- Email: [email protected]
or synaptonemal complex fragmentation at pachytene
RBMOnline - Vol 13. No 1. 2006 88–95 Reproductive BioMedicine Online; www.rbmonline.com/Article/ 2333 on web 16 May 2006
Article Altered patterns of meiotic recombination in human fetal oocytes with asynapsis and/ or synaptonemal complex fragmentation at pachytene Born and educated in Sweden, Maj Hultén (MD PhD FRCPath) has between 1975 and 1997 been the Clinical Director of the UK West Midlands Regional Genetics Services and is now Professor of Medical Genetics at the University of Warwick. Her main research interest concerns reproductive genetics, in particular the study of patterns of meiotic recombination and also techniques for improved diagnosis of chromosome disorders. Most recently, Professor Hultén is the founder of the EU Network of Excellence SAFE, which is working towards the routine introduction of non-invasive prenatal diagnosis.
Dr Maj Hultén Charles Tease, Geraldine Hartshorne, Maj Hultén1 Department of Biological Sciences, University of Warwick, Coventry, CV4 7AL, UK 1 Correspondence: Tel: +44 (0)2476 528976; Fax: +44 (0)2476 523701; e-mail: [email protected]
Abstract Meiotic recombination was analysed in human fetal oocytes to determine whether recombination errors are associated with abnormal chromosome synapsis. Immunostaining was used to identify the synaptonemal complex (SC, the meiosis-specific proteinaceous structure that binds homologous chromosomes) and the DNA mismatch repair protein, MLH1, that locates recombination foci. It was found that 57.1–74.2% of zygotene oocytes showed fragmentation and/or defective chromosome synapsis. Fewer such abnormal cells occurred at pachytene (15.8–28.9%). MLH1 foci were present from zygotene to diplotene in both normal and abnormal oocytes. However, the proportions of oocytes having MLH1 foci, and mean numbers of foci per oocyte, were both lower in abnormal oocytes. Oocytes with fragmented SC had more foci than those with synaptic anomalies. Analysis of chromosomes 13, 18, 21 and X by fluorescence in-situ hybridization (FISH) did not implicate particular chromosomes in recombination deficiency. These observations indicate that recombination is disturbed in oocytes with SC fragmentation and/or synaptic abnormalities during meiotic prophase I. Such disturbances might be a risk factor for selection of fetal oocytes for atresia, as occurs for homologous chromosome pairing. Recombination errors may potentially increase the risk of abnormal chromosome segregation in oocytes that survive and contribute to the reserve in the mature ovary. Keywords: crossing over, meiosis, MLH1, oocyte, recombination, synaptonemal complex
Introduction
88
Oogenesis is a very prolonged process in human females. Germ cells enter meiosis in the fetal ovary but they do not complete meiosis at this time. Instead, before birth, they become arrested at the diplotene stage of prophase I (see Baker, 1963; Tilly, 2001). Only the few oocytes that survive and remain arrested in diplotene until adulthood, and are then supported by a growing dominant follicle, will eventually complete meiosis. The first division of female meiosis takes place shortly before ovulation, as the oocyte prepares for fertilization, and the second occurs during fertilization. Less
than 0.1% of the germ cells initially present in the fetal ovary are eventually ovulated. The remainder are lost by atresia during fetal development or in the mature ovary. There is evidence that atresia can be apoptotic although this is not the only mechanism. Atresia particularly affects oocytes in the fetal ovary during the late second and third trimesters of pregnancy when over 70% of germ cells are lost (De Pol et al., 1997; De Felici et al., 1999; Morita and Tilly, 1999). There has been considerable speculation regarding factors that could predispose fetal oocytes to becoming atretic. One possible risk factor is error in meiotic chromosome behaviour.
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. Meiosis is a complex process and in the female, crucial aspects occur only in utero, in particular homologous chromosome synapsis and recombination. In an electron microscope study of human fetal oocytes, Speed (1988) found that almost 30% had some form of pairing anomaly at the pachytene stage of meiosis. In addition, a further 15% showed breakage of the proteinaceous scaffold for synapsis, known as the synaptonemal complex (SC); these cells were presumed to be the equivalent of the degenerating pachytene (‘Z’) cells previously described by Beaumont and Mandl (1962). Speed (1988) proposed that abnormal homologous pairing was an important determinant of risk of atresia. One possible way in which meiotic chromosome behaviour and atresia might be linked is suggested by the observation in yeast that errors in chromosome pairing and recombination trigger cell-cycle checkpoints and cause arrest of cell progression (see Roeder and Bailis, 2000). Cells that arrest, and that are unable to repair the defects that induced arrest, have an increased risk of becoming apoptotic. Meiotic chromosome synapsis is intimately associated with recombination (reviewed by Zickler and Kleckner, 1999). It is therefore likely that errors in synapsis in fetal oocytes have an impact on recombination. The development of an immunofluorescent staining methodology has made possible the visualisation of recombination events along meiotic chromosomes in fetal oocytes. The immunostaining method for recombination is based on the detection of the DNA mismatch repair protein (MutL homologue) MLH1. MLH1 forms discrete foci along the SC of paired homologous chromosomes at prophase I of meiosis. The numbers and distributions of these foci indicate the positions of recombination activity (Baker et al., 1996; Barlow and Hultén, 1998; Marcon and Moens, 2003) and they can therefore be used to assess recombination in human oocytes directly, while at the same time assessing the fidelity of pairing using the SC (see Barlow and Hultén, 1998; Tease et al., 2002; Tease and Hultén, 2004; Lenzi et al., 2005). In a previous study this analytical approach was applied to oocytes of a small number of human fetuses concerning the rate of meiotic recombination in pachytene oocytes with normal chromosome synapsis (Tease et al., 2002). A considerable number of oocytes with abnormal appearance were also identified in the course of that work. The present report provides a description of these cells and examines specifically the question of whether, in human fetal oocytes, SC fragmentation and errors in synapsis may be associated with recombination deficiency.
Materials and methods Cell preparation The oocytes described here were from human ovaries obtained from four second trimester fetuses (approximately 16–19 weeks gestational age) following prostaglandin-induced termination of pregnancy. The use of the tissue in this project was approved by the Coventry Research Ethics Committee, and the East Birmingham Research and Ethics Committee. The patients undergoing termination consented to the research use of the fetuses, in accordance with the recommendations of the Polkinghorne Report (Polkinghorne, 1989). Ovaries were removed within a few hours post-mortem and placed in Leibovitz L-15 medium (Life Technology, Paisley, UK) containing 0.3% bovine serum albumin (Sigma, Poole, UK).
A full description of the cell preparation and protein detection protocols has been published (Hultén et al., 2001). In brief, a suspension of fetal oocytes in medium was made, mixed with hypotonic solution (either 0.3% lipsol or 3% sucrose) and allowed to sediment onto clean glass slides for approximately 30 min. The cells were fixed using 2% formaldehyde (TAAB Laboratories, Aldermaston, UK). Following fixation, the slides were washed with phosphate-buffered saline/0.1% Tween 20 (PBT). Two primary antibodies were applied simultaneously to the cells: rabbit polyclonal anti-SYCP3 (kindly supplied by Professor Christa Heyting, University of Wageningen, The Netherlands) and mouse monoclonal anti-MLH1 (Pharmingen, Oxford, UK). The anti-SYCP3 antibody binds to a protein component of the SC and to the axial elements (cores) of unpaired meiotic chromosomes (Lammers et al., 1994). This antibody was used at a dilution of 1:2500 as advised by the supplier. The anti-MLH1 antibody was used at a dilution of 1:500. Primary antibodies were diluted in PBT containing 0.15% bovine serum albumin. The slides were incubated overnight at room temperature, washed with PBT and binding of the primary antibodies was detected using an anti-rabbit antibody conjugated with Texas red (Vector Laboratories, Peterborough, UK, 1:200 dilution) and an anti-mouse antibody conjugated with fluorescein isothiocyanate (FITC; Sigma, 1:500 dilution). The slides were mounted in Vectashield/DAPI (Vector Laboratories).
Fluorescence in-situ hybridization (FISH) FISH was applied to slides prepared from one of the fetuses. Repeat sequence DNA probes for the α-satellite regions of the centromeric heterochromatin were used to identify specific chromosomes. Chromosomes 21 and 13 were identified using a DNA probe common to both (Appligene Oncor, Livingstone, UK), and the chromosomes distinguished on the basis of size and arm ratios. Chromosome 18 and the X were identified using differentially labelled probes (Vysis, Richmond, UK). All the probes used were directly labelled with fluorochromes. FISH was carried out according to the manufacturers’ protocols. After FISH, the SYCP3 staining was refreshed by application of the Texas red conjugated anti-rabbit antibody in the same manner as used for the initial immunostaining.
Analysis The cells were analysed using a Zeiss Axioskop fluorescence microscope equipped with a cooled CCD camera (Photometrix, Australia) and a Quips SmartCapture™ image acquisition and analysis system (Vysis/Digital Scientific). The meiotic stage of the oocytes was determined using the degree of chromosome synapsis (SC formation) as the principal criterion (see Speed, 1988; Barlow and Hultén, 1998; Hartshorne et al., 1999; Tease et al., 2002). Axial element (AE) formation is first detectable at the leptotene stage and synapsis of homologous AE to form the SC marks the onset of the zygotene stage. By definition, pachytene cells had completed synapsis, and in diplotene the homologues had commenced desynapsis. The normality or otherwise of chromosome structure and pairing was assessed using previously described criteria (Speed, 1988; Barlow and Hultén, 1998; Hartshorne et al., 1999). Oocytes
89
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. were deemed to be abnormal for a variety of reasons: breakage (fragmentation) of the axial elements or SC; partial or complete failure of synapsis along one or more chromosome pairs; or non-homologous pairing. Previous studies of early prophase I in human fetal oocytes have shown that homologous chromosome pairing (SC formation) is predominantly initiated in segments near to the telomeres and then extends interstitially along the chromosomal axes (Bojko, 1983; Wallace and Hultén, 1985) although interstitial initiation of synapsis may also take place. In the present study, cells that deviated from this expected pattern were considered to show anomalous synapsis. Using this methodology, it was possible to analyse numbers of normal and abnormal cells on each slide. One potential problem with the methodology is that normal and abnormal cells may have different responses to the hypotonic and spreading conditions used to prepare the cells. Any such difference could potentially affect the ability to detect altered patterns of MLH1 in abnormal cells. To date, however, there is no available information on whether abnormal cells respond differently to the hypotonic and spreading
treatments compared with normal cells. No evidence was found to support differential sensitivity, and cells with fragmented and normal SC were found in close proximity. Hence, differences in staining patterns between normal and abnormal oocytes were considered to be reliable. The numbers of MLH1 foci were counted in cells that were well spread and flattened (in the present instance, where they lay in no more than two focal planes). Captured images were transferred as PICT files to IPLab and counting of foci was carried out using this software. Only foci unambiguously co-located with SC were included. Faint signals were excluded. The data on proportions of anomalous oocytes in zygotene and pachytene (Tables 1 and 2), and in the pooled data on normal and abnormal oocytes shown were analysed by chi-squared tests. The differences in numbers of MLH1 foci in normal and abnormal oocytes at the pachytene stage (Table 3), and between asynaptic and fragmented oocytes (Table 4) were analysed using Student’s t-test. A P-value <0.05 was considered significant.
Table 1. The numbers of oocytes at different stages of meiotic prophase I with or without MutL homologue (MLH1) foci. Sample
Stage
[C4, C8, C14]
Zygotene Pachytene Diplotene
7 88 4
C15
Zygotene Pachytene Diplotene
32 169 51 351
Total
Normal synapsis With foci Without foci 1 8 1 9 7 14 40 (10.2)c
Abnormal With foci Without foci 10 16 1 31 68 11 137
% Abnormal
Total oocytes
13 2 1
74.2a 15.8a 28.6
31 114 7
25 4 4 49 (26.3)c
57.1b 28.9 b 18.8
97 248 80
Values in parentheses are percentages. Values with the same superscripts are significantly different. a,b,c P < 0.001.
Table 2. The numbers of oocytes in meiotic prophase I having synaptonemal complex (SC) fragmentation and/or synaptic anomalies.
90
Sample
Stage
Number of oocytes with: SC SC fragmentation fragmentation (%) and synaptic anomalies (%)
[C4, C8, C14]
Zygotene Pachytene Diplotene
18 (58.1)a 16 (14.0)a 0
C15
Zygotene Pachytene Diplotene
40 (41.2)b 31 (12.5)b 12 (15.0)
3 (9.7) 0 2 (28.6) 4 (4.1) 20 (8.1) 0
Synaptic anomalies only (%)
Total no. oocytes
2 (6.5) 2 (1.8) 0
31 114 7
12 (12.4) 21 (8.5) 3 (3.8)
97 248 80
Values in parentheses are percentages. Values with the same superscripts are significantly different. a,b P < 0.001.
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. Table 3. The numbers of recombination foci in normal and abnormal oocytes at different stages of prophase I from fetus C15. Stage
Normal Mean no. of foci
SD
Range
No. of cells
Abnormal Mean no. of foci
SD
Zygotene Pachytene Early diplotene
51.4 70.3a,b 58.3b
26.0 10.5 13.3
2–90 48–102 35–89
17 95 24
30.2 55.1a 55.8
25.4 1–69 14.3 27–94 23.9 26–78
Range
No. of cells 13 33 4
Values with this superscript are significantly different: P < 0.001. Reported by Tease et al. (2002).
a
b
Table 4. The numbers of recombination foci in normal pachytene oocytes and those classified as abnormal because of asynapsis or synaptonemal complex fragmentation. Pachytene category
Normal Mean number of foci
Normal Asynaptic Fragmented
70.3 48.9a 59.6a
SD
Range of cells
Number
10.5 8.4 16.2
48–102 34–65 27–94
95b 14 19
Values with this superscript are significantly different: P < 0.02. Reported by Tease et al. (2002).
a
b
Results
Timing of MLH1 appearance
Four fetuses were examined and it was found that a substantial proportion of oocytes were anomalous in each (Table 1). Three of the four fetuses (C4, C8, C14) yielded relatively small numbers of analysable oocytes (48, 59 and 45 respectively); the fourth (C15) gave 425 cells. For statistical analyses, the samples from C4, C8 and C14 were combined, and tests were performed on C15 separately. The proportion of anomalous oocytes differed significantly between the zygotene and pachytene stages in both the combined group (χ21 = 29.37, P < 0.001) and C15 (χ21 = 15.67, P < 0.001). Thus there was a much larger rate of SC abnormality at zygotene (57.1–74.2%) than at the pachytene stage (15.8–28.9%). As shown by the selective analysis of C15, there was no difference in abnormality rates between the pachytene and diplotene stages (χ21 = 2.37, not significant).
MLH1 foci were first seen in some early zygotene oocytes, both normal and abnormal. In those oocytes in which they were present at this early stage, the foci did not appear on all stretches of SC. There was no apparent relationship between the extent of SC synapsis in early zygotene and the likelihood of a focus. Thus, some very short SC were seen with a focus in company with longer stretches that did not carry a focus. As expected, a higher proportion of pachytene than zygotene oocytes contained MLH1 foci (Table 1). Abnormal oocytes more often lacked MLH1 foci than normal oocytes. However, this difference was statistically significant only when the combined populations of zygotene to diplotene stage oocytes were compared in terms of whether their SC were normal or abnormal (χ23 = 16.77, P < 0.001 for the combined group; χ23 = 11.84, P < 0.01 for C15). Figures 1a and 1b show MLH1 foci in normally synapsed and partially fragmented oocytes, respectively.
Anomalous oocytes were further subdivided into those that contained SC fragmentation, synaptic anomalies, or both (Table 2). Significantly more oocytes with SC fragmentation were evident at the zygotene stage than at the pachytene stage for the combined group (χ21 = 11.43, P < 0.001) as well as for C15 (χ21 = 15.91, P < 0.001). In contrast, the proportion of oocytes with synaptic anomalies (Table 2) did not differ significantly between the zygotene and pachytene stages (χ21 = 2.16, not significant) in the C15 sample analysed in this respect.
Numbers of MLH1 foci Counting MLH1 foci was feasible only in fetus C15 where this was done for a small number of oocytes at the zygotene and diplotene stages and a larger number at pachytene (Table 3). In cells with normal SC, the average number of MLH1 foci per oocyte increased from the zygotene to pachytene stages and reduced at early diplotene (Tease et al., 2002). Oocytes
91
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al.
Figure 1. Human fetal oocytes in meiotic prophase I. Synaptonemal complexes (SC) are highlighted in red by immunostaining for SYCP3. In (a), (b) and (c), MutL homologue (MLH1) foci are highlighted in yellow. (d) shows the same oocyte as in (c), after fluorescence in-situ hybridization (FISH) with probes for chromosomes 13, 18, 21 and X. (a) Pachytene oocyte with full synapsis; 72 MLH1 foci are clearly well separated along the SC. (b) Abnormal pachytene oocyte with fragmented SC, demonstrating a reduced number of MLH1 foci (48). (c) Pachytene oocyte with complete failure of synapsis of some chromosomes. Unpaired SC axes have a thinner appearance and are longer than those that are synapsed, as exemplified by the unpaired X chromosomes. The 54 MLH1 foci are present on synapsed chromosomes only. (d) The same oocyte as in (c) after FISH with DNA probes for chromosomes 21, 18, 13 and X, showing unpaired SC axes for chromosome X, where the centromere X is highlighted in green.
92
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. with SC anomalies showed a similar but not identical pattern: the numbers of MLH1 foci remained fairly constant from zygotene to diplotene. Overall, such abnormal oocytes tended to have fewer MLH1 foci than normal ones at the same developmental stage (Table 2). This was clearest at the pachytene stage where an average of 55.1 MLH1 foci per cell was found in cells with SC abnormalities compared with 70.3 in cells with normal SC (P P < 0.001, t-test Table 3), however, the ranges were wide and overlapping. More than half of the 33 pachytene oocytes with abnormal SC (19/33) had fragmentation of the SC (and in some instances also asynapsis), the remaining 14/33 showed asynapsis of chromosome segments or whole chromosomes (e.g. Figure 1c). Those with synaptic abnormalities alone had a slightly higher mean number of MLH1 foci per oocyte (59.6) than those with fragmentation (48.9) as shown in Table 4. This difference was statistically significant (P P = 0.019), despite overlap in the ranges. Examination of some of these cells after FISH to identify chromosomes 21, 18, 13 and X showed that chromosome-pairing errors were not confined to a particular type of chromosome, e.g. the smallest members of the complement, but apparently could affect any sized chromosome. For example, Figure 1d shows asynapsis of chromosome X with the homologues widely separated. No MLH1 foci were observed on asynapsed chromosome segments or chromosomes at pachytene.
Distributions of MLH1 foci along chromosomes at pachytene Adjacent MLH1 foci were almost invariably separated by considerable distances along chromosomal axes (eg, Figure 1a) from zygotene to diplotene stages in oocytes with normal appearance (Tease et al., 2002). However, MLH1 foci positioned close together were occasionally noted in oocytes with both normal and abnormal SC. It is clearly not possible to comment definitively on distributions of MLH1 foci in oocytes with extensive SC fragmentation (see Figure 1b). Subjectively, fragmented SC did not appear to have foci closer together than in normal oocytes. In oocytes classified as anomalous because of synaptic errors, the MLH1 foci appeared to show a similar pattern to that of cells with normal SC (compare Figure 1c with 1a). Exact numbers and positions of foci were determined after FISH in a small number of oocytes having synaptic anomalies (eg, Figure 1d). Chromosome 21 had 1.17 ± 0.4 foci (6 bivalents), chromosome 18 had 2.67 ± 1.2 (n = 3), and chromosome 13 had 2.5 ± 0.6 (n = 4). There was no obvious difference in mean frequency of MLH1 foci in comparison to oocytes with normal SC. Thus the corresponding mean for chromosome 21 was 1.23 ± 0.5 (n = 86), chromosome 18 is 2.36 ± 0.7 (n = 67) and chromosome 13 is 2.5 ± 0.6 (n = 64) in normal pachytene oocytes (Tease et al., 2002). The distribution patterns of foci along these chromosomes in abnormal oocytes also fell within the range found in normal oocytes (data not shown).
Discussion A substantial proportion of oocytes identified in this series of fetuses showed anomalous SC at prophase I. This observation agrees with previous reports that have used SC formation to assess human fetal oocyte development (Bojko, 1983; Speed,
1988; Hartshorne et al., 1999). These observations were made on cells that were prepared for microscopy using a method that includes a hypo-osmotic swelling step, which may require membrane integrity of the oocytes. Only those cells that were well spread and sufficiently flattened for their chromosomes to be observed in two focal planes were included in the analyses. This selection criterion may have excluded oocytes that had already progressed substantially towards atresia or post mortem degeneration as the membrane function of these cells was presumably reduced. The population of oocytes studied here may therefore be considered to be mainly viable. The data reported here indicate a significant decrease in the proportion of oocytes with SC fragmentation between the zygotene and pachytene stages, which has not been reported previously in humans. This observation could be explained if such abnormal cells arrested in development before the pachytene stage and/or were eliminated preferentially. On the other hand, the range of anomalous configurations (both fragmentation and synaptic irregularities) seen at the pachytene stage was similar to that at the zygotene stage, which may be taken to indicate that many zygotene cells with fragmentation are indeed able to continue development into the pachytene stage. However, since the methodology used here precludes following individual oocytes through progression of the meiotic prophase, the alternative possibility cannot be excluded, i.e. that all anomalous zygotene cells are eliminated and that the abnormal oocytes present at pachytene result from de novo events. The initial study of MLH1 at meiosis in male mice showed that the protein was present only on fully paired chromosomes in cells classified as pachytene (Baker et al., 1996). In contrast, the work described here in human oocytes has demonstrated that MLH1 foci appear in early zygotene at the time when SC formation is occurring (Tease et al., 2002), as also observed by Lenzi et al., (2005). The present study suggests that the same is true in anomalous zygotene oocytes. Although the proportion of zygotene oocytes with MLH1 foci was lower in the group with SC abnormalities, this result was not significant. One of the main findings of this study was that pachytene oocytes with fragmentation and/or synaptic abnormalities had significantly fewer MLH1 foci on average than their normal counterparts. Thus it is proposed that consequent selection may take place against oocytes with low recombination frequency in association with SC fragmentation and/or synaptic irregularity. This hypothesis would concur with that of Kong et al. (2004), based upon studies of families, that a high recombination count increases the chance of a gamete becoming a live birth in older women. Their study also indicated that a high recombination rate is related to reproductive success, and that biological selection of gametes may therefore be occurring on the basis of recombination. Meiotic recombination is absolutely dependent on homologous chromosome synapsis in mammalian germ cells. Failure of pairing inevitably causes failure of recombination in the chromosome (or chromosomal segment) affected. In sample C15, 41 pachytene oocytes were identified with asynapsis of one or more chromosome pairs (Table 2; Figure 1d). As expected, the affected bivalents did not show MLH1 foci. It is also of interest that unsynapsed chromosome segments often
93
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. show a striking decondensation of the SC and appear much longer than when normally synapsed (Saadallah and Hultén, 1986; Barlow et al., 2002). This may be one of the factors that trigger selection against oocytes with synaptic abnormalities. It does not automatically follow, even where synapsis is achieved, that crossing over will ensue. It was also found that 7/176 (4.0%) pachytene oocytes in the C15 fetus and 8/96 (8.3%) in the combined group did not have MLH1 foci despite the presence of apparently normal and complete synapsis (Table 1). Thus even among cells with a normal appearance judged by complete homologous synapsis, there might be other errors of the meiotic process present. Oocytes with failure of recombination, whether affecting a limited number of chromosome pairs or the whole genome, would inevitably have considerable difficulty achieving correct chromosome segregation later in meiosis and may contribute to the high rate of aneuploidy in human female gametes (Lamb and Hassold, 2004). Other evidence from our laboratory shows that recombination may be disrupted by suboptimal environmental conditions (Lyrakou et al., 2002), suggesting that it may potentially be sensitive to local perturbations within the ovary, while Lenzi et al. (2005) have demonstrated significant differences among fetuses, despite studying relatively few oocytes per fetus. In male mice lacking MLH1, meiotic checkpoints eliminate pachytene cells lacking recombination before they progress through metaphase (Eaker et al., 2002), however, this is not the case in female mice, where oocytes can progress through an abnormal metaphase associated with unstable spindles and chaotic chromosome attachment and segregation (Woods et al., 1999). In humans, many aneuploid conditions in conceptuses may derive from susceptibilities acquired during female meiotic prophase, e.g. potentially arising from defects in recombination at prophase I (before birth) eventually causing malsegregation of chromosomes at oocyte maturation and fertilization (after puberty). The extensive attrition of oocytes before birth, relatively low fertility per ovulation of humans, high rates of infertility, miscarriage and genetic abnormality in offspring, are possible consequences of the unique features of human female meiosis. In contrast to males, human female meiosis has distinguishing features such as reduced checkpoint efficiency and a prolonged bouquet phase, possibly associated with compromised DNA repair mechanisms (Roig et al., 2004). The findings described here, combined with the permissive effect upon checkpoint systems potentially brought about by adverse conditions of development (Eichenlaub-Ritter et al., 2002) might predispose to subsequent problems with spindle alignment and function, with serious consequences for gamete formation.
94
Speed (1988) suggested that pachytene oocytes with broken SC were the equivalent of the ‘Z’ cells first described by Beaumont and Mandl (1962). These are usually presumed to be on an atretic pathway. The presence of some oocytes with fragmented axial elements and SC from early zygotene suggests that atresia can be present from the initiation of synapsis or earlier. It would therefore appear that errors in or failure of synapsis might not be the initiating event for atresia. MLH1 foci were present in some ‘pachytene’ cells with extensive SC fragmentation (Figure 1b). The numbers and distributions of MLH1 foci in these cells gave the impression of essentially normal MLH1 focus formation, i.e. there was no indication of excessive numbers or that the
MLH1 foci were more closely spaced than in normal pachytene cells. One explanation for this pattern of MLH1 foci is that these cells had undergone normal progression through the early part of prophase I and that the breakage of the SC was a late event. This still leaves the question of what causes cells with apparently normal chromosome pairing and recombination to undergo SC fragmentation. The application of additional molecular markers indicative of cellular control mechanisms including apoptosis may be informative. While only relatively small samples of cells could be assessed, subjective observations of MLH1 foci along SC of anomalous oocytes indicated similar patterns of distribution to those found in normal cells. Additional data were obtained when chromosomes 13, 18, 21 and X were identified using FISH. These chromosomes were selected, as they are among the smallest in humans and aneuploidy for these causes substantial numbers of intrauterine and perinatal losses, as well as abnormalities in live born children. Preliminary data from oocytes in which these chromosomes were identified offered no evidence that the presence of chromosome pairing anomalies within a cell disrupted the distribution pattern of MLH1 foci on chromosome pairs that had successfully synapsed. This preliminary conclusion requires verification in a larger sample of oocytes. However, constraints on the availability of material for research such as this may limit this possibility. In summary, this paper presents evidence for significantly compromised recombination in human pachytene oocytes showing features of SC fragmentation and deficient synapsis. Such abnormal meiosis may potentially predispose to atresia or to abnormal development during later oocyte maturation, with consequences for the ovarian reserve and children originating from these oocytes. On the other hand, the fate of the oocytes with normal pachytene SC but still showing reduced recombination, is not currently known.
Acknowledgements We wish to thank the staff and patients of the Calthorpe Clinic, Birmingham for the provision of tissue. This study was supported by WellBeing grant H1/98 and Wellcome Trust grant 061202/ZOOZ.
References Baker TG 1963 A quantitative and cytological study of germ cells in human ovaries. Proceedings of the Royal Society [Biology] 158, 417–433. Baker SM, Plug AW, Prolla TA et al. 1996 Involvement of mouse Mlh1 in DNA mismatch repair and meiotic crossing over. Nature Genetics 13, 336–342. Barlow AL, Hultén MA 1998 Crossing over analysis at pachytene in man. European Journal of Human Genetics 6, 330–358. Barlow AL, Tease C and Hultén MA 2002 Meiotic chromosome pairing in fetal oocytes of trisomy 21 human females. Cytogenetics and Genome Research 96, 45–51. Beaumont HM, Mandl AM 1962 A quantitative and cytological study of oogonia and oocytes in the fetal and neonatal rat. Proceedings of the Royal Society [Biology] 155, 557–579. Bojko M 1983 Human meiosis. VIII. Chromosome pairing and formation of the synaptonemal complex in oocytes. Carlsberg Research Communications 48, 457–483. De Felici M, Di Carlo A, Pesce M et al. 1999 Bcl-2 and bax regulation
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. of apoptosis in germ cells during prenatal oogenesis in the mouse embryo. Cell Death and Differentiation 6, 908–915. De Pol A, Vaccina F, Forabosco A et al. 1997 Apoptosis of germ cells during human prenatal oogenesis. Human Reproduction 10, 2235–2241. Eaker S, Cobb J, Pyle A, Handel MA 2002 Meiotic prophase abnormalities and metaphase cell death in MLH1 deficient mouse spermatocytes: insights into regulation of spermatogenic progress. Developmental Biology 249, 85–95. Eichenlaub-Ritter U, Shen Y, Tinneberg H-R 2002 Manipulation of the oocyte: possible damage to the spindle apparatus. Reproductive BioMedicine Online 5, 117–124. Hartshorne GM, Barlow AL, Child TJ et al. 1999 Immunocytogenetic detection of normal and abnormal oocytes in human fetal ovarian tissue in culture. Human Reproduction 14, 172–182. Hultén MA, Barlow AL, Tease C 2001 Meiotic studies in humans. In: Rooney DE (ed.) Human Cytogenetics: Constitutional Analysis (3rd edn): A Practical Approach. Oxford University Press, Oxford and New York. pp. 211–236. Kong A, Barnard J, Gudbjartsson DF et al. 2004 Recombination rate and reproductive success in humans. Nature Genetics 36, 1203–1206. Lamb NE, Hassold TJ 2004 Nondisjunction – a view from ringside. New England Journal of Medicine 351, 1931–1934. Lammers JHM, Offenberg HH, Van Aalderen M et al. 1994 The gene encoding a major component of synaptonemal complexes of the rat is related to X-linked lymphocyte-regulated genes. Molecular Cell Biology 14, 1137–1146. Lenzi ML, Smith J, Snowden T et al. 2005 Extreme heterogeneity in the molecular events leading to the establishment of chiasmata during meiosis 1 in human oocytes. American Journal of Human Genetics 76, 112–127. Lyrakou S, Hultén MA, Hartshorne GM 2002 Growth factors promote meiosis in vitro in mouse fetal ovaries. Molecular Human Reproduction 8, 906–911. Marcon E, Moens P 2003 MLH1p and MLH3p localize to precociously induced chiasmata of okadaic-acid-treated mouse spermatocytes. Genetics 165, 2283–2287.
Morita Y, Tilly JL 1999 Oocyte apoptosis: like sand through an hourglass. Developmental Biology 213, 1–17. Polkinghorne J 1989 Review of the Guidance of the Research Use of Fetuses and Fetal Material. Cm762, Her Majesty’s Stationery Office, London. Roeder GS, Bailis JM 2000 The pachytene checkpoint. Trends in Genetics 16, 395–403. Roig I, Liebe B, Egozcue J et al. 2004 Female-specific features of recombinational double-stranded DNA repair in relation to synapsis and telomere dynamics in human oocytes. Chromosoma 113, 22–33. Saadallah N, Hultén M 1986 EM investigations of surface spread synaptonemal complexes in a human male carrier of a pericentric inversion inv(13)(p12q14): the role of heterosynapsis for spermatocyte survival. Annals of Human Genetics 50, 369–383. Speed RM 1988 The possible role of meiotic pairing anomalies in the atresia of human fetal oocytes. Human Genetics 78, 260–266. Tease C, Hultén MA 2004 Inter-sex variation in synaptonemal complex lengths largely determine the different recombination rates in male and female germ cells. Cytogenetics and Genome Research 107, 208–215. Tease C, Hartshorne GM, Hultén MA 2002 Patterns of meiotic recombination in human fetal oocytes. American Journal of Human Genetics 70, 1469–1479. Tilly JL 2001 Commuting the death sentence: how oocytes strive to survive. Nature Reviews of Molecular Cell Biology 2, 838–848. Wallace BM, Hultén MA 1985 Meiotic chromosome pairing in the normal human female. Annals of Human Genetics 49, 215–226. Woods LM, Hodges CA, Baart E et al. 1999 Chromosomal influence on meiotic spindle assembly: abnormal meiosis I in female Mlh1 mutant mice. Journal of Cell Biology 145, 1395–1406. Zickler D, Kleckner N 1999 Meiotic chromosomes: integrating structure and function. Annual Review of Genetics 33, 603–754. Received 10 March 2006; refereed 21 March 2006; accepted 21 April 2006.
95
Article Altered patterns of meiotic recombination in human fetal oocytes with asynapsis and/ or synaptonemal complex fragmentation at pachytene Born and educated in Sweden, Maj Hultén (MD PhD FRCPath) has between 1975 and 1997 been the Clinical Director of the UK West Midlands Regional Genetics Services and is now Professor of Medical Genetics at the University of Warwick. Her main research interest concerns reproductive genetics, in particular the study of patterns of meiotic recombination and also techniques for improved diagnosis of chromosome disorders. Most recently, Professor Hultén is the founder of the EU Network of Excellence SAFE, which is working towards the routine introduction of non-invasive prenatal diagnosis.
Dr Maj Hultén Charles Tease, Geraldine Hartshorne, Maj Hultén1 Department of Biological Sciences, University of Warwick, Coventry, CV4 7AL, UK 1 Correspondence: Tel: +44 (0)2476 528976; Fax: +44 (0)2476 523701; e-mail: [email protected]
Abstract Meiotic recombination was analysed in human fetal oocytes to determine whether recombination errors are associated with abnormal chromosome synapsis. Immunostaining was used to identify the synaptonemal complex (SC, the meiosis-specific proteinaceous structure that binds homologous chromosomes) and the DNA mismatch repair protein, MLH1, that locates recombination foci. It was found that 57.1–74.2% of zygotene oocytes showed fragmentation and/or defective chromosome synapsis. Fewer such abnormal cells occurred at pachytene (15.8–28.9%). MLH1 foci were present from zygotene to diplotene in both normal and abnormal oocytes. However, the proportions of oocytes having MLH1 foci, and mean numbers of foci per oocyte, were both lower in abnormal oocytes. Oocytes with fragmented SC had more foci than those with synaptic anomalies. Analysis of chromosomes 13, 18, 21 and X by fluorescence in-situ hybridization (FISH) did not implicate particular chromosomes in recombination deficiency. These observations indicate that recombination is disturbed in oocytes with SC fragmentation and/or synaptic abnormalities during meiotic prophase I. Such disturbances might be a risk factor for selection of fetal oocytes for atresia, as occurs for homologous chromosome pairing. Recombination errors may potentially increase the risk of abnormal chromosome segregation in oocytes that survive and contribute to the reserve in the mature ovary. Keywords: crossing over, meiosis, MLH1, oocyte, recombination, synaptonemal complex
Introduction
88
Oogenesis is a very prolonged process in human females. Germ cells enter meiosis in the fetal ovary but they do not complete meiosis at this time. Instead, before birth, they become arrested at the diplotene stage of prophase I (see Baker, 1963; Tilly, 2001). Only the few oocytes that survive and remain arrested in diplotene until adulthood, and are then supported by a growing dominant follicle, will eventually complete meiosis. The first division of female meiosis takes place shortly before ovulation, as the oocyte prepares for fertilization, and the second occurs during fertilization. Less
than 0.1% of the germ cells initially present in the fetal ovary are eventually ovulated. The remainder are lost by atresia during fetal development or in the mature ovary. There is evidence that atresia can be apoptotic although this is not the only mechanism. Atresia particularly affects oocytes in the fetal ovary during the late second and third trimesters of pregnancy when over 70% of germ cells are lost (De Pol et al., 1997; De Felici et al., 1999; Morita and Tilly, 1999). There has been considerable speculation regarding factors that could predispose fetal oocytes to becoming atretic. One possible risk factor is error in meiotic chromosome behaviour.
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. Meiosis is a complex process and in the female, crucial aspects occur only in utero, in particular homologous chromosome synapsis and recombination. In an electron microscope study of human fetal oocytes, Speed (1988) found that almost 30% had some form of pairing anomaly at the pachytene stage of meiosis. In addition, a further 15% showed breakage of the proteinaceous scaffold for synapsis, known as the synaptonemal complex (SC); these cells were presumed to be the equivalent of the degenerating pachytene (‘Z’) cells previously described by Beaumont and Mandl (1962). Speed (1988) proposed that abnormal homologous pairing was an important determinant of risk of atresia. One possible way in which meiotic chromosome behaviour and atresia might be linked is suggested by the observation in yeast that errors in chromosome pairing and recombination trigger cell-cycle checkpoints and cause arrest of cell progression (see Roeder and Bailis, 2000). Cells that arrest, and that are unable to repair the defects that induced arrest, have an increased risk of becoming apoptotic. Meiotic chromosome synapsis is intimately associated with recombination (reviewed by Zickler and Kleckner, 1999). It is therefore likely that errors in synapsis in fetal oocytes have an impact on recombination. The development of an immunofluorescent staining methodology has made possible the visualisation of recombination events along meiotic chromosomes in fetal oocytes. The immunostaining method for recombination is based on the detection of the DNA mismatch repair protein (MutL homologue) MLH1. MLH1 forms discrete foci along the SC of paired homologous chromosomes at prophase I of meiosis. The numbers and distributions of these foci indicate the positions of recombination activity (Baker et al., 1996; Barlow and Hultén, 1998; Marcon and Moens, 2003) and they can therefore be used to assess recombination in human oocytes directly, while at the same time assessing the fidelity of pairing using the SC (see Barlow and Hultén, 1998; Tease et al., 2002; Tease and Hultén, 2004; Lenzi et al., 2005). In a previous study this analytical approach was applied to oocytes of a small number of human fetuses concerning the rate of meiotic recombination in pachytene oocytes with normal chromosome synapsis (Tease et al., 2002). A considerable number of oocytes with abnormal appearance were also identified in the course of that work. The present report provides a description of these cells and examines specifically the question of whether, in human fetal oocytes, SC fragmentation and errors in synapsis may be associated with recombination deficiency.
Materials and methods Cell preparation The oocytes described here were from human ovaries obtained from four second trimester fetuses (approximately 16–19 weeks gestational age) following prostaglandin-induced termination of pregnancy. The use of the tissue in this project was approved by the Coventry Research Ethics Committee, and the East Birmingham Research and Ethics Committee. The patients undergoing termination consented to the research use of the fetuses, in accordance with the recommendations of the Polkinghorne Report (Polkinghorne, 1989). Ovaries were removed within a few hours post-mortem and placed in Leibovitz L-15 medium (Life Technology, Paisley, UK) containing 0.3% bovine serum albumin (Sigma, Poole, UK).
A full description of the cell preparation and protein detection protocols has been published (Hultén et al., 2001). In brief, a suspension of fetal oocytes in medium was made, mixed with hypotonic solution (either 0.3% lipsol or 3% sucrose) and allowed to sediment onto clean glass slides for approximately 30 min. The cells were fixed using 2% formaldehyde (TAAB Laboratories, Aldermaston, UK). Following fixation, the slides were washed with phosphate-buffered saline/0.1% Tween 20 (PBT). Two primary antibodies were applied simultaneously to the cells: rabbit polyclonal anti-SYCP3 (kindly supplied by Professor Christa Heyting, University of Wageningen, The Netherlands) and mouse monoclonal anti-MLH1 (Pharmingen, Oxford, UK). The anti-SYCP3 antibody binds to a protein component of the SC and to the axial elements (cores) of unpaired meiotic chromosomes (Lammers et al., 1994). This antibody was used at a dilution of 1:2500 as advised by the supplier. The anti-MLH1 antibody was used at a dilution of 1:500. Primary antibodies were diluted in PBT containing 0.15% bovine serum albumin. The slides were incubated overnight at room temperature, washed with PBT and binding of the primary antibodies was detected using an anti-rabbit antibody conjugated with Texas red (Vector Laboratories, Peterborough, UK, 1:200 dilution) and an anti-mouse antibody conjugated with fluorescein isothiocyanate (FITC; Sigma, 1:500 dilution). The slides were mounted in Vectashield/DAPI (Vector Laboratories).
Fluorescence in-situ hybridization (FISH) FISH was applied to slides prepared from one of the fetuses. Repeat sequence DNA probes for the α-satellite regions of the centromeric heterochromatin were used to identify specific chromosomes. Chromosomes 21 and 13 were identified using a DNA probe common to both (Appligene Oncor, Livingstone, UK), and the chromosomes distinguished on the basis of size and arm ratios. Chromosome 18 and the X were identified using differentially labelled probes (Vysis, Richmond, UK). All the probes used were directly labelled with fluorochromes. FISH was carried out according to the manufacturers’ protocols. After FISH, the SYCP3 staining was refreshed by application of the Texas red conjugated anti-rabbit antibody in the same manner as used for the initial immunostaining.
Analysis The cells were analysed using a Zeiss Axioskop fluorescence microscope equipped with a cooled CCD camera (Photometrix, Australia) and a Quips SmartCapture™ image acquisition and analysis system (Vysis/Digital Scientific). The meiotic stage of the oocytes was determined using the degree of chromosome synapsis (SC formation) as the principal criterion (see Speed, 1988; Barlow and Hultén, 1998; Hartshorne et al., 1999; Tease et al., 2002). Axial element (AE) formation is first detectable at the leptotene stage and synapsis of homologous AE to form the SC marks the onset of the zygotene stage. By definition, pachytene cells had completed synapsis, and in diplotene the homologues had commenced desynapsis. The normality or otherwise of chromosome structure and pairing was assessed using previously described criteria (Speed, 1988; Barlow and Hultén, 1998; Hartshorne et al., 1999). Oocytes
89
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. were deemed to be abnormal for a variety of reasons: breakage (fragmentation) of the axial elements or SC; partial or complete failure of synapsis along one or more chromosome pairs; or non-homologous pairing. Previous studies of early prophase I in human fetal oocytes have shown that homologous chromosome pairing (SC formation) is predominantly initiated in segments near to the telomeres and then extends interstitially along the chromosomal axes (Bojko, 1983; Wallace and Hultén, 1985) although interstitial initiation of synapsis may also take place. In the present study, cells that deviated from this expected pattern were considered to show anomalous synapsis. Using this methodology, it was possible to analyse numbers of normal and abnormal cells on each slide. One potential problem with the methodology is that normal and abnormal cells may have different responses to the hypotonic and spreading conditions used to prepare the cells. Any such difference could potentially affect the ability to detect altered patterns of MLH1 in abnormal cells. To date, however, there is no available information on whether abnormal cells respond differently to the hypotonic and spreading
treatments compared with normal cells. No evidence was found to support differential sensitivity, and cells with fragmented and normal SC were found in close proximity. Hence, differences in staining patterns between normal and abnormal oocytes were considered to be reliable. The numbers of MLH1 foci were counted in cells that were well spread and flattened (in the present instance, where they lay in no more than two focal planes). Captured images were transferred as PICT files to IPLab and counting of foci was carried out using this software. Only foci unambiguously co-located with SC were included. Faint signals were excluded. The data on proportions of anomalous oocytes in zygotene and pachytene (Tables 1 and 2), and in the pooled data on normal and abnormal oocytes shown were analysed by chi-squared tests. The differences in numbers of MLH1 foci in normal and abnormal oocytes at the pachytene stage (Table 3), and between asynaptic and fragmented oocytes (Table 4) were analysed using Student’s t-test. A P-value <0.05 was considered significant.
Table 1. The numbers of oocytes at different stages of meiotic prophase I with or without MutL homologue (MLH1) foci. Sample
Stage
[C4, C8, C14]
Zygotene Pachytene Diplotene
7 88 4
C15
Zygotene Pachytene Diplotene
32 169 51 351
Total
Normal synapsis With foci Without foci 1 8 1 9 7 14 40 (10.2)c
Abnormal With foci Without foci 10 16 1 31 68 11 137
% Abnormal
Total oocytes
13 2 1
74.2a 15.8a 28.6
31 114 7
25 4 4 49 (26.3)c
57.1b 28.9 b 18.8
97 248 80
Values in parentheses are percentages. Values with the same superscripts are significantly different. a,b,c P < 0.001.
Table 2. The numbers of oocytes in meiotic prophase I having synaptonemal complex (SC) fragmentation and/or synaptic anomalies.
90
Sample
Stage
Number of oocytes with: SC SC fragmentation fragmentation (%) and synaptic anomalies (%)
[C4, C8, C14]
Zygotene Pachytene Diplotene
18 (58.1)a 16 (14.0)a 0
C15
Zygotene Pachytene Diplotene
40 (41.2)b 31 (12.5)b 12 (15.0)
3 (9.7) 0 2 (28.6) 4 (4.1) 20 (8.1) 0
Synaptic anomalies only (%)
Total no. oocytes
2 (6.5) 2 (1.8) 0
31 114 7
12 (12.4) 21 (8.5) 3 (3.8)
97 248 80
Values in parentheses are percentages. Values with the same superscripts are significantly different. a,b P < 0.001.
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. Table 3. The numbers of recombination foci in normal and abnormal oocytes at different stages of prophase I from fetus C15. Stage
Normal Mean no. of foci
SD
Range
No. of cells
Abnormal Mean no. of foci
SD
Zygotene Pachytene Early diplotene
51.4 70.3a,b 58.3b
26.0 10.5 13.3
2–90 48–102 35–89
17 95 24
30.2 55.1a 55.8
25.4 1–69 14.3 27–94 23.9 26–78
Range
No. of cells 13 33 4
Values with this superscript are significantly different: P < 0.001. Reported by Tease et al. (2002).
a
b
Table 4. The numbers of recombination foci in normal pachytene oocytes and those classified as abnormal because of asynapsis or synaptonemal complex fragmentation. Pachytene category
Normal Mean number of foci
Normal Asynaptic Fragmented
70.3 48.9a 59.6a
SD
Range of cells
Number
10.5 8.4 16.2
48–102 34–65 27–94
95b 14 19
Values with this superscript are significantly different: P < 0.02. Reported by Tease et al. (2002).
a
b
Results
Timing of MLH1 appearance
Four fetuses were examined and it was found that a substantial proportion of oocytes were anomalous in each (Table 1). Three of the four fetuses (C4, C8, C14) yielded relatively small numbers of analysable oocytes (48, 59 and 45 respectively); the fourth (C15) gave 425 cells. For statistical analyses, the samples from C4, C8 and C14 were combined, and tests were performed on C15 separately. The proportion of anomalous oocytes differed significantly between the zygotene and pachytene stages in both the combined group (χ21 = 29.37, P < 0.001) and C15 (χ21 = 15.67, P < 0.001). Thus there was a much larger rate of SC abnormality at zygotene (57.1–74.2%) than at the pachytene stage (15.8–28.9%). As shown by the selective analysis of C15, there was no difference in abnormality rates between the pachytene and diplotene stages (χ21 = 2.37, not significant).
MLH1 foci were first seen in some early zygotene oocytes, both normal and abnormal. In those oocytes in which they were present at this early stage, the foci did not appear on all stretches of SC. There was no apparent relationship between the extent of SC synapsis in early zygotene and the likelihood of a focus. Thus, some very short SC were seen with a focus in company with longer stretches that did not carry a focus. As expected, a higher proportion of pachytene than zygotene oocytes contained MLH1 foci (Table 1). Abnormal oocytes more often lacked MLH1 foci than normal oocytes. However, this difference was statistically significant only when the combined populations of zygotene to diplotene stage oocytes were compared in terms of whether their SC were normal or abnormal (χ23 = 16.77, P < 0.001 for the combined group; χ23 = 11.84, P < 0.01 for C15). Figures 1a and 1b show MLH1 foci in normally synapsed and partially fragmented oocytes, respectively.
Anomalous oocytes were further subdivided into those that contained SC fragmentation, synaptic anomalies, or both (Table 2). Significantly more oocytes with SC fragmentation were evident at the zygotene stage than at the pachytene stage for the combined group (χ21 = 11.43, P < 0.001) as well as for C15 (χ21 = 15.91, P < 0.001). In contrast, the proportion of oocytes with synaptic anomalies (Table 2) did not differ significantly between the zygotene and pachytene stages (χ21 = 2.16, not significant) in the C15 sample analysed in this respect.
Numbers of MLH1 foci Counting MLH1 foci was feasible only in fetus C15 where this was done for a small number of oocytes at the zygotene and diplotene stages and a larger number at pachytene (Table 3). In cells with normal SC, the average number of MLH1 foci per oocyte increased from the zygotene to pachytene stages and reduced at early diplotene (Tease et al., 2002). Oocytes
91
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al.
Figure 1. Human fetal oocytes in meiotic prophase I. Synaptonemal complexes (SC) are highlighted in red by immunostaining for SYCP3. In (a), (b) and (c), MutL homologue (MLH1) foci are highlighted in yellow. (d) shows the same oocyte as in (c), after fluorescence in-situ hybridization (FISH) with probes for chromosomes 13, 18, 21 and X. (a) Pachytene oocyte with full synapsis; 72 MLH1 foci are clearly well separated along the SC. (b) Abnormal pachytene oocyte with fragmented SC, demonstrating a reduced number of MLH1 foci (48). (c) Pachytene oocyte with complete failure of synapsis of some chromosomes. Unpaired SC axes have a thinner appearance and are longer than those that are synapsed, as exemplified by the unpaired X chromosomes. The 54 MLH1 foci are present on synapsed chromosomes only. (d) The same oocyte as in (c) after FISH with DNA probes for chromosomes 21, 18, 13 and X, showing unpaired SC axes for chromosome X, where the centromere X is highlighted in green.
92
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. with SC anomalies showed a similar but not identical pattern: the numbers of MLH1 foci remained fairly constant from zygotene to diplotene. Overall, such abnormal oocytes tended to have fewer MLH1 foci than normal ones at the same developmental stage (Table 2). This was clearest at the pachytene stage where an average of 55.1 MLH1 foci per cell was found in cells with SC abnormalities compared with 70.3 in cells with normal SC (P P < 0.001, t-test Table 3), however, the ranges were wide and overlapping. More than half of the 33 pachytene oocytes with abnormal SC (19/33) had fragmentation of the SC (and in some instances also asynapsis), the remaining 14/33 showed asynapsis of chromosome segments or whole chromosomes (e.g. Figure 1c). Those with synaptic abnormalities alone had a slightly higher mean number of MLH1 foci per oocyte (59.6) than those with fragmentation (48.9) as shown in Table 4. This difference was statistically significant (P P = 0.019), despite overlap in the ranges. Examination of some of these cells after FISH to identify chromosomes 21, 18, 13 and X showed that chromosome-pairing errors were not confined to a particular type of chromosome, e.g. the smallest members of the complement, but apparently could affect any sized chromosome. For example, Figure 1d shows asynapsis of chromosome X with the homologues widely separated. No MLH1 foci were observed on asynapsed chromosome segments or chromosomes at pachytene.
Distributions of MLH1 foci along chromosomes at pachytene Adjacent MLH1 foci were almost invariably separated by considerable distances along chromosomal axes (eg, Figure 1a) from zygotene to diplotene stages in oocytes with normal appearance (Tease et al., 2002). However, MLH1 foci positioned close together were occasionally noted in oocytes with both normal and abnormal SC. It is clearly not possible to comment definitively on distributions of MLH1 foci in oocytes with extensive SC fragmentation (see Figure 1b). Subjectively, fragmented SC did not appear to have foci closer together than in normal oocytes. In oocytes classified as anomalous because of synaptic errors, the MLH1 foci appeared to show a similar pattern to that of cells with normal SC (compare Figure 1c with 1a). Exact numbers and positions of foci were determined after FISH in a small number of oocytes having synaptic anomalies (eg, Figure 1d). Chromosome 21 had 1.17 ± 0.4 foci (6 bivalents), chromosome 18 had 2.67 ± 1.2 (n = 3), and chromosome 13 had 2.5 ± 0.6 (n = 4). There was no obvious difference in mean frequency of MLH1 foci in comparison to oocytes with normal SC. Thus the corresponding mean for chromosome 21 was 1.23 ± 0.5 (n = 86), chromosome 18 is 2.36 ± 0.7 (n = 67) and chromosome 13 is 2.5 ± 0.6 (n = 64) in normal pachytene oocytes (Tease et al., 2002). The distribution patterns of foci along these chromosomes in abnormal oocytes also fell within the range found in normal oocytes (data not shown).
Discussion A substantial proportion of oocytes identified in this series of fetuses showed anomalous SC at prophase I. This observation agrees with previous reports that have used SC formation to assess human fetal oocyte development (Bojko, 1983; Speed,
1988; Hartshorne et al., 1999). These observations were made on cells that were prepared for microscopy using a method that includes a hypo-osmotic swelling step, which may require membrane integrity of the oocytes. Only those cells that were well spread and sufficiently flattened for their chromosomes to be observed in two focal planes were included in the analyses. This selection criterion may have excluded oocytes that had already progressed substantially towards atresia or post mortem degeneration as the membrane function of these cells was presumably reduced. The population of oocytes studied here may therefore be considered to be mainly viable. The data reported here indicate a significant decrease in the proportion of oocytes with SC fragmentation between the zygotene and pachytene stages, which has not been reported previously in humans. This observation could be explained if such abnormal cells arrested in development before the pachytene stage and/or were eliminated preferentially. On the other hand, the range of anomalous configurations (both fragmentation and synaptic irregularities) seen at the pachytene stage was similar to that at the zygotene stage, which may be taken to indicate that many zygotene cells with fragmentation are indeed able to continue development into the pachytene stage. However, since the methodology used here precludes following individual oocytes through progression of the meiotic prophase, the alternative possibility cannot be excluded, i.e. that all anomalous zygotene cells are eliminated and that the abnormal oocytes present at pachytene result from de novo events. The initial study of MLH1 at meiosis in male mice showed that the protein was present only on fully paired chromosomes in cells classified as pachytene (Baker et al., 1996). In contrast, the work described here in human oocytes has demonstrated that MLH1 foci appear in early zygotene at the time when SC formation is occurring (Tease et al., 2002), as also observed by Lenzi et al., (2005). The present study suggests that the same is true in anomalous zygotene oocytes. Although the proportion of zygotene oocytes with MLH1 foci was lower in the group with SC abnormalities, this result was not significant. One of the main findings of this study was that pachytene oocytes with fragmentation and/or synaptic abnormalities had significantly fewer MLH1 foci on average than their normal counterparts. Thus it is proposed that consequent selection may take place against oocytes with low recombination frequency in association with SC fragmentation and/or synaptic irregularity. This hypothesis would concur with that of Kong et al. (2004), based upon studies of families, that a high recombination count increases the chance of a gamete becoming a live birth in older women. Their study also indicated that a high recombination rate is related to reproductive success, and that biological selection of gametes may therefore be occurring on the basis of recombination. Meiotic recombination is absolutely dependent on homologous chromosome synapsis in mammalian germ cells. Failure of pairing inevitably causes failure of recombination in the chromosome (or chromosomal segment) affected. In sample C15, 41 pachytene oocytes were identified with asynapsis of one or more chromosome pairs (Table 2; Figure 1d). As expected, the affected bivalents did not show MLH1 foci. It is also of interest that unsynapsed chromosome segments often
93
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. show a striking decondensation of the SC and appear much longer than when normally synapsed (Saadallah and Hultén, 1986; Barlow et al., 2002). This may be one of the factors that trigger selection against oocytes with synaptic abnormalities. It does not automatically follow, even where synapsis is achieved, that crossing over will ensue. It was also found that 7/176 (4.0%) pachytene oocytes in the C15 fetus and 8/96 (8.3%) in the combined group did not have MLH1 foci despite the presence of apparently normal and complete synapsis (Table 1). Thus even among cells with a normal appearance judged by complete homologous synapsis, there might be other errors of the meiotic process present. Oocytes with failure of recombination, whether affecting a limited number of chromosome pairs or the whole genome, would inevitably have considerable difficulty achieving correct chromosome segregation later in meiosis and may contribute to the high rate of aneuploidy in human female gametes (Lamb and Hassold, 2004). Other evidence from our laboratory shows that recombination may be disrupted by suboptimal environmental conditions (Lyrakou et al., 2002), suggesting that it may potentially be sensitive to local perturbations within the ovary, while Lenzi et al. (2005) have demonstrated significant differences among fetuses, despite studying relatively few oocytes per fetus. In male mice lacking MLH1, meiotic checkpoints eliminate pachytene cells lacking recombination before they progress through metaphase (Eaker et al., 2002), however, this is not the case in female mice, where oocytes can progress through an abnormal metaphase associated with unstable spindles and chaotic chromosome attachment and segregation (Woods et al., 1999). In humans, many aneuploid conditions in conceptuses may derive from susceptibilities acquired during female meiotic prophase, e.g. potentially arising from defects in recombination at prophase I (before birth) eventually causing malsegregation of chromosomes at oocyte maturation and fertilization (after puberty). The extensive attrition of oocytes before birth, relatively low fertility per ovulation of humans, high rates of infertility, miscarriage and genetic abnormality in offspring, are possible consequences of the unique features of human female meiosis. In contrast to males, human female meiosis has distinguishing features such as reduced checkpoint efficiency and a prolonged bouquet phase, possibly associated with compromised DNA repair mechanisms (Roig et al., 2004). The findings described here, combined with the permissive effect upon checkpoint systems potentially brought about by adverse conditions of development (Eichenlaub-Ritter et al., 2002) might predispose to subsequent problems with spindle alignment and function, with serious consequences for gamete formation.
94
Speed (1988) suggested that pachytene oocytes with broken SC were the equivalent of the ‘Z’ cells first described by Beaumont and Mandl (1962). These are usually presumed to be on an atretic pathway. The presence of some oocytes with fragmented axial elements and SC from early zygotene suggests that atresia can be present from the initiation of synapsis or earlier. It would therefore appear that errors in or failure of synapsis might not be the initiating event for atresia. MLH1 foci were present in some ‘pachytene’ cells with extensive SC fragmentation (Figure 1b). The numbers and distributions of MLH1 foci in these cells gave the impression of essentially normal MLH1 focus formation, i.e. there was no indication of excessive numbers or that the
MLH1 foci were more closely spaced than in normal pachytene cells. One explanation for this pattern of MLH1 foci is that these cells had undergone normal progression through the early part of prophase I and that the breakage of the SC was a late event. This still leaves the question of what causes cells with apparently normal chromosome pairing and recombination to undergo SC fragmentation. The application of additional molecular markers indicative of cellular control mechanisms including apoptosis may be informative. While only relatively small samples of cells could be assessed, subjective observations of MLH1 foci along SC of anomalous oocytes indicated similar patterns of distribution to those found in normal cells. Additional data were obtained when chromosomes 13, 18, 21 and X were identified using FISH. These chromosomes were selected, as they are among the smallest in humans and aneuploidy for these causes substantial numbers of intrauterine and perinatal losses, as well as abnormalities in live born children. Preliminary data from oocytes in which these chromosomes were identified offered no evidence that the presence of chromosome pairing anomalies within a cell disrupted the distribution pattern of MLH1 foci on chromosome pairs that had successfully synapsed. This preliminary conclusion requires verification in a larger sample of oocytes. However, constraints on the availability of material for research such as this may limit this possibility. In summary, this paper presents evidence for significantly compromised recombination in human pachytene oocytes showing features of SC fragmentation and deficient synapsis. Such abnormal meiosis may potentially predispose to atresia or to abnormal development during later oocyte maturation, with consequences for the ovarian reserve and children originating from these oocytes. On the other hand, the fate of the oocytes with normal pachytene SC but still showing reduced recombination, is not currently known.
Acknowledgements We wish to thank the staff and patients of the Calthorpe Clinic, Birmingham for the provision of tissue. This study was supported by WellBeing grant H1/98 and Wellcome Trust grant 061202/ZOOZ.
References Baker TG 1963 A quantitative and cytological study of germ cells in human ovaries. Proceedings of the Royal Society [Biology] 158, 417–433. Baker SM, Plug AW, Prolla TA et al. 1996 Involvement of mouse Mlh1 in DNA mismatch repair and meiotic crossing over. Nature Genetics 13, 336–342. Barlow AL, Hultén MA 1998 Crossing over analysis at pachytene in man. European Journal of Human Genetics 6, 330–358. Barlow AL, Tease C and Hultén MA 2002 Meiotic chromosome pairing in fetal oocytes of trisomy 21 human females. Cytogenetics and Genome Research 96, 45–51. Beaumont HM, Mandl AM 1962 A quantitative and cytological study of oogonia and oocytes in the fetal and neonatal rat. Proceedings of the Royal Society [Biology] 155, 557–579. Bojko M 1983 Human meiosis. VIII. Chromosome pairing and formation of the synaptonemal complex in oocytes. Carlsberg Research Communications 48, 457–483. De Felici M, Di Carlo A, Pesce M et al. 1999 Bcl-2 and bax regulation
Article - Meiotic recombination patterns in human fetal oocytes - C Tease et al. of apoptosis in germ cells during prenatal oogenesis in the mouse embryo. Cell Death and Differentiation 6, 908–915. De Pol A, Vaccina F, Forabosco A et al. 1997 Apoptosis of germ cells during human prenatal oogenesis. Human Reproduction 10, 2235–2241. Eaker S, Cobb J, Pyle A, Handel MA 2002 Meiotic prophase abnormalities and metaphase cell death in MLH1 deficient mouse spermatocytes: insights into regulation of spermatogenic progress. Developmental Biology 249, 85–95. Eichenlaub-Ritter U, Shen Y, Tinneberg H-R 2002 Manipulation of the oocyte: possible damage to the spindle apparatus. Reproductive BioMedicine Online 5, 117–124. Hartshorne GM, Barlow AL, Child TJ et al. 1999 Immunocytogenetic detection of normal and abnormal oocytes in human fetal ovarian tissue in culture. Human Reproduction 14, 172–182. Hultén MA, Barlow AL, Tease C 2001 Meiotic studies in humans. In: Rooney DE (ed.) Human Cytogenetics: Constitutional Analysis (3rd edn): A Practical Approach. Oxford University Press, Oxford and New York. pp. 211–236. Kong A, Barnard J, Gudbjartsson DF et al. 2004 Recombination rate and reproductive success in humans. Nature Genetics 36, 1203–1206. Lamb NE, Hassold TJ 2004 Nondisjunction – a view from ringside. New England Journal of Medicine 351, 1931–1934. Lammers JHM, Offenberg HH, Van Aalderen M et al. 1994 The gene encoding a major component of synaptonemal complexes of the rat is related to X-linked lymphocyte-regulated genes. Molecular Cell Biology 14, 1137–1146. Lenzi ML, Smith J, Snowden T et al. 2005 Extreme heterogeneity in the molecular events leading to the establishment of chiasmata during meiosis 1 in human oocytes. American Journal of Human Genetics 76, 112–127. Lyrakou S, Hultén MA, Hartshorne GM 2002 Growth factors promote meiosis in vitro in mouse fetal ovaries. Molecular Human Reproduction 8, 906–911. Marcon E, Moens P 2003 MLH1p and MLH3p localize to precociously induced chiasmata of okadaic-acid-treated mouse spermatocytes. Genetics 165, 2283–2287.
Morita Y, Tilly JL 1999 Oocyte apoptosis: like sand through an hourglass. Developmental Biology 213, 1–17. Polkinghorne J 1989 Review of the Guidance of the Research Use of Fetuses and Fetal Material. Cm762, Her Majesty’s Stationery Office, London. Roeder GS, Bailis JM 2000 The pachytene checkpoint. Trends in Genetics 16, 395–403. Roig I, Liebe B, Egozcue J et al. 2004 Female-specific features of recombinational double-stranded DNA repair in relation to synapsis and telomere dynamics in human oocytes. Chromosoma 113, 22–33. Saadallah N, Hultén M 1986 EM investigations of surface spread synaptonemal complexes in a human male carrier of a pericentric inversion inv(13)(p12q14): the role of heterosynapsis for spermatocyte survival. Annals of Human Genetics 50, 369–383. Speed RM 1988 The possible role of meiotic pairing anomalies in the atresia of human fetal oocytes. Human Genetics 78, 260–266. Tease C, Hultén MA 2004 Inter-sex variation in synaptonemal complex lengths largely determine the different recombination rates in male and female germ cells. Cytogenetics and Genome Research 107, 208–215. Tease C, Hartshorne GM, Hultén MA 2002 Patterns of meiotic recombination in human fetal oocytes. American Journal of Human Genetics 70, 1469–1479. Tilly JL 2001 Commuting the death sentence: how oocytes strive to survive. Nature Reviews of Molecular Cell Biology 2, 838–848. Wallace BM, Hultén MA 1985 Meiotic chromosome pairing in the normal human female. Annals of Human Genetics 49, 215–226. Woods LM, Hodges CA, Baart E et al. 1999 Chromosomal influence on meiotic spindle assembly: abnormal meiosis I in female Mlh1 mutant mice. Journal of Cell Biology 145, 1395–1406. Zickler D, Kleckner N 1999 Meiotic chromosomes: integrating structure and function. Annual Review of Genetics 33, 603–754. Received 10 March 2006; refereed 21 March 2006; accepted 21 April 2006.
95