Abstract
An in vitro system was used to analyze the effects of sex steroids on the development of primary (late perinucleolar stage) and early secondary, previtellogenic (early cortical alveolus stage) ovarian follicles of coho salmon cultured for up to 21 days. Late perinucleolar-stage follicles increased significantly in size after 7 days of treatment with low concentrations of 11-ketotestosterone (11-KT), a nonaromatizable androgen. An androgen receptor antagonist (flutamide) inhibited this growth-promoting effect, and the highest concentration resulted in atresia of follicles, implicating androgens as survival factors at this stage. Testosterone (T) was less effective than 11-KT in promoting growth, but blocking aromatization with exemestane resulted in a growth response similar to that of 11-KT. Estradiol-17beta (E2) had no effect on growth at this stage. After 21 days of culture, E2 was the most potent steroid in increasing the number of follicles containing cortical alveoli and the number of cortical alveoli within those follicles. At the early cortical alveolus stage, low doses of E2 promoted growth and strongly stimulated synthesis of cortical alveoli, actions that were inhibited by an estrogen receptor antagonist (tamoxifen). 11-KT displayed moderate growth-promoting effects, and 11-KT and T stimulated moderate to substantial increases in abundance of cortical alveoli. This study shows that the predominant role of androgens is the promotion of growth of late perinucleolar-stage follicles, while E2 stimulates both the growth and accumulation of cortical alveoli in early cortical alveolus-stage follicles.
Introduction
Progression of the ovarian follicle through development is regulated both by systemic hormones such as pituitary gonadotropins and by numerous intraovarian factors whose roles in many cases still remain to be definitively determined. The actions of gonadotropins and their main mediators, the sex steroid hormones, in promoting the later phases of ovarian follicle development are relatively well understood in both mammals and fish [1–4]: follicle-stimulating hormone (FSH) regulates follicular growth (tertiary growth in mammals, vitellogenesis in fish), whereas luteinizing hormone (LH) regulates the final stages of maturation and ovulation [5–11].
Studies on the regulation of earlier, pre-antral stages of follicular development in mammals using hypophysectomy or gene knockout models have led to the concept that the advancement of primary follicles into secondary growth is gonadotropin independent [9, 12–15]. In teleost fish, primary follicles are defined as those previtellogenic follicles that have not begun accumulating cortical alveoli (equivalent of cortical granules of mammals and other organisms) or yolk [4, 16]. A limited number of hypophysectomy studies on a few teleost species (e.g., [17–20], see review [4]) suggest that primary follicle growth can be completed in the absence of gonadotropins, with follicles able to develop to the late perinucleolar or early cortical alveolus stages.
Evidence from a variety of mammalian studies (reviewed by [21–23]) and a limited number of studies in teleost fish (e.g., [20, 24, 25], see review [4]) indicate that although the earlier stages of follicular growth appear to be gonadotropin independent, sex steroids may have important roles in the regulation of pre-antral growth in mammals and the completion of primary growth and transition into secondary growth (marked by cortical alveoli synthesis) in teleosts. In mammals, estradiol-17β (E2) regulates the size of the primordial follicle pool, appears to inhibit growth of the primordial oocyte, and is essential for stimulating the proliferation and differentiation of granulosa cells [21, 26, 27]. In teleosts, the primary role of E2 is stimulation of hepatically derived vitellogenin (yolk protein) production (reviewed by [4, 28]). However, prior to the onset of vitellogenesis, little is known about the role of E2 in fish, aside from several studies showing that estrogens can promote oogonial proliferation and the advancement of oogonia to the chromatin nucleolar stage in three species [29–31]. Treatment of hypophysectomized female goldfish (Carassius auratus), whose ovaries contained follicles arrested at the perinucleolar stage, with E2 induced the formation of cortical alveoli [20], and cortical alveoli accumulation has been associated with increased ovarian aromatase (cyp19a1a) expression and E2 production in zebrafish [32] and increased plasma FSH and E2 levels in coho salmon [33].
In addition to their role as precursors for E2 biosynthesis, several lines of evidence in mammals suggest that androgens are directly involved in regulating the earlier phases of ovarian follicle development [21, 22, 34]. Androgens promote ovarian follicle growth and the transition from primary to secondary growth [35–39], amplify FSH-stimulated follicle growth and granulosa cell differentiation [38, 40, 41], and have been intimately linked with other intraovarian regulators (e.g., IGF1 and IGF1 receptor expression, expression of members of the TGFβ superfamily [36, 37, 42, 43]). Testosterone (T) and/or nonaromatizable dihydrotestosterone stimulated primate and bovine primordial follicle development, resulting in an increase in the number of growing follicles [35–37, 44, 45]. Although studies on androgenic regulation of follicle development in teleosts are scarce, several in vivo and in vitro studies on freshwater eels have implicated a nonaromatizable [46, 47] testosterone derivative, 11-ketotestosterone (11-KT), in promoting growth of primary ovarian follicles. 11-KT is generally regarded as the major male-specific teleost androgen, but in female eels, high levels of 11-KT are associated with the initiation of ovarian growth [48], and both in vivo [49] and in vitro treatment with 11-KT advanced the development of primary ovarian follicles of short-finned eels [24]. 11-KT-induced follicle growth in eels has also been associated with increased lipid content and increased lipoprotein lipase [24, 49, 50]. Since 11-KT levels are generally very low in most female teleosts [51], it is not known if the results from eel are applicable to female teleosts in general, although very high (nonphysiological) concentrations of 11-KT have been reported to promote development of primary follicles of cod in vitro [52, 53].
Given the paucity of data on the regulation of primary growth and its transition into secondary growth in teleosts, a series of experiments were designed to directly investigate the effects of androgens and E2 on development of coho salmon (Oncorhynchus kisutch) ovarian follicles in vitro. Coho salmon are an ideal model for investigating early ovarian development since like other Pacific salmon they are semelparous, producing only a single clutch of synchronously developing ovarian follicles. Thus, the developmental stage of follicles at any given time is homogeneous.
Based on both mammalian studies and the few teleost studies available, the hypotheses addressed in this study are that androgens promote growth of primary (late perinucleolar stage) ovarian follicles and that E2 regulates the transition into early secondary growth (early cortical alveolus stage) marked by cortical alveoli synthesis. These hypotheses were tested using a previously established in vitro culture system in which the direct effects of steroids on follicle size and morphology were examined.
Materials and Methods
Chemicals
The following were purchased from Sigma-Aldrich (St. Louis, MO): l-proline, l-aspartic acid, l-glutamic acid, bovine serum albumin fraction V, HEPES, streptomycin sulfate, penicillin, tricaine methanesulfonate (MS-222), E2, 11-KT, T, tamoxifen, and flutamide. Exemestane was purchased from EDS Research, Inc. (Milwaukee, WI). Leibovitz L-15 medium was purchased from Gibco Cell Culture (Carlsbad, CA). Bouin fixative was purchased from Ricca Chemical Company (Arlington, TX).
Animals and Tissue Collection
Previtellogenic female coho salmon (brood years 2007–2009) were acquired from the Issaquah Creek Hatchery and the University of Washington Hatchery as eyed eggs and reared at the Northwest Fisheries Science Center and the University of Washington (Seattle, WA) in recirculated freshwater (10°C–15°C) under simulated natural photoperiod and fed a standard ration of a commercial diet (BioOregon, Longview, WA). Fish were euthanized using buffered MS-222 (250 mg/L). Fork length (cm) and weight (g) were recorded. Ovaries were dissected from the body cavity and weighed, and gonadosomatic index (GSI) was calculated using the following formula: gonad weight/body weight × 100. Ovarian follicles were staged based on morphological characteristics previously established for salmonids [54] and coho salmon [55]. Fish with follicles at the late perinucleolar stage, characterized by the absence of Balbiani bodies and prominent peripheral nucleoli, had a mean ovarian follicle diameter of 224.63 ± 22.62 μm (volume 0.012 ± 0.004 mm3). These fish had 18.1 ± 1.1 cm fork length, 68.1 ± 13.8 g body weight, and 0.30% ± 0.04% GSI (n = 15). Coho salmon with follicles at the early cortical alveolus stage, characterized by the accumulation of a few cortical alveoli (termed “yolk vesicles” in earlier literature and analogous to cortical granules in mammals and other organisms) within the ooplasm, had a mean ovarian follicle diameter of 363.50 ± 35.54 μm (volume 0.026 ± 0.008 mm3). The fish had 20.6 ± 1.2 cm fork length, 103.6 ± 21.7 g body weight, and 0.36% ± 0.09% GSI (n = 10). Fish were reared and handled under an approved protocol that followed the policies and guidelines of the University of Washington's Institutional Animal Care and Use Committee.
Culture of Ovarian Fragments
In vitro cultures were performed utilizing a floating agarose gel method [31, 56, 57]. Briefly, ovarian tissue fragments with ovarian follicles at the late perinucleolar or early cortical alveolus stages were placed on a 2% agarose gel cylinder covered with a nitrocellulose membrane in a 48-well culture plate (Becton Dickinson Labware, Franklin Lakes, NJ). The culture medium consisted of Leibovitz L-15 medium supplemented with the following: 1.7 mM l-proline, 0.1 mM l-aspartic acid, 0.1 mM l-glutamic acid, 0.5% bovine serum albumin fraction V, 10 mM Hepes, 0.1 g/L streptomycin sulfate, and 100 000 IU/L penicillin. The medium was adjusted to pH 7.3 and the same osmolality (300 mmol/kg; Vapor Osmometer 5520; Wescor ELITech, Berkhamsted, Herts, U.K.) as the plasma of coho salmon. Approximately 0.5 ml medium was added to each well so that a thin film of medium covered the ovarian fragments. Ovarian fragments were incubated in a humidified, refrigerated incubator at 15°C (100% air). Culture media were changed every 3 days throughout the culture period (7, 14, and 21 days).
In order to determine whether increased volume of the ovarian follicle was associated with altered water content, the percent water content was determined after culture with 11-KT or E2. Late perinucleolar-stage ovarian fragments from five different fish were blotted dry and weighed (wet weight) prior to being cultured in media containing 0 (control) or 0.003–30 ng/ml 11-KT or E2. After 7 days, ovarian tissues were weighed, dried in an oven (50°C) for 24 h and then weighed again (dry weight). The percent water content was calculated ([wet weight – dry weight]/wet weight)*100). Water content was not determined after longer exposure periods because the maximum increase in follicle size in response to steroid treatment was observed at 7 days. Early cortical alveolus-stage fragments were not analyzed for water content since there was no change in the water content of late perinucleolar-stage follicles and because in addition to size, increases in the incidence and abundance of cortical alveoli were used as morphological indicators of steroid responsiveness in these follicles.
Effects of Long-Term Culture with Sex Steroids on Development of Ovarian Follicles
Late perinucleolar-stage ovarian fragments were cultured in media with 0 (control), 0.03–30 ng/ml (approximately 100 pM-100 nM) 11-KT, 0.03–30 ng/ml (100 pM-100 nM) T, or 0.03–30 ng/ml (100 pM-100 nM) E2. Early cortical alveolus-stage ovarian fragments were cultured in media with 0 (control), 0.003–30 ng/ml (approximately 10 pM-100 nM) 11-KT, 0.003–30 ng/ml (10 pM-100 nM) T, or 0.003–30 ng/ml (10 pM-100 nM) E2. Ovarian tissue from five female fish was used for the experiment. The ovaries were large enough so that pooling tissue from individuals was not necessary; therefore five independent replicates (each replicate from the ovary of a different female) were used for each experiment such that tissue from the same five individuals was used for all steroid treatments and controls.
Effects of an Aromatase Inhibitor and Sex Steroid Receptor Antagonists on Steroid-Induced Growth and Development of Ovarian Follicles
T was less effective than 11-KT in promoting growth of late perinucleolar follicles in the experiments described above (see Results), but compared to controls, a greater percentage of follicles containing cortical alveoli were observed after culture of these follicles with T for 21 days. To determine whether the difference in effect of 11-KT and T was due to aromatization of T to E2, late perinucleolar ovarian fragments were incubated with exemestane, a non-reversible aromatase inhibitor [58]. Fragments were treated with exemestane (100 and 200 nM) alone and in combination with an optimal growth-promoting concentration of T (3 ng/ml, from the experiment above) for 21 days. Exemestane concentrations were based on those used in mammalian in vitro experiments (e.g., [58, 59]) and on the almost complete inhibition by these concentrations of exemestane of E2 production by vitellogenic coho salmon follicles incubated with T (Forsgren et al. unpublished).
Because androgens promoted growth of late perinucleolar follicles in the experiments described above (see Results), late perinucleolar fragments were incubated with flutamide, an androgen receptor antagonist, to confirm that the androgen effects were mediated via androgen receptors. Ovarian fragments were incubated in steroid or flutamide-free media, with flutamide (5, 50, and 100 nM) alone, with an optimal growth-promoting concentration of 11-KT (3 ng/ml from the experiment above), or with a combination of 11-KT and flutamide for 7 days. The flutamide concentrations were chosen based on concentrations used in vitro in teleost and mammalian studies (e.g., [39, 60, 61]).
The experiments described above showed that E2 increased both the size of early cortical alveolus-stage follicles and the number of cortical alveoli within the ooplasm (see Results). To confirm that the actions of E2 were mediated through estrogen receptors, tamoxifen, an estrogen receptor antagonist was utilized. Ovarian fragments were incubated for 7 days in steroid or agonist-free media, with tamoxifen (5 and 50 μM) alone, with a growth-promoting concentration of E2 (3 ng/ml from the experiment above), or with a combination of E2 and tamoxifen for 7 days. Tamoxifen concentrations were based on previous studies on teleosts (e.g., [60, 62]). For each of these experiments, ovarian tissue from five female fish was utilized as described above.
General Histological Procedures
Ovarian fragments were placed in Bouin fixative for 48 h prior to being stored in 70% ethanol. Fixed ovarian fragments were dehydrated in a series of graded ethanol, cleared with xylene, and infiltrated and embedded in paraffin wax. Ovarian fragments were sectioned to a thickness of 5 μm and stained with hematoxylin and eosin (Richard Allen Scientific, Kalamazoo, MI). Sectioned ovarian fragments were examined using a light microscope (Nikon Eclipse 50i; Nikon, Melville, NY). Micrograph images were captured using a digital camera attached to the microscope (Nikon Digital Sight DS-U1; Nikon). The circumference (μm) of follicles was measured using NIS-Elements Imaging Software (version 2.1; Nikon). Only follicles sectioned through the approximate center of the nucleus were measured. Initially, 30 follicles per fish replicate were measured. However, because of the very low coefficient of variation (6.16%, n = 30), the number of follicles measured was reduced in later experiments to 15 since that number resulted in a similar coefficient of variation (6.53%, n = 15). A total of 30 (early experiments) and 15 (later experiments) ovarian follicles were measured (n) for each replicate with a total of five replicates per experiment (N = number of fish). For late perinucleolar follicles, the presence and abundance of cortical alveoli within the ooplasm of ovarian follicles was determined by first scoring follicles for the presence (one or more) or absence of cortical alveoli, achieved by examining 15 randomly chosen ovarian follicles (sectioned through the approximate center of the nucleus) per replicate, for a total of 75 follicles scored for each treatment (five replicates). Data are presented as the percentage of follicles containing cortical alveoli. The number of cortical alveoli was counted in those follicles displaying one or more cortical alveoli (cortical alveoli-positive follicles). Cortical alveolus-stage follicles by definition contain cortical alveoli. Therefore, the number of cortical alveoli in the ooplasm of 15 randomly chosen ovarian follicles (sectioned through the approximate center of the nucleus) per replicate was counted, for a total of 75 follicles scored for each treatment.
Statistics
Statistical analyses were performed using a two-way or one-way ANOVA followed by Tukey multiple means comparison test. Additional analysis was performed using a Student t-test (Prism 4 for Macintosh; GraphPad Software, Inc., La Jolla, CA). The level of significance was determined at P < 0.05.
Results
Effects of Long-Term Culture with Sex Steroids on Development of Ovarian Follicles
All concentrations of 11-KT significantly (P < 0.0001) increased the volume of late perinucleolar-stage follicles in a concentration-dependent manner by Day 7 of culture (Figs. 1 and 2). The lowest concentration of 11-KT (0.03 ng/ml) increased follicle volume by 21%, whereas the highest concentration (30 ng/ml) increased volume by 62% compared to control follicles after 7 days. There was no evidence for further 11-KT-induced increases in volume after 7 days in culture (Fig. 1). For incubations examining the effects of T and E2, a slight but significant decrease (P < 0.05) in the volume of control follicles occurred at Days 7 and 14 compared to time 0 follicles (Fig. 1). After 7 days in culture, T also significantly (P < 0.0001) stimulated late perinucleolar follicle growth, but was much less effective than 11-KT. After 7 days in culture, a significant (P < 0.001) 14% increase in volume compared to controls was induced by the lowest concentration of T (0.03 ng/ml). Late perinucleolar-stage follicles treated with 0.3 and 3 ng/ml T increased 24% and 28% in volume, respectively, compared to controls (P < 0.001). After 14 days in culture, only the volumes of follicles exposed to the highest concentrations of T (3 and 30 ng/ml) were significantly (P < 0.001) greater than controls (Fig. 1). After 21 days in culture, follicles treated with 0.03, 0.3, and 30 ng/ml T were 13% larger in volume than the controls, whereas follicles exposed to 3 ng/ml increased by 29% compared to controls (P < 0.001). E2 did not increase the volume of late perinucleolar-stage follicles, but significantly (P < 0.05) reduced the volume of follicles exposed to 0.03 and 0.3 ng/ml at Day 7 by 17% and 10%, respectively (Fig. 1). At Day 14, there were no significant (P > 0.05) differences between treatments, and by Day 21, the only significant difference (P = 0.0089) was the smaller volume (by 5%) of follicles exposed to 0.03 ng/ml E2 versus controls (Fig. 1). The percentage of follicles containing cortical alveoli and the numbers of cortical alveoli were quantified histologically after culture of late perinucleolar follicles for 21 days (Fig. 3). Cortical alveoli were not present in late perinucleolar treated follicles at Days 7 and 14. After 21 days in culture, cortical alveoli were present in only 1.33% ± 1.3% of control follicles (Fig. 3A). A significant increase (P < 0.0001) in the percentage of follicles (approximately 16%) containing cortical alveoli in response to 11-KT occurred only with the 30 ng/ml treatment. Culture with T resulted in a significant (P = 0.0019) increase in the percentage of follicles containing cortical alveoli, ranging from 12% to 38%, but there was no significant (P = 0.3878) difference in the response to different concentrations of T. E2 at all concentrations significantly increased (P = 0.0459) the proportion of follicles containing cortical alveoli to 73%–88%, a significantly greater (P < 0.0001) response than those to 11-KT or T; there was no significant (P = 1.00) difference in response to different concentrations of E2 (Fig. 3A).
Changes in the volume of late perinucleolar-stage ovarian follicles of coho salmon exposed to various concentrations of 11-KT, T, and E2 in vitro for 7, 14, and 21 days (mean ± SEM). N = ovarian tissue from five fish with 30 follicle measurements made for each replicate. Different letters indicate significant differences (P < 0.05).
Micrographs of late perinucleolar-stage ovarian follicles of coho salmon exposed to sex steroids in vitro after 21 days showing the following culture conditions: time 0, control (no steroid), follicles exposed to 30 ng/ml 11-KT, follicles exposed to 30 ng/ml T, and follicles exposed to 30 ng/ml E2. Bars = 200 μm. Insets of each condition at higher magnification show details of the follicle. Bars = 100 μm. Arrows point to cortical alveoli within the ooplasm of follicles.
A) The percentage of late perinucleolar-stage follicles of coho salmon containing cortical alveoli within the ooplasm. B) The number of cortical alveoli counted in cortical alveoli-positive follicles after being treated with 11-KT, T, and E2 for 21 days. Different letters indicate significant differences (P < 0.05). N = ovarian tissue from five fish; cortical alveoli were counted in 15 follicles for each replicate.
Only a single cortical alveolus was found in two control late perinucleolar follicles at Day 21. Culture of follicles with 30 ng/ml 11-KT, but not lower concentrations, resulted in a significant increase (P = 0.0045) in the mean number of cortical alveoli (4.0 ± 0.3) in cortical alveoli-positive follicles (Fig. 3B). Similarly, only 30 ng/ml T significantly increased (P = 0.0019) the mean number of cortical alveoli per follicle (7.8 ± 0.3), a significantly greater (P = 0.0256) response than that to 30 ng/ml 11-KT. Individual cortical alveoli observed in 11-KT-treated follicles appeared to be larger than those in T- or E2-treated follicles, although this was not quantified. Although there were indications of a concentration-dependent response to E2, only follicles cultured with 3 (5.6 ± 0.5 cortical alveoli per follicle) and 30 ng/ml (18.1 ± 1.0 cortical alveoli per follicle) displayed a significant increase (P < 0.0001) in cortical alveoli-positive follicles versus control follicles, with the response to 30 ng/ml E2 being a significantly greater (P = 0.001) response than that to any concentration of 11-KT or T or to 3 ng/ml E2 (Fig. 3B).
At Day 7, all concentrations of 11-KT significantly (P < 0.0001) increased the volume of early cortical alveolus-stage follicles between 35% and 45% above controls (Figs. 4 and 5). However after 14 days in culture, the magnitude of the maximum increase in follicle volume was less than that seen at Day 7 (17% above controls; P < 0.0001), and after 21 days, the growth-promoting effect of 11-KT was lost, with the follicle volume of 11-KT-exposed follicles either being no different (0.03, 3 ng/ml; P > 0.05), or lower (0.03, 0.3, 30 ng/ml; P < 0.0001) than the volume of control follicles. Similarly, in vitro treatment with T for 7 days increased (P < 0.01) early cortical alveolus-stage follicle volume by 20%–34% (0.03–3 ng/ml) relative to controls, but there was no difference (P > 0.05) between the volume of follicles exposed to 30 ng/ml T and controls (Fig. 5). After 14 days in culture, no significant (P = 0.1794) effect of T on follicle volume was observed, and after 21 days of culture, volume was slightly but significantly higher (17%; P = 0.0219) higher after incubation with 3 ng/ml T (Figs. 4 and 5). All concentrations of E2 promoted a significant (P < 0.0001) increase in follicle volume at Day 7, with the lowest dose of E2 (0.003 ng/ml) increasing follicle volume by 57% compared to controls (Fig. 5). At Day 14, follicles exposed to all doses of E2 had significantly (P < 0.0001) greater volumes than controls, with follicles exposed to 0.03 ng/ml being 62% larger than control follicles. By Day 21, follicle volumes were significantly greater (P < 0.001) than controls after exposure to all concentrations of E2, with the greatest increase being in response to 0.3 ng/ml E2 (49%).
Changes in volume of early cortical alveolus-stage ovarian follicles of coho salmon exposed to various concentrations of 11-KT, T, and E2 in vitro for 7, 14, and 21 days (mean ± SEM). N = ovarian tissue from five fish with 15 follicle measurements made for each replicate. Different letters indicate significant differences (P < 0.05).
Micrographs of early cortical alveolus-stage ovarian follicles of coho salmon exposed to sex steroids in vitro for 21 days showing the following culture conditions: time 0 follicles, control (no steroid), follicles exposed to 0.03 ng/ml 11-KT, follicles exposed to 0.03 ng/ml T, and follicles exposed to 0.03 ng/ml E2. Bars = 200 μm. Insets of each condition at higher magnification show details of the follicle. Bars = 100 μm. Arrows point to cortical alveoli within the ooplasm of follicles.
The mean number of cortical alveoli in control early cortical alveolus-stage follicles increased (52%) from Day 7 to Day 21 (P < 0.001; Fig. 6). By Day 21, all follicles treated with all concentrations of steroids contained significantly (P < 0.0001) more cortical alveoli compared to control follicles (Fig. 6). For cultures with 11-KT, follicles exposed to 3 or 30 ng/ml had significantly greater (P < 0.0001) numbers of cortical alveoli, a 37% and 41% increase, respectively, than did controls at Day 7, but only follicles exposed to 0.3 and 3 ng/ml for 14 days had significantly greater (P < 0.0001) numbers of cortical alveoli than did controls, an increase of 27% and 53% respectively, which differed significantly from one another (P < 0.0001). By Day 21, follicles exposed to all concentrations of 11-KT had approximately 17%–20% more cortical alveoli (P < 0.0001) than controls, with no significant difference (P = 0.5034) in response to the various concentrations of 11-KT. Follicles exposed to 0.3–30 ng/ml T for 7 days had significantly greater numbers (P < 0.0001) of cortical alveoli compared to controls, increasing approximately 36%–71%, with the response to 0.3 ng/ml T being significantly (P < 0.0001) less than to 3 or 3 ng/ml T. Culture of follicles with 0.03–30 ng/ml T resulted in a significant (P < 0.0001) 34%–65% increase at Day 14, with the response to 0.03 and 0.3 ng/ml being significantly less (P = 0.0012 and P = 0.0017, respectively) than that to 3 ng/ml T. The response to 30 ng/ml T was significantly less (P < 0.0001) than that to 3 ng/ml T. After 21 days of culture with T, significant (P < 0.0001) 69%–84% increases over controls in the mean number of cortical alveoli per follicle occurred, with no significant difference in the response between concentrations of T.
The number of cortical alveoli counted in the ooplasm of early cortical alveolus-stage ovarian follicles of coho salmon exposed to various concentrations of 11-KT, T, and E2 in vitro for 7, 14, and 21 days (mean ± SEM). Different letters indicate significant differences (P < 0.05); additional statistical comparisons are described in Results. N = ovarian tissue from five fish; cortical alveoli were counted in 15 follicles for each replicate.
Significant increases (P < 0.0001) in the numbers of cortical alveoli in follicles treated with all doses of E2 compared to controls occurred after treatment for 7, 14, and 21 days (Fig. 6). Follicles exposed to 0.03–30 ng/ml E2 for 7 days had significantly (P < 0.0001) more cortical alveoli than controls, increases of between 54% and 130%. There was no significant difference (P = 0.0706) in the response to 0.03 and 0.3 ng/ml E2, a significant difference (P < 0.0001) in response to 0.3 and 3 ng/ml, and no significant difference (P = 0.1514) in response to 3 and 30 ng/ml E2. After 14 days in culture, follicles exposed to 0.03–3 ng/ml responded in a concentration-dependent fashion to E2, with follicles exposed to all concentrations of E2 containing significantly (P < 0.0001) more cortical alveoli than controls; increases ranged between 40% and 146%. The response to each concentration of E2 varied significantly (P < 0.0001) from the other concentrations, except for the response to 0.3 and 30 ng/ml E2 (P > 0.05). Follicles cultured with all concentrations of E2 for 21 days contained significantly more (P < 0.0001) cortical alveoli than control follicles, with increases of 73%–91%. The only significant difference (P = 0.0013) in response between concentrations of E2 occurred with 0.03 and 30 ng/ml.
Comparisons of the response to the same concentration of each steroid revealed that on Day 7, the response to 3 ng/ml E2 induced a significantly greater (P < 0.0001) increase in the number of cortical alveoli per follicle than 3 ng/ml 11-KT, and 30 ng/ml T or E2 evoked a significantly greater (P < 0.0001) response than 30 ng/ml 11-KT. Also on Day 7, the responses to 0.3, 3, or 30 ng/ml E2 were significantly greater (P < 0.0001) than to the same concentrations of T. By Day 14, the only difference between 11-KT and T treatments was that the number of cortical alveoli significantly (P = 0.0003) increased in response to 30 ng/ml T. At Day 14, all concentrations of E2 stimulated a greater response (P < 0.0001) than the equivalent concentrations of 11-KT, and with the exception of 0.03 ng/ml, the same was true for comparisons of cortical alveoli numbers in response to E2 and T. By Day 21, all concentrations of T (P < 0.0001) or E2 (P < 0.0001) stimulated a significantly greater response than the equivalent concentrations of 11-KT, and the increase in the number of cortical alveoli was significantly greater (P < 0.0001) in response to E2 compared to T.
The percent water content of late perinucleolar follicles (time 0, E2 experiment, 86.7% ± 0.3%) did not change after culture (E2 control, 87.7% ± 0.4%; P = 0.0858). Follicle water content of the E2 control follicles did not differ significantly (P = 0.8220) from the water content of follicles treated with 0.03 ng/ml (87.7% ± 0.4%), 0.3 ng/ml (88.1% ± 0.2%), or 30 ng/ml (88.6% ± 0.3%) E2 (P = 0.8220). Similarly, there was no significant difference (P = 0.0885) in the water content of the time 0 sample (86.5% ± 0.8%) and control follicles (88.3% ± 1.6%) in the 11-KT experiment, and no difference between the water content of control follicles and follicles exposed to 0.003 ng/ml (87.6% ± 0.2%), 0.3 ng/ml (88.3% ± 0.3%), or 30 ng/ml (86.9% ± 0.4%) 11-KT (P = 0.2214).
Effects of an Aromatase Inhibitor and Sex Steroid Receptor Antagonists on Steroid-Induced Growth and Development of Ovarian Follicles
The experiments described above showed that T was less effective than 11-KT in promoting growth of late perinucleolar-stage follicles, but was more effective than 11-KT in stimulating an increase in both the percentage of follicles containing cortical alveoli and the number of cortical alveoli present in cortical alveoli-positive follicles. To determine if the relatively poor growth effect and the positive effect on accumulation of cortical alveoli was due to aromatization of T to E2, late perinucleolar follicles were exposed to either T (3 ng/ml), the aromatase inhibitor exemestane (100 and 200 nM), or combinations of T and exemestane (Fig. 7). Treatment with 3 ng/ml of T alone for 7 days resulted in a significant (P < 0.0001) 19% increase in follicle volume compared to controls (Fig. 7A). Exposure to 100 nM exemestane alone had no significant (P > 0.05) effect on follicle volume, whereas 200 nM exemestane promoted a significant (P < 0.01) 26% increase in follicle volume compared to the control. There was no significant (P > 0.05) difference in volume between follicles exposed to T alone, T in combination with 100 nM exemestane, or 200 nM exemestane alone. However, the volume of follicles exposed to T and 200 nM exemestane was significantly (P < 0.001) greater (35%) than the volume of follicles exposed to only T, T in combination with 100 nM exemestane, or to 200 nM exemestane alone (Fig. 7A). In histological sections, cortical alveoli were not present in any of the control follicles after culture for 21 days. Treatment with 3 ng/ml T for 21 days resulted in a significantly (P < 0.0001) greater percentage of follicles containing cortical alveoli after treatment with T (22.0% ± 5.0%), as seen previously (Fig. 7B). Significant increases (P < 0.0001) in the percentage of late perinucleolar follicles with cortical alveoli occurred after culture with 100 (13.7 ± 2.6%) or 200 nM (12.7 ± 2.4%) exemestane when compared with control follicles (Fig. 7B).
Late perinucleolar-stage ovarian follicles of coho salmon exposed to 3 ng/ml T, the aromatase inhibitor exemestane (EXEM; 100 nM and 200 nM), or combinations of T and EXEM in vitro for 21 days (mean ± SEM). A) Changes in the volume of ovarian follicles. N = ovarian tissue from five fish with 15 follicle measurements made for each replicate. B) The number of cortical alveoli counted in the ooplasm of cortical alveoli-positive ovarian follicles. C) The percentage of follicles containing cortical alveoli within the ooplasm. N = ovarian tissue from five fish; cortical alveoli were counted in 15 follicles for each replicate. Different letters indicate significant differences (P < 0.05).
Follicles that contained cortical alveoli after culture with T for 21 days had a mean of 5.1 ± 0.6 cortical alveoli per follicle (Fig. 7C). Treatment with 100 or 200 nM exemestane alone significantly increased (to 2.5 ± 0.1 and 2.2 ± 0.2 cortical alveoli per follicle, respectively; P < 0.0001) the number of cortical alveoli in cortical-alveoli-positive follicles compared to controls. The presence of exemstane in cultures with T also significantly increased (P < 0.0001) the number of cortical alveoli per follicle compared to controls (T with 100 nM exemestane: 2.0 ± 0.3 cortical alveoli per follicle; T with 200 nM exemestane: 2.2 ± 0.3 cortical alveoli per follicle). However, the number of cortical alveoli per follicle was significantly reduced (P < 0.0001) in treatments with exemestane alone and in combination with T compared to the response of T alone (Fig. 7C). There was no significant difference (P = 0.4898) in the number of cortical alveoli per follicle in cortical alveoli-positive follicles treated with exemestane alone or exemestane in combination with T (Fig. 7C).
Follicles were exposed to steroid receptor antagonists to determine if the action of steroids identified in the previous experiments was mediated via classical steroid receptors (Figs. 8 and 9). Late perinucleolar-stage follicles treated with 11-KT (3 ng/ml) for 7 days were 28% greater in volume (P < 0.001) than controls (Fig. 8A). Exposure to flutamide (5 and 50 nM), an androgen receptor antagonist, completely inhibited the growth-promoting effects of 11-KT (P < 0.05; Fig. 8A). The highest concentration of flutamide (100 nM) resulted in widespread atresia of late perinucleolar follicles; therefore, no follicle measurements were made for this treatment (Fig. 8B). As seen previously with this concentration of 11-KT in 7-day cultures of late perinucleolar follicles, cortical alveoli were absent from all follicles, irrespective of treatment (data not shown).
Late perinucleolar ovarian follicles of coho salmon exposed to 3 ng/ml 11-KT, the androgen receptor antagonist flutamide (FLUT; 5, 50, or 100 nM), or combinations of 11-KT and FLUT in vitro for 7 days. A) Changes in the volume of ovarian follicles (mean ± SEM). N = ovarian tissue from five fish with 15 follicle measurements made for each replicate. Different letters indicate significant differences (P < 0.05). B) Micrographs of ovarian follicles exposed to 5, 50, and 100 nM FLUT. Bars = 200 μm.
Early cortical alveolus-stage ovarian follicles of coho salmon exposed to 3 ng/ml E2, the estrogen receptor antagonist tamoxifen (TAM; 5 or 50 μM), or combinations of E2 and TAM in vitro for 7 days (mean ± SEM). A) Changes in the volume of ovarian follicles. N = ovarian tissue from five fish with 15 follicle measurements made for each replicate. B) The number of cortical alveoli counted in the ooplasm of ovarian follicles. N = ovarian tissue from five fish; cortical alveoli were counted in 15 follicles for each replicate. Different letters indicate significant differences (P < 0.05).
As observed previously, early cortical alveolus-stage follicles exposed to 3 ng/ml E2 for 7 days had a significantly greater volume (25%; P < 0.001) compared to control follicles (Fig. 9A). Treatment with the lowest concentration of tamoxifen (5 μM), an estrogen receptor antagonist, alone resulted in follicle volumes that were slightly but significantly (P < 0.01) less than control volumes, but volumes after culture with the highest concentration of tamoxifen (50 μM) were no different from controls (P = 0.773). The growth-promoting effect of E2 was completely abolished in the presence of tamoxifen (P < 0.001; Fig. 9A). Similarly, as shown previously, early cortical alveolus-stage follicles exposed to E2 for 7 days displayed a 24.9% ± 0.3% increase in the number of cortical alveoli compared to controls (P < 0.001; Fig. 9B). Tamoxifen abolished the effect of E2 on increasing numbers of cortical alveoli. Numbers of cortical alveoli in follicles from cultures treated with either concentration of tamoxifen, alone or in combination with E2, were not significantly different (P = 0.2683) from those in control follicles (Fig. 9B).
Discussion
This study addressed the hypothesis that androgens and E2 have disparate, stage-specific effects on early ovarian follicle growth and development in coho salmon. During long-term culture, the response of late primary and early secondary growth follicles to sex steroids differed substantially between steroids and stages of follicle development.
Follicles in primary growth at the late perinucleolar stage were highly responsive to 11-KT, with low 11-KT concentrations being extremely potent in stimulating follicle growth after 7 days of exposure and with a maximal increase in follicle volume of over 60% (at 30 ng/ml 11-KT). 11-KT had no further effects on follicle growth beyond 7 days of culture. T was much less effective in stimulating increases in follicle volume, and E2 was not potent in this regard and even appeared to inhibit follicle growth after 7 days of culture, when follicles exposed to E2 (0.03 and 0.3 ng/ml) were smaller than controls. The effect of 11-KT on follicle growth was similar to the 10%–20% increase in diameter reported for late perinucleolar-stage follicles of shortfinned eel (Anguilla australis) cultured for 18 days in the presence of exogenous 11-KT [24]; in this same study, E2 was without effect on follicle diameter. Kortner et al. [52, 53] reported that exposure to 11-KT and T increased the proportion of “advanced previtellogenic” follicles in cod ovarian fragments in vitro, although the description of the follicle stages and the micrographs presented appear not to match. The “advanced previtellogenic” follicles that appeared after treatment seem equivalent to the early perinucleolar stage. The extremely high pharmacological concentrations used in this study (up to 1000 μM or ∼ 300 μg/ml) were much greater than the physiological levels observed in a survey of females of over 30 teleost species [51] or the steroid concentrations used in the present study and, thus, the physiological relevance of these results for cod are unclear. From the results of the present study and the study of eel follicles exposed to sex steroids [24], it appears that physiological concentrations of the nonaromatizable androgen, 11-KT, stimulate the growth of late perinucleolar follicles of eels and Pacific salmon. Although the processes underlying steroid-induced increases in follicle size are not yet known, the growth-promoting actions of 11-KT on late perinucleolar follicles were not associated with a change in water content of the follicle.
11-KT was used in the current study because it is a teleost nonaromatizable androgen usually regarded as male specific [51]. 11-KT is not considered to be a major circulating androgen in most female teleosts, but levels of up to 2 ng/ml have been detected in previtellogenic salmonids [51]. Although we have principally used 11-KT as a model nonaromatizable androgen without assumption about levels during early ovarian development, the significant low concentration effects (30 pg/ml) found in vitro suggest that such low levels in vivo may have physiological effects. Neither the source of 11-KT nor the factor(s) controlling 11-KT synthesis in female teleosts are known.
11-KT only slightly increased the percentage of follicles transitioning from primary growth into secondary growth, as evidenced by the presence of cortical alveoli in the ooplasm, and it did so only at high concentrations (3 and 30 ng/ml) after 21 days of exposure. T was much more effective than 11-KT in increasing the percentage of follicles with cortical alveoli (38% vs. 16% maximum increase), but much less effective in this regard than E2, which induced a 73%–88% increase in the incidence of follicles containing cortical alveoli. Similarly, 30 ng/ml 11-KT (but not lower concentrations) only moderately increased the number of cortical aveoli in the ooplasm of oocytes bearing some cortical alveoli and, again, T (albeit only at 30 ng/ml) was more effective than 11-KT in this regard, but much less effective than E2, which evoked a strong concentration-dependent increase in cortical alveoli abundance.
Follicles at the early cortical alveolus stage, the first stage of secondary growth, responded to 11-KT at all concentrations, with an increase in follicle volume after 7 days of exposure. 11-KT had no further effects on follicle growth after 7 days. T was less effective in stimulating increases in follicle volume, in terms of the concentration needed to achieve a significant growth effect. In contrast to the late perinucleolar stage, early cortical alveolus-stage follicles were highly responsive to E2, with very low concentrations substantially stimulating follicle growth after 7 days of exposure. Unlike 11-KT and T, follicles continued to be responsive to E2 after 7 days of exposure. The enhanced sensitivity of the early cortical alveolus follicle to E2 in terms of growth promotion may be due to increases in estrogen receptor expression at this stage and/or to changes in mechanisms downstream of ligand binding.
A feature of the experiments with early cortical alveolus-stage follicles was the relative decrease in the size of steroid-treated follicles, but not controls, at Days 14 and 21 in comparison to Day 7. This phenomenon was more apparent in 11-KT-treated follicles, and unlike the results for 11-KT and T exposure, E2-treated follicles were still significantly larger than controls after 14 and 21 days of culture and at all concentrations of E2. The reduction in size of treated follicles at Days 14 and 21 of culture was relatively minor in late perinucleolar follicles; therefore, this phenomenon appears to be both steroid- and stage-specific, but the underlying cause is unknown. However, although reductions in the size of early cortical alveolus follicles occurred, the numbers of cortical alveoli increased progressively from Day 7 to 21 in follicles exposed to all steroids (discussed below), an observation that gives reassurance that the longer-term cultured follicles were able to maintain synthesis of cortical alveoli and suggests that follicle growth and synthesis of cortical alveoli are not tightly coupled processes.
Experiments using early cortical alveolus-stage follicles provided further confirmation that E2 is the most potent steroid in promoting cortical alveoli synthesis. All steroids promoted an increase in the number of cortical alveoli in early cortical alveolus-stage follicles, although with marked differences in relative potency and in the speed and magnitude of the response. Although final numbers of cortical alveoli per follicle on Day 21 were only slightly higher for cultures with E2 versus T, results of shorter cultures clearly show that E2 was more effective than T (in terms of response to the same concentration of steroid) and in the relative increase in the number of cortical alveoli per follicle between Days 7 and 14. Again, the lesser potency of T to stimulate cortical alveoli synthesis may be due to its aromatization to E2, although since the nonaromatizable 11-KT had modest effects on number of cortical alveoli per follicle, androgens may directly participate in promoting cortical alveoli synthesis. Androgens could potentially also upregulate the E2-synthesizing capacity of the follicle through effects on the expression of steroidogenic proteins, particularly cyp19a1a.
These data indicate that E2 is a major regulator in the transition of the coho salmon ovarian follicle from late primary growth into early secondary growth and is essential for continued growth and development of the cortical alveolus-stage follicle. This conclusion is consistent with the appearance of cortical alveoli after treatment of hypophysectomized goldfish with E2 [20] and the correlation between cortical alveoli accumulation and cyp19a1a mRNA expression and E2 production in zebrafish [32]. Similarly, in coho salmon, the transition to the cortical alveolus stage was associated with activation of the brain-pituitary-gonad axis, including increased levels of plasma and pituitary FSH, plasma E2, and transcripts encoding the protein that controls the key rate-limiting step in steroidogenesis, steroidogenic acute regulatory protein [33].
Together, these results indicate that a key role of androgens during the late perinucleolar stage of the coho salmon ovary is to promote growth of the ovarian follicle, while the transition to the next stage, characterized by synthesis of cortical alveoli, appears to largely be an E2-driven process, although 11-KT was still able to promote significant short-term growth of the follicle at this stage. E2 had no growth-promoting effects on late perinucleolar-stage follicles; it was much more potent than the androgens in stimulating increases in the proportion of follicles with cortical alveoli and increases in the numbers of cortical alveoli in such follicles. T moderately exerted both the androgenic effect (increase in size of follicles) and the estrogenic effect (proportion of follicles with cortical alveoli and number of cortical alveoli per follicle) on these late perinuculeolar-stage follicles. These dual effects of T on follicle primary growth and transition of follicles into early secondary growth appear to be due, at least in part, to its aromatization to E2, because the growth-promoting action of T was significantly enhanced when the aromatase inhibitor, exemestane, was present.
Interestingly, the highest concentration of exemestane caused a significant increase in follicle volume in the absence of exogenous T, most likely because it inhibited conversion of endogenous T to E2 by follicle cells. The effects of the aromatase inhibitor on T-stimulated increases in the proportion of follicles displaying cortical alveoli and on the numbers of cortical alveoli in such follicles are less easy to interpret, because exemestane alone increased the number of follicles containing cortical alveoli and because the presence of exemestane did not reduce the proportion of follicles with cortical alveoli in cultures with exogenous T. Similarly, when administered alone, exemestane stimulated a modest increase in the number of cortical alveoli in such follicles, although to a lesser extent than did T alone. Nonetheless, the aromatase inhibitor significantly reduced the number of cortical alveoli per follicle, presumably by inhibiting the conversion of T to E2, reinforcing other findings (discussed below) that indicate that E2 is the most potent steroid regulating the formation of cortical alveoli.
Experiments employing highly specific androgen and estrogen receptor antagonists confirmed that 11-KT and E2 interact with their cognate receptors in promoting follicle growth, and, for E2, cortical alveoli synthesis. Of particular interest was the observation that late perinucleolar follicles underwent atresia when exposed to a high concentration of flutamide (100 nM) after 7 days of exposure, suggesting that at this stage, androgens are critical ovarian follicle survival factors, and suppression of the interaction of endogenous androgens with their receptors results in entry of follicles into an atretic pathway. The concentration of flutamide that resulted in atresia has been employed in both mammalian and teleost studies previously [39, 60, 61], with no reports of toxic effects. In mammals, androgens have been shown to either inhibit or increase atresia of preantral follicles [35, 36, 40, 63, 64].
In summary, this study demonstrates that androgens promote growth of late stage primary follicles of coho salmon, in agreement with previous studies on eels, and provides the first direct evidence that synthesis of cortical alveoli is under the control of E2, although with the caveat that androgens may also directly or indirectly participate in the regulation of development of early cortical alveolus-stage follicles. The coho salmon ovarian follicle in vitro system provides a robust experimental platform to further explore the differential morphological effects of steroids on previtellogenic stages of follicle development. Identification of the intracellular mediators involved, and the determination of steroid-sensitive genes at different stages of follicle development, will provide a mechanistic basis for understanding the stage-specific effects of androgens and E2.
Acknowledgment
We thank Dr. Brian Beckman and Abby Tillotson for their generous gift of coho salmon, Amanda Bruner and Larissa Felli for help with fish maintenance, Mollie Middleton for assistance with histology, and Drs. Penny Swanson, Walt Dickhoff, and Robert Steiner for constructive advice and comments on an early draft of the manuscript. We are especially grateful to Drs. Takeshi and Chiemi Miura for their guidance in the use of the culture system employed and for helpful comments.
References
Author notes
Supported by National Research Initiative Competitive Grant 2003–35203-13602 from the U.S. Department of Agriculture Cooperative State Research, Education and Extension Service, and National Science Foundation grants OISE-0914009 and IOS-0949765.
