| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
e
a
a Department of Biomedical Sciences, The University of Edinburgh, Edinburgh EH8 9XD, United Kingdom
ABSTRACT
Follicle selection occurs throughout an adult female's reproductive life, with selected, dominant follicle(s) developing to the preovulatory stage whereas the remaining, subordinate follicles within the growing cohort instead undergo atresia and die. To date, most research into follicle dominance has concentrated on its endocrine regulation, although it seems likely that intraovarian mechanisms are also involved in its regulation. We demonstrate here that the response of singly cultured murine follicles to declining concentrations of FSH depends on their developmental stage, with follicles at an earlier stage of development being much more susceptible than mature follicles to a lowering of FSH levels. We then extrapolate this information to follicle cocultures, in which a large dominant follicle was grown with a small subordinate follicle in a manner that maintained a dominant/subordinate relationship, with follicle health assessed by a terminal transferase-mediated 2'-deoxyuracil 5'-triphosphate nick end-labeled reaction on whole-follicle mounts. Our investigations show a combined negative effect of coculture and FSH withdrawal on small subordinate follicles, such that subordinate follicles cocultured with dominant follicles and subjected to a lowering of FSH levels during the culture period exhibit a greatly increased incidence of apoptosis in the granulosa cells (750% increase) compared with that exhibited by the dominant follicles (97% increase). We suggest that a similar interaction between endocrine and intraovarian factors regulates follicular dominance in vivo, such that dominant follicles, in addition to bringing about a fall in FSH levels via the hypothalamic-pituitary axis, exert local, direct effects on subordinate follicles, with both of these influences combining to induce atresia in subordinate follicles.
apoptosis, developmental biology, follicle, follicle-stimulating hormone, follicular development
INTRODUCTION
The prime influence on ovarian follicle selection is a decline in the systemic concentration of follicle-stimulating hormone (FSH) brought about by the action of estradiol and inhibin (produced in the ovary) on the hypothalamic-pituitary axis [1, 2]. This provides a relatively straightforward model for follicle selection: as follicles become more advanced in their development, the amount of estradiol and inhibin that they produce increases, resulting in a decline of FSH levels to concentrations that are insufficient to support the growth of the less-advanced, most gonadotropin-dependent follicles [3]. The more-developed antral follicles are able to survive both by sequestering the little FSH that is available [4] and by shifting dependence to luteinizing hormone (LH) [5, 6]. The less-developed follicles are pushed down the atretic pathway, reducing the number of follicles in the growing cohort—possibly to the desired ovulatory number in the mouse and other multiparous mammals. This mechanism is likely to form the broad framework by which ovulatory number is regulated, but it fails to adequately explain how developmental differences between follicles arise in the first place. One additional method by which follicle dominance could be regulated is via direct follicle–follicle interactions. Contact-dependent communication between equivalent follicles could select a dominant population from among the growing cohort [7]. Other interfollicular processes may then maintain these differences [8], either independently or in conjunction with this mechanism, holding nonselected follicles in a retarded stage of development until the declining FSH concentration pushes them down the atretic pathway.
During earlier experiments in which "follicle dominance" was established in cultures via a contact-dependent mechanism, subordinate follicles remained healthy throughout the culture period, despite exhibiting markedly retarded growth and development. Furthermore, subordinate follicles were capable of increased growth rates if the dominant follicle was removed [9]. These experiments were carried out with high levels of FSH in the culture medium, and the high FSH concentration may have masked any atresia-inducing effect of the dominant follicles.
The experiments described here investigate the effect of lowering FSH levels during culture on both dominant and subordinate cocultured follicles. First, we characterize the response of individual follicles at different developmental stages to a decline in FSH level. To our knowledge, this has not previously been shown in the mouse, nor has it been shown in any species in vitro. Second, we show that when pairs of follicles with an initial size difference are cocultured, the larger follicle becomes dominant over the smaller one, and we use this paradigm to examine the maintenance of a dominant/subordinate relationship in vitro. Finally, we test the hypothesis that when dominant follicles inhibit the development of subordinate follicles, they also render them increasingly susceptible to a lowering of FSH concentrations.
MATERIALS AND METHODS
Media and Chemicals
The Sigma Chemical Company (St. Louis, MO) supplied all chemicals with the exception of those noted here. 
-Minimum essential medium (-MEM) and Liebovitz L-15 dissection medium was obtained from Gibco BRL (Renfrew, UK). Recombinant human (rh) FSH and rhLH were supplied by Serono (Geneva, Switzerland).
Animals
Three-week-old CBA/C57Bl mice (F1) were used for all culture experiments. They were housed under environmentally controlled conditions (14L:10D) with food and water ad libitum.
Standard Follicle Culture
A whole-follicle culture technique was used, as described previously, that supports the growth of follicles in a highly physiological manner, including allowing the oocytes obtained at the end of culture to be fertilized and to develop to live young [10]. Briefly, late preantral follicles (160–200 µm in diameter) were manually dissected out of ovaries using acupuncture needles in Liebovitz L-15 medium supplemented with 0.3% (w/v) bovine serum albumin under aseptic conditions. Follicles were cultured in -MEM supplemented with 5% (v/v) mature F1 mouse serum, 0.01 IU ml-1 of rhLH (except in experiment 2) and rhFSH at various concentrations as indicated below. Incubations were at 37°C, 5% CO2, with high humidity. When indicated, early antral follicles (
230 µm in diameter) were also dissected for culture.
Apoptotic DNA Fragmentation Analysis
Follicles were snap-frozen on dry ice in microcentrifuge tubes. Samples were lysed with 100 µg ml-1 of Proteinase K in buffer containing 0.5% SDS, 0.1 M NaCl, 0.05 M Tris (pH 8.0), and 2.4 mM EDTA at 55°C (5–6 h). The DNA was isolated by chloroform extraction and isopropanol precipitation before washing with 80% ethanol and resuspension in double-distilled (dd) H2O. Absorbance at 260 nm was used to calculate DNA concentration. A digoxygenin (DIG) 3' end-labeling kit was used to tag the extracted DNA strand breaks (Roche, Lewes, UK), which were then cleaned and resuspended in ddH2O, both according to the manufacturer's instructions. Labeled samples were fractionated through 2% agarose gels by electrophoresis at 75 V. The DNA was Southern blotted overnight onto positively charged nylon membranes and subsequently baked at 120°C for 30 min. The DIG-labeled DNA bands were detected using the DIG-detection kit (Roche) according to manufacturer's instructions. The color-reacted membrane was scanned using an optical densitometer (Bio-Rad, Hemel Hempstead, UK), and the band density calculated using Molecular Analyst (Bio-Rad) software. The density of the bands in each experimental group was calculated relative to the 1 IU ml-1 of FSH (control) group, which was assigned an arbitrary value of one.
Terminal Transferase-Mediated dUTP Nick End-Labeled Analysis of DNA from Whole Follicles Using Confocal Microscopy
Follicles were washed in PBS (pH 7.2–7.4) at 37°C (10 min), transferred into 0.5% Triton X-100 and 0.25% paraformaldehyde in PBS (37°C, 40 min), fixed in 4% paraformaldehyde (30 min), and then washed twice in PBS (10 min each wash). At this point, follicles could be stored in PBS with 0.02% sodium azide at 4°C. After removal from storage, follicles were washed in PBS (10 min), transferred into 17.1 µg ml-1 of Proteinase K at 37°C (30 min), washed in PBS with 0.01% Triton X-100 (10 min), fixed in 3% paraformaldehyde (30 min), and then washed twice in PBS (20 min each wash). Following 10 min of preincubation in terminal deoxynucleotidyl transferase buffer (30 mM Tris-HCl [pH 7.2], 140 mM sodium cacodylate, and 1 mM cobalt chloride), each follicle/double-follicle unit was placed in a commercially available terminal transferase-mediated dUTP (2'-deoxyuracil 5'-triphosphate) nick end-labeled (TUNEL) reaction mixture (Roche) for 2.5 h. Follicles were then washed in PBS (10 min) and moved into RNase buffer containing 200 µg ml-1 of DNase-free RNase and 2.5 µg ml-1 of propidium iodide (1 h), washed in PBS containing 0.01% Triton X-100 (20 min), and followed by two washes in PBS (20 min each wash). To preserve fluorescence, the follicles were equilibrated in 50% Vectashield (Vector Laboratories Ltd., Burlingame, CA). They could be stored overnight at 4°C at this stage or transferred into 100% Vectashield on a concave microscope slide, cover-slipped, and sealed with nail polish for microscopic analysis. Follicles were examined using the Leica TCSNT Confocal system (Leica Microsystems, Milton Keynes, UK) using a 63x, water-corrected, PL APO lens. A single scan was taken through the center of each follicle as determined by central positioning of the propidium iodide-stained germinal vesicle in the oocyte. Simultaneous scans at 488 and 568 nm were taken to produce an amalgamated, true-color RGB image. Each channel could also be viewed separately. Four accumulations were taken for each image saved, which averaged the fluorescent signal and removed electronic noise to produce a sharper image. Images were saved for later analysis.
Analysis of Confocal Images
Each entire follicle cross-section was saved as a mosaic of images. Looking at each image in turn, the complete number of cells sharply in focus in the cross-section was counted. The number of TUNEL-labeled cells that were in focus was also counted. In this way, the proportion of apoptotic to healthy cells could be calculated. Analysis was performed on a Gateway 2000 computer (Sioux City, SD) using the free UTHSCA ImageTool program (developed at the University of Texas Health Science Center at San Antonio, TX, and available on the Internet by anonymous FTP from ftp://maxrad6.uthsca.edu).
Experiment 1: Effect of Varying FSH Concentrations on Individually Cultured Follicles at the Early and Late Antral Stages
Follicles were cultured singly for either 2 or 5 days. As described above, all follicles were cultured in medium with 0.01 IU ml-1 of rhLH. The 2-day cultured follicles were immediately placed in medium with either control levels (1 IU ml-1) or varying FSH concentrations (0.5, 0.25, 0.1, and 0 IU ml-1). At the end of the 2-day culture period, follicles had developed to the early antral stage; these follicles were termed the "early antral" group (treatment 1). In the longer-term cultures, follicles were maintained at high levels of FSH (1 IU ml-1) for the first 3 days of culture, during which time they developed to the midantral stage. They were then maintained at control levels (1 IU ml-1) or transferred to varying levels of FSH (0.5, 0.25, 0.1, and 0 IU ml-1) on Day 3 for the last 2 days of the culture period, during which time further growth and development occurred. These follicles were termed the "late antral" group (treatment 2). At the end of culture, apoptotic DNA fragmentation analysis was performed according to the protocol outlined above.
Experiment 2: Coculture Paradigm to Investigate the Maintenance of Dominance In Vitro
Early antral follicles (230 µm in diameter; "large/dominant follicles"—treatment 1) and late preantral follicles (160 µm in diameter; "small/subordinate follicles"—treatment 2) were dissected individually. Cocultured pairs of follicles that had been separately dissected from ovaries were set up with one large/dominant follicle in contact with one small/subordinate follicle, whereas singly cultured follicles of each size provided controls ("large control" and "small control" follicles, respectively). Control and experimental follicles were cultured during a 4-day period in standard medium supplemented with 1 IU ml-1 of rhFSH. Follicles were measured every other day, with an average obtained from two cross-sections of follicle diameter (taken at 90° to each other).
Experiment 3: Effect of Follicle–Follicle Interactions and Varying FSH Concentrations on Follicle Development
Different-sized follicles (230 and 160 µm in diameter) were dissected and placed in contact in the culture dish, as described for experiment 2. Cocultured dominant/subordinate pairs were cultured for 3 days in standard culture medium containing 1 IU ml-1 of rhFSH, then for a further 48 h in either 1 IU ml-1 (high FSH) or 0.1 IU ml-1 (low FSH) of rhFSH. The concentration of 0.1 IU ml-1 of rhFSH was chosen as the low level of FSH, because the results of experiment 1 showed that both early and late antral follicles were significantly affected by lowering FSH concentrations to this level. Thus, four treatment groups were used in this experiment: "dominant/high-FSH" (treatment 1), "dominant/low-FSH" (treatment 2), "subordinate/high-FSH" (treatment 3), and "subordinate/low-FSH" (treatment 4). Single follicles, dissected at 230 µm in diameter and cultured in 1 IU ml-1 of rhFSH throughout, provided a control for the dominant/high-FSH follicles ("large controls"). Although the subordinate follicles had been in culture for 5 days, at the end of the culture period they were only at the late preantral/early antral stage, with just small patches of antral cavities. The control group for these follicles was, therefore, composed of single follicles dissected at 160 µm in diameter but cultured in 1 IU ml-1 of rhFSH for 2 days only, by which time they were at the same stage of development ("small controls"). As described for experiment 1, all follicles were cultured with 0.01 IU ml-1 of rhLH throughout. At the end of the culture period, follicles were fixed and prepared for whole-follicle TUNEL analysis using the confocal microscope. This technique was used for analysis of the double follicles rather than the "laddering" analysis used in experiment 1, because separating the dominant and subordinate follicles is technically difficult and because it is not possible to be certain that the cell populations have remained distinct.
Statistical Analysis
Statistical significance between mean values was determined by ANOVA, followed by Student t-test analysis. The chi-square test was used to compare follicles between different treatments that had been analyzed using the TUNEL-labeling protocol.
RESULTS
Experiment 1: Effect of Varying FSH Concentrations on Individually Cultured Follicles at the Early and Late Antral Stages
Figure 1 shows the increase in apoptosis in treatment 1 (early antral group) (unhatched bars) and treatment 2 (late antral group) (hatched bars) as the FSH concentration in the culture medium was decreased relative to the 1 IU ml-1 of rhFSH control groups. In each case, the density of the control group cultured in 1 IU ml-1 of FSH throughout was assigned an arbitrary value of one, so that follicles cultured in varying levels of FSH could be compared directly with their respective high-FSH controls. However, the absolute levels of laddering in the early and late antral control groups differed; thus, the level of apoptosis could not be compared across experiments. When each experimental concentration was compared with its control (same-sized follicles cultured at 1 IU ml-1 of FSH throughout), the 0.1 and 0 IU ml-1 groups showed significantly higher levels of apoptosis compared to their respective control groups in both treatment 1 (early antral group) and treatment 2 (late antral group) (early antral group: P < 0.05 and P < 0.01, respectively; late antral group: P < 0.05 and P < 0.001, respectively). However, the percentage increase in apoptotic ladder density was much greater in treatment 1 (early antral group) (150% in the 0.1 IU ml-1 of rhFSH and 310% in the 0 IU ml-1 of rhFSH group) than in treatment 2 (late antral group) (40% in the 0.1 IU ml-1 of rhFSH and 70% in the 0 IU ml-1 of rhFSH group). Thus, a much less marked increase was observed in the degree of apoptosis as FSH levels were decreased in treatment 2 (late antral group) compared to when FSH was decreased in treatment 1 (early antral group). No significant increase was found in the incidence of apoptosis compared with that in the respective control groups as the result of culture in 0.5 or 0.25 IU ml-1 of rhFSH for either treatment 1 (early antral group) or treatment 2 (late antral group).
|
Experiment 2: Coculture Paradigm to Investigate the Maintenance of Dominance In Vitro
The growth of contacting, different-sized follicles during 4 days of coculture and that of their singly cultured controls is shown in Figure 2. No difference was found in growth between treatment 1 (large/dominant follicles) (cocultured in contact with the smaller follicles) and the singly cultured large controls. When compared with the small controls, treatment 2 (small/subordinate follicles) (cocultured in contact with the larger follicle) exhibited significantly retarded growth (P < 0.01). These results illustrate that a dominance effect can be observed in vitro when follicles are cultured in contact with differences in starting size, with the larger follicle at the start of culture exerting dominance over the smaller one and inhibiting its growth. This experimental paradigm was then used in experiment 3.
|
Experiment 3: Effect of Follicle–Follicle Interactions and Varying FSH Concentrations on Follicle Development
Figure 3 shows examples of treatment 1 (dominant/high-FSH) and treatment 3 (subordinant/high-FSH) cocultured follicles cultured for 5 days in 1 IU ml-1 of FSH (Fig. 3A), treatment 2 (dominant/low-FSH) and treatment 4 (subordinate/low-FSH) co-cultured follicles cultured for 3 days in 1 IU ml-1 of rhFSH and then for a further 2 days in 0.1 IU ml-1 of rhFSH (Fig. 3B), and single follicles cultured in 1 IU ml-1 of FSH either for 5 days (large control follicle) (Fig. 3C) or for 2 days (small control follicle) (Fig. 3D). Cells showing green fluorescence have a high incidence of DNA fragmentation, which is likely to be apoptotic. As can be seen in Figure 3, A–C, cocultured follicles rapidly develop an extensive and extremely close area of contact, with only a few thecal cells between the follicles. This architecture occurred in all pairs of follicles cocultured in this manner.
|
The large control follicles in 1 IU ml-1 of FSH, such as that shown in Figure 3C, contain apoptotic cells; by far the majority of these are found among the granulosa cells surrounding the antral cavities of all antral follicles, including freshly dissected healthy antral follicles (results not shown). In contrast, it was much rarer to find apoptotic cells in the small control follicles, such as that shown in Figure 3D, which had a significantly lower incidence of apoptosis than the large controls (P < 0.05).
Dominant follicles maintained in high levels of FSH throughout the experiment (treatment 1—dominant/high-FSH group) had significantly higher levels of apoptotic cells than the subordinate follicles maintained in high levels throughout (treatment 3—subordinate/high-FSH; 8.9 and 2.2 apoptotic cells per 1000 cells, respectively; P < 0.01) (Fig. 4). As can be seen in Figure 3B, this is because the dominant follicles, at a later stage of follicle development and with a much larger antral cavity, have apoptotic cells around the antral cavity. These same cells can be seen around the antral cavity of the control follicle (Fig. 3C), and no significant difference was found between the level of apoptosis in treatment 1 (dominant/high-FSH) follicles and that in large control follicles. The level of apoptosis in the dominant/high-FSH follicles is, therefore, not an indication of poor follicle health but, instead, is a normal phenomenon in antral follicles.
|
Treatment 3 (subordinate/high-FSH group), at an earlier stage of follicle development and with only small, if any, antral cavities, have a lower baseline of apoptosis. However, no significant difference was found between the level of apoptosis in treatment 3 and that in the small control follicles (Fig. 4). Thus, it also appears that the subordinate/high-FSH follicles are healthy, with a typical level of apoptosis for follicles at that developmental stage.
Both dominant and subordinate follicles reacted to the reduction in FSH levels with an increase in the number of apoptotic cells per 1000 cells (P < 0.001 for both). The incidence of apoptosis in treatment 2 (dominant/low-FSH) and treatment 4 (subordinate/low-FSH) was similar (17.5 and 18.7 apoptotic cells per 1000 cells, respectively). However, this level represents only a relatively small increase in the level of apoptosis in the dominant follicles (from 8.9 to 17.5 apoptotic cells per 1000 cells, an increase of 97%) but a massive increase in the level of apoptosis in the subordinate follicles (from 2.2 to 18.7 apoptotic cells per 1000 cells, an increase of 750%) (Fig. 4). With the presence of apoptotic cells in healthy late preantral/early antral follicles being such a rare event, an increase of 750% is likely to indicate poor follicle health in treatment 4 (subordinate/low-FSH group). A chi-square test showed that subordinate follicles were more susceptible to reduced levels of FSH than the cocultured dominant follicles (P < 0.05). Thus, the subordinate follicles are rendered particularly susceptible to a fall in FSH by the presence of the dominant follicle.
DISCUSSION
The results from these experiments clearly show that late preantral/early antral follicles in vitro have a requirement for FSH as a survival factor and undergo apoptosis as concentrations of this gonadotropin are lowered—despite the continued presence of low levels of LH in the culture medium. When follicles were maintained in high levels of FSH until they had developed to the midantral stage and were then transferred to varying levels of FSH, the experiments showed that these mid to late antral follicles also have a requirement for FSH and become increasingly apoptotic as it is withdrawn. However, they display a markedly lower increase in apoptosis relative to their controls than do the less-developed preantral/early antral follicles. This could be as a result of their ability to transfer their gonadotrophic dependence from FSH to LH [11], responding to LH via granulosa LH receptors [5, 6]. These findings support those of previous work, which demonstrated that the gonadotrophins are effective inhibitors of apoptosis in cultured rat antral follicles [12].
Given the hypothesis that LH dependence is the principle way in which late antral follicles are able to withstand the decline in FSH levels in vivo, it is perhaps surprising that any increase in apoptosis is detectable in these more-developed follicles. Several factors could explain this. First, the DIG 3' end-labeling assay for apoptosis is very sensitive and able to detect very small changes in DNA laddering. Second, the follicles in these cultures were transferred into the experimental FSH concentrations after 3 days in culture with 1 IU ml-1, at which point the follicles would have been at the early to mid stages of antral development. Although LH receptors should start being manufactured on granulosa cells at this developmental stage [13], they might not be present in sufficiently high numbers to allow the follicles to transfer their dependence to LH as the FSH concentration is lowered. In addition, at least 10 follicles are pooled for each ladder analysis, and these follicles may be at different stages of maturity. It would only take one less-developed, more acutely FSH-dependent follicle within that group to increase the combined ladder intensity for the whole group.
This series of experiments investigating the responses of singly cultured follicles to their FSH environment has shown that follicles are developmentally competent to undergo apoptosis-mediated cell death from the preantral/early antral stage onward, at least in vitro, substantiating previous reports [14]. Furthermore, LH in the culture medium is likely to provide a survival factor that follicles may be able to respond to, depending on their developmental stage. These findings tie in with the gonadotrophin hypothesis for follicle selection, and they provide the groundwork for investigating the combined action of follicle–follicle communication and the systemic environment.
Earlier work [9] had shown that where two similarly sized follicles were cultured in contact, a dominant/subordinate relationship developed during the culture period, with only the dominant follicle growing well in coculture. In that work, it was not possible to predict at the start of the culture period which follicle in the cocultured pair would become dominant, and both follicles would grow well if cultured singly. It can, therefore, be considered that the pairing of two like-sized follicles in coculture encompasses both the initiation and the maintenance of follicle dominance in vitro. When the results of this experiment and those of Spears et al. [9] are considered together, it appears that the establishment and maintenance of dominance are two distinct and separable events, each of which can be studied in isolation. Thus, it is possible to bypass the initial establishment of dominance in culture, by pairing a large and a small follicle which had been individually dissected from ovaries, together in contact. With this paradigm, the larger follicle will assume a dominant role over the smaller follicle and inhibit its growth (maintenance of dominance only). In both this and the previous experiments [9], cocultured follicles rapidly developed an extensive area of contact between the follicles, with the whole two-follicle unit becoming spherical and with a remarkably thin area of thecal tissue between the two follicles. This phenomenon is found in all follicles that are cocultured in this way. Preantral follicles are also found with similarly close associations in vivo in many species [7], and it would be interesting to examine how this is developing in vitro. The method described here has the advantage that the dominant cocultured follicle grows as well as singly cultured control follicles over the culture period, whereas in the original method, the growth of the dominant follicle was inhibited compared to that of singly cultured control follicles. We have used this method of investigating the maintenance of dominance to examine the relative responses of dominant and subordinate follicles to a fall in FSH levels. Our analyses involved the development of a whole-follicle TUNEL-labeling protocol with confocal microscopic visualization, which has provided a powerful tool for rapid follicle analysis. To our knowledge, the literature currently contains no reports documenting the use of this protocol for cultured ovarian follicles, although a similar technique has been reported for the analysis of blastocysts [15]. The use of this method for analyzing cultured follicles offers a rapid, accurate, and relatively simple indication of follicle health. The TUNEL-labeling protocol was particularly useful for the analysis of cocultured follicles, which would be difficult to separate; additionally, it provides information about the spatial location within the follicle of apoptotic cells.
Dominant and subordinate follicles have a different baseline in the level of apoptosis in the granulosa cells when cultured in high levels of FSH, with dominant follicles containing a higher proportion of apoptotic cells than the subordinate ones. It has been reported that preantral/early antral follicle cells are capable of undergoing apoptosis in vitro [14], but this may happen less readily than in more mature follicles. Apoptosis is a normal process of cell elimination in most healthy tissues [16]. Large ovarian follicles have more cells than less-developed follicles, many of which may play different roles in the follicular syncytium. Consequently, a greater requirement exists for strict follicular organization, including the elimination of unhealthy, inappropriately located, or unwanted cells. Figure 3C shows a healthy single follicle cultured in high levels of FSH; several granulosa cells that have undergone cell death can be clearly seen bordering the antral region. This is a frequent observation and appears to be a normal developmental process, because the same pattern can be seen in freshly dissected, healthy antral follicles (unpublished observation). This may account for a higher baseline of apoptosis in the more developmentally advanced follicles, such as the dominant follicles in the experiments presented here. We have used the term "apoptosis" when reporting the results of our TUNEL assay, but it is important to note that the reaction detects DNA fragmentation, which can also occur in forms of cell death other than apoptosis. Van Wezel et al. [17] describe the death of granulosa cells in the antral region as being more consistent with cells that die as a consequence of terminal differentiation rather than apoptosis.
Of interest here was the proportional increase in the incidence of apoptosis in follicles. Follicles cultured in low FSH levels were compared with "normal" follicles maintained in high FSH levels at the same developmental stage. A comparison between the response of subordinate and dominant cocultured follicles to a lowering of FSH levels showed that subordinate follicles exhibited an 750% increase in the proportion of apoptotic cells, compared to a 97% increase in the dominant follicles. This marked increase in apoptosis in the response of subordinate follicles to a lowering of FSH levels was the key result in this series of experiments, indicating that, in vitro, dominant follicles not only inhibit the growth of subordinate follicles, they also induce increased susceptibility to the apoptotic-inducing effect of lowered FSH levels. In summary, contact with a dominant follicle, in addition to FSH withdrawal, has significant ramifications for subordinate follicle health.
The findings of these experiments offer support to the hypothesis that contact-mediated communication between follicles may play a role in follicle selection, both by holding back the development of smaller follicles and, in addition, by rendering them more vulnerable to systemic changes in FSH concentration.
Much of the research into follicle dominance to date has been focused on endocrine regulation. Whereas endocrine control of follicle dominance can explain many of the later processes that occur, such as ensuring that subordinate follicles become atretic, intraovarian mechanisms also exert an important influence [7]. The best studied of these is the way in which differences that emerge between dominant and subordinate follicles can bring about the subsequent atresia of the subordinate follicles. Bioavailability of insulin-like growth factors (IGFs) [18–20], changes in the secretion of IGF-binding protein [21, 22], and differences in the production of estradiol [6, 23], among many other factors, are all likely to play an important role. Such studies do not, however, address how these differences could first have arisen in a species-specific number of dominant follicles.
Less studied is the variety of ways in which direct follicle–follicle interactions could influence follicle fate in the first place. Such interactions could enhance the effect of endocrine regulation of follicle dominance, with dominant follicles producing factors that directly influence the development of subordinate follicles [7, 24]. In part, these interactions could be regulated by components of bovine and ovine follicular fluid shown to inhibit granulosa cell proliferation and differentiation: low molecular fractions of bovine follicular fluid inhibit proliferation of cow and rat granulosa cells in vitro [25, 26], whereas steroid-free extracts of ovine follicular fluid inhibit estrogen secretion and P450 aromatase expression [27].
Recently, several studies have used culture systems to examine follicle–follicle interactions. Some of these have concentrated on preantral follicle development. Mizunuma et al. [28] showed that antral follicles inhibited the in vitro development of mouse early preantral follicles via the secretion of activin, whereas Zhao et al. [29] demonstrated that rat preantral follicle development was stimulated by the coculture of other preantral follicles. Previous experiments from this laboratory illustrated that like-sized mouse follicles establish a dominant/subordinate relationship when cocultured in contact from the late preantral stage onward [9]. Further work revealed that close contact between follicles is common in vivo [7], offering support to the hypothesis that follicle–follicle contact plays a role in determining follicle fate.
In the experiments presented here, a large (dominant) follicle retarded the growth of smaller (subordinate) follicles when cocultured in contact. To create an in vitro paradigm that encompasses the combined influence of follicle–follicle interactions and a changing systemic environment, we cocultured different-sized, contacting follicles in falling concentrations of FSH. Our experiments have clearly illustrated a combined, negative effect of coculture and lowering of FSH concentrations on small (subordinate) follicles that is significantly greater than the effect of either factor alone. We believe that direct follicle–follicle interactions are likely to play an important role in the determination of follicle fate, perhaps underpinning broadly acting systemic influences. As such, these current experiments could be viewed as a starting point for a much larger investigation into this little-understood component of ovarian function.
ACKNOWLEDGMENTS
The authors would like to thank Vivian Allison for her assistance with follicle cultures and Linda Sharp for help with confocal microscopy. The rhFSH and rhLH were kindly supplied by Serono-Ares, Geneva, Switzerland.
FOOTNOTES
First decision: 18 December 2001.
1 Funded by the MRC. N.S. is a Royal Society University Research Fellow. 
2 Correspondence: N. Spears, Dept. of Biomedical Sciences, The University of Edinburgh, Hugh Robson Building, George Square, Edinburgh EH8 9XD, UK. FAX: 44 131 651 1706; norah.spears{at}ed.ac.uk
Accepted: April 30, 2001.
Received: November 15, 2001.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  A. Pesty, O. Broca, C. Poirot, and B. Lefevre The Role of PLC{beta}1 in the Control of Oocyte Meiosis During Folliculogenesis Reproductive Sciences, September 1, 2008; 15(7): 661 - 672. [Abstract] [PDF]
|
|
|
|
 |
|
 A. A. Murray, A. K.E. Swales, R. E. Smith, M. D. Molinek, S. G. Hillier, and N. Spears Follicular growth and oocyte competence in the in vitro cultured mouse follicle: effects of gonadotrophins and steroids Mol. Hum. Reprod., February 1, 2008; 14(2): 75 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A Fowler and N. Spears The cultured rodent follicle as a model for investigations of gonadotrophin surge-attenuating factor (GnSAF) production Reproduction, June 1, 2004; 127(6): 679 - 688. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Sun, I. Betzendahl, Y. Shen, R. Cortvrindt, J. Smitz, and U. Eichenlaub-Ritter Preantral follicle culture as a novel in vitro assay in reproductive toxicology testing in mammalian oocytes Mutagenesis, January 1, 2004; 19(1): 13 - 25. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Spears, S. Baker, V. Srsen, R. Lapping, J. Mullan, R. Nelson, and V. Allison Mouse Ovarian Follicles Secrete Factors Affecting the Growth and Development of Like-Sized Ovarian Follicles In Vitro Biol Reprod, December 1, 2002; 67(6): 1726 - 1733. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
