| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Reproductive Biology, German Primate Centre, 37077, Göttingen, Germany
b The Reproductive Medicine Unit, Department of Obstetrics and Gynaecology, University of Adelaide, The Queen Elizabeth Hospital, Woodville, 5011, Adelaide, Australia
ABSTRACT
The common marmoset (Callithrix jacchus) belongs to the family Callitrichidae, the only anthropoid primates with a high and variable number of ovulations (one to four). An understanding of folliculogenesis in this species may provide some insight into factors regulating multiple follicular growth in primates. The aims of this study were to characterize in detail changes in the antral follicle population at different stages of the ovarian cycle, to characterize the marmoset FSH profile, and to relate cyclic changes in FSH to changes in follicle sizes and circulating estradiol concentrations. Fifty-five pairs of ovaries were collected (32 of which were at five distinct stages of the cycle) from adult marmosets, and antral follicles were manually excised and separated into four size groups. Daily urinary FSH and plasma estradiol and progesterone concentrations from Day 0 of the follicular phase to 2 days postovulation were measured in 22 marmosets using enzyme immunoassays. The FSH profile revealed two distinct peaks, on Days 2 and 6, during the 10-day follicular phase, with a marginal periovulatory increase on Days 9 and 10. Estradiol levels rose significantly (P < 0.05) above baseline (Days 1–4) on Day 5 and continuously increased to a peak on the day preceding ovulation (Days 8 and 9). Follicle dissection revealed a high (mean = 68) and variable (range, 14–158) total number of antral follicles >0.6 mm. The number of antral follicles significantly declined (P < 0.001) with age. The number of preovulatory follicles (>2 mm) was positively correlated with the number of antral follicles (P < 0.001) and tended to be negatively related to age (P = 0.06). The number of antral follicles did not vary significantly with stage of the ovarian cycle, although the follicle size distribution was cycle-stage dependent (P < 0.05). Follicles >1.0 mm appeared only in the follicular phase, and preovulatory follicles (>2.0 mm) appeared only at the end of the follicular phase (Days 7–9). The Day 2 FSH peak corresponded to emergence of a population of medium-size antral follicles, and the Day 6 peak was consistent with rising estradiol levels and appearance of the preovulatory follicles. These results suggest that some aspects of marmoset folliculogenesis are comparable to those in Old World primates, including the absence of multiple follicular waves and the appearance of an identifiable dominant follicle in the midfollicular phase. However, the midphase FSH peak, multiple dominant follicles, and abundance of nonovulatory antral follicles differ strongly from the pattern in Old World primates and humans. The findings are discussed in relation to the regulation of growth of multiple ovulatory follicles and provide the basis for further studies on factors influencing the dynamics of follicular growth and development in this species.
follicle, follicle development, follicle-stimulating hormone, ovary
INTRODUCTION
Our basic understanding of the regulation of mammalian folliculogenesis and ovulation rate comes largely from work with rodents and some domestic ruminants. These studies have revealed that the earliest stages of follicle growth and progression up to approximately antrum formation occur largely independently of gonadotropins and are under the control of local peptide growth factors [1]. However, LH and in particular FSH are obligatory for healthy progression of follicles beyond antrum formation, and these hormones are central in driving folliculogenesis through to ovulation. In most ungulates, rats, and primates, there is a small elevation in basal FSH secretion early in the follicular phase that recruits a species-specific number of follicles for further development [2]. The selected follicles appear to maintain dominance by secreting negative feedback factors suppressing circulating FSH levels. In mono-ovular species such as humans, rhesus monkeys, and cows, only one follicle is selected from the pool of small antral follicles, but in polyovular species such as pigs and rats, multiple follicles are selected and grow synchronously through to ovulation. Factors determining the species-specific number of selected follicles and determining ovulation rates are poorly understood, although studies with highly fecund breeds of sheep have provided some insight into the effects of negative feedback factors in determining multiple ovulations [3].
The New World callitrichid monkeys (marmosets and tamarins) present unique opportunities to study the dynamics of primate folliculogenesis because they are the only primates that routinely ovulate more than one follicle per cycle. They also have a highly variable ovulation rate, ranging from one to four follicles per cyle, and as such marmoset litter sizes typically vary in captivity from singletons to quadruplets [4]. The common marmoset (Callithrix jacchus) is the most widely used New World primate in reproductive research, and in general marmoset ovarian endocrinology is well characterized. Marmosets do not menstruate but have a 
28-day ovarian cycle consisting of an 8- to 10-day follicular phase and an 18- to 20-day luteal phase [5]. As in other primates, the follicular phase is characterized by low levels of circulating progesterone, rising levels of estradiol, and a prominent preovulatory surge in LH [6, 7]. However, despite nearly two decades of reproductive research using the marmoset, FSH profiles have not been comprehensively documented at any stage of the ovarian cycle.
A rudimentary understanding of folliculogenesis in the marmoset does exist, although details remain lacking. Marmoset ovaries contain a multitude of small antral follicles (0.6–1.5 mm), averaging approximately 80–90/ovary pair [8, 9] and ranging from 20 to 150. The factors influencing this variation are not yet known. From Day 7 of the follicular phase, typically two or three 2-mm follicles emerge from the pool of smaller follicles [8, 9] and go on to ovulate 3 days later at 3.5 mm [4, 10, 11]. By Day 7, these preovulatory follicles are morphologically and functionally distinct from the nonovulatory antral follicles, with higher levels of aromatase activity and progesterone secretory potential [12–15], down-regulated expression of the androgen receptor [16], and negligible levels of granulosa cell proliferation [17]. Nothing, however, is known of follicle development at stages other than the late follicular phase, including when the dominant follicles are selected and how long these follicles take to grow. It is also not clear why the marmoset has such a high number of small nonovulatory antral follicles and whether the highly variable number is of functional significance. Preliminary data have indicated that on Day 7 of the follicular phase the number of nonovulatory antral follicles is positively correlated with the number of preovulatory follicles [9]. This finding suggests that animals with fewer small antral follicles would have a lower ovulation rate and hence smaller litter sizes, but factors determining the total number of antral follicles and hence ovulation rate remain largely unknown.
In an effort to improve the basic understanding of folliculogenesis and the factors regulating this process in the marmoset, this study was conducted to 1) characterize in detail changes in the antral follicle population at different stages of the cycle, 2) analyze factors that may be related to follicle numbers, such as animal age and weight and ovary size and weight, 3) characterize the FSH profile throughout the follicular phase and the beginning of the luteal phase, and 4) relate cyclic changes in FSH to changes in follicle sizes and circulating estradiol concentrations.
MATERIALS AND METHODS
Animals
Fifty-five sexually mature captive-bred female common marmosets were used for this study. They were housed with intact male partners at either the German Primate Centre, Göttingen, or The Queen Elizabeth Hospital, Adelaide. This study was approved by local animal ethics committees and was conducted in accordance with the Guiding Principals for the Care and Use of Research Animals. Blood was collected from all animals once or twice weekly to monitor the ovarian cycle via plasma progesterone concentrations. Luteolysis and onset of a follicular phase were induced by i.m. injection of 0.8 µg of cloprostenol, an analogue of prostaglandin F2
(PGF2) (Estrumate; Pitman-Moore, Burgwedel, Germany), administrated 12–15 days postovulation (Day 0 = day of PGF2 administration). Plasma progesterone levels typically declined to <10 ng/ml within 24 h of PGF2 administration and remained low throughout the follicular phase. A subsequent rise above a defined threshold of 10 ng/ml [5] was detected 8–11 days after PGF2, indicating the occurrence and approximate timing of ovulation.
Collection and Preparation of Blood and Urine Samples
In animals used to monitor FSH, E2 and P4 profiles (n = 10; age range, 1.5–6.8 yr), blood and urine samples were collected daily between 0700 and 1000 h for a period of 13 days starting on the day of PGF2 administration. Of the 34 animals (mean age = 3.29 yr; range, 1.56–7.1 yr) used for characterization of the follicle distribution, 12 were chosen for daily blood sampling and morning urine collection throughout the follicular phase until the day of ovary collection to monitor endocrine changes in relation to follicle development. Ovaries from the remaining 21 animals (age range, 1.8–11.21 yr) were collected at various stages of the follicular phase from PGF2-regulated cycles. Blood samples (0.3–0.4 ml) were collected from the femoral vein in unsedated animals into heparinized syringes and centrifuged, and the plasma was frozen at -20°C for later use. Daily morning urine was collected on plastic mats positioned under each cage before lights were switched on. These samples were mixed with preservative (3% glycerol, v/v) and stored at -20°C until assayed. Creatinine content of each urine sample was measured [18] to compensate for differences in urine concentration and volume, and urinary hormone concentrations were calculated per milligram of creatinine. Urine was prepared for enzyme immunoassay (EIA) of FSH using a method previously described [19, 20]. Thawed urine samples were mixed with ammonium bicarbonate buffer (0.05 mol/L) containing benzamidine HCl (5 mmol/L) and phenylmethylsulfonyl fluoride (1 mmol/L) as protease inhibitors. Samples were immediately desalted on a 1.2- x 8-cm Sephadex G-25 column (PD 10; Pharmacia, Heidelberg, Germany) using ammonium bicarbonate (0.05 mol/L) as elution buffer. Urine samples were lyophilized and resuspended in 1 ml of PBS.
Follicle Dissection
To characterize changes in the follicle size distribution throughout the ovarian cycle, pairs of ovaries were collected at five different stages: during the follicular phase on Days 1 or 2 (n = 5 pairs of ovaries), Days 3 or 4 (n = 7), Days 6 or 7 (n = 10), and Days 8 or 9 (n = 7) and during the luteal phase on Day 12 postovulation (Day 22 after PGF2 administration; n = 5). Ovaries were collected in Leibovitz L-15 medium (37°C) with supplements, cleaned of excess tissue, and weighed, and the dimensions were measured using a calibrated ocular micrometer. Ovaries were mechanically dissected using ultrafine forceps and 28-gauge insulin needles in medium on a heated microscope stage. In contrast to livestock and Old World primate ovaries, marmoset ovaries are relatively nonfibrous, which allows for complete mechanical dissection of the ovaries and accurate excision of whole follicles with comparative ease. Typically, <5 follicles would be lost during dissection of a pair of ovaries containing 70 antral follicles. This approach has been utilized previously to obtain an accurate assessment of follicle populations in the marmoset [9]. Antrum formation in marmoset follicles occurs relatively late in folliculogenesis, at 0.50 ± 0.013 mm (mean ± SEM; unpublished results); therefore, for the collection of antral follicles a minimum size was arbitrarily set at 0.6 mm. All follicles >0.6 mm were excised. Dark follicles with uneven granulosa and theca layers were classified as grossly atretic and were excluded from this study (<10% of small antral follicles). Follicle diameters were then measured in two dimensions using an ocular micrometer and were separated into four groups according to the mean diameter: small antral (0.6–1.0 mm), medium antral (1.0–1.5 mm), large antral (1.5–2.0 mm), and preovulatory (>2.0 mm). Follicles >2.0 mm, which normally appear on Days 6–7 in the marmoset follicular phase, are defined as preovulatory follicles because they are functionally distinct [14] and will normally ovulate 3 days later [4].
EIA for FSH
Because of the small blood volume of the marmoset and the need for frequent sampling, insufficient plasma was available (100–150 µl/sample) for simultaneous measurement of FSH, estradiol, and progesterone. Because of the relatively large volume of plasma needed for FSH determinations (50–100 µl in duplicate) and associated matrix effects, follicular phase FSH profiles were determined in urine samples only. Urinary FSH (uFSH) was measured by EIA using a monoclonal antibody (Mab) against human FSH with a biotinylated FSH preparation as label. The antibody was characterized, and the specific FSH EIA was validated in detail by Rosenbusch et al [20]. The Mab 46.3h6.b7 (J. Dias, Wadsworth Center, New York State Department of Health, Albany, NY) was used at a final concentration of 3.12 ng 50 µl-1 well-1, equivalent to a dilution of 1:16 000. Highly purified human uFSH (Fertinorm, Serono, Unterschleissheim, Germany; 75 IU/10 mg FSH) was used as the assay standard and as label for the biotinylation. The uFSH was biotinylated according to a method previously described [21] and was used at a concentration of 25 ng 50 µl-1 well-1.
Assays were performed in 96-well microtiter plates (immuno Maxisorb F96; Nunc, Roskilda, Denmark) precoated with sheep anti-mouse immunoglobulin G for 24 h at 4°C (1 ng 300 µl-1 well-1) followed by addition of peroxidase free casein (0.5 mg 350 µl-1 well-1; I-Block; Serva/Tropix, Heidelberg, Germany) and incubated for 24 h at 4°C [20]. After three washes with 350 µl/well of washing buffer (2.72 mM NaCl, 0.16 mM Na2HPO4, 0.6 mM KCl, 0.03 mM KH2PO4 + 0.05% Tween 80), urine samples (50 µl/well), FSH standard (11.7–1500 mIU 50 µl-1 well-1), and biotin-labeled FSH (label) were added and incubated for 18–24 h at 4°C with the FSH antibody (50 µl/well). FSH standard and label were diluted in PBS (0.58 M Na2HPO4, 0.17 M NaH2PO4 x H2O, 0.68 M NaCl + 0.1% casein). The plates were washed three times with washing buffer prior to incubation with streptavidin-peroxidase (20 ng 150 µl-1 well-1) for 1 h at room temperature. After a final wash, 150 µl/well of substrate solution consisting of 25 ml of substrate buffer (100 mmol/L CH3COONa, pH 5.5, with 0.18 M citric acid) with 100 µl of CH2N2O x H2O2 (0.94 M) and 400 µl of 0.6% tetramethylbenzidine in dimethylsulfoxide were added. The plates were incubated for 45–60 min in the dark, and the reaction was stopped by adding 50 µl of 4 M H2SO4. Absorbance was measured at 450 nm using a reference filter of 630 nm. Within- and between-assay coefficients of variation were 10.3% and 12.4%, respectively. The minimum detectable concentration of FSH was 23.4 mIU/50 µl.
EIA for Estradiol
Estradiol (E2) was determined by extracting plasma (100 µl) twice with 10 volumes of diethyl ether. The combined ether phase (separated from the aqueous phase by snap-freezing in methanol/dry ice) was evaporated, and the extract was reconstituted in assay buffer (PBS, pH 7.2, 150 µl). Immunoreactive E2 was measured by EIA using an antiserum raised against estradiol-6-CMO-BSA (S. Klinger, St. Albans, Hertfordshire, UK) together with estradiol-17-glucuronide coupled to alkaline phosphatase as label. The antibody was highly specific, with a cross-reactivity of 1.2% with estrone and <1% for all other steroids tested. For E2 determination, duplicate 50-µl aliquots of samples and E2 standard (range, 0.49–125 pg/well) were combined with the label (50 µl) and antiserum (50 µl), mixed thoroughly, and incubated overnight at 4°C. After incubation, the plates were washed four times and blotted dry, 150 µl of phosphatase substrate (Sigma 104; Sigma, Deisenhofen, Germany; 20 mg/15 ml substrate buffer) was added to each well, and the plates were incubated for a further 45–60 min by shaking in the dark at room temperature before absorbance was measured at 405 nm. Sensitivity of the assay at 90% binding was 1 pg/well. Serial dilutions of plasma extracts from the follicular and luteal phase gave displacement curves parallel to that of the E2 standard. Intra- and interassay coefficients of variation were 6.4% and 10.2%, respectively. Extraction efficiency (mean ± SD), determined by the recovery of 3H-E2 (1000 counts/min) added to the samples prior to extraction, was 89.4% ± 3.7%.
EIA for Progesterone
Plasma progesterone was determined by EIA as previously described by Heistermann et al. [22]. Inter- and intra-assay coefficients of variation for the assays carried out in this study were 6.7% and 9.4%, respectively.
Statistical Analyses
One-way analysis of variance (ANOVA) or Kruskal-Wallis ANOVA on ranks (when normality tests failed) were used to assess differences throughout the ovarian cycle in total follicle numbers and in follicle numbers from each class, respectively. The relationships between follicle numbers and several animal parameters were assessed using linear, inverse, and logarithmic regression analyses. The most appropriate regression models for a given data set were determined using the curve estimation procedure (SPSS, Chicago, IL). Changes in hormone concentrations throughout the ovarian cycle were examined using repeated measures ANOVAs (E2, one-way repeated measures ANOVA; FSH, Friedman repeated measures ANOVA on ranks [normality test failed]), with comparisons between individual means examined using paired t-tests.
RESULTS
Total Number of Antral Follicles
Marmoset ovary pairs contained on average 68 antral follicles that were >0.6 mm. Mean antral follicle number did not vary significantly (P > 0.05) throughout the ovarian cycle, although numbers were lowest in the luteal phase and highest in the early follicular phase, 3–4 days after PGF2 administration (Fig. 1). The number of antral follicles did however vary substantially between individual marmosets, with numbers ranging from 14 to 158. This variability in follicle numbers was significantly related to animal age (P < 0.0001), with numbers declining (best described logarithmically) from an average of 87 in 1–3 yr olds to 30 in animals >9 yr of age (Fig. 2A). There was nevertheless substantial variation in follicle numbers among young postpubertal animals. The total number of antral follicles was unrelated to animal weight (P > 0.05; data not shown) but was positively correlated with the weight (P < 0.001; Fig. 2B) and volume (P < 0.001; Fig. 2C) of the ovary.
|
|
Cycle-Dependent Changes in the Follicle Size Distribution
The size distribution of antral follicles in five stages of the ovarian cycle is presented in Figure 3. Small antral follicles (
1.0 mm) represented the vast majority of follicles during all stages of the cycle and were the only size class in the luteal phase, although their mean number did vary significantly throughout the cycle (P > 0.05). However, the average number of follicles in all three size classes >1 mm was cycle-stage dependent (P < 0.05). Follicular growth beyond the 1 mm threshold seemed to be initiated at the start of the follicular phase, with the appearance of a small number (n = 4) of medium-size (1–1.5 mm) follicles on Days 1–2. The increase in total number of follicles seen on Days 3–4 was due to an increase in the number of both the small (n = 76) and medium (n = 19) antral follicles and by the appearance of small numbers of large antral follicles (n = 1.5). Follicles of preovulatory size were seen from Day 7 of the follicular phase, averaging 2.5 follicles/ovary pair (range, 1–4). Emergence of the preovulatory follicles in the late follicular phase was associated with falling numbers of small antral follicles (P > 0.05) and very few large nonovulatory follicles (one or two). The proportion of the total population represented by small antral follicles fell from 100% in the luteal phase progressively throughout the follicular phase (Days 1–2, 94%; Days 3–4, 80%; Days 6–7, 79%; Days 8–9, 77%). After Day 7 of the follicular phase, the total number of nonovulatory antral follicles (0.6–2 mm) was significantly correlated (P = 0.001; inverse regression) with the number of preovulatory follicles (Fig. 4), suggesting that a higher ovulation rate may be expected in animals with higher total numbers of antral follicles. There was also a strong tendency for a higher number of preovulatory follicles in younger animals (P = 0.062) and in females with a higher body weight (P = 0.080).
|
|
Cyclic Changes in Immunoreactive FSH and Steroid Secretion
Figure 5A shows the profile of uFSH throughout the follicular phase (mean ± SEM). One-way ANOVA of the complete data set revealed a significant variation in FSH content between cycle days (P < 0.01). The profile shows two clear peaks: the first in the early follicular phase (Day 2) and a second major peak on Day 6, with a third rise around the time of ovulation on Day 9 or 10. FSH levels on Days 2 and 6 rose significantly (P < 0.05) above the values on the respective preceding days and then fell again on the following day. The lowest mean level of FSH (<22 mIU/mg creatinine) was measured on Day 0. After achieving mean values of 57 mIU/mg on Day 2, FSH levels decreased (lowest mean value = 24.3 mIU/mg) and then increased to a maximum mean value of 87.6 mIU/mg on Day 6. Eleven of the 19 animals monitored up to this day showed maximum FSH values on Day 6. Decreasing levels on Days 7 and 8 preceded a third peak coinciding with the days when ovulation typically occurred (Day 9 or 10 after PGF2). A similar profile displaying three distinct FSH peaks was also obtained after normalizaing data to the day of the E2 peak, except that as would be expected, there was less variation about the mean during the later stages of the follicular phase (data not shown).
|
Follicular phase E2 concentrations in 10 animals are depicted in Figure 5B, showing a notable rise in the mid-late follicular phase. Mean E2 levels were low (70 pg/ml) until Day 4 of the cycle, after which there was a significant (P < 0.05) and progressive increase to maximum concentrations on Day 8 (mean = 205.6 pg/ml). From Day 8, E2 concentrations decreased continuously to reach baseline levels by Day 12 (85 pg/ml).
All animals responded to PGF2 with a marked decline in plasma progesterone concentrations to <10 ng/ml within 24 h (Fig. 5C). According to progesterone profiles, most animals ovulated between Days 9 and 11, although one animal ovulated on Day 6, explaining why mean progesterone values on Day 9 were slightly above the 10 ng/ml threshold. On average, ovulation occurred 9.4 days after PGF2 administration.
Relationships Between Number of Follicles and Hormone Levels
Based on the low number of cycles studied here, no relationship was found between FSH or E2 concentrations on any day of the follicular phase and either the number of preovulatory follicles or the total number of antral follicles (P > 0.05). Although from just seven observations, there was a strong positive tendency for animals with higher Day 6 FSH concentrations to have more preovulatory follicles (r = 0.713, P = 0.072).
DISCUSSION
In this study, we characterized changes in the in vivo ovarian antral follicle population in relation to stage of cycle, pattern of FSH, and other parameters potentially related to multiple follicular development in the marmoset. The results reveal some features of follicle population dynamics that appear to be unique among the primates studied to date and provide possible new insights into the factors influencing multiple ovulation.
This study demonstrates that marmosets have a high number of antral follicles at all stages of the cycle (averaging 68/ovary pair across all age groups) as compared with other primate species. Apparently, marmoset ovaries contain more antral follicles than do those of tamarins (mean = 20–30); tamarins are also polyovular and belong to the same family (Callitrichidae) [23, 24]. Ovaries of humans and of mono-ovular Old World primates such as rhesus monkeys contain considerably fewer antral follicles, averaging just 10–20 [25, 26]. It is unclear whether the high number of nonovulatory antral follicles is of functional significance, but it indicates that there are probably unusual features of the mechanisms regulating marmoset folliculogenesis. Perhaps the large pool of antral follicles in marmosets is related to their high level of fertility and reproductive efficiency [11]. Marmosets are highly fecund, typically producing two sets of twins or triplets per year in captivity [27]. Data from the present study indicate that animals with a higher number of antral follicles are likely to have a higher ovulation rate and hence a larger litter size. Although direct evidence for a functional relationship between size of the antral follicle pool and ovulation rate is lacking, the abundance of small antral follicles typically found in marmoset ovaries may nevertheless be a factor contributing to the high fecundity of this species.
The present study confirms that marmosets, like most other mammals, experience a decline in total number of antral follicles with advancing age. Numbers decline from around 90 in 2 yr olds to approximately 30 in animals >9 yr of age. This reduction in antral follicle population size with age is comparable to what occurs in tamarins [23] and humans [26]. The decline in follicle numbers in the marmoset is best described as a logarithmic rather than a linear function, demonstrating that the loss of follicles is greater in early postpubertal life than in later stages. Older marmosets with fewer nonovulatory antral follicles also have a tendency to have fewer preovulatory follicles, implying a lower ovulation rate. This finding contradicts that of Tardif et al. [28], who reported no relationship between ovulation rate and age in marmosets, although this discrepancy may be a consequence of differences in sampling techniques and the low number of subjects in that analysis (n = 11). The decrease in numbers of preovulatory follicles in older marmosets observed in the present study implies that these animals will have smaller litters in the second half of their reproductive life, before reaching senescence. To our knowledge, this information is not available for the marmoset. Captive marmosets are however capable of producing viable offspring up to at least 11 yr of age (personal observations). Tardif et al. [28] reported that the greatest source of variability of ovulation rate in marmosets is within females (i.e., between successive cycles) rather than between females. The variation in ovulation rate in individual marmosets is positively correlated with body weight, suggesting that reproductive output is related to energy availability [28]. The results of the present study confirm this relationship and suggest that it is most likely manifest in the later stages of follicular development; we were unable to demonstrate any correlation between body weight and total number of antral follicles within the size range available in our study population.
Although age (but not body weight or cycle stage) accounts for some of the variability between animals in total number of antral follicles, a substantial proportion of the variability remains unexplained. The absolute extremes in the number of follicles collected (13 vs. 158) were both from young (2 yr old) monkeys. Ovarian inactivity also cannot account for the variable number of follicles because all the animals used for this study displayed regular cycles. An understanding of the factors determining this variability may provide some insight into factors regulating natural variability in ovulation rate. Furthermore, it would provide an indication of the number of follicular cells and oocytes that can be expected from an animal for in vitro experiments. Currently, ovary size and weight are useful tools for estimating the number of follicles that can be collected.
This study has also documented for the first time changes in the follicle distribution throughout the ovarian cycle in the marmoset monkey. There was an abundance of small antral follicles at all stages of the cycle, although progression beyond 1 mm was cycle-stage dependent. Development of larger antral follicles (>1 mm) does not appear to occur during the luteal phase but seems to be initiated early in the follicular phase. By Day 3, there is a slight increase in total follicle number and a notable increase in medium-size antral follicles. No information on mechanisms of dominant follicle selection in callitrichids has been available, but because these animals typically have relatively short follicular phases and long luteal phases (e.g., luteal phase is 2–3 times longer than the follicular phase in natural cycles in the marmoset), selection was speculated to occur in the luteal phase. However, the data from the present study indicate the absence of follicular development beyond 1 mm during the luteal phase, suggesting that selection of the dominant follicles in the marmoset occurs early in the follicular phase. More complete examination of FSH and inhibin profiles and follicle sizes during the luteal period would be required to confirm this prediction. However, in natural cycles (non-PGF2-regulated cycles), where progesterone levels decline gradually over a period of several days, some degree of follicular growth above 1 mm might be predicted to take place toward the end of the luteal phase. It is likely that marmosets, like Old World primates, do not have follicular waves during the luteal phase, as is observed in some ruminants.
This study provided information on uFSH profiles throughout the follicular phase for the first time in the marmoset or any other New World primate. In comparison to Old World primates, the marmoset FSH profile is unusual in being characterized by three peaks: one on Day 2 of the follicular phase, a second on Day 6, and a third around the time of ovulation. This profile differs to some extent from the preliminary observations made for two marmosets by Rosenbusch et al. [20], in that the Day 6 FSH peak was less evident. Reasons for this discrepancy are unclear, and although two peaks are more typically seen in Old World primates (early intercycle rise and late preovulatory peak), three follicular-phase FSH peaks have been described following measurements of bioactive uFSH in the lowland gorilla [29]. The authors of the gorilla report proposed that the second FSH peak, 6–7 days before ovulation, was responsible for follicular selection (and equivalent to the intercycle FSH peak in women), whereas the initial peak, which was more variable in timing, was probably related to an ovarian priming function. Evidence on which to propose a function for the individual FSH peaks in the marmoset is not available, although the timing of the second (major) peak on Day 6 coincides with the time (Days 6–7) that follicles of ovulatory size (2 mm) can first be physically and functionally distinguished from the nonovulatory follicles.
As in other primates, the early rise in FSH is likely due to a loss of negative feedback effects of products of the corpus luteum as a consequence of PGF2-induced luteolysis. Marmoset corpora lutea secrete substantial amounts of E2 and inhibin and have progesterone concentrations 10- to 30-fold higher than those in Old World primates [5, 22, 30], although much of the inhibin is apparently secreted as the free -subunit [31]. Nevertheless, it would seem reasonable to assume that the rise in FSH on Day 2 is a consequence of abrupt (PGF2 induced) luteal regression and that this rise is responsible for the initiation of growth of a new cohort of follicles (>1 mm). This idea is supported by data for Days 3–4, which show a rise in the total number of antral follicles and a notable increase in the population of those of medium size (1–1.5 mm). The fall in FSH concentrations on Day 3 may be related to increased inhibin B levels; follicles at this stage are known to express high levels of ßB-mRNA [32]. As in humans, aromatase activity in marmosets is FSH dependent, being demonstrable in granulosa cells of medium-size antral follicles (<1.5 mm) from around Day 6 [14]. Thus, the middle to second half of the marmoset follicular phase is characterized by growing presumptive preovulatory follicles and rising E2 levels. By Day 6 or 7, the total number of antral follicles has fallen again and the ovary is dominated by the presence of typically two or three large, distinct follicles. From studies using repeated laparoscopic [10] and ultrasound examination [4, 11], it is known that follicles that are >2 mm on Day 7 represent the preovulatory follicles, 90% of which will go on to ovulate 3 days later at a size of 3.5 mm. The factors responsible for the FSH peak on Day 6 are not clear, but the lower levels of FSH from Days 3–5 may lead to decreased inhibin and increased activin secretion by granulosa cells, overriding any tendency for early marginal increases in E2 levels to suppress FSH secretion. Increased follicular development by Day 7 leads to increasing E2 secretion, which in turn would suppress FSH levels again on Days 7–8.
Examination of the patterns of E2 and FSH secretion from individual marmosets reveals a considerable degree of variability between animals. Although the significance of this variation is not clear, the analysis suggested that estrogen concentrations were more closely related to total numbers of antral follicles than to the number of preovulatory structures and that there may be a relationship between number of preovulatory follicles and FSH levels on Day 6 (as opposed to earlier in the cycle). Thus, despite the low number of animals, the data suggest that the Day 6 peak may provide a stimulus important in triggering the final stages of antral follicular growth and to some extent in determining the number of preovulatory follicles that continue development. Variation in FSH secretion may also be attributed to differences in age between animals. Circulating FSH levels in women >40 yr of age are significantly higher than FSH levels in women 20–25 yr old [33]. These higher levels in older women are thought to be due to decreased inhibin B secretion from the depleted pool of preantral and small antral follicles in aged ovaries [34]. In the present study, we were unable to demonstrate a correlation between FSH level and age because of the low number of older marmosets used to monitor FSH (only two animals >5 yr of age). Nevertheless, there is a tendency for increased FSH secretion in older monkeys, and the maximum FSH peak measured was in the oldest monkey.
This study has demonstrated that marmoset folliculogenesis shares some features with follicular development in Old World primates but also has some unusual characteristics. The variability in ovulation rate both within and between females is an interesting feature that can only partly be explained by age and body weight and may be related in some way to the unusually large pool of nonovulatory antral follicles. The exact mechanisms regulating cyclic changes in marmoset FSH secretion in general and the Day 6 peak in particular remain unknown; however, the unusual FSH profile indicates that there may be important differences between marmosets and Old World primates in the regulation of folliculogenesis.
ACKNOWLEDGMENTS
The FSH antibody used in this study was generously donated by J. Dias. The authors thank Dr. Evgenia Isachenko for assistance with ovarian dissections and Jutta Hagedorn, Nicole Nüsse, Petra Kiesel, and Andrea Heistermann (Göttingen) and Lesley J. Ritter (Adelaide) for expert technical assistance. Dorothea Blank and Cornelia Casper (Göttingen) and Ken Porter and Bronwyn Hutchens (Adelaide) are also acknowledged for care and maintenance of animals. The authors are also grateful to Professor David T. Armstrong (Adelaide) for critical review of the manuscript and thoughtful suggestions and to Professor Robert J. Norman for providing animal and support facilities in Adelaide.
FOOTNOTES
1 Support for the Adelaide component of this project was provided by The National Health and Medical Research Council (NHMRC) of Australia Project Grant 981200 to R.B.G. 
2 Correspondence: R.B. Gilchrist, The Reproductive Medicine Unit, Department of Obstetrics and Gynaecology, University of Adelaide, The Queen Elizabeth Hospital, Woodville, 5011, Adelaide, Australia. FAX: 61 8 82227521; robert.gilchrist{at}adelaide.edu.au
Accepted: August 15, 2000.
Received: May 11, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  F. H. Thomas, H. Wilson, A. Silvestri, and H. M. Fraser Thrombospondin-1 Expression Is Increased during Follicular Atresia in the Primate Ovary Endocrinology, January 1, 2008; 149(1): 185 - 192. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P.D. Taylor, H. Wilson, S.G. Hillier, S.J. Wiegand, and H.M. Fraser Effects of inhibition of vascular endothelial growth factor at time of selection on follicular angiogenesis, expansion, development and atresia in the marmoset Mol. Hum. Reprod., October 1, 2007; 13(10): 729 - 736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Silvestri and H. M Fraser Oestrogen and progesterone receptors in the marmoset endometrium: changes during the ovulatory cycle, early pregnancy and after inhibition of vascular endothelial growth factor, GnRH or ovariectomy Reproduction, August 1, 2007; 134(2): 341 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. H. Thomas, E. E. Telfer, and H. M. Fraser Expression of Anti-Mullerian Hormone Protein during Early Follicular Development in the Primate Ovary in Vivo Is Influenced by Suppression of Gonadotropin Secretion and Inhibition of Vascular Endothelial Growth Factor Endocrinology, May 1, 2007; 148(5): 2273 - 2281. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. M Fraser, H. Wilson, C. Wulff, J. S Rudge, and S. J Wiegand Administration of vascular endothelial growth factor Trap during the 'post-angiogenic' period of the luteal phase causes rapid functional luteolysis and selective endothelial cell death in the marmoset. Reproduction, October 1, 2006; 132(4): 589 - 600. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P D Taylor, S G Hillier, and H M Fraser Effects of GnRH antagonist treatment on follicular development and angiogenesis in the primate ovary J. Endocrinol., October 1, 2004; 183(1): 1 - 17. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Zuhlke and G. Weinbauer The Common Marmoset (Callithrix jacchus) as a Model in Toxicology Toxicol Pathol, January 1, 2003; 31(1_suppl): 123 - 127. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
