Exposure of Female Macaques to Western-Style Diet With or Without Chronic T In Vivo Alters Secondary Follicle Function During Encapsulated 3-Dimensional Culture

Abstract

Increased adiposity and hyperandrogenemia alter reproductive parameters in both animal models and women, but their effects on preantral follicles in the ovary remain unknown. We recently reported that Western-style diet (WSD) consumption over 1 year, with or without chronic exposure to elevated circulating T, increased the body fat percentage, elicited insulin resistance, suppressed estradiol and progesterone production, as well as altered the numbers, size, and dynamics of antral follicles in the ovary during the menstrual cycle in female macaques. Therefore, experiments were designed to compare the WSD and WSD+T effects to age-matched controls on the survival, growth, and function of isolated secondary follicles during 5 weeks of encapsulated 3-dimensional culture. Follicle survival significantly declined in the WSD and WSD+T groups compared with the control (CTRL) group. Although media progesterone levels were comparable among groups, androstenedione and estradiol levels were markedly reduced in the WSD and WSD+T groups compared with the CTRL group at week 5. Anti-Müllerian hormone levels peaked at week 3 and were lower in the WSD+T group compared with the WSD or CTRL group. Vascular endothelial growth factor levels also decreased at week 5 in the WSD+T group compared with the WSD or CTRL group. After human chorionic gonadotropin exposure, only antral follicles developed from the CTRL group yielded metaphase II oocytes. Thus, WSD with or without T exposure affects the cohort of secondary follicles in vivo, suppressing their subsequent survival, production of steroid hormones and local factors, as well as oocyte maturation in vitro.

As recently reviewed by Robker and colleagues (1), female fertility is negatively correlated with increasing body weight (body mass index [BMI], kg/m²), especially obesity (BMI ≥ 30). The infertility associated with obesity is of growing concern, considering the increase in the percent of population that is overweight worldwide, predominantly in developed countries (2). Moreover, a number of studies over the past decade reported detrimental responses in overweight and obese patients to infertility therapy, including i) higher doses of gonadotropin required for ovarian stimulation, ii) decreased numbers of large antral follicles and high-quality oocytes, iii) increased cancellation rates for assisted reproductive technology cycles, iv) greater rates of pregnancy failure, and hence v) reduced live birth rates (3, 4). Studies, particularly in rodent and domestic animal models, are investigating the effect of high levels of lipids (eg, free fatty acids) and adipokines (eg, adiponectin and leptin) on ovarian function (5–7). Collectively, there is mounting evidence that “lipotoxicity” and elevated levels of adipokines can impair follicular development, oocyte quality, and early embryonic development (5, 7). Correlative data from clinical infertility patients support this concept (8). However, to date, research on adipose-ovarian interactions in primates focused primarily on granulosa cells or cumulus-oocyte complexes from antral follicles (9–11), or the relationship between follicular fluid content and fertility outcomes (6, 10). Information regarding the effects of adiposity on preantral follicle development and function is lacking.

Adiposity, and its metabolic or endocrine effects, also reportedly exacerbates various features of polycystic ovarian syndrome (PCOS) (12, 13). A key feature of PCOS is hyperandrogenemia, and changes during obesity may enhance androgen production or action (13). Conversely, androgen excess may influence the level and distribution of adipose tissue (14). Investigators have attempted to evaluate the role of elevated androgens in the etiology and characteristics of PCOS using rodent, domestic animal, and, to a lesser extent, primate models (15–18). However, the combination of hyperandrogenemia with increased adiposity awaits exploration. Also, as noted above, ovarian research has primarily focused on the numerous small-to-medium antral follicles characteristic of PCOS. Information regarding their antecedent preantral follicles is very limited, except for a few studies on primate ovaries analyzing follicles in tissue slices or sections (19, 20), due to a lack of adequate in vitro models.

We performed a pilot study to examine the effects of chronic exposure to elevated T levels, beginning prepubertally (1 year of age), on the hypothalamic-pituitary-ovarian axis and insulin-sensitive glucose metabolism in female macaques (21). Following timely menarche, neuroendocrine effects were noted, with a higher LH response to GnRH and a greater LH pulse frequency during the early follicular phase in T-treated animals at 4 and 5 years of age, respectively, relative to the controls. There were no remarkable differences in insulin sensitivity or antral follicle numbers, but only half the monkeys displayed ovulatory cycles at this age (21). To consider the effects of adiposity and its possible interaction with hyperandrogenemia, these monkeys were switched to a typical Western-style diet (WSD) and experimental parameters monitored for another 18 months (22). As reported recently, not only did the monkeys' percent body fat increase remarkably to 14–19%, the numbers, size, and dynamics of the antral follicle pool also changed, with the level of circulating estradiol (E2) during the follicular phase and diameter of the largest antral follicle diminished in both the WSD and WSD+T groups. Chronic exposure to WSD+T also reduced peak progesterone (P4) levels during the luteal phase and notably decreased insulin sensitivity compared with the WSD group (22). After 18 months of WSD±T, a variety of tissues were collected for additional studies, eg, adipose tissue (23). The ovaries were collected for analysis of tissue sections (24), as well as isolation of preantral and antral follicles for culture and molecular evaluation (25). The current report describes the results of studies examining the survival, growth, and function (production of steroid hormones and paracrine factors, oocyte maturation) of secondary follicles isolated from WSD, WSD+T, and age-matched control monkeys during encapsulated 3-dimensional culture for 5 weeks.

Materials and Methods

Animals and ovary collection

The general care and housing of rhesus macaques was provided by the Division of Comparative Medicine, Oregon National Primate Research Center (ONPRC). Animals were housed in a temperature-controlled (24 ± 2°C) light-regulated (12 h light:12 h dark) room. Animals were treated according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and protocols were approved by the ONPRC Institutional Animal Care and Use Committee (21).

T implants and diet protocol were described in our previous reports (21, 22). Briefly, female rhesus macaques received sc T implants (n = 6) prepubertally beginning at 1 year of age, maintaining a 3.7-fold increase (P = .001) in circulating T levels over cholesterol-implant controls (n = 6) until the end of the study (7 years of age; young adults), based on the clinical evidence that PCOS patients have T levels approximately 3–4 times higher than controls. At 5.5 years, all 12 monkeys were placed on a high-fat, high-fructose diet (WSD), which was maintained for 18 months. Age-matched monkeys (n = 6) on a typical diet (Purina monkey chow supplemented with fresh fruit or vegetables; 26, 27) served as controls. Body weight, percent body fat, and BMI of the monkeys were not different between WSD and WSD+T groups as previously reported (22, 23).

Ovaries were collected from three randomly selected monkeys in each treatment group at their reproductive ages, 7 years for WSD and WSD+T-treated monkeys, 7.3 years for age-matched controls. All nine monkeys exhibited regular menstrual cycles and were evaluated daily for menstruation with the first day of menses termed day 1 of the cycle. Ovariectomies were conducted on anesthetized monkeys by laparoscopy at early follicular phase, day 1–4 of the cycle, as previously described (28). Ovaries were immediately transferred into SAGE OFC Holding Medium (CooperSurgical) at 37°C (26, 27).

Follicle isolation, encapsulation, and culture

Follicle isolation and encapsulation were performed as previously described (26, 27). All procedures were conducted at 37°C. Briefly, ovarian cortex was cut into 1 mm³ cubes. Follicles were mechanically isolated in the Holding Medium using 31-gauge needles. Secondary follicles (diameter, 130–220 μm) that displayed the following characteristics were selected for encapsulation: i) an intact basement membrane, i) healthy granulosa cells, and iii) a visible, healthy oocyte that was round and centrally located within the follicle, without vacuoles or dark cytoplasm. Follicles from each monkey (26 ± 3 follicles/monkey) were assigned for subsequent encapsulation.

Follicles were encapsulated individually into 5-μl alginate matrix containing 0.25% (w/v) sterile sodium alginate (FMC BioPolymers) PBS (137mM NaCl, 10mM phosphate, 2.7mM KCl; Invitrogen), which was crosslinked in 50mM CaCl₂, 140mM NaCl, 10mM HEPES solution (pH, 7.2). Each encapsulated follicle was transferred into individual wells of 48-well plates containing 300 μl α-MEM (Invitrogen) supplemented with recombinant human FSH (3 ng/mL from weeks 1–3 and 0.3 ng/mL from weeks 4–5; NV Organon), 0.3% (v/v) human serum protein supplement (CooperSurgical), 5 μg/mL insulin, 0.5 mg/mL purified bovine fetuin, 5 μg/mL transferrin, and 5 ng/mL sodium selenite (Sigma-Aldrich) (26, 27).

Encapsulated follicles were cultured at 37°C in a 5% (v/v) O₂ environment (in 6% CO₂/89% N₂) for 5 weeks. Follicles that reached the antral stage were treated with 100 ng/mL recombinant human chorionic gonadotropin (hCG; Merck Serono) for 34 hours. Oocytes were retrieved to determine their size and stage of meiotic maturation. Half of the culture media (150 μl) was collected and replaced every other day, and stored at −20°C. The media samples from each culture week were assigned to ovarian steroids, anti-Müllerian hormone (AMH), and vascular endothelial growth factor (VEGF) assays (27).

Follicle survival and growth

Follicle survival, diameter, and antrum formation were assessed weekly using an Olympus CK40 inverted microscope and an Olympus DP11 digital camera (Olympus Imaging America) as described previously (26, 27). Follicle photographs were imported into ImageJ 1.42 software (National Institutes of Health) and the diameter of each follicle was measured. Follicles were measured from the outer layer of cells that included a measurement at the widest diameter of the follicle and a second measurement perpendicular to the first. The mean of the values was calculated and reported as the follicle diameter. Follicles were considered to be undergoing atresia if i) the oocyte was dark or not surrounded by a layer of granulosa cells, ii) the granulosa cells became dark and fragmented, or iii) the diameter of the follicle decreased (26).

Follicle histology

Selected in vitro–developed antral follicles were fixed in 4% paraformaldehyde-PBS solution for 3 hours at room temperature. Follicles were embedded in HistoGel (Thermo Scientific) before being dehydrated in ascending concentrations of ethanol (70–100%) and embedded in paraffin. Five micrometer sections were cut by the Imaging and Morphology Support Core at Oregon National Primate Research Center (ONPRC), and stained with hematoxylin and eosin (27).

Ovarian steroids, AMH, and VEGF assays

One media sample collected weekly was analyzed for ovarian steroids by the Endocrine Technology Support Core at ONPRC. P4 and E2 concentrations were assayed using an Immulite 2000, a chemiluminescence-based automatic clinical platform (Siemens Healthcare Diagnostics) validated for macaque follicle culture media (29). Androstenedione (A4) levels were measured by RIA using a DSL-3800 kit (Diagnostic Systems Laboratories) also validated for macaque follicle culture media (30).

Another two media samples collected weekly were analyzed for AMH and VEGF concentrations by ELISA using a DSL-10–14400 kit (Diagnostic Systems Laboratories) and a Human VEGF Quantikine ELISA Kit (R&D Systems), respectively, based on the manufacturers' instructions (31, 32). Both kits were validated for macaque follicle culture media (30, 32). Due to the cross-reaction of fetuin with the AMH antibody, levels assayed in media without cultured follicles were subtracted from AMH levels in media samples as previously described (30).

Oocyte retrieval, maturation, and fertilization

Oocyte retrieval and evaluation were performed on a 37°C warming plate. Cumulus-oocyte complexes were released by breaking the follicle wall using 31-gauge needles in Tyrode's albumin lactate pyruvate (TALP)-HEPES-BSA (0.3%) medium. Cumulus-oocyte complexes were treated with 2 mg/mL hyaluronidase (Sigma-Aldrich) in TALP-HEPES-BSA for 30 seconds to dissociate cumulus cells and obtain denuded oocytes. Retrieved oocytes were transferred to TALP medium and photographed. Oocyte diameters (excluding the zona pellucida) and meiotic status were assessed using the same camera and software as described above (27).

Metaphase II (MII) oocytes were maintained in TALP medium at 37°C in a 20% O₂/5% CO₂/75% N₂ environment for conventional in vitro fertilization (IVF) as previously described (33) within 3 hours of oocyte retrieval. Semen collection was performed by the Assisted Reproductive Technologies (ART) Support Core at ONPRC as previously reported (34). The resulting zygotes were transferred to 500 μl hamster embryo culture medium-9 with 5% fetal bovine serum and cultured at 37°C in 5% O₂/6% CO₂/89% N₂ as previously described (35). Embryos were photographed daily to document development. Reagents and protocols for embryo culture were provided by the Assisted Reproductive Technologies Support Core.

Statistical analysis

Statistical analysis was performed with SAS v9.3 software (SAS Institute). Due to the limited sample size for formal statistical hypothesis testing, a permutation test (randomization test) was used to compare follicle survival rates between groups. A permutation test is a statistical tool that provides statistical inferences tied to chance mechanism in random assignment. Data represent three individual animals in each of the three treatment groups. There were a total of 1680 possible permutations to assign the data to three groups with group size of three. The1680 one-way ANOVAs were performed followed by pair-wise comparison between groups to generate pseudo distributions of groups differences. Follicle growth, steroid/AMH/VEGF production, and oocyte sizes were analyzed for each individual follicle with total follicle numbers indicated in the figure legends, and represent follicles obtained from nine individual animals. Mixed effect model was used with treatment group as between group factor and monkeys nested within the group as random effect. Tukey-Kramer correction for the multiple comparisons was used to correct overall type I error rate. Due to their skewed distribution, the data were transformed using logarithmic function with base 2. Differences were considered significant at P ≤ .05 and values are presented as mean ± SEM.

Results

During manual dissection, we observed that the stroma of ovaries from monkeys with the WSD±T treatment was less compact. The tissue was spongy and clearly atypical wherein follicles were not tightly associated with the stroma compared with ovaries from monkeys fed a typical diet.

Follicle survival

Although their diameters did not increase remarkably during the first 2 weeks of culture, greater than 90% of the secondary follicles from the control group survived, remaining intact with a round oocyte and healthy granulosa cells (Figure 1A). In contrast, greater than 50% of the secondary follicles from the WSD or WSD+T groups degenerated within 2 weeks of culture with a dark and collapsed oocyte (Figure 1A), or oocyte denuded from granulosa cells.

Figure 1.

The effects of WSD with and without in vivo T treatment on follicle survival in vitro. A, Follicle morphology at the beginning of culture (week 0) and culture week 2 for the CTRL and WSD groups. Scale bar = 100 μm. B, Follicle survival rate (surviving/total cultured follicles) at week 5 for the CTRL (37/52), WSD (24/108), and WSD+T (27/77) groups. Data are presented as the mean ± SEM with three animals per group. *, significant difference with P < .05; **, significant difference with P < .01.

Follicle survival was markedly lower (P < .05, permutation test) in both the WSD and WSD+T groups compared with controls at week 5 (Figure 1B). However, the WSD+T-treated follicles had a higher (P < .01, permutation test) 5-week survival rate than those of WSD alone (Figure 1B).

Follicle growth and histology

At the beginning of culture, diameters of the surviving follicles did not differ between experimental groups (control vs WSD vs WSD+T = 181 ± 5 vs 180 ± 4 vs 172 ± 5 μm). After 5 weeks, distinct cohorts of surviving follicles were observed based on their growth rate (30). Greater than 80% of surviving follicles from the WSD or WSD+T groups, whereas approximately 40% from the control group, increased their diameters by a minimum of 3-fold (> 500 μm; termed fast-grow follicles) (30). Because there were no differences among the three experimental groups in any other parameters analyzed for the no- or slow-grow follicle cohorts (30; data not shown), only data from the fast-grow follicles are described below.

Diameters increased over the culture period for fast-grow follicles (data not shown), with antrum formation within 3 weeks of culture in all groups (Figure 2A). Diameters of fast-grow follicles were comparable among all experimental groups at week 5 (Figure 2B). Hematoxylin and eosin staining of antral follicles from the control group revealed morphology similar to that observed in primate small antral follicles in vivo (36): a spherical shape with an antrum, a healthy germinal vesicle (GV)-stage oocyte with surrounding cumulus cells, an intact granulosa layer, and presumptive theca cells (Figure 2A). However, follicles in the WSD (Figure 2A) and WSD+T (data not shown) groups had a less-developed granulosa layer surrounded by an extracellular matrix-like structure.

Figure 2.

The effects of WSD with and without in vivo T treatment on follicle growth in vitro. A, Morphology of representative small antral follicles at culture week 4 from the CTRL and WSD groups. Scale bar = 100 μm. H&E, hematoxylin and eosin staining; OO, oocyte; AT, antrum; GC, granulosa cells. B, Diameters of fast-grow follicles at week 5 for the CTRL (n = 15), WSD (n = 21), and WSD+T (n = 21) groups. Data are presented as the mean ± SEM with three animals per group.

Follicular steroids

Cultured follicles secreted ovarian steroids into the media beginning at antrum formation with levels increasing through week 5, as observed in our previous studies (26, 27), regardless of the initial in vivo treatment (data not shown). At culture week 5, P4 levels were not different in the WSD or WSD+T groups compared with controls (Figure 3A). In contrast, A4 concentrations produced by fast-grow follicles decreased (P < .05) in the WSD+T, but not WSD, group relative to controls (Figure 3B). Follicles in both the WSD and WSD+T groups produced remarkably lower (P < .01) levels of E2 than controls (Figure. 3C).

Figure 3.

The effects of WSD with and without in vivo T treatment on follicle steroid production in vitro. A, Progesterone; B, androstenedione; and C, estradiol levels produced by fast-grow follicles at week 5 for the CTRL (n = 12), WSD (n = 21), and WSD+T (n = 21) groups. Data are presented as the mean ± SEM with three animals per group. *, significant difference with P < .05; **, significant difference with P < .01.

AMH and VEGF

AMH levels produced by fast-grow follicles were detectable in the media at the beginning of culture and peaked at week 3 after antrum formation, a similar observation to our previous studies (26, 27), in all experimental groups (data not shown). Although there was no difference between the WSD and control groups, WSD+T greatly decreased (P < .01) peak AMH concentrations compared with controls at week 3 (Figure 4A). AMH levels produced by WSD+T-treated follicles at week 3 were also lower (P < .01) than those of WSD follicles.

Figure 4.

The effects of WSD with and without in vivo T treatment on follicle AMH and VEGF production in vitro. A, Peak AMH levels produced by fast-grow follicles at week 3 for the CTRL (n = 12), WSD (n = 20), and WSD+T (n = 21) groups. B, Peak VEGF levels produced by fast-grow follicles at week 5 for the CTRL (n = 12), WSD (n = 21), and WSD+T (n = 21) groups. Data are presented as the mean ± SEM with three animals per group. *, significant difference with P < .05; **, significant difference with P < .01.

Fast-grow follicles from all experimental groups also produced VEGF, an angiogenic factor generated by follicles after antrum formation, with levels increasing over the 5 weeks of culture (data not shown) as previously reported (27, 30). WSD alone did not alter the VEGF levels at week 5 compared with the controls (Figure 4B). VEGF concentrations produced by WSD+T-treated follicles were lower than those of the WSD (P < .01) and control (P = .05) groups (Figure 4B).

Oocyte diameter, maturation, and fertilization

Following exposure of antral follicles to hCG, healthy, as well as degenerate (dark and condensed cytoplasm), oocytes were retrieved from all experimental groups (Figure 5A). Most of the healthy oocytes remained at the GV stage. Although there was no distinction between the WSD+T and control groups, diameters of GV oocytes were smaller (P < .05) in the WSD group compared with those of the control and WSD+T groups (Figure 5A). MII oocytes were only obtained from the control group with diameters over 120 μm, which were greater than any of the GV oocytes (Figure 5B). After insemination, the MII oocytes fertilized, cleaved, and the embryos developed to the morula stage at day 5 post-IVF (Figure 5C).

Figure 5.

Oocyte retrieval and embryonic development after IVF. A, Characteristics of oocytes retrieved from antral follicles at week 5 from the CTRL, WSD, and WSD + T groups. B, One of two MII oocytes retrieved from the CTRL group. PB, polar body. C, A morula-stage embryo developed in vitro from the inseminated MII oocyte by day 5 post-IVF. Scale bar = 60 μm.

Discussion

We recently developed an encapsulated 3-dimensional culture system that supports the growth of macaque preantral (primary and secondary) follicles to the small antral stage (1–2 mm in diameter), with associated steroid and paracrine factor production, as well as oocyte maturation (26, 27). This in vitro follicle maturation technique allows one to monitor individual follicles and their response to in vitro manipulations (eg, steroid depletion and repletion) (37) or in vivo conditions (eg, stage of the menstrual cycle or aging) (29, 30). This report describes the first study in primates to determine the effects of WSD alone and in combination with chronic exposure to elevated androgen levels in vivo on preantral follicle development in vitro. The data suggest that the WSD, presumably through the diet or increased adiposity (22, 23), negatively affects the cohort of secondary follicles in macaques in terms of their survival and steroidogenic function during culture. Moreover, the addition of chronic hyperandrogenemia, with a 3–4-fold increase in T levels reminiscent of those observed in adolescent girls at risk for PCOS (38), further reduced follicular AMH and VEGF production. Furthermore, such treatments also reduced oocyte growth (diameter, WSD group) and possibly meiotic maturation (WSD and WSD+T groups).

Some alterations in preantral follicles were comparable in both the WSD and WSD+T groups, suggesting a primary effect of diet or adiposity. Whereas secondary follicles from control monkeys typically survived during the first 2 weeks of culture, more than 50% of those in the WSD and WSD+T groups degenerated. We previously reported that macaque secondary follicles require FSH to survive during culture, and fewer follicles survive when cultured in low-dose (10% of current concentration) FSH (26, 30). Given that FSH was present in the culture media, one hypothesis is that the diet/adiposity leads to insufficient FSH receptor (FSHR) expression or signaling in the secondary follicles. Alternatively, the altered hormonal or local milieu caused by diet/adiposity may promote expression of local factors that induce apoptosis and preantral follicle atresia as noted in the rodent ovary (39, 40).

Nevertheless, a subcohort of the secondary follicles survived and grew in a timely manner to small antral follicles at approximately 1 mm diameter by week 5 of culture. However, the ability of these follicles originating from the WSD or WSD+T groups to produce the primary steroid E2 was markedly suppressed. Because P4 levels in the WSD and WSD+T groups did not differ from controls, the defect seems to be in the cytochrome P450 family 17 subfamily A polypeptide 1 and/or 19 subfamily A polypeptide 1 (CYP19A1; aromatase) enzymatic steps. Based on the two-cell model for estrogen production, in which theca cells generate androgen as a precursor (41), insufficient estrogen synthesis could be due to limited development of the theca layer or its function, including cytochrome P450 family 17 subfamily A polypeptide 1expression. Somewhat reduced (40%) A4 levels in the culture media could support such a lesion. However, the reduction in precursor A4 levels was markedly less than that (90%) for E2 levels, suggesting that the primary lesion is in the granulosa cell layer, which results in reduced expression or activity of CYP19A1. Granulosa cell proliferation could be suppressed in the WSD and WSD+T group as suggested by the histologic data. Furthermore, because it is generally accepted that a major action of FSH is to induce CYP19A1 in granulosa cells of the maturing antral follicle (42), one can again hypothesize that a deficit in FSHR or its signaling by adiposity (9, 11), or gonadotropin-regulated paracrine factors (43, 44) leads to the subsequent loss of estrogen synthesis by granulosa cells.

Other alterations in preantral follicles were more pronounced in the WSD+T, compared with the WSD, group, suggesting that elevated androgen alone or acting synergistically with WSD impaired follicular maturation. Notably, AMH and VEGF levels produced by cultured follicles were greatly suppressed by the WSD+T exposure, compared with WSD alone or controls. Our previous culture studies suggested that AMH production increased in growing preantral follicles, and declined or plateaued after antrum formation, whereas VEGF production increased during antral follicle development (27, 30). Thus, T exposure in vivo suppressed the production, and presumably action, of paracrine factors at both the preantral and antral stage in primate follicles in vitro. The decline in AMH levels produced by WSD+T-treated follicles is consistent with previous studies in which obesity (45) and hyperandrogenism (46) are associated with a decreased AMH production by growing follicles. We also noted that FSH promoted AMH and VEGF production by macaque follicles during culture (26, 47). Data from human granulosa cell culture suggested that androgen-receptor signaling increases granulosa cell sensitivity to FSH (48). Therefore, the hypothesized decrease in FSHR signaling and A4 action may limit AMH and VEGF production in WSD+T-exposed follicles.

After secondary follicles from control monkeys develop to the small antral stage in vitro, a small fraction (10–20%) of oocytes reinitiate meiosis and reach the MII stage in response to hCG treatment (26, 27). These oocytes typically achieve a diameter of greater than 115 μm. However, none of the small antral follicles from the WSD or WSD+T groups yielded MII-stage oocytes. This may be related to the inadequate E2 production by WSD- and WSD+T-exposed follicles, which is consistent with the observation that E2 improved the in vitro development of bovine cumulus-oocyte-complex and oocyte nuclear maturation (49). Notably, the mean diameter of the healthy (GV) oocytes from the WSD group was less than that of controls, suggesting that diet or adiposity impaired oocyte growth as noted in obese mice (50). However, diameters of GV oocytes from the WSD+T group were comparable to controls. Thus, androgen levels seem to counteract the effect of WSD on oocyte growth, but still impair their maturation. However, the number of oocytes analyzed from the WSD±T groups was relatively small, due in part to reduced follicle survival. Further investigation is needed on the effects of diet/adiposity with and without hyperandrogenemia on oocyte growth and maturation in primates.

The discovery of altered potential for secondary follicles from WSD and WSD+T groups to survive, grow, and mature in vitro broadens our perspective on the effects of these treatments on primate folliculogenesis. Our prior analyses focused on the antral follicle pool and its steroidogenic potential. Initially, ultrasound analyses determined that WSD±T increased the number of antral follicles but decreased the diameter of the largest antral follicle during the early follicular phase of the menstrual cycle (22). The decreased size of the largest antral follicle (1.0–1.5 mm) was important, because our earlier report noted that a follicle at least 2 mm typically present at or soon after menses was destined to ovulate at midcycle (51). Mean E2 levels during the early follicular phase in WSD and WSD+T animals were also reduced compared with pretreatment values. Ongoing histologic and immunohistochemical studies on sections of excised ovaries (n = 3) removed at the same time as the current studies, confirmed the greater number of antral follicles, primarily due to an increase in degenerating (atretic) antral follicles (24). The current data suggest that defects in the developmental potential of preantral (secondary) follicles precede and portend the dysfunctional antral follicles observed with WSD ± T. Indeed, it is possible that follicles at earlier stages (ie, primordial or primary stages) are also altered by the in vivo milieu (52, 53), leading to dysfunctional secondary and, subsequently, antral follicles. There are limited reports that, although preantral follicles of patients with PCOS seem similar histologically to those of normal women (54), their dynamics in terms of rate of growth or loss may be altered (55). Such alterations in preantral dynamics are likely important for subsequent abnormal antral follicle structure-function.

In conclusion, diet/adiposity seems to modulate preantral (secondary) follicle development and function, as monitored in vitro, which may be exacerbated by chronic hyperandrogenemia. Additional studies are warranted to determine whether treatment-induced deficits in endocrine or paracrine factors and/or their receptor signaling pathways result in impaired follicle survival and reduced steroidogenic function at the subsequent antral stage. Based on other reports, lesions in gonadotropin (particular FSH) or insulin receptor-signaling pathways (56) may be critical. Also, changes in paracrine factor expression or action, including those identified in this study (AMH and VEGF), could be important determinants. The specific features associated with WSD/adiposity, such as lipids, adipokines, inflammatory agents, as well as excess androgen action in vivo, that modify the subsequent developmental potential of secondary follies in vitro await elucidation. The direct effects of steroids and adipokines on macaque follicular development in vitro are under investigation. The current report describes an extended pilot study based on the experimental setting from previous work with limited animal resources (21–23). The findings provide the basis for an ongoing in-depth study (10 monkeys per group), including four animal groups in a 2 × 2 factorial as a function of control vs WSD and with or without chronic T exposure, to discern the effects of WSD/adiposity vs excess androgen exposure in peripubertal to young-adult female monkeys (age 2.5–7.5 y). Our findings may be relevant to understanding the onset of various features of PCOS in overweight adolescent girls and possible treatment.