Midgestation origins of androgenic and estrogenic pathophysiology in three major women's health disorders

David Barker first associated low birth weight with subsequent cardiovascular disease over 30 years ago, framing mechanistic understanding of sub-optimal gestational environments as major contributors to developmental origins of adult disease. Rapid onset, maternal immunological tolerance of the semi-allogeneic fetus is essential for optimal endothelial cell reprogramming within uterine spiral arteries, enabling optimal placental function. Not surprisingly, therefore, sub-optimal gestational environments arise when pre-existing maternal immunological (inflammatory), metabolic and endocrine disorders impair placental development and function. Increasing mechanistic insights into sub-optimal gestation are now inspiring technological innovations targeting the placenta, so as to engender a paradigm shift in clinical management towards early preventive intervention. The steroid hormones testosterone and estradiol play vital roles in differentiating sex-specific fetal characteristics. Sufficient placental aromatization of maternal and fetal androgens is crucial, not only for placental vascular health and function, but also for protection of female fetuses against maternal androgen-mediated virilization. In excess, testosterone can oppose estradiol action in females, diminishing estrogen receptor alpha (ESR1)-mediated progesterone receptor expression, thus diminishing progesterone action and compromising gestational environments. In addition, excess testosterone in female fetuses, from mice to nonhuman primates and humans, increases the likelihood of polycystic ovary syndrome (PCOS) at adolescence and throughout adult life, faithfully reconfirming developmental origins of adult disease, with altered anogenital and finger length ratio parameters providing quantifiable infant and adult biomarkers of fetal androgen and estrogen exposure. Human and nonhuman primate mid-gestational fetal ovaries express the full complement of steroidogenic enzymes and androgen receptors necessary to generate fetal female hyperandrogenism ( 1 ). In this issue, Peters and Lambalk describe a case-control study in which they demonstrate several associations between adult female anogenital parameters in 172 normal-to-overweight Dutch women in a control group (n=43), and those diagnosed with one of three female reproductive disorders with complex genetic-endocrine interactions: Mayer-Rokitansky-Kuster-Hauser (MRKH) syndrome (n=43), PCOS (n=43) and severe endometriosis (n=43) ( 2 ). The authors propose gestational steroid-mediated contributions to the developmental origins of each of these disorders. Their study examined women with MRKH, characterized by incomplete mullerian duct development with (type 1, 74% of MRKH subjects) or without (type 2, 26%) concomitant renal, skeletal or other anomalies; PCOS defined by Rotterdam criteria; and severe endometriosis of whom ~58% had undergone surgery ( 2 ). The control group comprised women undergoing IVF/ICSI due to male infertility. The authors measured anogenital lengths and finger length ratios as two accepted biomarkers of fetal androgen ( 1 ) and estrogen ( 3 ) exposure, demonstrating a high (>90%) degree of intra- and inter-observer reliability. Anogenital measurements were from the center of the anus to either the anterior labial commissure (AGDac), or the posterior fourchette (AGDaf). Caliper-assisted or digitally-imaged measurements of the 2 nd and 4 th digit length ratios (2D:4D) were taken from the ventral hand between the basal crease and fingertip, along the finger midline ( 2 ). AGDaf was longer in women with MKRH compared to the other three groups, while AGDac was longer in women with PCOS versus those with endometriosis. The longer AGDaf measure in MRKH subjects, however, was significantly associated with natural vaginal dilation due to sexual intercourse ( 2 ), therefore limiting applicability of these findings to MRKH in general, since vaginal construction, sexual activity and Frank’s method of vaginal dilation can affect the AGDaf. In this regard, it would be interesting to know the AGDaf of the 20.9% of women with MKRH who did not use any therapy for the creation of a vagina. Despite non-significant differences in AGDac lengths between women with these female disorders and controls, those with endometriosis tended to have shorter AGDac lengths, while hyperandrogenic women with PCOS tended to have longer AGDac lengths, with the women with MRKH in between compared to controls. AGDaf positively correlated with hirsutism score in women with MRKH, as did AGDac with hirsutism score across the entire study population, and also with biochemical hyperandrogenism in women with PCOS. 2D:4D finger length ratios were not associated with female group differences or reproductive measures. The significant female-type differences in anogenital measures, but not 2D:4D finger length ratios, suggest a transient midgestational exposure to androgen or estrogen excess. This is because anogenital distance and 2D:4D finger length ratios are unaffected by later gestation androgen excess, while earlier gestation onset of androgen excess virilizes external female genitalia and increases the 2D:4D finger length ratio ( 1 ). Such findings agree with previous reports implicating mid-gestational androgen excess exposure as a contributing factor in the pathogenesis PCOS, and is reinforced by multiple animal models ( 1 ). The inclusion of women with MRKH within the concept of prenatal androgen origins of disease, however, is unexpected. The authors speculate that hyperandrogenic women with MRKH, possibly from WNT4 gene variants ( 4 ), represent a discrete MRKH phenotype, possibly through prenatal androgen excess and overexpression of AMH. In support of this idea, diminished functional WNT4 gene variants among women with MRKH might prevent Mullerian duct development and impair inhibition of prenatal androgen production in a manner analogous to findings in female 46XX individuals with WNT4 deficiencies ( 5 ). It is unlikely, however, that exaggerated fetal ovarian AMH production can inhibit Mullerian duct development by nine weeks of gestation, since ovarian AMH is produced in late fetal life. More probable, any prenatal androgen excess in women with MRKH due to diminished functional WNT4 gene variants ( 4 , 5 ) may represent an epiphenomon rather than an underlying causal mechanism. The findings of a relatively diminutive anogenital length in women with endometriosis agrees with previous reports and the possibility of a mid-gestational estrogenic environment from androgen deficiency ( 3 ), as exemplified by both an increased risk of endometriosis in women prenatally exposed to diethylstilbestrol, and the postmortem findings of endometriosis in midgestational human female fetuses ( 3 ). Taken together, these findings raise environmental concerns about fetal female exposure not only to common estrogenic chemical disruptors in packaging and consumables as developmental programmers of endometriosis, but also to androgenic chemical disruptors such as the organometallic xenobiotic tributyltin in wood preservatives, fungicides, molluscicides, rodent repellants and antifouling paints as analogous programmers of PCOS and MRKH-associated hyperandogenemia. In closing, Peters and Lambalk’s paper ( 2 ) provides additional insight into the potential importance of gestational hormonal environments in contributing pathogenic origins to major reproductive disorders in women. Although the technological abilities to safely quantify circulating hormone concentrations in, or to obtain tissue from, human fetuses do not yet exist, improvements in measuring postnatal biomarkers of prenatal hormone exposure may provide clinically relevant indicators to target preventative interventions during maternal-placental-fetal development, infancy and/or childhood.