A theory of follicle selection: I. Hypotheses and examples

A theory of follicle selection is developed in which all follicles, ovulatory and atretic, inherit the same developmental plan for responding to circulating concentrations of estradiol and gonadotropins. In the model, this plan is represented by a maturation surface that determines the rate of follicle growth as a function of follicle maturity and circulating hormone concentration. Examples of maturation surfaces are constructed for single and multiple spontaneous ovulators. When model follicles are activated from the reserve pool at random times, spontaneous and coordinated cycles of maturation develop in which the number of ovulatory follicles is controlled within a small and predictable range. For a given maturation surface, this range is nearly independent of the number of follicles in the interacting population, their activation times, and their initial maturities. The theory is therefore consistent with Lipschütz's Law of Follicular Constancy. The theory can account for certain statistical effects that occur with age on the timing of ovulation and the control of ovulation number. Maturation surfaces are also constructed that exhibit spontaneous transitions from cyclic to steady anovulatory behavior. The theory predicts an important regulatory role for atretic follicles in controlling both the timing of maturation and the number of follicles that reach ovulatory maturity.     

Lacker's model: control of follicular growth and ovulation in domestic species

Lacker (1981) and Lacker & Akin (1988) developed a mathematical model of follicular maturation and ovulation; this model of only four parameters accounts for a large number of results obtained over the past decade or more on the control of follicular growth and ovulation in mammals. It establishes a single law of maturation for each follicle which describes the interactions between growing follicles. The function put forward is sufficient to explain the constancy of the number of ovulations or large follicles in a female as well as the variability of this number among strains or species and for either induced or spontaneous ovulators. According to the model, the number of ovulations or large follicles lies between two limits that are themselves simple functions of two parameters of the model. Moreover, Lacker's model exhibits interesting characteristics in agreement with results obtained by physiologists: in particular, it predicts that the number of ovulating or large follicles is independent of:

1. the total number of maturing follicles,
2. the process of recruitment of newly maturing follicles towards the terminal maturation (Poisson or other),
3. the form of the LH or FSH secretion curves as functions of the systemic level of oestradiol. The model further predicts that
4. selection and dominance of follicles result from the feedback between the ovary and the hypophysis through the interactions between follicles; these interactions are expressed by the maturation function of the model.
5. recovery from atresia is possible for a follicle: from decreasing, the rate of secretion of oestradiol may increase.
6. the revised model suggests a renewal of follicles during the sexual cycle, as “waves of follicular growth”.

Lacker's model is a model of strict dominance; it maintains a hierarchy of the follicles as soon as they start their final maturation to the ovulations as that is observed in bird or reptile ovary. Such a strict hierarchy is possible but it is probably not a general situation in all mammals. We therefore modified the maturing function of the follicle in order to make it compatible with the observations of physiologists: follicles always interact as in the initial model but they individually become old, the hierarchy of follicles can be modified with time and the largest follicles do not indefinitely grow as in the initial model.     

Intra-ovarian regulation of follicular development and oocyte competence in farm animals

In both mono-ovulatory species, such as cattle, and poly-ovulatory species, such as pigs, the interactions among extra-ovarian gonadotropins, metabolic hormones and intra-ovarian growth factors determine the continued development of follicles, the number of follicles that ovulate and the developmental competence of the ovulated oocyte. FSH and then subsequently LH are the main hormones regulating antral follicle growth in both mono- and poly-ovular species. However, a range of intra-ovarian growth factors, such as insulin-like growth factors (IGFs) and bone morphogenetic proteins (BMPs), are expressed throughout follicle and oocyte development and interact with gonadotropins to control follicle maturation. In addition, environmental factors such as nutrition, including both the amount and composition of the diet consumed prior to ovulation, can influence follicle development and the quality of the oocyte. Recent progress in our understanding has resulted in the development of diets that enhance oocyte quality and improve pregnancy rate in both pigs and cattle. In conclusion, despite some species-specific differences, similar interacting mechanisms control follicular development and influence oocyte quality.     

Ovarian follicle growth and consequences for fertility in sheep

Understanding the pattern of ovarian follicle development is seen as an important step leading to the development of techniques that maximise fertility in sheep. Repeated observations of the growth of individual follicles have led to the understanding that follicles develop in a wave-like pattern during the oestrous cycle, with two to four waves per cycle being the most common. The ease with which follicle waves are described seems to depend on the their frequency and the number of follicles per wave. There is evidence for the largest follicle(s) of a follicle wave inhibiting the development of other follicles; however, in some cases this is not apparent as other follicle waves emerge when a previous large, healthy follicle is still present. Follicle development can be manipulated using exogenous gonadotrophins or progestagens and these have been shown to alter the number or age profile of developing follicles. The ovulation of aged follicles in cattle clearly has a detrimental effect on fertility, but this relationship is less clear and seems to be less critical in sheep. Recent findings at the molecular level show that the bone morphogenetic proteins (BMP) and their receptors are critically involved in the control of ovulation rate, but fully understanding their mechanism remains to be described. This highlights the potential for the integration of molecular and physiological findings to better develop methods to manipulate follicle development and reproduction in sheep.     

Ultrasonography of the bovine ovary

A linear-array ultrasound scanner with a 5-MHz transducer was evaluated for the study of follicular and luteal status in heifers. The ovaries of five heifers were monitored daily until all heifers were examined for a period from three days before an ovulation to three days after the next ovulation. There was a significant difference among days for diameter of the largest follicle and second largest follicle and for the number of follicles 4-6 mm and >10 mm. Differences seemed to be caused by the presence of several 4- to 6-mm follicles in early diestrus, growth to an ostensibly ovulatory size and subsequent regression of a follicle during mid-cycle, and selective accelerated growth of the ovulatory follicle four days before ovulation. The corpus luteum became visible approximately three days after ovulation and was identifiable throughout the rest of the interovulatory interval. In two of the five heifers, the corresponding corpus albicans was identified for three days after the second ovulation. Two heifers were induced to superovulate and follicular growth was monitored. The results indicated that the follicles which ovulated originated from the population present when the superovulation treatment was initiated. The ultrasound instrument was judged effective for monitoring and evaluating ovarian follicles and corpora lutea in normal and superovulated heifers.     

Pattern of follicular growth and resumption of ovarian activity in post-partum beef suckler cows

The ovaries of 18 post-partum beef suckler cows were examined daily, using ultrasound, from Day 5 post partum until a normal oestrous cycle was completed. Periods of growth and regression of medium-sized (5-9 mm) follicles were identified before one medium follicle became dominant (single large follicle greater than or equal to 10 mm). The mean (+/- s.e.m.) number of days from parturition to detection of the first post-partum dominant follicle was 10.2 +/- 0.5. The first post-partum dominant follicle ovulated in 2/18 (11%) cows. The interval from calving to first ovulation (mean +/- s.e.m. = 35.9 +/- 3.3 days) was characterized by the growth and regression of a variable number (mean = 3.2 +/- 0.2; range 1-6) of dominant follicles. The maximum diameter of the dominant follicle increased as the cows approached first ovulation (P less than 0.05). Behavioural oestrus was not detected in 16/18 (89%) cows at first ovulation. Following first ovulation, the length of the subsequent cycle was short (mean = 9.7 +/- 0.5 days; range 8-15 days) in 14/18 (78%) cows and was characterized by the development and ovulation of a single dominant follicle. During oestrous cycles of normal length (mean = 20.6 +/- 0.5 days; range 18-23 days) one (N = 2), two (N = 7) or three (N = 8) dominant follicles were identified. The growth rate, maximum diameter or persistence of non-ovulatory dominant follicles before first ovulation or during oestrous cycles were not different (P greater than 0.05). These data show that, in beef suckler cows, follicular development and formation of a dominant follicle occur early after parturition and the incidence of ovulation of the first dominant follicle is low. The number of dominant follicles that develop before first ovulation is variable; first ovulation is rarely associated with oestrus and short cycles are common after first ovulation. It is concluded that prolonged anoestrus in post-partum beef suckler cows is due to lack of ovulation of a dominant follicle rather than delayed development of dominant follicles.     

The ability of subordinate follicles of the second follicular wave to become dominant is lost by day 15 of the estrous cycle in cattle

Generally, unilateral ovariectomy before a critical period in the latter part of the estrous cycle induces a transitory increase in plasma FSH, which causes subordinate follicles to develop and maintain ovulation rates characteristic of the species. A limiting period for subordinate follicles to assume dominance and from which ovulation occurs has not been shown for cattle. Growth and/or regression of subordinate follicles were characterized following removal of the dominant follicle at different days of the luteal phase of the estrous cycle in cattle in this study. In the mid-luteal phase (Day 13 or 15), the ovary with the dominant follicle of the second wave was ablated via unilateral ovariectomy; the corpus luteum also was removed. In the late luteal phase (Day 17 or 19), the dominant follicle was ablated with an ultrasonically guided 20 gauge needle. When the dominant follicle was removed on Day 13, the largest subordinate follicle of the second wave of follicular development became dominant and ovulation occurred from this follicle in 4 of 4 animals. However, when the dominant follicle was removed on Day 15, 17 or 19, a new wave of follicular development was induced in 14 of 15 animals. Moreover, the recovered subordinate follicle of the second wave of follicular development had similar growth characteristics to naturally occurring dominant follicles. In conclusion, the subordinate follicle in the second follicular wave in cattle retained the ability to become dominant, but this ability was lost by Day 15 of the estrous cycle. However, cattle then were able to maintain ovulation by developing a new wave of follicular growth.     

Matrix metalloproteinase-2 and -9 secretion by the equine ovary during follicular growth and prior to ovulation

Profound hormonally controlled tissue remodelling occurs in the equine ovary for follicle growth and development, and also for the alteration in follicle shape directed towards the ovulation fossa, the site where ovulation occurs. The aim of this study was to examine the spatial and temporal regulation of matrix metalloproteinases (MMP)-2 and MMP-9, important enzymes in tissue remodelling, during follicle growth, and ovulation. Using gelatin substrate zymography, we measured these MMPs in follicular fluid of large anovulatory follicles collected during spring transition, early dominant follicles (> 23 mm), and at oestrus in follicles approximately 3 days prior to ovulation, and post-hCG treatment when ovulation was predicted in approximately 4 h. The most abundant activity detected in follicular fluid was MMP-2, although there were no changes in secretion or activation in association with ovulation. The activity of MMP-9 was detected in lower amounts, with no changes prior to ovulation, although it decreased significantly (P < 0.05) post-hCG treatment. At oestrus, when different regions of the ovary were maintained in explant culture for 24 h, there were no significant changes in either MMP-2 or MMP-9 secretion by stromal tissues collected at the ovarian fossa, adjacent to the preovulatory follicle but away from the fossa, and a further site remote from the preovulatory follicle. Over this same time period, follicular progesterone (P < 0.01) and oestradiol (P < 0.05) increased significantly, although oestradiol tended to decrease after hCG administration. These findings indicate that MMP-2 and MMP-9 are not key acute regulators for the changes in follicle shape immediately prior to ovulation.     

Ultrasonic anatomy of equine ovaries

A linear-array ultrasound scanner with a 5-MHz transducer was evaluated for studying follicular and luteal status in mares, and the ultrasonic properties of equine ovaries were characterized. Follicular diameters were estimated in vivo and after removing and slicing six ovaries. Correlation coefficients between the two kinds of determinations were 0.91 for number of follicles >/=2 mm in diameter and 0.95 for diameter of largest follicle. The ovaries of five mares were examined daily until all mares had been examined from three days before an ovulation to three days after the next ovulation. There was a significant difference among days for diameter of largest follicle and second largest follicle and for number of follicles 2-5 mm, 16-20 mm, and >20 mm. Differences seemed to be caused by the presence of many 2- to 5-mm follicles during early diestrus, initiation of growth of large follicles at mid-cycle, selective accelerated growth of an ovulatory follicle beginning five days before ovulation, and regression of large nonovulatory follicles a few days before ovulation. In one of the five mares, the corpus luteum was identified throughout the interovulatory interval, and the corresponding corpus albicans was identified for three days after the second ovulation. In the other four mares, the corpus luteum was last identified an average of 16 days after ovulation or five days before the next ovulation. In a blind study, the location of the corpus luteum (left or right ovary) as determined by ultrasonography agreed with a previous determination of side of ovulation by palpation in 88% of 40 mares on days 0-14. In the remaining 12% and in all of 12 estrous mares, the location was recorded as uncertain. The ultrasound instrument was judged effective for monitoring and evaluating follicles and corpora lutea.     

The kinetics of pre-antral follicle development in ovaries of CBA/Ca mice during the first 14 weeks of life

Abstract The kinetics of ovarian follicle growth and death have been estimated in virgin inbred mice using a compartmental model and data obtained from differential follicle counts in histologically sectioned ovaries. The results showed that both growth and death rates are dependent on stage of development, defined by the compartments, and age, indicated in the model by step functions with transitions at 20 and 60 days of age. During the initial phase of postnatal ovarian development, large numbers of follicles disappeared from the non-growing reserve as a result of the combined effects of follicle death and recruitment into the growing population. The reduced death rate after 20 days led to a secondary peak in the numbers of follicles at intermediate stages. In contrast to these fluctuations, the number of large follicles, including pre-ovulatory types, were remarkably constant after this age and the rate of outflow stabilized at two to three follicles per day after an initially high value. This rate is sufficient for the normal ovulation rate in a 4-day oestrous cycle with a small surplus of follicles undergoing atresia. The rates of migration through successive stages of development decreased during ageing as large follicles began to emerge at the approach of puberty: this result may indicate that the recruitment of small growing follicles is influenced by a feedback effect.     

Apparent involvement of plasmin in early-stage follicle rupture during ovulation in medaka

Until recently, the role of the proteolytic system involving serine proteases in follicle rupture during ovulation in mammalian species has been a subject of controversy. We undertook the present study to examine whether proteases play a role in follicle rupture using the teleost medaka (Oryzias latipes) model. Various serine protease inhibitors, including a specific plasmin inhibitor, drastically reduced the rate of ovulation, as assessed by an in vitro ovulation assay, which was established for the fish. Biochemical, molecular biological, and immunological analyses demonstrated that plasminogen/plasmin was present in large follicles destined to ovulate. The active protease, plasmin, was detected in follicles approximately 3-7 h before the expected time of ovulation. Specific antibodies against the medaka plasmin light chain suppressed the ovulation rate of the follicles when antibodies were added to the medium during the period in which active plasmin was generated. This finding was an indication that a plasmin-like protease similar if not identical to plasmin plays a role in follicle rupture during ovulation in the medaka. Our data also indicate that this serine protease participates in the rupture for only a few hours prior to the activation of matrix metalloproteinase (Mmp)-mediated hydrolysis at ovulation. Based on our previous and current data, we propose a follicle rupture model involving two different proteolytic enzyme systems, serine protease and Mmp, in medaka ovulation. The current study is the first to provide evidence of the indispensable role of plasmin or a plasmin-like protease in the ovulation of a nonmammalian vertebrate species.     

The kinetics of pre-antral follicle development in ovaries of CBA/Ca mice during the first 14 weeks of life

The kinetics of ovarian follicle growth and death have been estimated in virgin inbred mice using a compartmental model and data obtained from differential follicle counts in histologically sectioned ovaries. The results showed that both growth and death rates are dependent on stage of development, defined by the compartments, and age, indicated in the model by step functions with transitions at 20 and 60 days of age. During the initial phase of postnatal ovarian development, large numbers of follicles disappeared from the non-growing reserve as a result of the combined effects of follicle death and recruitment into the growing population. The reduced death rate after 20 days led to a secondary peak in the numbers of follicles at intermediate stages. In contrast to these fluctuations, the number of large follicles, including pre-ovulatory types, were remarkably constant after this age and the rate of outflow stabilized at two to three follicles per day after an initially high value. This rate is sufficient for the normal ovulation rate in a 4-day oestrous cycle with a small surplus of follicles undergoing atresia. The rates of migration through successive stages of development decreased during ageing as large follicles began to emerge at the approach of puberty: this result may indicate that the recruitment of small growing follicles is influenced by a feedback effect.     

Differentiation of dominant versus subordinate follicles in cattle

Selection of a dominant follicle, capable of ovulating, from among a cohort of similarly sized follicles is a critical transition in follicular development. The mechanisms that regulate the selection of a species-specific number of dominant follicles for ovulation are not well understood. Cattle provide a very useful animal model for studies on follicular selection and dominance. During the bovine estrous cycle, two or three sequential waves of follicular development occur, each producing a dominant follicle capable of ovulating if luteal regression occurs. Follicles are large enough to allow analysis of multiple endpoints within a single follicle, and follicular development and regression can be followed via ultrasonographic imaging. Characteristics of recruited and selected follicles, obtained at various times during the first follicular wave, have been determined in some studies, whereas dominant and subordinate follicles have been compared around the time of selection in others. As follicular recruitment proceeds, mRNA for P450 aromatase increases. By the time of morphological selection, the dominant follicle has much higher concentrations of estradiol in follicular fluid, and its granulosa cells produce more estradiol in vitro than cells from subordinate follicles. Shortly after selection, dominant follicles have higher levels of mRNAs for gonadotropin receptors and steroidogenic enzymes. It has been hypothesized that granulosa cells of the selected follicle acquire LH receptors (LHr) to allow them to increase aromatization in response to LH, as well as FSH. However, LH does not appear to stimulate estradiol production by bovine granulosa cells, and the role of LHr acquisition remains to be determined. Recent evidence suggests a key role for changes in the intrafollicular insulin-like growth factor (IGF) system in selection of the dominant follicle. When follicular fluid was sampled in vivo before morphological selection, the lowest concentration of IGF binding protein-4 (IGFBP-4) was more predictive of future dominance than size or estradiol concentration. Consistent with this finding, dominant follicles acquire an FSH-induced IGFBP-4 protease activity. Thus, a decrease in IGFBP-4, which would make more IGF available to interact with its receptors and synergize with FSH to promote follicular growth and aromatization, appears to be a critical determinant of follicular selection for dominance.     

Association of uterine edema with follicle waves around the onset of the breeding season in pony mares

During spring transition, when estrus may be exhibited for prolonged periods, it is important for veterinarians and stud farm personnel to be able to predict whether a large follicle will ovulate or regress. It is thought that the presence of ultrasonically detectable uterine edema indicates that a follicle will ovulate, however, there is little evidence to support this. In the present study, 16 mares were regularly examined by transrectal ultrasonography to follow growth and regression of follicles from seasonal anestrus in February until second ovulation. Blood samples were collected daily for measurement of estradiol concentrations when a large ovarian follicle was present. Estrous-like uterine edema was detected during 7 of 11 (64%) anovulatory follicle waves, in 12 of 14 (86%) mares before their first ovulation, and in 100% of mares before their second ovulation. Uterine edema was first detected 43+/-6.7 days before first ovulation. Large anovulatory follicles tended to be present for longer periods of time than ovulatory follicles. Uterine edema was present for a significantly greater proportion of time in the presence of a large follicle at second ovulation than at first ovulation (P<0.05) or for anovulatory follicles (P<0.01). Peak plasma estradiol concentrations and mean plasma estradiol concentrations were significantly higher (P<0.001) when a dominant preovulatory follicle was present compared with a dominant anovulatory follicle, but there was no difference in estradiol concentrations between first and second ovulations. It was apparent, therefore, that uterine edema was not a reliable indicator of follicular steroidogenic competence, or of whether the follicle would ovulate.     

Ovarian follicle dynamics of female Greater Scaup during egg production

ABSTRACT.   Studies of female waterfowl nutrient reserve use during egg production require a precise understanding of ovarian follicle dynamics to correctly interpret breeding status, and, therefore, derive proper inference. Concerns over numerical declines of North American scaup have increased the need to better understand the role of female condition in reproductive performance. We quantified ovarian follicle dynamics of female Greater Scaup ( Aythya marila ) breeding on the Yukon–Kuskokwim Delta, Alaska, using a method that accounts for within day variation in follicle size. We considered several models for describing changes in follicle growth with the best supported model estimating the duration of rapid follicle growth (RFG) to be 5.20 ± 0.52 days (±95% confidence intervals) for each developing follicle. Average diameter and dry mass of preovulatory follicles were estimated to be 9.36 mm and 0.26 g, respectively, at the onset of RFG, and these follicle characteristics were 41.47 mm and 15.57 g, respectively, at ovulation. The average diameter of postovulatory follicles immediately following ovulation was estimated to be 17.35 mm, regressing quickly over several days. In addition, we derived predictive equations using diameter and dry mass to estimate the number of days before, and after, ovulation for pre- and postovulatory follicles, as well as an equation to estimate dry mass of damaged follicles. Our results allow precise definition of RFG and nest initiation dates, clutch size, and the daily energetic and nutritional demands of egg production at the individual level. This study provides the necessary foundation for additional work on Greater Scaup reproductive energetics and physiology, and offers an approach for quantifying ovarian follicle dynamics in other species.     

Relationship between more follicles in right than left ovary in recently born calves and right ovary propensity for ovulation in cattle

The mechanism for more frequent ovulation from right ovary (RO) than left ovary (LO) was considered by determining if RO of recently born calves had a propensity for more follicles. Rationale was from reports that RO in heifers has more 6-mm follicles before selection of the future ovulatory follicle as well as greater frequency of RO ovulation. Dimensions, weight, and number of follicles per ovary were compared between LO and RO in 10 Holstein calves (age, 1 to 7 days). Weight of an ovary was greater (P < 0.05) for RO (0.393 ± 0.04 g) than LO (0.355 ± 0.05 g). Follicles were delineated by translucency of follicular fluid from transmitted light. Follicles from 0.3 mm diameter (smallest identified) to 4.8 mm (largest present) were counted. Mean number of translucent antral follicles (8.1 ± 1.8 vs 5.3 ± 1.2 follicles) and means for follicle diameter, fluid volume, and surface area were each greater (P < 0.01) for RO than LO. Combined for all diameters (0.3–4.8 mm), the hypothesis was supported that more follicles are present in RO than LO in calves 1 to 7 days of age. Although follicle activity in the fetus has not been compared between LO and RO, more follicles in RO than LO in recently born calves is consistent with the concept that the propensity for RO ovulation is congenital.     

Bone morphogenetic protein-4 acts as an ovarian follicle survival factor and promotes primordial follicle development

The growth and development of follicles within the ovary are highly dependent on autocrine and paracrine signaling involving growth factors from granulosa cells, theca cells, stromal interstitial cells, and the oocytes. The growth factor bone morphogenetic protein-4 (BMP-4) and its receptor (BMPR-IB) have been detected in ovaries, and a mutation in BMPR-IB has been associated with abnormal ovulation rate. The objective of the current study was to examine the role that BMP-4 plays in the early stages of primordial follicle development. Ovaries from 4-day-old rats were placed into a whole-ovary organ culture system for 2 wk to investigate the effect that treatment with exogenous BMP-4 has on early follicle development. BMP-4-treated ovaries had a significantly higher proportion of developing primary follicles and fewer arrested primordial follicles than did untreated controls. This indicates that BMP-4 promotes primordial follicle development and the primordial-to-primary follicle transition. Ovaries were also treated with neutralizing antibody against BMP-4 to determine effects of removing endogenously produced BMP-4. Interestingly, ovaries treated with BMP-4 antibody were markedly smaller than controls. This was associated with a progressive loss of oocytes and primordial follicles, a progressive increase in cellular apoptosis, and an accompanying loss of normal ovarian tissue morphology over time. Immunocytochemistry localized BMP-4 protein to isolated stromal cell populations, selected stromal cells (i.e., pretheca cells) associated with developing primordial follicles, and the basement membrane of follicles. Ovaries were treated with BMP-4 and RNA collected after organ culture to determine whether BMP-4 signaling affects expression of other growth factors. Kit ligand and basic fibroblast growth factor expression was unchanged, but TGFalpha expression was decreased in whole ovaries. Taken together, these data suggest that BMP-4 plays an important role in promoting the survival and development of primordial follicles in the neonatal ovary.     

Laparoscopic investigation of the bovine ovary in the periovulatory phase of the cycle

Laparoscopy, in combination with a rapid radioimmunoassay for plasma-LH determination, has been used to predict and observe ovulation in heifers. Experiments on three animals with typical progesterone levels, LH levels and apex formation are described. Ovulation was observed from 25 h to 29 h after the beginning of LH rise and from 17 h to 19 h after LH peak. The LH peak lasted for 9 h to 11 h. Cow ovulation was observed and photographed. The preovulatory follicle, apex formation, ovulation and freshly ruptured follicles are illustrated. The results presented here demonstrate that laparoscopy could offer valuable diagnostic assistance in clinical veterinary medicine.     

A model for follicular selection and ovulation: lessons from superovulation

A model for selection of the preovulatory follicle during the normal ovarian cycle is proposed. During menstruation the concentration of FSH rises to a level high enough to "activate" a single small antral follicle (2-4 mm dia.) so that it can produce large amounts of oestradiol. As the follicle develops, the concentration of FSH is suppressed below this threshold level by the secretion of oestradiol and inhibin. The dominant follicle becomes increasingly sensitive to FSH so that it continues to develop in an environment which inhibits development of other follicles. Multiple ovulation can be achieved by extending the period during which the level of FSH remains above this threshold level (e.g. during treatment with clomiphene or gonadotrophins). Although multiple ovulation occurs when the gate is widened in this way, the follicles are never completely synchronous as they continue to grow at approximately the same rate. Current evidence suggests that ovulation occurs at random between the two ovaries in successive cycles and that the corpus luteum exerts an inhibitory effect on folliculogenesis by suppressing the secretion of FSH and LH. These observations are compatible with the hypothesis that while small antral follicles are recruited continuously, at all stages of reproductive life, selection of the dominant follicle requires the unique gonadotrophic environment which is only present in the early follicular phase. The follicle of the month is, therefore, selected by chance because it is at the right place at the right time.