to adult, prenatally DHT-treated female mice normalizes neuronal connectivity between GABA and GnRH neurons, as well as circulating LH levels, negative feedback regulation of GnRH/LH, ovarian cyclicity and follicle morphology [88], suggesting that extant, extra-ovarian androgen excess is crucial to maintain neuronal reprogramming and its PCOS-like sequelae. Figure 2 illustrates hypothetical sites, implicated by animal models, for specific, androgen receptor-mediated, in utero programming of PCOS-like traits. The implication, emphasized by sheep and transgenic mouse studies, is that abrogation of androgen production and action in specific organ systems may effectively normalize PCOS traits.
Fig. 2. Hypothetical target sites for androgen receptor mediated in utero androgen excess female fetal reprogramming for reproductive and metabolic PCOS-like traits.
Naturally Occurring in utero Androgen Excess and Female Hyperandrogenism: Origins of PCOS beyond Humans?
Naturally occurring high levels of T confer athletic advantage in women [89] and accompany female behavioral dominance in some, but not all, nonhuman primates [90, 91]. About 15–25% of women in their reproductive years exhibit androgen excess, but most have PCOS [92]. Naturally occurring, hyperandrogenic (high T) female macaque monkeys emulate PCOS women in the co-occurrence of PCOS-like traits [93, 94]. In a multi-Primate Research Center study, adult female macaques with average BMI and in prime reproductive years were identified with high levels of T, 1SD above each population mean. Naturally occurring high T levels were observed in 3 separate laboratory populations of macaques. In the Wisconsin rhesus monkey population [94], PCOS-like traits of high T females exhibited generalized hyperandrogenism and increased steroidogenesis, including elevated circulating levels of unbound, “free” testosterone (unpublished results), hypersecretion of both LH and AMH, as well as uterine endometrial hyperplasia and infertility (Table 1). Intriguingly, a predominance of infertility and insulin resistance is found among high T monkeys with the most extreme elevations of T, 2SD above the monkey population mean (unpublished results), and equivalent to hyperandrogenic criteria required for peer-reviewed clinical studies of PCOS [95].
Table 2. Classification of PCOS-like phenotypes in PCOS-like adult female rhesus monkeys and in women with PCOS
High T, in and of itself, may not impair either fertility or metabolic function. Subtle changes in anatomical biomarkers of prior T exposure suggest fetal origins for naturally hyperandrogenic female rhesus monkeys [94], analogous to findings in behaviorally dominant female Malagasy lemurs [91], a more ancient branch of nonhuman primates. Interestingly, mild-to-moderate PCOS-like phenotypes were typically those identified among naturally hyperandrogenic female monkeys, emulating PCOS phenotype prevalence in studies of women recruited from local populations, and not from clinical referrals (Table 2). The contrast of predominantly mild-to-moderate phenotypes in naturally occurring PCOS-like monkeys and PCOS women recruited from local human populations to the predominantly severe and classic phenotypes of early-to-mid gestation T-exposed monkeys and clinically referred PCOS women suggests commonality in PCOS phenotype may include duration or degree of gestational T exposure, age, and BMI (younger age and normal BMI with more mild-to-moderate phenotype). Naturally occurring PCOS-like phenotypes beyond humans certainly support increasing speculation of survival and reproductive advantages from hyperandrogenic, energy-conserving, insulin resistant, delayed fecundity, female phenotypes [1, 96].
In utero Androgen Excess and Androgen Receptor: Developmental Commonality and Molecular Gateway to PCOS?
Mounting evidence from human and animal studies repeatedly implicates appropriately timed in utero androgen excess, from either maternal and/or fetal sources, as high risk for PCOS emerging at adolescence. Figure 1 provides a diagrammatic representation of maternal and fetal sources of gestational androgen excess, taken together with relevant PCOS risk genes, may programme for ovarian androgen excess. Figure 2 illustrates hypothetical sites for female reprogramming mediated by androgen receptor, as identified by genetically manipulated mouse studies [17, 18]. Such a unified hypothetical model is compatible with postnatal androgen-activated reprogrammed functions, such that anti-androgens or androgen-diminishing consequences of weight loss interventions, including lifestyle, diet, bariatric surgery, and insulin-sensitizing treatments, ameliorate PCOS traits in adulthood. Increasing sophistication of bioinformatics to assess risk for functional outcome of genomic and epigenomic variants vulnerable to in utero androgen excess, hold promise for identification of PCOS risk in newborn, enabling early intervention.
Acknowledgment
We thank Suzanne M. Moenter, PhD, for her comments on an earlier draft of this manuscript.
Funding Source
This work was funded, in part, by NIH grants P50 HD028934 (PI: Marshall), P50 HD044405 (PI: Dunaif), and P50 HD071836 (PI: Stouffer).
Disclosure Statement
The authors declare that they have no potential conflicts of interest to disclose.
References
1Escobar-Morreale HF: Polycystic ovary syndrome: definition, aetiology, diagnosis and treatment. Nat Rev Endocrinol 2018;14:270–284.
2Foecking EM, McDevitt MA, Acosta-Martínez M, Horton TH, Levine JE: Neuroendocrine consequences of androgen excess in female rodents. Horm Behav 2008;53:673–692.
3Moore AM, Campbell RE: Polycystic ovary syndrome: understanding the role of the brain. Front Neuroendocrinol 2017;46:1–14.
4Burt Solorzano CM, Beller JP, Abshire MY, Collins JS, McCartney CR, Marshall JC: Neuroendocrine dysfunction in polycystic ovary syndrome. Steroids 2012;77:332–337.
5Lie Fong S, Laven