Immunocytogenetic detection of normal and abnormal oocytes in human fetal ovarian tissue in culture*

G.M. Hartshorne1,7, A.L. Barlow2,4, T.J. Child3,6, D.H. Barlow3 and M.A. Hultén2,5

1 Sir Quinton Hazell Molecular Medicine Research Centre, Department of Biological Sciences, University of Warwick, 2 LSF Research Unit, Regional Genetic Services, Heartlands Hospital, Birmingham and 3 Nuffield Department of Obstetrics and Gynaecology, University of Oxford, Oxford, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This study aimed to: (i) determine whether oocytes are present in cultures of human fetal ovary; (ii) identify whether meiotic anomalies are evident; and (iii) assess whether preparation or culture conditions influence oocyte survival and meiotic progression. Ovaries were collected from fetuses after termination at 13–16 weeks. Oocyte assessment utilized antibodies specific for synaptonemal complex proteins (associated with chromosomes only during meiosis), and antibodies to centromeric proteins. Fragments of tissue were cultured in minimal essential medium + 10% serum ± follicle stimulating hormone (100 mIU/ml). The sera were fetal calf serum (FCS), FCS for embryonic stem cells (ES-FCS) and human female serum. The numbers and stages of oocytes were assessed after 7–40 days, and particular arrangements of chromosome synapsis identified. Results in fresh tissue included oocytes at leptotene, zygotene, pachytene and diplotene in three of five samples. Four specimens remained viable in vitro, and three had detectable oocytes after culture. The numbers of oocytes and the proportions of zygotene and pachytene cells increased with time in culture. The proportion of degenerate cells in culture was initially higher than in fresh samples, but declined subsequently. More oocytes were detected in ES-FCS and human serum than in FCS. We conclude that human oocytes survive in culture and that progression through prophase I continues.

Key words: culture/fetus/meiosis/oocyte/ovary


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The study of human female gametes as they develop in the ovary in utero began 50 years ago with the classic histological studies of Witschi (1948). Traditionally, meiotic cells have been identified by histology (Ohno et al, 1961Go; Baker, 1963Go; Manotaya and Potter, 1963Go). More recently, cytological analysis has been performed, using light and electron microscopy, and new information has been obtained by fluorescent in-situ hybridization (FISH) (e.g. Wallace and Hultén, 1983Go, 1985Go; Speed, 1985Go, 1988Go; Garcia et al., 1989Go; Cheng and Gartler, 1994Go). Here, we describe a novel application of an immunocytogenetic method to stage meiotic prophase in freshly collected fetal oocytes and after in-vitro culture. This method allows screening of large numbers of oocytes while providing excellent detail of chromosome synapsis (Heng et al., 1994Go, 1996Go; Barlow and Hultén, 1996Go, 1997Go). A human autoimmune serum known to label centromeres (Earnshaw and Rothfield, 1985Go) provides additional information which is invaluable in interpreting the degree of chromosome synapsis in microspread oocyte nuclei.

In human females, primordial germ cells arise in the yolk sac very early in gestation and migrate to the developing gonad where they undergo repeated mitosis (for review see Motta et al., 1997Go). The resulting oogonia are believed to start entering meiosis at about 12 weeks gestation, beginning with those in the medulla of the ovary. There is a longstanding debate as to the factors controlling entry into meiosis, the control of meiotic arrest, and the subsequent utilization of the pool of oocytes formed (Byskov et al., 1997Go). In contrast to some other species, e.g. mice, the process of oogenesis in the human female is not synchronous. Thus, diploid primordial germ cells are present concurrently with various stages of meiotic prophase I between leptotene and diplotene. Atresia can occur during meiosis, as an apoptotic phenomenon (DePol et al., 1997Go), and it has been proposed that certain stages may be more susceptible than others (Baker, 1963Go; Baker and Neal, 1974Go; Speed, 1988Go).

The developmental competence of oocytes from prepubertal animals is lower than in maturity (Gandolfi et al., 1998Go). Nevertheless, normal offspring have been generated from prepubertal or fetal oocytes in many species (e.g. Hashimoto et al., 1992Go; Ledda et al., 1997Go). The developmental competence of oocytes matured in vitro from juvenile animals may be improved by altering the endocrine environment in vivo before oocyte collection (Presicce et al., 1997Go). Such experimentation may offer important information about oocyte function. We are culturing human fetal ovaries in order to provide an alternative system with which to study oocyte formation and development. Experiments in this area began in the early years of this century (see Martinovitch, 1937Go) and have continued sporadically (reviewed in Hartshorne, 1996Go).

Our previous data have shown that human fetal ovarian cells can be cultured for many weeks using simple techniques (Hartshorne et al., 1994aGo,bGo), as also reported by others (Blandau, 1969Go; Baker and Neal, 1974Go; Zhang et al., 1995Go). The primary purpose of this study was to determine whether oocytes survive and progress in our culture system, and to assess whether preparation or culture conditions might influence the survival and progression of oocytes. We have also examined to what extent errors in chromosome synapsis occur during prophase of meiosis I. Such errors may have serious consequences for subsequent development; however, the control of prophase I and the potential for generation of anomalies has not been explored previously in any detail. An improved culture system for fetal oocytes would be valuable to increase our understanding of a variety of areas by offering new opportunities to study meiosis in developing human oocytes. Subjects of interest might include the acquisition of cytoplasmic and maturational competence, and the assessment of factors which may affect meiotic development and recombination, such as genetic or cell cycle modifiers, teratogens or environmental toxins.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Tissue supply
Patients undergoing second-trimester termination of pregnancy at the John Radcliffe Hospital, Oxford, UK, gave consent to research in accordance with the Guidelines of the British Government (Polkinghorne, 1989Go). The study was approved by the Central Oxford Research Ethics Committee and the Coventry Research Ethics Committee.

Ovarian tissue was collected from five human female fetuses after termination of singleton pregnancies induced by priming with RU 486 (mifepristone, 200 mg, orally), followed 48 and 51 h later by a prostaglandin E1 analogue (Cervagem pessaries, 2 mg). At 54 h, an intravenous drip of syntocinon (50 IU) in 500 ml dextrose/saline was given at a rate of 62 ml/h. There was no known genetic indication for termination and the fetuses appeared morphologically normal. The gestational age was assessed by ultrasound scanning during pregnancy, and confirmed by measurement of foot length at the post-mortem examination.

Tissue culture
After delivery, the products of termination were kept at 4°C until a routine post-mortem examination. At this time, the ovaries were removed by the perinatal pathologist and placed in Leibovitz medium (Life Technology, Paisley, UK) containing 0.3% bovine serum albumin (BSA; Sigma, Poole, Dorset, UK) for transport by car approximately 50 miles to the University of Warwick.

Using sterile techniques, the ovaries were dissected free of adherent tissues and samples taken for histology and analysis of meiotic cells in the fresh specimen (see below). The remainder was cut into small pieces, transversely across the gonad, and then into wedges (approximately 0.3x0.3x0.3 mm) to control for the medullary/cortical arrangement of oocytes in the ovary (Byskov et al., 1997Go). Approximately 14 to 38 fragments per pair of ovaries were obtained for tissue culture. The number was variable dependent upon the size of the ovaries and the amount dissected off for other analyses. The tissue fragments were cultured individually on 0.3 µm pore polyester membranes over 0.5–1 ml medium in 24- or 12-well plates (Transwell, Costar, High Wycombe, UK). The level of medium was just sufficient to cover the tissue fragment. Any remaining wells were partially filled with medium to maintain humidity in the culture vessel. In our experience, this method provides a suitable environment for culture of the tissue fragments.

The medium comprised minimal essential medium alpha (MEM{alpha}) + 10% serum ± follicle stimulating hormone (FSH) (100 mIU/ml; HP Metrodin, Serono, Welwyn Garden City, UK). The sera used were fetal calf serum (FCS, Life Technology), FCS suitable for embryonic stem cells (ES-FCS, Life Technology) and human female serum [locally produced pool from patients on day 8 of a cycle of pituitary down-regulation using a superactive agonist of gonadotrophin releasing hormone (goserelin) and gonadotrophin stimulation in preparation for in-vitro fertilization treatment]. The cultures were maintained at 37°C in a humidified incubator under 5% CO2 in air. Every 2–4 days the medium under the membrane was removed and replaced with fresh medium. After 7–40 days, individual cultures were prepared for analysis of meiotic cells.

In addition, 26 cultures from three fetal samples were set up on gas-permeable tissue culture dishes, either in open medium as above, or in drops of medium under paraffin oil (Petriperm, Philip Harris, London, UK) as recommended for fragment cultures of mouse fetal gonads (McLaren and Buehr, 1990Go).

Analysis of oocytes
Oocytes were analysed using a novel immunocytochemical technique, as described previously (Barlow and Hultén, 1997Go), but with slight modification. For each specimen received, a sample of the freshly collected tissue (time 0) was assessed as well as fragments cultured under various conditions. For cultured tissue, the bulk of the tissue fragment was removed manually from the top of the transwell membrane using a Pasteur pipette and teased apart using fine scalpel blades. One drop of cell suspension was mixed with one drop of 0.3% Lipsol (LIP, Shipley, UK) on a clean glass slide and allowed to stand for 30 min. Cells were fixed with 2 drops of 1% Ultrapure formaldehyde (TAAB, Aldermaston, UK) in 0.04% sodium dodecyl sulphate (SDS) for 20 min. The cells were blocked for 30 min in PBT buffer [phosphate-buffered saline (PBS), 0.1% BSA, 0.1% Tween 20]. The oocytes were labelled using antibodies specific to lateral elements, which are components of the meiosis-specific synaptonemal complex (SC) and centromeres. The anti-SC antibody (A1), raised in rabbits, was a gift from Christa Heyting, University of Wageningen, The Netherlands (Lammers et al., 1994Go), and the anti-centromeric antibody (GS), a human autoimmune serum, was a gift from William Earnshaw, University of Edinburgh, Scotland (Earnshaw and Rothfield, 1985Go). Primary antibody A1 was used at 1:1000 dilution and GS at 1:5000 dilution in an overnight incubation at room temperature. The primary antibodies were detected using fluorescence-labelled secondary antibodies; FITC-labelled anti-rabbit IgG (Sigma) and TRITC-labelled anti-human IgG. Individual oocytes were identified by fluorescence microscopy using a Zeiss Axioplan microscope and a Vysis image analysis system on a Power Macintosh 8100/ computer. The images obtained were stored on optical disks. For maximum contrast, the illustrations of oocytes have been presented with FITC labelling shown in white.

For fresh specimens, at least 100 oocytes were analysed and for cultured fragments, all the oocytes in each preparation were counted. The developmental stage of oocytes in prophase I was assessed based on the degree of formation of SC, together with the number and location of centromeric signals. During leptotene, proteinaceous cores known as axial elements form along the length of chromosomes. SC are formed by the synapsis of axial elements which, in the context of the SC, are called `lateral elements'. The anti-SC antibody used recognizes both axial and lateral elements. Thus, in leptotene, the chromosomes condense and axial elements form between the chromatids of parental chromosomes. The anti-SC antibody gives a thin linear signal alongside the chromatids in regions where the axial elements have formed. The number of centromeric signals expected is 46, since chromosome synapsis has not begun. In zygotene, homologous chromosomes start to synapse with the formation of SC, observed with anti-SC antibody as a signal of double thickness along the synapsed part of the bivalent. Between 23 and 46 centromeric signals are observed, depending upon the extent of synapsis. In pachytene, the homologous chromosomes are completely synapsed, resulting in 23 linear signals of double thickness and 23 centromeric signals. In diplotene, the SC break down and the homologous chromosomes separate, except where chiasmata (crossing over points) exist. The SC have a chain-like appearance at this time and 46 centromeric signals are normally evident. Diplotene is the stage at which prophase I oocytes arrest during development to form primary (germinal vesicle stage, immature) oocytes. If such oocytes are surrounded by pregranulosa cells they become primordial follicles, in which form the majority of oocytes within the ovary are stored.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Tissue supply
The time between delivery and removal of the ovaries at post-mortem examination varied between 1 and 18 h, and the time from post-mortem to completion of setting up in culture was 3–5 h. The time from delivery to setting up was 4–22 h.

Tissue culture
Viable cultures resulted from four of the five specimens received. The sample which did not survive in vitro had an interval of 9 h between delivery and setting up in culture. During culture for 13 days, the tissue fragments remained unattached to the culture vessel, and the periphery appeared rough and ragged throughout. No viable cells migrated out of the tissue fragments. For the viable cultures, the appearance of the tissue fragment depended upon the culture conditions. During the first 2 days, the fragment rounded up and formed smooth surface contours, but subsequently attached to the membrane by about day 4–8 in most cultures. The degree of outgrowth on the membrane was greatest in the presence of FCS or ES-FCS. In human serum, the tissue usually remained partially or completely rounded up. At the end of the culture period, the main bulk of cells remained on the membrane's upper surface. However, clusters of cells were frequently found on the underside of the membrane, on the transwell support, or attached to the surface of the well. The membranes always appeared intact.

Petriperm culture dishes were unsuitable for culture due to the medium leaking through the porous surface within a week of initiation of culture. Since the membranes appeared intact and control dishes lacking cells did not leak, we believe that migration of cells through the membrane resulted in loss of surface tension, increased capillary action and consequent loss of medium through the membrane pores.

Analysis of oocytes
Meiotic prophase I cells were unequivocally present and clearly labelled with the anti-SC (A1) antibody, with superimposed centromeric signals generated by antibody GS. Leptotene cells of two types were observed. One type had fully developed axial elements. Other cells, however, showed highly fragmented axial elements. Zygotene and pachytene nuclei were readily identified by their incomplete and fully formed SC respectively. Somatic cells and primordial germ cells remained unlabelled with the anti-SC antibody.

A summary of the proportions of oocytes ascribed to the various stages of meiotic prophase I is given in Figure 1Go. No oocytes were detected in two fresh samples at time 0 (H10 and H6; 13 and 14 weeks gestation respectively). In contrast, oocytes at all of the prophase I stages (leptotene, zygotene, pachytene and diplotene) were readily identified in the remaining samples (H7, H8 and H9; 15, 16 and 16 weeks gestation respectively). In one of these cases (H9) the SC of all cells visualized appeared `fuzzy', though this feature was not seen in any of the other cases investigated.



View larger version (40K):
[in this window]
[in a new window]
 
Figure 1. Distribution of meiotic stages of oocytes (%) detected in fresh samples of fetal ovarian tissue. Four viable specimens were obtained, three of which contained oocytes. At least 100 oocytes were assessed for specimens H7, H8 and H9. No oocytes were found in H10.

 
In one of the cases showing no meiotic cells at time 0 (H6), no cells at all survived in culture. This specimen was probably not viable at collection. Remarkably, however, the other case in which no oocytes could be identified at time 0 (H10), did show clear cells at leptotene or zygotene in 5/12 cultures after 18–40 days in culture with ES-FCS or human female serum. In addition, the two cases (H7 and H8; 15 and 16 weeks gestation) with oocytes containing normal-appearing SC at time 0 had a higher proportion of cultures (8/11, 11/12) where oocytes were present. On the other hand, the case (H9; 16 weeks gestation) which was characterized by a `fuzzy' appearance of the SC at time 0, did not have any oocytes following in-vitro culture for 16 or 34 days. In this case, following in-vitro culture, there was evidence of high numbers of viable cells which did not show any SC labelling. These were classified as somatic or premeiotic cells, as there was no indication of any cells having entered meiosis. These results are summarized in Table IGo.


View this table:
[in this window]
[in a new window]
 
Table I. Summary of oocyte detection in fresh and cultured specimens of human fetal ovary
 
Examples of the recognizable stages of meiotic prophase found in cultured human fetal ovaries are shown in Figure 2Go. Various other meiotic chromosome configurations were also observed in fresh and/or cultured tissue. A summary of these is presented in Table IIGo and some examples from cultured cells are shown in Figures 3 and 4GoGo.



View larger version (45K):
[in this window]
[in a new window]
 
Figure 2. Microspread cultured human oocytes labelled with antisera A1 (anti-lateral element, white) and GS (centromeres, red). (a) Early zygotene nucleus. The axial elements of terminal regions have a well-defined solid appearance, while interstitial regions are less well-defined and speckled. Many terminal regions of axial elements are arranged in pairs (e.g. yellow arrows) and some have already initiated synapsis (e.g. green arrows). (b) Zygotene nucleus in which the process of synapsis is under way. (c) Pachytene nucleus with a full complement of complete synaptonemal complexes. Note the presence of a bivalent in which sera GS has produced two regions of labelling separated by a short stretch of synaptonemal complex (green arrow). (d) Diplotene nucleus containing regions of both interstitial (green arrows) and terminal (yellow arrow) desynapsis. Scale bar = 10 µm.

 

View this table:
[in this window]
[in a new window]
 
Table II. Examples of unusual cytogenetic features noted in oocytes
 


View larger version (66K):
[in this window]
[in a new window]
 
Figure 3. (a,b,c,e,f) Microspread cultured human oocytes labelled with antisera A1 (anti-lateral element, white) and GS (anti-centromeres, red). (d) Microspread nucleus labelled with antisera A1 (anti-lateral element, green), GS (anti-centromeres, red) and DAPI (grey). (a) Multiply asynaptic nucleus containing a full complement of well-defined axial elements. (b) `Stellar' nucleus in which angular axial/lateral elements emanate from a central cluster of centromeres. (c) Multiply asynaptic nucleus in which short stretches of fold-back self-synapsis have been initiated. Regions of synapsis (e.g. green arrows) can be distinguished from regions of asynapsis (e.g. yellow arrow) on the grounds of thickness and labelling intensity. (d) Nucleus in which chromatin is divided into distinct chromosome-like domains that label with antibody A1 to produce a flocculated pattern. (e) Multiply asynaptic nucleus containing a full complement of well-defined axial elements, many of which are apparently arranged into homologous pairs, the example (green arrows) also shows a centromeric association (yellow arrow). (f) Late zygotene nucleus containing an apparent quadrivalent (yellow arrow). Scale bar = 10 µm.

 


View larger version (63K):
[in this window]
[in a new window]
 
Figure 4. Microspread cultured human oocytes labelled with antisera A1 (anti-lateral element, white) and GS (anti-centromeres, red). (a) Highly aberrant zygotene nucleus in which synapsis appears to be progressing with little regard for homology. There is a clear region of non-homologous synapsis (yellow arrow). (b) Highly aberrant nucleus containing asynapsed axial elements with a `wiggly' appearance. (c) Highly aberrant nucleus in which axial elements are arranged in pairs that show frequent points of convergence, appearing `twisted'. Scale bar = 10 µm.

 
Where meiosis was detected in fresh tissue samples, the oocytes were abundant and at least 100 cells were analysed in detail from each specimen. The numbers of oocytes detected in individual samples after culture ranged from 0–63 (mean of positive results = 19). The results of meiotic analysis after culture have been presented for sample H8 (16 weeks gestation), for which the most complete data are available. Figure 5Go shows the absolute numbers of oocytes detected in H8 after culture for 7 and 14 days under various culture conditions. The numbers of meiotic cells increased between 7 and 14 days of culture, except in FCS. A similar trend was evident for sample H7, which was cultured for up to 25 days.



View larger version (33K):
[in this window]
[in a new window]
 
Figure 5. Numbers of oocytes (n) detected in cultures of human fetal ovarian fragments after 7 and 14 days of culture under three different serum conditions. Results are presented for specimen H8 (16 weeks gestation). The sera were fetal calf serum (FCS), embryo stem cell-tested FCS (ES-FCS) and a pool of human female serum from patients undergoing ovarian stimulation. Bars represent average and range of two data points.

 
Figure 6Go shows the proportions of oocytes in H8 under various conditions of culture. The proportion of degenerate cells was higher after 7 days in culture than at the start of culture (compare Figures 1 and 6GoGo), but had declined to similar levels under all conditions by day 14. The proportions in zygotene after 14 days were higher than at the start of culture. Pachytene cells were not found on day 7, but were evident in five of six cultures on day 14. The proportion of pachytene cells after culture for 14 days was, however, less than that in the fresh tissue sample. The results of cultures with and without FSH were similar in terms of the numbers and proportions of cells in the various meiotic stages, and so data for cultures with and without supplementary FSH have been combined.





View larger version (94K):
[in this window]
[in a new window]
 
Figure 6. Proportions of oocyte meiotic stages (%) identified in fragments of human fetal ovarian tissue cultured under different serum conditions for 7 or 14 days. The numbers of cells evaluated for each culture are presented in the legend. (a) Fetal calf serum (FCS). (b) Embryo stem cell-tested fetal calf serum (ES-FCS). (c) Human female serum (locally produced pool from patients undergoing ovarian stimulation).

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The results presented in this paper show clearly that oocytes derived from human fetal ovaries can survive in culture, and furthermore, that meiosis can progress in vitro. The presence of oocytes in cultures derived from a fresh specimen in which meiotic figures were not detected strongly suggests that meiosis can also be initiated in vitro. This work confirms, with unequivocal cytogenetic evidence, the results of others who have cultured human fetal ovaries (Blandau, 1969Go; Baker and Neal, 1974Go; Zhang et al., 1995Go). The techniques presented offer new opportunities to study meiosis in developing human oocytes, and to assess factors which may affect meiotic development and recombination, such as genetic or cell cycle modifiers, teratogens or environmental toxins.

The immunocytogenetic technique used provides evidence of the presence of oocytes, and greatly enhances the reliability of meiotic staging, while retaining the potential to analyse individual cells further via FISH and/or electron microscopy if necessary (Barlow and Hultén, 1997Go). The SC is found only in meiotic prophase I cells, and the additional inclusion of a centromeric marker provides invaluable information for identifying synapsis and to assess the degree of synapsis of bivalents; i.e. 46 spots in leptotene, when no synapsis has taken place; 23 in pachytene, when full synapsis has been achieved; and between 23 and 46 for zygotene, with partial synapsis, and diplotene, with desynapsis. Previous evidence of the reliability of this technique was gained examining testicular specimens (Barlow and Hultén, 1996Go) and ovarian samples from genetically abnormal fetuses (Barlow and Hultén, 1997Go).

Culture
Previous investigators have used a variety of methods of culture, as well as tissue from various sources (miscarriage, termination for genetic or other anomaly or termination for non-genetic reasons). Blandau (1969) reported that living oogonia and oocytes were maintained in some human fetal ovary samples cultured for up to 79 days in Rose chambers. Active migration of oogonia, repeated mitotic divisions, and growth and differentiation of fetal ovarian tissue in culture, were observed by direct microscopic visualization of `squash' preparations and time-lapse cinematography. Blandau concluded that the oogonia entered meiosis, formed primordial follicles, and some oocytes achieved diameters of 80 µm. His starting material was obtained from therapeutic terminations or spontaneous abortions, having crown–rump lengths of 22–171 mm, although the gestational age was not reported.

Zhang et al. (1995) cultured fragments of frozen or fresh ovaries from fetuses of 16–20 weeks gestation, terminated for either therapeutic or elective reasons in China. These authors report the persistence of primordial follicles in culture, which grew to >60 µm diameter, and subsequently released oocytes. A quarter of the oocytes extruded a polar body, implying that they had reached metaphase II. There are a number of differences between the study by Zhang et al. (1995) and our own. It is important to note, firstly, that Zhang et al. classified their developing oocytes only according to the size and appearance of the cells, whereas we have used specific markers for meiotic chromosomes. Secondly, the gestational ages of material used in the studies may differ substantially. Zhang et al. (1995) did not verify gestational age by ultrasound, which dates pregnancies to within 3–7 days at 12–16 weeks gestation, with the most accurate results being obtained early in the second trimester. Instead, they assessed gestational age based upon a combination of crown–rump and foot measurements and the dates of the last menstrual period, which is usually estimated to the nearest gestational week. The methods of termination of pregnancy used in China are not described. In addition, their starting material already included many primordial follicles, identified histologically, which contain oocytes in the diplotene stage of meiosis. In contrast, no primordial follicles were evident in our starting material, and only a small proportion of diplotene nuclei were present. Most of the oocytes present in our freshly collected tissue were in earlier stages of prophase I, and primordial germ cells apparently persisted, based on the presence of alkaline phosphatase-positive cells. Moreover, 6% of oocytes which Zhang et al. (1995) identified at the start of culture were at least 80 µm diameter, whereas we identified no such enlarged oocytes in either our starting material or after culture. Variation between different fetuses might be expected; nevertheless, the extensive differences noted here suggest that the specimens used by Zhang et al. (1995) were substantially advanced compared with the material used in our study. This may account for the apparently greater degree of oocyte development observed in their study.

Factors which normally arrest or regulate the further growth of most primordial follicles in vivo may be lacking in vitro after tissue disruption, enabling their further development to a greater extent than in vivo (Gosden, 1990Go; Byskov et al., 1997Go; Wandji et al., 1997Go). We have not seen any clearly identifiable primordial follicles after in-vitro culture, although structures which appear superficially similar have been observed in cultures in this series and previously (Hartshorne et al., 1994aGo,bGo), though these have not been confirmed histologically. Primordial follicles contain immature oocytes arrested in the diplotene stages of meiosis, and we saw few diplotene nuclei in our fresh specimens and an even smaller proportion were present in cultured tissue. The reasons for this are uncertain, but may relate to aspects of either the termination procedure or suboptimal conditions in vitro.

This study has shown that the collection of second-trimester termination specimens is compatible with the use of this material in experiments to study oocyte meiosis in culture. Four of the five ovarian tissue samples remained viable, despite the inevitable long delay between termination and tissue supply. This confirms our previous results where five of six similar samples resulted in viable cultures, as assessed by the morphological appearance of the cultured tissue fragments (Hartshorne et al., 1994bGo). Meiotic cells were known to remain suitable for microspreading for at least 24 h, although a shorter time was preferred (Wallace and Hultén, 1985Go; Speed, 1988Go). However, the feasibility of tissue culture after such a delay was uncertain. Post-mortem degeneration of mouse oocytes at the dictyotene stage (prophase I arrest, germinal vesicle) was irreversible from about 6 h after death when the oocytes were left in intact carcasses at room temperature (Schroeder et al., 1991Go). This was shown by their reduced ability to fertilize and cleave to yield viable embryos. Oocytes which were collected from large antral follicles 6 h after death contained densely staining granules in the cytoplasm, and the surrounding follicular structure had already begun to degenerate, as shown by pyknotic cells and deteriorating intercellular contacts (Schroeder et al., 1991Go). The exact time of death of fetuses in our study may be considered as the time of delivery at the latest. They were kept at 4°C until post-mortem, and this temperature may have contributed to the prolonged survival of cells, yet, it is still remarkable that viable meiotic cells could be maintained after so many hours delay before culture.

In one sample, despite the presence of oocytes in the freshly collected tissue, only somatic cells grew in culture. This suggests a differential sensitivity of oocytes to some aspect of their handling, possibly including the drug exposure, storage at 4°C, time of demise of the fetus, or an in-vitro selection pressure. The high sensitivity of germ cells to various insults during mitosis or meiosis has been reported previously (Peters, 1969Go; Byskov et al., 1997Go). Notably, cultures of spermatocytes are prone to revert to fibroblasts under in-vitro conditions (Steinberger et al., 1970Go).

The drugs used to induce termination included RU 486 and a prostaglandin E1 analogue. RU 486 is a potent anti-progestin and anti-glucocorticoid, which also has anti-oestrogenic effects (Teutsch and Philibert, 1994Go). RU 486 is steroidal and an approximately 40% reduction in activity occurs due to first-pass metabolism in humans; however, it has a long plasma half life (30–48 h), possibly due to extensive binding to {alpha}1-acid glycoprotein in the serum (Heikinheimo et al., 1994Go). Metabolic breakdown of RU 486 might be expected in the placenta and fetal liver; however, its extent is uncertain. Concentrations of RU 486 reaching the chorionic villi are low compared with those in the serum and decidua (Wang et al., 1994Go). The fetal ovary is known to contain low levels of steroidogenic enzymes during the mid-trimester (Voutilainen and Miller, 1988Go). Therefore, factors affecting steroid function which are used during termination might possibly affect this tissue, although the nature and extent of any such effects are currently unknown.

Exogenous prostaglandin E1 is used to induce uterine contractions and promote cervical ripening to facilitate uterine evacuation at termination. It is known to affect uterine contractility and decidual prostaglandin F2{alpha} production during subsequent tissue culture (Norman et al., 1991Go). Histology has shown detectable effects of prostaglandin E1 used during second-trimester termination upon fetal organs, notably the liver, but data on the ovary are lacking (Lazda and Sams, 1995Go). Prostaglandins are involved in many aspects of normal ovarian function, including ovulation and luteal function. It is uncertain whether the concentrations employed at termination might have an effect upon the subsequent behaviour of fetal ovarian tissue in vitro.

The ability of cultured fetal ovarian cells to migrate through small apertures suggests that some of their inherent motility, evident in migration to the developing gonad and subsequently within the developing ovary (Motta and Makabe, 1986Go), may be maintained in vitro. Some, but not all of the cells found below the culture membrane stained positive for alkaline phosphatase, which is a commonly used marker of primordial germ cells (see Buehr and McLaren, 1993Go). Our observations support the movements of human oogonia and oocytes that have been visualized previously under in-vitro conditions (Blandau, 1969Go).

The culture conditions which we applied in these experiments have not yet been optimised, but are based upon a synthesis of the available literature on culture of ovarian tissue in animals and man. We utilized a complex medium and varied the type of serum supplement, as a starting point for identifying suitable culture conditions for human fetal ovarian tissue, and have shown that the serum can clearly affect the survival of oocytes, as well as the appearance of the tissue in vitro. The presence or absence of supplementary FSH had no discernible effect on oocytes, in agreement with reports by others. Second-trimester fetal ovaries probably lack FSH receptors (Huhtaniemi et al., 1987Go), despite being exposed to very high levels of FSH at this time in vivo (Beck-Peccoz et al., 1991Go). Thus, Blandau (1969) tried various concentrations of homologous human sera, but preferred heat-inactivated horse serum (10%) for culture. Wandji et al. (1997) have shown that serum-free conditions were sufficient for baboon primordial follicles to initiate growth in vitro, and that FCS inhibited some follicular growth effects. Baker and Neal (1974) showed that oogenesis occurs at the same rate with or without gonadotrophins, in vitro or in vivo. Much more work is necessary to refine the medium and its supplements, and this may have important influences upon the results obtained in this and other studies. Work in mice has demonstrated the potential for advanced stages of oocyte development from fetal ovarian specimens cultured in vitro (McLaren and Buehr, 1990Go; Eppig and O'Brien, 1996Go) and offers hope that a defined culture system for human fetal oocytes may be feasible.

Meiotic analysis
The proportions of meiotic stages which we found before culture differ from those reported by Cheng and Gartler (1994) for similar-aged fetuses, using FISH. At 15–16 weeks' gestation, our data showed zygotene or pachytene stages as more prevalent than leptotene, whereas Cheng et al. (1994) found leptotene to be the predominant stage. Speed (1988), using electron microscopy of microspread cells, found a consistently lower proportion of zygotene nuclei in three fetuses of 17 weeks gestation than we found in 16-week-old fetuses. Speed (1988) also identified cells containing multiply asynapsed SC which he regarded as being anomalous leptotene cells. Many of the previous investigators of human fetal ovaries have reached different conclusions about the prevalence of different stages of meiotic prophase at different gestational ages (Ohno et al., 1962Go; Baker, 1963Go; Manotaya and Potter, 1963Go; Gondos et al., 1971Go; Kurilo, 1981Go; Speed, 1985Go, 1988Go; Rabinovici and Jaffe, 1990Go). This is generally believed to result from differences in interpretation of the meiotic figures viewed, but differences between subjects are also a likely source of variation. Hopefully, the availability of more easily interpretable methodology, as we have demonstrated here, will improve the consistency of data.

Oocytes were detected under all culture conditions, despite the tissue samples for in-vitro culture being considerably smaller than those analysed fresh; thus, relatively lower numbers of oocytes were available. In addition, the different culture conditions are expected to be of importance, not only for oocytes' survival, but also for their capability of developing through meiotic prophase in vitro. Finally, the tendency for spreading and migration of cells in vitro may have influenced the efficiency of collection of cultured tissue fragments for meiotic analysis. We found that there was an initially high level of degeneration of oocytes in culture, in agreement with others (Blandau, 1969Go; Baker and Neal, 1974Go). After culture, the proportions of oocytes in leptotene and zygotene were similar to or higher than those in fresh tissue, whereas a lower proportion of pachytene cells was found. It is not clear from our data whether any of the pre-existing oocytes survived the transition to culture, or whether what we observed was de novo meiosis, or a mixture of both. Baker and Neal (1974) measured [3H]thymidine incorporation into oocytes in organ-cultured fetal ovary after a maximum delay before culture of 2 h. This method detects active DNA synthesis. Labelled zygotene cells were evident after 10–12 days of culture, pachytene from day 20, and diplotene from 28 days in vitro, suggesting that these were the times required for progress to these stages of meiotic prophase. However, we observed pachytene figures after 14 days, though not 7 days, of culture. This may suggest that either some leptotene or zygotene cells were surviving, or that pachytene occurred more rapidly in our system. This might be feasible since other aspects of development, such as oocyte growth rates, may be accelerated in vitro (Byskov et al., 1997Go).

The lack of diplotene stages which we observed was quite probably due to an insufficient duration of culture; however, there are alternative explanations. Selective degeneration in pachytene may occur (Baker and Neal, 1974Go; Speed, 1988Go), although we found no obvious evidence of degenerate pachytene figures in the present data. Indeed, many anomalies were evident in zygotene nuclei, but pachytene nuclei usually appeared normally synapsed, more in agreement with Wallace and Hultén (1985) and Cheng and Gartler (1994). Alternatively, if primordial follicles are resistant to or remain intact during the spreading procedure, diplotene cells may be masked. However, this appears unlikely in view of the hypo-osmotic conditions employed. These possibilities require further investigation.

The proportion of cells containing a full complement of normally synapsed chromosomes was estimated as 54% by Speed (1988). We found that it was not always possible to determine whether a cell's chromosomes were synapsed homologously throughout, particularly where synapsis was not complete. Many of the abnormalities which Speed described in fresh ovaries were seen in our cultured material, as well as some others (see Table IIGo). In view of the comparatively low numbers of meioses which we have so far analysed in cultured material, no attempt at quantification has been made.

One feature commonly observed in both fresh and cultured cells was the apparent staggering of the centromeres in otherwise normally synapsed bivalents (e.g. see Figure 2cGo, green arrow). No explanation can currently be offered for this phenomenon. FISH evidence with alpha satellite probes confirms that the antibodies label centromeres, so artefactual labelling has been excluded (data not shown). Some oocytes present in culture also showed clearly abnormal synapsis, either in isolated bivalents, or generalized throughout (Figure 4aGo).

In summary, this work has confirmed that oocytes survive in cultures of human fetal ovaries maintained on transwell membranes in a complex medium, and in a variety of sera, for up to 40 days after termination of second-trimester pregnancy. Antibody-mediated identification of the SC has shown clearly that both normal and abnormal oocytes are present. The proportions of oocytes in the different stages of meiosis appeared similar in the different sera evaluated (FCS, ES-FCS or human female); however, the total number of oocytes appeared lower in FCS. This study provides a basis from which to work towards further characterization of the culture conditions and an assessment of factors which may influence meiosis.


    Acknowledgments
 
We wish to thank Dr Stephen Gould and the perinatal pathology team at the John Radcliffe Hospital, Oxford, for their assistance with supplying specimens. We are most grateful to Christa Heyting, University of Wageningen, The Netherlands, and William Earnshaw, University of Edinburgh, UK, for their generous provision of A1 antibodies and serum GS, respectively. This work was supported in part by a Sanofi Winthrop Foundation grant to G.M.H. and a Wellcome Trust grant to M.A.H.


    Notes
 
*Presented in part at the 13th Annual Meeting of ESHRE, Edinburgh, June 1997, and the British Fertility Society Annual Meeting, Sheffield, April 1998.

4 Present address: Department of Anatomy, University of Birmingham, UK Back

5 Present address: Sir Quinton Hazell Molecular Medicine Research Centre, Department of Biological Sciences, University of Warwick, UK Back

6 Present address: Horton Hospital, Banbury, UK Back

7 To whom correspondence should be addressed at: Department of Biological Sciences, University of Warwick, CV4 7AL, UK e-mail: hx{at}dna.bio.warwick.ac.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Baker, T.G. (1963) A quantitative and cytological study of germ cells in human ovaries. Proc. R. Soc. B., 158, 417–433.[ISI]

Baker, T.G. and Neal, P. (1974) Oogenesis in human fetal ovaries maintained in organ culture. J. Anat., 117, 591–604.[ISI][Medline]

Barlow, A. and Hultén, M. (1996) Combined immunocytogenetic and molecular cytogenetic analysis of meiosis I human spermatocytes. Chromosome Res., 4, 562–573.[ISI][Medline]

Barlow, A. and Hultén, M. (1997) Sequential immunocytogenetics, molecular cytogenetics and transmission electron microscopy of microspread meiosis I oocytes from a human fetal carrier of an unbalanced translocation. Chromosoma, 106, 293–303.[ISI][Medline]

Beck-Peccoz, P., Padmanabhan, V., Baggiani, A.M. et al. (1991) Maturation of hypothalamic-pituitary-gonadal function in normal human fetuses: circulating levels of gonadotrophins, their common {alpha}-subunit and free testosterone, and discrepancy between immunological and biological activities of circulating follicle stimulation hormone. J. Clin. Endocrinol. Metab., 73, 525–532.[Abstract]

Blandau, R.J. (1969) Observations on living oogonia and oocytes from human embryonic and fetal ovaries. Am. J. Obstet. Gynecol., 104, 310–319.[ISI][Medline]

Buehr, M. and McLaren, A. (1993) Isolation and culture of primordial germ cells. Methods Enzymol., 225, 58–77.[ISI][Medline]

Byskov, A.G., Guoliang, Z. and Yding Andersen, C. (1997) The cortex-medulla oocyte growth pattern is organized during fetal life: an in-vitro study of the mouse ovary. Mol. Hum. Reprod., 3, 795–800.[Abstract]

Cheng, E.Y. and Gartler, S.M. (1994) A FISH analysis of X chromosome pairing in early human female meiosis. Hum. Genet., 94, 389–394.[ISI][Medline]

DePol, A., Vaccina, F., Forabosco, A. et al. (1997) Apoptosis of germ cells during human prenatal oogenesis. Hum. Reprod., 12, 2235–2241.[Abstract]

Earnshaw, W.C. and Rothfield, N.F. (1985) Identification of a family of human centromere proteins using an autoimmune sera from patients with scleroderma. Chromosoma, 91, 313–321.[ISI][Medline]

Eppig, J.J. and O'Brien, M.J. (1996) Development in vitro of mouse oocytes from primordial follicles. Biol. Reprod., 54, 197–207.[Abstract]

Gandolfi, F., Milanesi, E., Pocar, P. et al. (1998) Comparative analysis of calf and cow oocytes during in vitro maturation. Mol. Reprod. Dev., 49, 168–175.[ISI][Medline]

Garcia, M., Dietrich, A., Pujol, R. and Egozcue, J. (1989) Nucleolar structures in chromosomes and SC preparation from human oocytes at first meiotic prophase. Hum. Genet., 82, 147–153.[ISI][Medline]

Gondos, B., Bhiraleus, P. and Hobel, C.J. (1971) Ultrastructural observations of germ cells in human fetal ovaries. Am. J. Obstet. Gynecol., 110, 644–652.[ISI][Medline]

Gosden, R.G. (1990) Restoration of fertility in sterilised mice by transferring primordial ovarian follicles. Hum. Reprod., 5, 499–504.[Abstract]

Hartshorne, G.M. (1996) Fetal ovarian tissue in vitro. Assist. Reprod. Rev., 6, 72–82.

Hartshorne, G.M., Sargent, I.L. and Barlow, D.H. (1994a) In vitro maturation as a source of human oocytes and embryos for research. Hum. Reprod., 9, 970–972.[ISI][Medline]

Hartshorne, G.M., Ferguson, D., Sargent, I.L. and Barlow, D.H. (1994b) Preliminary data culturing human fetal ovarian tissue in vitro. Hum. Reprod., 9 (suppl. 4), abstract 112.

Hashimoto, K., Noguchi, M. and Nakatsuji, N. (1992) Mouse offspring derived from fetal ovaries or reaggregates which were cultured and transplanted into adult females. Dev. Growth Differ., 34, 233–238.[ISI]

Heikinheimo, O., Pesonen, U., Huuponen, R. et al. (1994) Hepatic metabolism and distribution of mifepristone and its metabolites in rats. Hum. Reprod., 9 (suppl. 1), 40–46.[ISI][Medline]

Heng, H.H.Q., Tsui, L.C. and Moens, P.B. (1994) Organization of heterologous DNA inserts on the mouse meiotic chromosome core. Chromosoma, 103, 401–407.[ISI][Medline]

Heng, H.H.Q., Chamberlain, J.W., Shi, X.M. et al. (1996) Regulation of meiotic chromatin loop size by chromosomal position. Proc. Natl. Acad. Sci. USA, 93, 2795–2800.[Abstract/Free Full Text]

Huhtaniemi, I.T., Yamamoto, M., Ranta, T. et al. (1987) Follicle stimulating hormone receptors appear earlier in the primate fetal testis than in the ovary. J. Clin. Endocrinol. Metab., 65, 1210–1214.[Abstract]

Kurilo, L.F. (1981) Oogenesis and antenatal development in man. Hum. Genet., 57, 86–92.[ISI][Medline]

Lammers, J.H.M., Offenberg, H.H., van Aalderen, M. et al. (1994) The gene encoding a major component of the lateral elements of synaptonemal complexes of the rat is related to X-linked lymphocyte-regulated genes. Mol. Cell. Biol., 14, 1137–1146.[Abstract]

Lazda, E.J. and Sams, V.R. (1995) The effects of gemeprost on the second trimester fetus. Br. J. Obstet. Gynaecol., 102, 731–734.[ISI][Medline]

Ledda, S., Bogliolo, L., Calvia, P. et al. (1997) Meiotic progression and developmental competence of oocytes collected from juvenile and adult ewes. J. Reprod. Fertil., 109, 73–78.[Abstract]

Manotaya, T. and Potter, E.L. (1963) Oocytes in prophase of meiosis from squash preparation of human fetal ovaries. Fertil. Steril., 14, 378–392.[ISI][Medline]

Martinovitch, P.N. (1937) The development in vitro of the mammalian gonad. Proc. R. Soc. B., 125, 232–249.

McLaren, A. and Buehr, M. (1990) Development of mouse germ cells in cultures of fetal gonads. Cell Differ. Dev., 31, 185–190.[ISI][Medline]

Motta, P.M. and Makabe, S. (1986) Elimination of germ cells during differentiation of the human ovary: an electron microscopic study. Eur. J. Obstet. Gynecol. Reprod. Biol., 22, 271–286.[ISI][Medline]

Motta, P.M., Makabe, S. and Nottola, S.A. (1997) The ultrastructure of human reproduction. I. The natural history of the female germ cell: origin, migration and differentiation inside the developing ovary. Hum. Reprod. Update, 3, 281–295.[Abstract/Free Full Text]

Norman, J.E., Kelly, R.W. and Baird, D.T. (1991) Uterine activity and decidual prostaglandin production in women in early pregnancy in response to mifepristone with or without indomethacin in vivo. Hum. Reprod., 6, 740–744.[Abstract]

Ohno, S., Makino, S., Kaplan, W.D. and Kinosita, R. (1961) Female germ cells of man. Exp. Cell Res., 24, 106–110.[ISI][Medline]

Ohno, S., Klinger, H.P. and Atkin, N.B. (1962) Human oogenesis. Cytogenetics, 1, 43–58.

Peters, H. (1969) The development of the mouse ovary from birth to maturity. Acta Endocrinol., 62, 98–116.[ISI][Medline]

Polkinghorne, J. (1989) Review of the Guidance on the Research Use of Fetuses and Fetal Material. Cm762, Her Majesty's Stationery Office, London.

Presicce, G.A., Jiang, S., Simkin, M. et al. (1997) Age and hormonal dependence of acquisition of oocyte competence for embryogenesis in prepubertal calves. Biol. Reprod., 56, 386–392.[Abstract]

Rabinovici, J. and Jaffe, R.G. (1990) Development and regulation of growth and differentiated function in human and subhuman primate fetal gonads. Endocr. Rev., 11, 532–557.[Abstract]

Schroeder, A.C., Johnston, D. and Eppig, J.J. (1991) Reversal of post-mortem degeneration of mouse oocytes during meiotic maturation in vitro. J. Exp. Zool., 258, 240–245.[ISI][Medline]

Speed, R.M. (1985) The prophase stages in human fetal oocytes studied by light and electron microscopy. Hum. Genet., 69, 69–75.[ISI][Medline]

Speed, R.M. (1988) The possible role of meiotic pairing anomalies in the atresia of human fetal oocytes. Hum. Genet., 78, 260–266.[ISI][Medline]

Steinberger, A., Ficher, M. and Steinberger, E. (1970) Studies of spermatogenesis and steroid metabolism in cultures of human testicular tissue. In Rosemberg, E. and Paulsen, C.A. (eds), The Human Testis. Plenum Press, New York, pp. 333–354.

Teutsch, G. and Philibert, D. (1994) History and perspectives of anti-progestin from the chemist's point of view. Hum. Reprod., 9 (suppl. 1), 12–31.[ISI][Medline]

Voutilainen, R. and Miller, W.L. (1988) Developmental and hormonal regulation of mRNAs for insulin-like growth factor II and steroidogenic enzymes in human fetal adrenals and gonads. DNA, 7, 9–15.[ISI][Medline]

Wallace, B.M.N. and Hultén, M.A. (1983) Triple chromosome synapsis in oocytes from a human fetus with trisomy 21. Ann. Hum. Genet., 47, 271–276.[ISI][Medline]

Wallace, B.M.N. and Hultén, M.A. (1985) Meiotic chromosome pairing in the normal human female. Ann. Hum. Genet., 49, 215–226.[ISI][Medline]

Wandji, S.A., Srsen, V., Nathanielsz, P.W. et al. (1997) Initiation of growth of baboon primordial follicles in vitro. Hum. Reprod., 12, 1993–2001.[Abstract]

Wang, J.D., Shi, W.L., Zhang, G.Q. and Bai, X.M. (1994) Tissue and serum levels of steroid hormones and RU 486 after administration of mifepristone. Contraception, 49, 245–253.[ISI][Medline]

Witschi, E. (1948) Migration of the germ cells of human embryos from the yolk sac to the primitive gonadal folds. Contrib. Embryol., 32, 67.

Zhang, J., Liu, J., Xu, P. et al. (1995) Extracorporeal development and ultrarapid freezing of human fetal ova. J. Assist. Reprod. Genet., 12, 361–368.[ISI][Medline]

Submitted on May 22, 1998; accepted on September 29, 1998.