 16 h of culture. Oocytes matured in modified M2 medium were cultured in an ungassed incubator at 37°C for 3, 6 or 16 h and then spread for chromosomal analysis. In the rest of the experiments, oocytes matured in -MEM for 16 h in an atmosphere of 5% CO2 in air in the absence or presence of 10 µM FF-MAS. The concentration of FF-MAS was chosen because our previous study showed efficient induction of maturation to meiosis II without detectable adverse side effects on chromosome segregation whereas spindle formation and cytoplasmic maturation events appeared improved (Hegele-Hartung et al., 1999; Cukurcam et al., 2003).
In the last set of experiments, FF-MAS-induced maturation was analysed in oocytes from young and aged mice isolated and cultured in -MEM plus 3 mM HX (Hegele-Hartung et al., 1999; Downs et al., 2001; Cukurcam et al., 2003). After culture for 22 h, the rate of PB formation and maturation to metaphase II was determined. The -MEM + HX group served as a negative control. Only those oocytes from experiments with <20% spontaneous maturation in the HX control group were considered blocked by HX and were therefore included in the analysis.
Analysis of meiotic maturation
Meiotic progression was analysed in in vitro matured oocytes by determining the number of meiotically blocked or incompetent oocytes with an intact germinal vesicle (GV). From those oocytes resuming maturation, the percentage of oocytes blocked at GVBD or progressing to meiosis II and extruding a first PB was analysed. Activation of oocytes was rarely observed and not further considered in the statistical evaluation. Maturation was analysed after 16 h of spontaneous maturation without HX or after 22 h of FF-MAS-stimulated maturation in the presence of HX.
Analysis of nuclear maturation and chromosomal constitution
Spreading, hypotonic treatment, fixation and C-banding methods have been previously described (Eichenlaub-Ritter and Boll, 1989; Soewarto et al., 1995; Cukurcam et al., 2003). Chromosomes were analysed by phase-contrast with an Axiophot microscope (Zeiss, Oberkochem, Germany) equipped with a sensitive CCD camera (Sensi Cam; PCO CCD Imaging, Kelheim, Germany).
Meiotic status of oocytes resuming maturation was analysed by counting the numbers of oocytes with bivalent chromosomes or dyads without or with chromatids, independent of ploidy.
Oocytes matured in modified M2 medium and spread at 3 or 6 h of maturation were analysed for the presence of univalent chromosomes, which lacked exchanges or precociously lost cohesion. Only when homologues were separated for more than their length, they were considered as univalents. Furthermore, the numbers of oocytes with chromosomes with a single distal chiasma and the percentage of homologous chromosomes with single distal chiasma of all bivalents were analysed. In addition, the percentage of bivalents with single distal chiasma, which were separated visibly from each other and had an unstained gap between telomeres of homologous chromosomes after spreading, was assessed. This was analysed to identify chromosomes possibly highly susceptible to loose cohesion.
All oocytes, which had matured for 16 h, were spread for analysis of non-disjunction and predivision. Hyperploid oocytes with chromosomes in the haploid range containing >20 dyads or the respective numbers of chromatids were determined (Figure 1D). Oocytes containing at least 36 individually recognizable dyads were designated as polyploids (Figure 1C). Percentage of polyploids (‘diploid’ metaphase II oocytes with twice the number of dyads or the respective number of chromatids that progressed to meiosis II in the absence of cytokinesis and PB formation) was determined in relation to all oocytes with dyads.
|
Precocious separation of chromatids (with single or pairs of monads) at metaphase II (Figure 1E and F, arrows) was separately analysed for all oocytes with unambiguously determined number of chromosomes possessing a PB. Separately, we also analysed the numbers of polyploid oocytes with chromatids and calculated predivision rate in all oocytes containing dyads, irrespective of ploidy. Only those oocytes with chromatids separated from each other for more than their lengths were included in the group with ‘predivision’, whereas separation of chromatids for a shorter distance was regarded as spreading-associated loss of contact (arrowhead in Figure 1F). The numbers of oocytes with one to two chromatids, three to four chromatids or more than four chromatids were analysed from all oocytes containing chromatids to assess relative susceptibility to predivision in comparison with age and FF-MAS exposure.
Labelling of mitochondria in maturing mouse oocytes
Oocytes were cultured in -MEM without or with FF-MAS for 6.5 h. Oocytes were vital stained by 200 ng/ml MitoTracker TM (Molecular Probes) and 0.5 ng/ml Hoechst 33342 during the last 30 min before analysis by fluorescent microscope. The numbers of GVBD oocytes with distinct mitochondrial ring around the spindle (clustering) with congressing chromosomes were analysed in oocytes of both groups, and the numbers of oocytes with only partial clustering or with more random, dispersed distribution of mitochondria (‘no clustering at spindle’) were determined.
Quantitative real-time PCR
Oocytes from sexually mature, unprimed young (5 months) and aged (9 months) mice that matured spontaneously in -MEM without or with 10 µM FF-MAS for 16 h were collected for analysis of expression of Mad2 and Smc1 mRNA with -actin (Actb) as standard. Five-month-old females rather than 6- to 16-week-old ones were chosen in this part of the experiment in the young age group because mRNA levels increase from 2 months of age and are closer to maximal levels of smc1 mRNA in GV-stage oocytes of this age, according to reports in the literature (Hodges et al., 2005). To assess the influence of FF-MAS, we isolated total RNA from 15 to 20 metaphase II mouse oocytes in each age group matured without or with FF-MAS using the NucleoSpin RNA II Kit with 0.1 µg/µl glycogen as a carrier (Macherey-Nagel, Düren, Germany). Real-time RT–PCR was used in a one-step procedure by LightCycler (Roche Diagnostics) and the QuantiTect SYBR Green RT–PCR Kit (Qiagen) with 20–30 ng/µl template RNA in a volume of 20 µl according to the manufacturer’s protocol. Primers for Smc1 were the same as those reported by Hodges et al. (2005). Primers for Mad2 and -actin were designed using Primer Select analysis software [DNASTAR, Madison, USA; Accession ID (NCBI): Mad2 U83902
[GenBank]
; Actb NM-007393]. The following primers were used: Smc1 (5'-CACAGTTTTCGGCCTGCTCCA-3' and 5'-CCTGCTGCCTTCTTGGTCTTCG-3'), Mad2 (5'-GCATTTTGTATCAGCGTGGCAT-3' and 5'-GGCTTTCTGGGACTTTTCTCTACG-3') and -actin (5'-TGCGTGACATCAAAGAGAAG-3' and 5'-GATGCCACAGGATTCCATA-3'). Reverse transcription was at 50°C (20 min), followed by denaturation of the complementary DNA/RNA hybrid and activation with HotStarTaq DNA Polymerase at 95°C (15 min), followed by amplification and quantification at 55 cycles (denaturation at 94°C for 15 s, annealing at 58°C for 45 s and elongation at 72°C for 60 s) and a melting curve programme. The specificity of RT–PCR products was analysed by agarose gel electrophoresis. Quantification was by LightCycler Software Version 3.5.3. Crossing point values were determined by the second derivative method (Rasmussen, 2001). Smc1 and Mad2 levels were normalized to -actin for comparison of expression between groups matured with or without FF-MAS in the young and old age group. Experiments were repeated twice, using all three probes for each experimental group in repeat runs. Relative expression levels and statistical significance were calculated using the REST software (Pfaffl et al., 2002), and according to these methods, relative quantification results are expressed as the log2 Smc1:-actin ratio of the sample from groups exposed to FF-MAS normalized to the Smc1:-actin ratio of the sample obtained from culture without FF-MAS.
Statistical analysis
Statistical analysis was by 
2-test with Yates correction. P < 0.05 was considered significant in comparisons between young and aged groups. For comparisons between age and groups matured with and without FF-MAS, P < 0.025 or P < 0.005 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Spontaneous maturation of young and aged oocytes cultured in modified M2 medium
The pool of follicles becomes depleted with advanced maternal age in the ovaries of CBA/Ca mice. Accordingly, the numbers of oocytes, which could be retrieved from large antral follicles of hormonally unstimulated mice at diestrus, were significantly lower in the aged as compared with the young group (P < 0.001). About 70% of the oocytes from young and aged mice emitted a first PB (upper part of Table I), not significantly different from each other. A similar percentage of young and aged oocytes resuming maturation possessed bivalents or dyads, respectively (last column, upper part of Table I).
|
Chromosomal constitution of young and aged, meiosis I and metaphase II oocytes cultured in modified M2 medium
There was only one young oocyte that contained a pair of univalents after GVBD, after 3 h of maturation in vitro, whereas all other young and aged oocytes spread after 3 or 6 h possessed bivalent chromosomes. About 90% of all oocytes from young and aged mice contained at least one bivalent with a single distal chiasma (Figure 1 arrowheads and arrows; left columns of Table II). However, the percentage of oocytes that contained a bivalent with a distal chiasma with a visible, unstained gap between telomeres of homologous chromosomes (Figure 1A) was significantly higher in the aged as compared with the young group (Table II). There was no indication that increased tension on chromosomes during spindle formation during the 3- to 6-h culture period increased the number of oocytes with more loosely attached homologous chromosomes. However, the number of bivalents with only one distal chiasma per oocyte (right panel of Table II) was higher in the aged as compared with the young group (about one more bivalent with such a configuration, Table II). The percentage of tightly attached chromosomes was lower in the aged group whereas that with presumably loosely attached homologous chromosomes with unstained gap was significantly higher in the aged as compared with the young mouse oocytes (P < 0.05 and 0.001, for 3 and 6 h, respectively; Table II).
|
Analysis of control oocytes from young and aged females, which matured to metaphase II during 16 h of culture in modified M2 medium, revealed that more aged oocytes possessed two sets of dyads (polyploidy, Figure 1D) as compared with the young group, although the differences did not reach statistical significance (Table III). Hyperploidy was significantly higher (P < 0.05) in aged as compared with young oocytes matured in modified M2 medium.
|
In addition, predivision was significantly higher in the aged as compared with the young oocytes (P < 0.001; Table IV). Predivision was only once found in a metaphase II oocyte of the young group matured in modified M2 medium but occurred in 14 of 176 metaphase II oocytes of the aged group with PB and in 4 aged polyploid meiosis II oocytes (Table IV). The majority of aged oocytes that matured in M2 medium and underwent predivision contained one to two chromatids and a few oocytes possessed three to four chromatids. None of the young oocytes had more than two chromatids at metaphase II.
|
Of the oocytes failing to emit a PB and still arrested in meiosis I after 16 h of culture that were chromosomally analysed (n = 86 in the young and 44 in the aged group), only a few possessed univalents (data not summarized in Table II). Thus, 0.4% of bivalents in the young group had separated into dyads (7 of 1720), in contrast to 2.8% in the aged oocytes (25 of 880), significantly different from each other (
2-test: P < 0.001).
Spontaneous maturation of young and aged oocytes cultured in -MEM
About 4–5% of the young and aged cumulus-cell denuded oocytes isolated from large antral follicles of diestrous CBA/Ca mice did not resume maturation after 16 h of culture in -MEM (left panel, Table I). Only a low percentage of oocytes arrested after GVBD and failed to emit a PB. Accordingly, only few young and aged oocytes maturing without FF-MAS contained bivalent chromosomes (right columns of Table I). There was no difference in nuclear maturation rate between the age groups cultured without FF-MAS, and most oocytes (>94%) reached meiosis II, emitted a PB and possessed dyads (Table I, right column).
Chromosomal constitution of young and aged, metaphase II oocytes cultured in -MEM
Most oocytes cultured in -MEM without FF-MAS were euploid and contained 20 dyads (Figure 1C). The number of polyploid oocytes with two sets of dyads (Figure 1D), which underwent anaphase I in the absence of cytokinesis, was low in both age groups cultured in -MEM (Table III). The percentage of hyperploid oocytes was higher in the aged control compared with the young control group, but the difference did not reach statistical significance (P = 0.4). About 10% more oocytes of the aged compared with the young mice cultured in -MEM without FF-MAS contained chromatids (P < 0.025; Table IV). On average, 
40% of all young and aged oocytes with predivision, which matured in -MEM, had one to two individual chromatids, 20% three to four chromatids and 35% of the young and aged oocytes had more than four chromatids. Separation of all chromatids (anaphase II) was only found in a single aged oocyte matured spontaneously in -MEM.
Influence of FF-MAS on spontaneous maturation in -MEM
There was no influence of 10 µM FF-MAS on the percentage of maturation incompetent oocytes remaining in GV stage when culture was in -MEM (Table I). The numbers of oocytes that failed to emit a PB after GVBD was increased slightly in both age groups, but the increase reached significance only in the aged group (P < 0.001; middle panel of Table I). Accordingly, there was an increase in the numbers of oocytes containing bivalents in spread oocytes of the aged FF-MAS group as compared with that cultured without FF-MAS (Tables I; P < 0.005). The numbers of polyploid oocytes, which failed to emit a PB, were slightly but significantly increased by in vitro maturation in the presence of FF-MAS in the aged group (P < 0.005; Table III), with a similar trend for an increase in polyploidy in the young group (Table III).
FF-MAS did not appear to influence non-disjunction in young or aged oocytes. The percentage of hyperploid oocytes tended to be higher in aged as compared with the young oocytes, but this was not significant (Table III).
There was a pronounced effect of FF-MAS on predivision rate (Table IV). The percentage of all oocytes with chromatids (with the numbers of dyads in haploid and/or polyploid range) dropped significantly in the young group (P < 0.01) and even more dramatically in the aged group (P < 0.001; Table IV). Predivision rate was decreased in oocytes with dyads in the haploid or polyploid range upon culture in FF-MAS (middle columns of Table IV). The percentage of oocytes with only one to two chromatids increased and that with more than four chromatids decreased in the aged oocytes matured with FF-MAS. Unlike in the control, the number of oocytes with more than four chromatids was significantly lower in the aged as compared with the young FF-MAS group.
Maturation and chromosomal constitution of young and aged oocytes induced to resume maturation by FF-MAS
HX was effective in preventing spontaneous maturation of young and aged mouse oocytes (Table I). Most oocytes escaping the HX arrest without FF-MAS matured to metaphase II (Table I). Hyperploidy was high in young oocytes escaping the HX arrest without FF-MAS, but the numbers of oocytes were too low to draw any conclusion (Table III). It was noted that the percentage of oocytes with predivision was again higher in the aged group cultured without FF-MAS that escaped meiotic arrest by HX as compared with the young group (Table IV).
The numbers of oocytes cultured in -MEM with HX that contained a GV 16 h after isolation dropped significantly when young or aged oocytes were cultured in the presence of FF-MAS (P < 0.001), confirming the meiosis-inducing properties of FF-MAS (lower part of Table I). The number of GV stage oocytes was significantly lower in the FF-MAS-stimulated aged compared with the young oocytes (P < 0.001), suggesting a particularly strong meiosis-inducing effect of FF-MAS in the aged population. In contrast, the number of oocytes with GVBD was higher and with PB was significantly lower in the FF-MAS-stimulated aged as compared with the young group of oocytes (P < 0.001). Although there were many oocytes with GVBD in the aged group, the numbers of oocytes with bivalents were not higher in the aged as compared with the young group (right, lower column in Table I). Most oocytes resuming maturation contained dyads. The significant differences in the numbers of GVBD oocytes between the age groups were apparently related to a failure of a large number of aged oocytes to emit a PB at anaphase I. They therefore became polyploid (Table III). Only one of the aged oocytes with PB was hyperploid, whereas 4% of the FF-MAS-stimulated young oocytes possessed >20 dyads (Table III).
Maturation in -MEM plus HX in the presence of FF-MAS did not result in extremely high rates of predivision (Table IV) and was significantly lower than the predivision rate in oocytes spontaneously maturing to metaphase II with FF-MAS without HX (P < 0.001). Predivision rate was not different between age groups and mostly involved only one to two chromatids.
Behaviour of mitochondria in young meiosis I oocytes matured with and without FF-MAS
Mitochondrial distribution characteristically changes during meiotic maturation of mammalian oocytes (Calarco, 1995), such that many mitochondria accumulate in dense clusters in the vicinity of the spindle during meiosis I and II(Figure 2a–a'). However, mitochondria of most denuded mouse oocytes that matured spontaneously in vitro in -MEM for 6.5 h failed to cluster and were either only partially associated with the ooplasm in the vicinity of the spindle (Figure 2b–b') or were even still dispersed throughout the ooplasm of the meiosis I oocyte (Figure 2c–c'). Only a small percentage of the oocytes (20%) matured sub-optimally in -MEM without FF-MAS possessed typical ring-like mitochondrial aggregates at the periphery of the spindle with the congressing chromosomes (‘clustering at spindle’, Table V). In >50% of the oocytes matured without FF-MAS, mitochondria appeared dispersed (‘no clustering at spindle’, Table V). In contrast, the percentage of oocytes with mitochondrial clustering was significantly higher in the oocytes cultured in -MEM in the presence of FF-MAS (Table V) compared with oocytes maturing without FF-MAS.
|
|
Expression of Mad2 and Smc1
mRNA in young and aged, metaphase II oocytes matured spontaneously without or with FF-MASThe regulation factor in the younger group (y) corresponded to an absolute gene regulation (2-log) of plus 0.259 ± 0.21 and minus 0.106 ± 0.08 for Smc1
and Mad2, respectively. The absolute gene regulation (2-log) in the aged population (a) was plus 0.817 ± 0.22 and plus 0.305 ± 0.26 for Smc1 and Mad2, respectively (Figure 2d). Accordingly, Smc1 mRNA levels appeared increased by culture in the presence of FF-MAS by 25% in the young group and by 81% in the aged oocytes. Although the increase in Smc1 was particularly high and consistently found in repeat analysis of the aged and young groups, differences to culture without FF-MAS did not reach statistical significance. Mad2 mRNA was insignificantly reduced in the young oocytes cultured with FF-MAS compared with control and increased in the aged group (Figure 2d); again differences did not reach statistical significance. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Chromosomal constitution of aged meiosis I oocytes
High fidelity in meiotic chromosome segregation depends on a sequential loss of cohesion between the arms of sister chromatids at meiosis I leading to chiasma resolution followed by the loss of cohesion between the centromeres of the sister chromatids at anaphase of meiosis II (Lee and Orr-Weaver, 2001
; Nasmyth, 2001; Petronczki et al., 2003; Eichenlaub-Ritter et al., 2004; Watanabe, 2004). Some of the molecules and genes regulating this sequential loss have only recently been discovered (Terret et al., 2003; Rabitsch et al., 2004; Revenkova et al., 2004; Penkner et al., 2005; Watanabe and Katajima, 2005). Sequential loss of cohesion between arms and centromeres of sister chromatids requires activity of the protease separase (Kudo et al., 2006; Lee et al., 2006) and appears regulated by post-translational phosphorylation (Kitajima et al., 2006). At meiosis II arrest, EMI1 (early mitotic inhibitor 1) and EMI2 (Emi-related protein 1, Erp1/Emi2) inhibitors retain anaphase-promoting factor inactive and prevent the loss of cohesion between centromeres of sister chromatids until fertilization (Paronetto et al., 2004; Shoji et al., 2006). Cohesion proteins are initially synthesized in prophase I of meiosis in the fetal ovary and appear to be at least partially renewed during the long meiotic arrest, as suggested by analysis of relative levels of the meiosis-specific cohesion mRNA of Smc1 (Hodges et al., 2005). It has been speculated that aged oocytes may gradually lose cohesion molecules during the sustained time in G2-phase arrest resulting in human oocytes with univalents at meiosis I (Angel, 1995) and meiosis II oocytes with chromatids (Angell, 1991, 1995; Dailey et al., 1996; Clyde et al., 2001; Sandalinas et al., 2002; Gutierrez-Mateo et al., 2004; Kuliev and Verlinsky, 2004). When cohesion is lost precociously, or when univalents or chromosomes with only one distal attachment are present, sister chromatids may be susceptible to attach to opposite instead of one spindle pole at metaphase I (Wolstenholme and Angell, 2000; Pellestor et al., 2002; Lynn et al., 2004). This may lead to errors in chromosome segregation as shown in oogenesis of the X0 mouse (Hunt et al., 1995). This study did not provide evidence that chromosomes were achiasmatic and totally lacked exchanges in either age group, as originally discussed according to the ‘production line hypothesis’ (Henderson and Edwards, 1968; Speed, 1977). Attachment to spindle fibres and increasing tension on chromosomes did not cause total resolution of chiasmata and loss of attachments at distal and proximal sites during the meiotic prometaphase I stage in the mouse oocytes. However, the reduction in Smc1 mRNA and, presumably, the respective cohesion protein during ageing, before resumption of meiosis, may explain the increase in univalents and susceptibility to a precocious loss of cohesion observed in the meiosis I-blocked aged compared with young oocytes cultured for 16 h in M2 medium. Aged mammalian oocytes may therefore possess increased numbers of susceptible bivalents, and these are prone to precociously divide to form univalents. This also explains why aged oocytes are prone to predivision and may form bivalents with only telomeric attachment. SMC1 is a meiotic cohesion that mediates attachment at arms and centromeres of sister chromatids in female meiosis (Hodges et al., 2005). The phenotype in slightly aged oocytes (6 months) of transgenic Smc1–/– knockout mice is very similar to what we detected in aged oocytes (9 months) of CBA/Ca mice, although the age-effect (partial or total loss of cohesion) was much more profound in the knockout mice as compared with aged, wild-type oocytes of the CBA/Ca strain.
Predivision in aged oocytes cultured in M2 medium
Aged oocytes cultured for only 3 or 6 h had frequently bivalents with a single distal, presumably ‘terminalized’ chiasma close to the telomeres. They may be susceptible to stretching during spreading and therefore frequently separated by an unstained gap. A reduction in cohesion might contribute to the loss of physical attachment before anaphase II. Analysis of recombination in chromosomally normal or unbalanced offspring of aged women suggests that a production line (Henderson and Edwards, 1968) does not exist in human oogenesis (Lynn et al., 2004), nor in the mouse (Hodges et al., 2005). High levels of recombination appear to protect the chromosomes, which reside in the cytoplasm of an aged oocyte, from random and untimely segregation (Kong et al., 2004; Lynn et al., 2004). Predivision appeared to affect mostly only one chromosome (one to two chromatids at meiosis II) when maturation of CBA/Ca oocytes occurred in M2 medium while it increased by culture in -MEM. Number and position of exchanges might play a role in susceptibility to predivsion of individual chromosomes. Chromosome size per se did not appear to be predict predivision in diazepam-induced sustained meiotic arrest (Yin et al., 1998; Sun et al., 2001). The present observations support the notion that homologous chromosomes in aged oocytes are generally more prone to predivision, presumably because they are connected less tightly to each other compared with chromosomes in oocytes of younger females supporting the predictions by the Smc1–/– mouse model (Hodges et al., 2005).
Induction of high rates of predivision in aged oocytes cultured in sub-optimal media (-MEM)
Predivision rate was, in fact, much higher in the aged compared with the young oocytes matured without follicle cells in -MEM. Loss of cumulus and sub-optimal maturation media may therefore synergistically affect susceptibility to meiotic error in human oocytes obtained from reproductively aged patients.
Post-ovulatory ageing at metaphase II caused separation of sister chromatids, irrespective of maternal age (Dailey et al., 1996; Mailhes et al., 1998), associated with a reduction in MAD2 mRNA levels (Ma et al., 2005; Steuerwald et al., 2005). We show here that FF-MAS protects from predivision but does not significantly influence Mad2 mRNA levels. The minor increase of Mad2 in aged oocytes could possibly contribute to improvement of checkpoint control. Relative increase in mRNA coding for meiotic cohesion protein SMC1 by FF-MAS appears of greater significance, especially in the aged group. Ablation of expression of genes by injection of small inhibitory RNA showed that translation from existing mRNAs is essential for normal meiotic maturation and checkpoint control (Stein et al., 2003, e.g. translation of Mad2 mRNA; Homer et al., 2005; Vogt et al., 2005). FF-MAS may exert some of its positive effects on oocyte quality by subtly and differentially influencing gene expression during maturation.
Protective effect of FF-MAS on predivision in aged oocytes
This study provides further evidence that FF-MAS, the natural component of follicular fluid, protects oocytes from predivision (Cukurcam et al., 2003). The present work extends the knowledge on pathways that may be influenced by FF-MAS showing that FF-MAS supported normal clustering of mitochondria around the spindle. Now, further studies are required to explore the molecular basis of this activity.
FF-MAS reduced the overall incidence of predivision of individual chromosomes in aged oocytes, because the numbers of meiosis II oocytes with only one to two chromatids increased whereas that with several chromatids decreased. Previous studies reported that oocyte and embryo quality of several species was enhanced by the presence of FF-MAS or its synthetic analogues during maturation (Andersen et al., 1999; Hegele-Hartung et al., 1999; Grondahl et al., 2000a; Cavilla et al., 2001; Donnay et al., 2004; Marin Bivens et al., 2004a,b). It is feasible that the protective effect on predivision contributed to the formation of euploid oocytes, with higher developmental potential of the FF-MAS-exposed groups compared with controls.
Influence of FF-MAS on non-disjunction rate
Initial data implied that FF-MAS might interfere with development and chromosome segregation of human embryos after exposure of human oocytes at metaphase II. However, the chromosomal aberrations were possibly related to the time of treatment and the solvent rather than the sterol (Ziebe et al., 2003; Bergh et al., 2004; Loft et al., 2004, 2005). Hyperploidy rate involving whole dyads was unaffected by maturation in the presence of FF-MAS in the present study. Our hopes that FF-MAS protects from age-related non-disjunction were therefore not substantiated. Hyperploidy was also not significantly increased in FF-MAS-stimulated maturation comparable with a study with human oocytes in which FF-MAS at metaphase II (4 h) did not adversely affect chromosomal constitution of the embryo (Loft et al., 2005). If non-disjunction is related to the presence of increased numbers of susceptible chromosomal configurations in the aged oocytes before resumption of maturation, due to the loss of attachment before maturation, the aberration may be irreversible. FF-MAS during maturation might not be able to eliminate such earlier age-related processes. The number of oocytes with dyads in the haploid range was too low to draw any firm conclusions on FF-MAS protective effects when culture was with HX. The pronounced asynchrony in nuclear maturation and cytokinesis in the FF-MAS-stimulated maturation by far outweigh the effects of FF-MAS on non-disjunction under these conditions (presence of HX).
Basis of activity of FF-MAS
Maturation by FF-MAS depends on the genetic background (Griffin et al., 2004), questioning a universal role of FF-MAS in meiotic resumption and progression to meiosis II (Byskov et al., 1997; Leonardsen et al., 2000; Lu et al., 2000; Baltsen, 2001; Downs et al., 2001; Tsafriri et al., 2002; Cao et al., 2004). Rather, other signalling pathways appear essential for resumption of maturation in vivo and in vitro (Richard et al., 2001; Conti et al., 2002; Mehlmann et al., 2002, 2004; Coticchio et al., 2004; Kawamura et al., 2004; Park et al., 2004; Ashkenazi et al., 2005; Hinckley et al., 2005). Phosphatidylinositol 3 kinase and Akt/protein kinase B activity is required for GVBD and cytokinesis in FSH- and FF-MAS-induced maturation of cumulus-enclosed oocytes (Hoshino et al., 2004; Kalous et al., 2006) but not in FF-MAS-stimulated maturation of denuded oocytes (Hoshino et al., 2004). Transcription is also not essential for the FF-MAS-induced resumption of meiosis, so that non-genomic activities may be important (Grondahl et al., 2000b, 2003; Faerge et al., 2001b). Our data confirm that FF-MAS stimulates nuclear maturation potently in young and aged oocytes of the CBA/Ca mouse.
Balanced predivision causing presence of pairs of chromatids in metaphase II oocytes was frequently associated with mitochondria with reduced redox potential in human oocytes and in embryos with low developmental potential (Wilding et al., 2003). Distribution and activity of mitochondria with high redox potential are tightly controlled during maturation and can influence oocyte fate (for discussion see, Wilding et al., 2003; Eichenlaub-Ritter et al., 2004; Van Blerkom, 2004). FF-MAS supports normal mitochondrial distribution and could thus contribute to normal mitochondrial functionality. Thouas et al. (2005) provided evidence that aged oocytes are more developmentally sensitive to mitochondrial damage. Further studies are on the way to assess mitochondrial activity when oocytes mature in sub-optimal conditions in -MEM or in other media with or without FF-MAS.
The spindle assembly/attachment checkpoints (SACs) ensure that separation of homologues occurs only when all chromosomes are under tension from spindle fibres (Eichenlaub-Ritter et al., 2004; Watanabe, 2004; Homer, 2006). The SACs may be leaky and permissive, especially in aged oocytes (for discussion, see Homer, 2006). Checkpoint signalling requires sufficient expression of checkpoint genes (Homer et al., 2005; Vogt et al., 2005). FF-MAS did not significantly affect the concentrations of Mad2 mRNA whereas increases in mRNA of genes in chromatid cohesion might be of relevance. Influences on mitochondrial distribution might be upstream of the availability of high-energy substrates to support gene expression and to prevent mRNA degradation at maturation. Improvement of local mitochondria-dependent signalling and availability of high-energy substrates by FF-MAS may thereby protect centromeric cohesion proteins from precocious dissociation from centromeres of chromatids in a direct and indirect fashion (Mailhes et al., 2003).
FF-MAS-induced polyploidy in aged oocytes
Maturation rates were consistently higher in -MEM compared with M2 medium. Nuclear maturation and cytokinesis appear well supported by culture in -MEM but may also be less guarded by checkpoint controls. This may be the reason why polyploidy was enhanced in the FF-MAS groups and particularly in the FF-MAS-stimulated maturation in HX (Cukurcam et al., 2003). Maturation to meiosis II in the absence of cytokinesis was much more frequent in aged as compared with young oocytes stimulated to mature in the presence of HX. The accumulation of cortical vesicles at the oolemma was delayed in FF-MAS-stimulated maturation (Hegele-Hartung et al., 1999), possibly due to a delay in actin-dependent processes. Cytokinesis requires actin polymerization (Soewarto et al., 1995). Delayed cytokinesis and rapid anaphase I progression that is characteristic for aged mouse oocytes (Eichenlaub-Ritter and Boll, 1989) and may relate to reduced expression of checkpoint genes (Steuerwald et al., 2001; Hamatani et al., 2004; Homer et al., 2005; Vogt et al., 2005) can therefore synergistically act to increase the risk of polyploidy in aged oocytes. The present study confirms that aged oocytes have permissive cell cycle controls affecting synchrony in sequential loss of cohesion between arms and centromeres of sister chromatids at meiosis I and II and co-ordination between anaphase I and cytokinesis. FF-MAS can only partially restore the loss of control and may even increase the risk of polyploidy, particularly in the aged oocytes.
Beneficial influence of FF-MAS to improve oocyte quality
Reduced synthesis of FF-MAS (Byskov et al., 1997; Eppig et al., 2002; Grondahl et al., 2003) may contribute to high susceptibility to predivision of aged oocytes in vivo when auto- and paracrine communication is compromised (Volarcik et al., 1998; Combelles et al., 2004). It is of relevance that predivision not only drops significantly by maturation in FF-MAS but reaches levels in aged oocytes that are no more significantly different from young oocytes. Our observations are therefore of clinical relevance for assisted reproduction because they suggest that addition of FF-MAS to isolation and culture media might have a protective effect in preventing precocious chromatid segregation in human oocytes. This study emphasizes that protection from predivision may be particularly important when oocytes are derived from aged patients. FF-MAS may therefore be able to improve developmental competence after in vitro maturation and IVF.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The study was supported by EU (QCRT-2000-00058) and grants from Schering AG, D-13342 Berlin, Germany. Preliminary data on chiasma localization in aged and young oocytes were presented at the 10th Reinier de Graaf Symposium, Zeist, the Netherlands in 1999 (Eichenlaub-Ritter, 2000
). The technical assistance of Ilse Betzendahl and Manuela Grützner is gratefully acknowledged. We thank R. Eichenlaub for critical reading of the manuscript. |
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Andersen CY, Baltsen M, Byskov AG. (1999) Gonadotropin-induced resumption of oocyte meiosis and meiosis-activating sterols. Curr Top Dev Biol 41:163–185.