Endocrine Disruptors of Sex Hormone Activities

Abstract

Sex hormones, such as androgens, estrogens and progestins are naturally occurring compounds that tightly regulate endocrine systems in a variety of living organisms. Uncontrolled environmental exposure to these hormones or their biological and synthetic mimetics has been widely documented. Furthermore, water contaminants penetrate soil to affect flora, fauna and ultimately humans. Because endocrine systems evolved to respond to very small changes in hormone levels, the low levels found in the environment cannot be ignored. The combined actions of sex hormones with glucocorticoids and other nuclear receptors disruptors creates additional level of complexity including the newly described “dynamic assisted loading” mechanism. We reviewed the extensive literature pertaining to world-wide detection of these disruptors and created a detailed Table on the development and current status of methods used for their analysis.

INTRODUCTION

Environmental pollutants, many of which are man-made chemicals, can affect the endocrine (hormonal) systems and some endocrine disruptors (EDCs) interfere with normal developmental processes in humans and wildlife species. The World Health Organization (WHO) defines EDCs as an “exogenous substance or a mixture that alters function(s) of the endocrine system and consequently causes adverse health effects in an intact organism, its progeny, or (sub) populations” (WHO, 2011). The Scientific Statement of the Endocrine Society postulates that EDCs have effects on female and male reproduction, metabolism, obesity, neuronal function, endocrine and cardiovascular systems, and several cancer types (Diamanti-Kandarakis et al., 2009). Diverse pathways contribute to the ubiquitous presence of EDCs in aquatic ecosystems. Water in rivers and lakes, discharge from wastewater treatment plants, direct runoff from urban and agricultural sites, and the atmosphere can be contaminated with hormones and chemicals which bind to and activate or block hormonal receptor activities in humans and other species (Ciparis et al., 2012; Gavrilescu et al., 2015). PMID: 32707118). In 2018, the Lancet Commission on Pollution and Health used 2015 Global Burden of Disease data and calculated that the combined pollution of air, water and soil by chemicals was responsible for 268 million disability-adjusted life-years and 9 million premature deaths each year—more than 3 times the total annual deaths caused by HIV/AIDS, tuberculosis, and malaria—with the greatest burden in low- and middle-income countries (Landrigan et al., 2018a; Landrigan et al., 2018b).

The endocrine system evolved to respond to very low levels of hormones (Eick and Thornton, 2011; Vandenberg et al., 2012; Welshons et al., 2003). Because of common receptor-mediated mechanisms in humans and other species, EDCs that mimic natural hormones are likely to have biological effects at the low levels documented for several contaminants (Hayes et al., 2011; Sheehan, 2000; Welshons et al., 2003). Similarly, EDCs that influence the synthesis, metabolism, uptake or release of the natural hormones, also have effects at low doses, because small changes in hormonal concentrations have biologically important consequences (Vandenberg et al., 2012; Welshons et al., 2003). The finding that chemicals have adverse effects on wildlife and humans and the range of environmental exposures clearly indicate that the low doses of EDCs cannot be ignored.

In this review we focus on sex hormones and their mimetics whose effects reach far beyond reproductive health and development: androgen (A), estrogen (E) and progesterone (P), because substantial experimental and epidemiologic data support their presence in the environment and potential effects as EDCs. We reviewed extensive literature on steroids in the environment and methods used for their detection, studies on variety of organisms, including fish and amphibians (Scholz et al., 2013). We have included glucocorticoids (G) in this review because activation of glucocorticoid receptor (GR) substantially alters the biological effects of other nuclear receptors, as described in the GR section. This information underscores the complexity in establishing guidelines for “acceptable” levels of EDCs, especially when a mixture of EDCs is detected in environmental samples (see Table I). In China alone, it has been estimated that ~3,000 t/yr of steroids are released annually in the environment and are only marginally reduced to >2,000 t/yr by various recently established waste treatments. From these contaminants ~ 80% are discharged into the surface water and ~20% into the soil (Zhang et al., 2014). For example, bisphenol A (BPA), originally synthesized as an estrogenic compound, was later found to activate not only ER but also pregnane X receptor and estrogen related receptor γ. BPA also activate or repress AR and thyroid hormone receptors (TR) in different systems, and these effects can be transgenerational (Rochester, 2013; Watts et al., 2001). BPA is a high production volume chemical widely used in manufacturing polycarbonate plastics and epoxy resins for industrial use, and it is frequently found in environmental samples (EPA; https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-bisphenol-bpa).

Table I:

Methods of Detection in the environment

Assay type	Endpoint	Advantages	Disadvantages	References	Notes
Cell Proliferation Assay
E2 and estrogenic pesticides screen using MCF-7 cell line. ER: Use of E2 dependent cell line MCF-7 to quantify physiological response.	Count nuclei after E2 or xenobiotics for 6 days. PR expression added. Expression of selected 4 marker genes in response to endogenous ER activation	Sensitive: E2 at 10 pg/mL. Mixtures found to have cumulative effect. Comparative analysis of agonist and antagonist activity to E2 and phytoestrogens at pM concentration.	Proliferation is an indirect effect, has high variation and potential toxicity. Cells cultured for 6 days in charcoal free media. High variation and potential toxicity.	(Soto et al., 1994) (Soto et al., 1995) (Jorgensen et al., 2000)	First assay for EDCs
HPLC/LC/MS - Single Hormone, slow and expensive
AR: Large-volume direct injection by HPLC and tandem mass spectrometry (LVI-HPLC-MSCMS).	Detection of known androgens in wastewater inflow site collected over a day.	Sensitive, from 1.2ng to 360 ng/L). Temporal profile in ng/hour.	Identifies and quantifies only known structures.	(Backe et al., 2011)	Precise and reproducible, at environmentally relevant concentrations.
ER: Automated solid phase extraction coupled to HPLC-tandem MS detecting sub-ng/L ER: Automated solid phase extraction coupled to HPLC/MS with fluorescence detection.	Detects known estrogens in surface water Detects known estrogens and their metabolites.	Sensitive, at sub-ng/L levels. Automated sample handling. Identification sub- ng/L of E in water. Structure-function relationships.	Low throughput. Identifies and quantifies only known structures Low throughput and expensive.	(Ciofi et al., 2013) (Guedes-Alonso et al., 2015)	Precise and reproducible, at environmentally relevant concentrations. Useful for confirmation with known compounds.
HPLC/LC/MS multiple targets
A, P: UPLC/MS + Solid phase extraction A, E, G, P: UHPLC-HR- Q-Orbirtrap-MS multiresidue method A, P: Fabric phase sorptive extraction followed by UHPLC- MS/MS A, G, P: Solid Phaseextraction-rapid resolution tandem RRLC-MS/MS with electrospray ionization. A, E: Liquid chromatography /RIA for simultaneous detection.	Quantification of known androgenic and progestins. Measurement of 70 known steroidal EDCs Detection of 4 progestins and 6 androgens Water, fish food and sediment samples surrounding Hailing Island, China Detection in small plasma sample in a clinical laboratory.	Detection in the influent, effluent and surface water. Sensitivity in ng/L range Simultaneous quantitation in sea and freshwater. Sensitivity in ng/L range Detects 24 known steroids in aquatic organisms and age- dependent human intake calculated up to >8,000ng/d Reverse phase chromatography microcolumns.	Quantitative simultaneous detection of 9 known compounds Detects only compounds with known chemical structure Detects 10 compounds with known chemical structure. Not suitable for compounds with unknown chemical structure. Sensitivity improved in later studies.	(Chang et al., 2008) (Huysman et al., 2017) (Guedes-Alonso et al., 2016) (Liu et al., 2015) (Hamalainen, 1982)	Overall recoveries from 78 to 100%, with no apparent signal suppression. Excellent reproducibility and recovery.
Ligand Binding Receptor Assays - Single Hormone
AR: Chimpanzee (chAR) in cell-binding assay with transcriptional activation and nuclear translocation of the complex.	AR binding and nuclear translocation in several cell lines.	Human and chAR translocation was indistinguishable. Suitable for screening.	Chimpanzee AR affinity to other ligands has not been determined.	(Hartig et al., 2008)
ER: Binding to GST-ER LBD derived from different species using cellular or nuclear extracts of tissues to radiolabeled H3 17β-estradiol. ER: Fluorescence polarization, competition binding of human recombinant estrogen receptors to fluorescein- labeled EREs.	Competitive binding and Kd of ER from different species Compared estrogen with phytoestrogens binding to ERE from the Xenopus and human ERα and Erβ.	Sensitive, quantifies differential ligand preferences from each species Performed at RT, fast, ligand- receptor and receptor-response element interactions.	Use of radiolabeled ligand; detects affinities of all chemicals to ER Sensitivity in high nM range.	(Matthews et al., 2000) (Nikov et al., 2000)
Ligand Binding Assays on Multiple Targets
AR, ER: Competitive ligand binding confirmed by GC-MS and/or LC- MS/MS. Continuous and simultaneous sampling of water inflow and outflow.	Fluorescent Polarization. Shows removal efficiency of 2 wastewater treatment plants to be 51–55%.	Large fraction of contaminants with unknown structures. Few confirmed by GC- MS.	Higher level of contaminants is required to establish IC50	(Liu et al., 2009)
ERα, Erβ: GST pull- down/FRET/ two-hybrid assay	3H labeled - association with coactivators	ER agonists and antagonists	Low throughput, does not measure direct ER activity.	(Mueller et al., 2003; Routledge et al., 2000; Zhou et al., 1998)
Yeast Reporter assays
ER: Yeast based reporter assay AR: 2 screens: β-gal reporter and yEGFP assay does not require cell lysis. AR: hARE β-gal reporter. Screen for agonists and antagonists papermill effluents. ERα, Erβ, PR: newly established mutant yeast strains lacking drug efflux pumps + SRC1. ER, AR, RXR, and TR: Bacillus subtilis Rec assay and used GC/MS to identify 51 PAH known compounds in local fish muscles.	Hormone binding in MCF-7 cells. 96-well format for yEGFP; 24h incubation. β-gal activity over 16–24h time range. 96-well format. β-gal activity after 18h incubation. 96- well plate format. 96-well based assay overnight incubation, ligands at EC50 for antagonist’s screen.	Dextran-coated charcoal assay EC50 = 50nM by testosterone. AR activity at 0.1 to 1.2 μg/L eq. of testosterone. ER EC50 in pM and AR nM range. Agonist/antagonist activities screened. Fish muscle levels replicated water and soil.	Low sensitivity Screen for antagonists using known ligand. Not as sensitive but specific for AR; slow Cross-reactivity with androgen and thyroid ligands; slow Genotoxicity in yeast may not be comparable to mammalian cells.	(Metzger et al., 1988) (Bovee et al., 2007) (Chatterjee et al., 2007) (Ito-Harashima et al., 2020) (Chou et al., 2006)	First reports of yeast with functional ER
In Vitro Transcriptional Activation Assays - Single Hormone
AR: Stable cells CV-1 express hAR with MMTV promoter-driven luciferase gene.	Luminescence in response to AR activation	Sensitivity nM range. CV-1 cells express low level endogenous receptor.	Maximal Luminescence in 18–24 hours	(Hartig et al., 2002; Legler et al., 1999)
ER-CALUX assay in T47D cells. Stable transfection of pERE- TATA-Luc.	Luminescence screen for inhibitors and effect of additive mixtures	Detection of 0.5 pM, EC50 6 pM, 100-fold induction.	Exposure for 24 hours; potential changes in cellular context.	(Legler et al., 1999)
In Vitro Transcriptional Activation Assays - Multiple targets
AR; ER; GR: Luciferase activity in cells stably expressing hormone response element- luciferase promoterreporter gene construct. ER; ERα, ERβ, GR and PR: CALUX assay multimerized response elements coupled to the TATA box/luciferase reporter gene. AR, ERα, ERβ, GR, MR and PR: human line transformed with the MMTV/luciferase/neo reporter gene. (MDA-kb2) AR, ERα, ERα, GR, MR and PR: Gal4 with DBD and LBD chimeras in HEK293T cells with β-lactamase reporter gene	96-well standard curve luciferase assay in stable cell lines responsive to A(MDA), E(T47D) and G(CV-1). Measure luciferase activity U2-OS cells. Specific for AR, ERα, ERβ, GR and PR. Luciferase in response to androgen- and glucocorticoid- receptor activation. Overnight incubation. Fluorescence by LiveBLAzer substrate	Sensitivity in pM- nM range and 3080-fold induction Stable cell line, 96- well format. 2 days assay. 384-well format; pM range; Independent measurement for each receptor. Fast, receptor- based screen of 35 compounds for agonist/antagonists	Detection of antagonists requires second passage. Detection of antagonists with second passage. Indirect effects may arise in over 2 days. Requires specific inhibitors to distinguish each receptor activity. Some compounds may have different affinities for full length receptors.	(Conley et al., 2017) (Sonneveld et al., 2005) (Wilson et al., 2002) (Wilkinson et al., 2008)	Effect-based assay to quantify and screen for chemicals with NR activity in environmental samples. Considered as one of the most sensitive and specific reporter assays developed in a mammalian cell line. Can be used for screening of chemicals with agonist and antagonist activities. Useful for rapid screen of individual HR
High-throughput in vitro Assays (HTS) - Single Receptor
AR: qHTS screen of Tox21 library for activators using 2 cell	Identified 75 compounds; 16 also confirmed by	Identified potentially novel AR agonists	Antagonist’s screen was not performed but is	(Lynch et al., 2017)
lines: full length hAR-luc (MDAkb2) and rat AR- LBD-β-gal (HEK293). GR: Enzyme fragment complementation in CHO- K1 cells/β-galactosidase activity detected by chemiluminescent substrate. ER: qHTS screen for agonists/antagonists from Tox21 library in 2 cells: full length ER-luc (BG1); ER-LBD-β-gal HEK293	biochemical binding and nuclear translocation. Indirect; detects GR translocation to the nucleus. 1536-well plate format, no washes. Chemiluminescence is more sensitive; β-gal is more specific. Triplicates and 15 dilutions of each.	confirmed by independent approaches. Rapid ~3h, large- scale HTS, screened 9,920 compounds at 7–15 concentrations; Dex EC50 of 6nM Structure-activity relationships of known and potentially novel ERa active classes.	possible in current settings. Screen requires additional studies depending on cell line used for each assay. Multiplexed assay has potential cytotoxicity among 2 cell lines.	(Zhou et al., 1998) (Zhu et al., 2008) (Huang et al., 2014)
High-throughput in vitro Assays (HTS) - Multiple Receptors
AR, GR: Cell lines expressing GFP-tagged receptors. Few identified by HPLC/MS and transcriptional activation confirmed in living cells. AR, GR, TR and AhR: Cell lines expressing GFP- tagged receptors. ER, AR, RXR, and TR: lacZ reporter assay and Bacillus subtilis Rec-assay	qHTS by GFP imaging of nuclear to cytoplasmic ratio Contamination found in 27–35% of river samples. qHTS by GFP imaging of nuclear to cytoplasmic ratio Wide level of contamination. Genotoxic potential in samples from water, soil and local fish extracts.	Rapid: 30min, pM sensitivity, wide range of doses. No nonmonotonic dose-response effects. As above Rapid and inexpensive	Identifies activity in complex mixtures no need for known chemicals from biodiversification. As above Genotoxic potential of chemicals	(Stavreva et al., 2016) (Stavreva et al., 2021) (Ito-Harashima et al., 2020)	Rapid and reproducible live cell assay with concentrations estimated from standard curve. As above Prioritizes compounds for further in-depth evaluation
Low Throughput
ERα, PPARγ: FRET/two- hybrid assay (FRET-based coactivator association)	Saturation kinetics for coactivators: strong ERα - SRC 1 interaction.	Differences depending on cellular background.	Low throughput. Additional studies needed depending on cell lines used.	(Zhou et al., 1998)
Whole Organisms assays
AR: An alternative, in vivo test based on a fluorescent reporter gene driven by the spiggin promoter in transgenic Medaka for detection of androgenic and antiandrogenic chemicals AR: Changes in sexual behavior in male Xenopus laevis by p,p'- Dichlordiphenyldichloroet hylene. Sexual differentiation embryonic or larval life- stages, usually extended through periods of gonadal differentiation and development Eight assay types on aquatic organisms reviewed: ligand binding, cell-free assays, yeast	GFP expression in the kidney of newly hatched stickleback 3–4 days of exposure. Suitable for detection of androgens and antiandrogens. DDE displays both, estrogenic and antiandrogenic effe cts Tested on fathead minnows, medaka, and zebrafish Extensive review of multiple assay types for EDC and effects on fish and	AR-specific screening for environmental samples (3–4 days). No need for identification of ligand chemical structure. Antiandrogenic mode of action Water exposure depends on solubility in water. Can be administered ip/oral feed Fish and amphibians, some assays have been adapted to	Highest response to relatively high levels of hormone, at 1.5 μg/L of 17MT. Not explored for screening or for other contaminants. Expensive and long-term studies require large amount of starting material. Only applicable to compounds with known hormonal effects.	(Sebillot et al., 2014) (Hoffmann and Kloas, 2016) (Ankley and Johnson, 2004) (Scholz et al., 2013)	Alternatives to other in vivo tests. Review: Small fish models for identifying and assessing the effects
reporter assays, cell line reporter assays, primary cells, embryos, organ culture and tail assays.	amphibians.	mammalian tissues and cells.			of endocrine-disrupting chemicals

Abbreviations:

A - Androgen; AR - androgen receptor; AhR - aryl hydrocarbon receptor; E and E2 - estrogen; ER (ERα, Erβ) - estrogen receptors; G - glucocorticoid; GR – glucocorticoid receptor; MR - mineralocorticoid receptor; P – progestins; PR - progesterone receptor; PPARγ - Peroxisome proliferator-activated receptor gamma; RXR - retinoid X receptor; TR - thyroid receptor; CALUX® (Chemically Activated LUciferase eXpression), DHT dihydroxy testosterone; FRET – fluorescence resonance energy transfer; GFP - green fluorescent protein; GC – gas chromatography; HPLC - high pressure liquid chromatography; LBD - ligan binding domain; LC - liquid chromatography; MMTV - mouse mammary tumor virus promoter; MS – mass spectroscopy; 17MT - 17α methyltestosterone; qHIT – quantitative high throughput; UHPLC-HR-Q-Orbirtrap-MS - ultra-high-performance liquid chromatographic high-resolution Q-Orbitrap mass spectrometer; UPLC/MS - ultra performance liquid chromatography-electrospray tandem mass spectroscopy.

The most sensitive window for effects induced by sex EDCs is during fetal development, which may have long-lasting consequences in humans and other organisms (Philippat et al., 2017). Some EDCs affect fetal development in late pregnancy (Ohtani et al., 2018) whereas others are harmful early on, even before the woman is aware of her pregnancy (Fucic et al., 2018; Grindler et al., 2018). The harmful effects of prenatal exposure to EDCs may not be immediately apparent and could manifest later in life, making it difficult to discern from other causes and some EDCs lead to harmful traits that are carried over to future generations (transgenerational effects) (Brehm and Flaws, 2019). These transgenerational effects frequently have epigenetic origin. Although they do not induce changes in DNA sequence, their effects are, nonetheless, inheritable (Ankolkar and Balasinor, 2016).

Reproduction in vertebrates is under tight control by the hypothalamus-pituitary-gonadal axis which is likely to be disrupted by inappropriate exposure to hormones. In seasonal spawning fish and amphibians external factors, such as light and water temperature, control the timing of gonad recrudescence and maturation via the brain. Thus, long-lasting effects on male and female fertility in variety of species are particularly of concern (Rattan et al., 2017; Zoeller et al., 2014). Changes in social structure and later age at first pregnancy in the last 3–4 decades in humans may account for the decline in fertility rates which has fallen below replacement levels in Japan and Europe. However, other evidence suggests additional causes including biological decline in world-wide male and female human and animal fertility (Vander Borght and Wyns, 2018).

Disproportionally higher exposure to hazardous chemicals among ethnic minorities, described as “environmental injustice”, is a special risk factor in adverse health outcomes (Attina et al., 2019; Fiscella et al., 2000; Landrigan et al., 2018a; Morello-Frosch et al., 2011). Environmental injustice is defined as the inequitable exposure of poor, minority, and disenfranchised populations to EDCs, toxic chemicals, contaminated air and water, unsafe workplaces and other forms of pollution which result in disproportionate health decline in these populations (Fiscella et al., 2000).

Environmental injustice has been characterized as a form of violence linked to so-called “structural racism” (Bailey et al., 2017). One example is exposure of Hispanic farm workers in California to organophosphate pesticides, such as parathion, an EDC linked to sex organ dysfunction (Landrigan et al., 2018a). Organophosphate pesticides are the largest contributor to costs associated with EDC exposure, calculated as $121 billion in the European Union, and $42 billion in the US (Attina et al., 2016). Other studies documented that non-whites have higher exposure to phthalates, which increases the incidence of endometriosis and other conditions associated with female reproductive tract disfunction (James-Todd et al., 2016). A significantly higher exposure to diabetogenic EDCs, such as BPA and phthalates has been reported in Latino and Black populations in the US (Attina et al., 2019; Ruiz et al., 2018). Some of these workers are undocumented immigrants and, hence, unable to protest environmental injustice and pollution (Cushing et al., 2015). Factors cited by the UN Environmental Program and the WHO that contribute to high exposure to EDCs in these populations include differences in quality of food, consumer products and other environmental conditions http://www.who.int/ceh/publications/endocrine/en/; all of which are driven, at least in part, by socioeconomic status (Bailey et al., 2017; Hicken et al., 2012). In addition, increased overall cancer risk is found in workers and residents living in the vicinity of cement, lime, and plaster plants (Garcia-Perez et al., 2015; Jansson et al., 2005). Testicular cancer, specifically linked to exposure to androgens in utero, is significantly increased in families working in construction and plastic industries (Garcia-Perez et al., 2015; Jansson et al., 2005). All of these cancer types are associated with disturbances of the estrogen/testosterone ratio or the expression and distribution of their receptors (Fucic et al., 2018; Jukic et al., 2017; Nagaraja and Eslick, 2014; Wu et al., 2018).

At present, EDCs in the environment are not efficiently monitored partly because their detection has previously relied on laborious chemical methods. More importantly, when the identity of a candidate EDCs is unknown and therefore absent from chemical libraries prepared for analysis, the specific EDC remains undetected. Because environmental microorganisms, algae and other biological processes can metabolize and transform the contaminants initially released into the environment, their derivatives may not be known and absent from chemical libraries, and their structures undetectable by chemical methods (Kinani et al., 2010; Stavreva et al., 2012b). Of critical importance is that it is difficult, if at all possible, to establish whether EDCs detected by chemical methods elicit biological responses in mammalian systems and wildlife. Thus, the existing chemical approaches for detection and identification of EDCs are of little biological value, expensive and incompatible with large-scale screening.

Many novel approaches for detection of EDCs have been developed in the last decade to test samples without knowledge of their chemical structures. These use reporter systems, receptor activation and nuclear translocation assays for high-throughput screening. These methods are less expensive, faster than chemical detection methods and can be adapted for large-scale screening. Some do not distinguish between agonist and antagonist activities and require a second pass testing. However, biological activities of EDCs detected by these methods can be quantitated by comparison with activities generated by natural hormones. We review these approaches and provide information on their relative sensitivities, cost and time involved in obtaining results. We also review the abundant literature over the last several decades on the potential consequences of environmental contamination with sex EDCs and provide possible solutions for efficient monitoring and establishing consensus to develop guidelines to protect human and animal life.

The emerging overwhelming concept is that contamination of water, air and soil with sex hormones EDCs threatens the integrity of aquatic and other ecosystems and poses a serious concern for human and animal health (Deblonde et al., 2011; Diamanti-Kandarakis et al., 2009).

ANDROGENS

Testosterone and its more active metabolite converted in target tissues, 5α-dihydrotestosterone (5α-DHT) (Goodman et al., 2015; La Spada et al., 1991; Matsumoto et al., 2013; Ortiz-Flores et al., 2019; Rosenfield and Ehrmann, 2016), are naturally occurring sex hormones critical for development and maintenance of male and female fertility. Human testosterone is mainly produced by testis in males and the ovary in females. In addition to male sexuality and reproductive health, androgens are also critical for female reproduction. Dysregulation of androgens in females results in polycystic ovary, which is associated with infertility, obesity, insulin resistance and other disorders (Goodman et al., 2015; La Spada et al., 1991; Matsumoto et al., 2013; Ortiz-Flores et al., 2019; Rosenfield and Ehrmann, 2016). Exposure to androgens and drugs with androgenic activity in utero leads to female and animal pseudohermaphrodism (Grumbach and Ducharme, 1960; Saunders, 1968). In addition, androgens have been linked to development and progression of prostate cancer, and most prostate cancer patients respond to androgen deprivation. The link to human malignancies is not unique for male population: exposure to androgens is also linked to breast and ovarian cancers (Bleach and McIlroy, 2018; Mizushima and Miyamoto, 2019). The androgen receptor (AR), a canonical nuclear receptor, is responsible for all physiological effects of androgens. Following ligand binding, AR becomes a powerful transcription factor (TF) that regulates expression of many genes including those that determine sexual differentiation, growth and survival of prostate cells. AR activity has been implicated in the pathogenesis of prostate and breast cancers.

Testosterone and many synthetic compounds which were developed in the 1970’s for their androgenic effects, also cross-react with other nuclear receptors, including estrogen, progesterone and thyroid receptors (Gonsioroski et al., 2020). The classic example is BPA, developed as a synthetic analogue for estrogen, but at higher concentrations activates AR and the thyroid receptor. BPA is one of the highest production-volume chemicals. In spite of recent restrictions, BPA is widely used in manufacturing polycarbonate plastics and epoxy resins for industrial use (https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/risk-management-bisphenol-bpa).

The increased rate of prostate and breast cancers in humans and dysregulation of human and animal fertility has been repeatedly linked to environmental contaminations with compounds having androgenic or anti-androgenic activities (Di Nisio and Foresta, 2019; Eertmans et al., 2003). These contaminants are especially prevalent downstream of intense urbanization and livestock production. In recent decades, many studies revealed world-wide contamination of water, soil and other environmental sources with agonists and antagonists of androgen alone or in combination with compounds which also mimic estrogen, thyroid hormones and/or progesterone activities (Hotchkiss et al., 2008; Schug et al., 2016; Soto et al., 2004).

The US EPA (United States Environmental Protection Agency) and WHO have performed screening studies to detect EDCs but have done little to widely monitor and establish guidelines to ameliorate the contaminations. The regulations are, at best, a patchwork with little regard for the existing publications and scientific evidence. For example, DDT metabolite p,p’-DDE, a potent androgen receptor antagonist, was banned by the Stockholm Convention according to WHO guidelines over a decade ago (http://chm.pops.int/TheConvention/Overview/TextoftheConvention/tabid/2232/Default.aspx). However, because no other effective and inexpensive chemicals are available to date, DDT is still widely used for control of malaria and other parasites in Africa, India, China and South America (Kelce et al., 1995; van den Berg et al., 2017). Thus, there is an urgent need to develop other malarial mosquito controlling agents. Other examples of potent anti-androgenic compounds include α-zearalanol (α-ZA) and its derivatives, pesticide metabolite M2, cosmetics such as benzophenone 2, fungicide vinclozolin and some bisphenols, such as chlorinated BPA and BPC (Delfosse et al., 2012; Molina-Molina et al., 2013).

Of importance, monitoring androgenic activities in the environment by previously reported methods has been problematic because most of the efforts have been directed to the identification of compounds with known chemical structure which may not be present in the chemical libraries and requires substantial amounts of starting material. In addition, the contaminants are present in small amounts the water and other sources which sometimes are insufficient for laborious chemical identification, and the effects may not correlate with the in vivo activities and thus not be physiologically relevant (Backe et al., 2011; Chang et al., 2008; Huysman et al., 2017; Scholz et al., 2013). Only recently studies have shifted to functional detection of agonists and antagonists activities using cell-based yeast and mammalian cell reporter assays, quantitative receptor translocation and whole organisms assays in amphibians, fish and mammalian cells (Ankley et al., 2003; Ankley and Johnson, 2004; Hoffmann and Kloas, 2016; Liu et al., 2009; Scholz et al., 2013; Stavreva et al., 2019; Stavreva et al., 2012b). These approaches, summarized in Table I, are not limited by identification of known chemical structures.

ESTROGENS

Estrogens and their synthetic mimetics are steroid hormones characterized by the shared ability to bind to and activate Estrogen Receptors (ERα and ERβ). These receptors have differential affinities and expression patterns in target organs, and thus regulate many functions including bone mineralization, immunity, metabolism and other vital biological processes in addition to female and male reproduction (Hamilton et al., 2017). Estradiol-17β (E2) and estriol (E3) are produced from testosterone in the ovaries, whereas the weaker estrone (E1) is derived from androstenedione in adipocytes in mammals. The levels of endogenous estrogens in the blood vary dramatically depending on sex, age, and physiological state (such as stage of the menstrual cycle, adiposity, pregnancy, etc.).

In addition to the naturally occurring estrogens, many other natural and synthetic substances are estrogen mimetics which can bind ERs and modulate their activities. The presence of such substances in the environment has been known for over a century. For example, farmers noted a relationship between clover-rich diet and significant reproductive disorders in cows and sheep. It was later discovered (Adams et al., 1995) that certain clover species contain high amounts of estrogen-activating compounds. As industrial hormonal synthesis and world-wide use exploded during the 1940’s, toxicologists began to notice their presence in the environment and report their effects on organisms. Naturally occurring and synthetic estrogenic compounds were found in United States wastewater in 1965 (Stumm-Zollinger and Fair, 1965) and reported again in 1970 (Tabak and Bunch, 1970). The public concern for estrogenic chemicals in the environment increased in the mid-1990s, after a study using yeast ER transcriptional activation assay claimed that estrogen-receptor activating chemicals can be synergistic. Although these findings were not reproduced, and the study was later retracted, this paper prompted a shift in attention by scientists and the public to understand better how these chemicals affect humans and wildlife. The same year that this study was published, congress updated the U.S. Safe Drinking Water Act to include screening programs to detect these contaminants.

The detection of synthetic estrogens and estrogenic chemicals in the environment has increased significantly across the globe since the 1940s. The environmental fate of these chemicals is vast and depends largely on the unique physicochemical properties of each chemical. Generally, estrogens absorb to sediments because of their relatively high molecular weight and moderate hydrophobicity. Their solubility in water depends on temperature and pH. The addition of polar groups, such as hydroxyl or carbonyl groups, as well as overall chemical stability contribute to their presence and persistence in the water. For example, estrogens on average have a relatively short half-life: one study calculated half-lives of 0.2 to 9 days at 20 °C for 17β-E2 and E1 in English rivers (Benijts et al., 2002). However, biodegradation and processing by algae and other marine organisms may lead to chemically undetectable levels of estrogens whereas these mimetics are still capable of biological activity.

There are many sources of estrogens and estrogen-mimetics in the environment. The largest contaminations are detected in sewage treatment plants, runoff from livestock farms, especially cow farms, and human excrements (Adeel et al., 2017). One study estimated that, in 2002, farm animals in the US excreted 49 tons of estrogens (Anderson et al., 2012; Bartelt-Hunt et al., 2012). Sewage treatment plants outflow is another major site of detected contaminations. Estrogens and estrogen mimics originating from industrial, hospital, and domestic waste enter these facilities and, because the currently used treatment processes cannot completely remove sterols, hormonal activities remain in the water flowing into open waterways and reservoirs. While the effects of exposure to physiological levels of E1 and E2 have been well documented, there are few long-term studies on exposure to trace amounts of estrogens and estrogenic EDCs.

Metals also interfere with the function of steroid receptors, specifically ERs. For example, some nonessential metals, such as cadmium (Cd), lead (Pb), barium (Ba), chromium (Cr) and a few essential microelements, copper (Cu) and zinc (Zn), exhibit estrogenic activities as determined by ER-dependent transcription and estradiol-induced responses (Martin et al., 2003; Zhang et al., 2019).

There are several options for detecting the presence of estrogens and estrogenic EDCs in environmental samples. As with techniques used for other hormones, the methods vary in throughput, endpoint, cost and difficulty. Determining the most appropriate assay(s) to use involves evaluating each feature in relation to the overall reason for testing (Table 1). Below are a few examples.

Liquid chromatography coupled with mass spectrometry (LC-MS) has been a traditional approach to identify chemical structure of chemicals in environmental samples. Samples are prepared in a manner similar to other assays; however, after solid-phase extraction and evaporation, the concentrated sample is resuspended in a solvent that is compatible with the solvent system used by the LC-MS instrument (Huysman et al., 2017). While robust, this method fails to identify environmentally modified estrogenic and other hormonal mimetics that are physiologically active but whose structures have not been previously reported (Stavreva et al., 2012b).

Utherogenic binding assays test the ability of a compound at different concentrations to interfere with the binding of radiolabeled 17β-estradiol to rat uterine ER. This type of assay permits measurement of IC₅₀ for a given sample or specific compound. A limitation is that the assay cannot distinguish between ER agonists and antagonists or detect a mixture of compounds (Nikov et al., 2000; Zhou et al., 1998). A similar method, fluorescence polarization, works at room temperature and uses purified recombinant ERα instead of a lysate. This assay measures the ability of a substance to displace a fluorescent ligand from ER. While this approach is more versatile and scalable, it can only measure ERs activation and not physiological responses (Mueller et al., 2003; Routledge et al., 2000; Zhou et al., 1998).

The ER transcriptional activation assay in living cells has been adapted by many laboratories as the main screening method. This cell-based assay measures the ability of a substance to “activate” human ERs through binding to chromatin. The first widely utilized version of this assay utilized cells engineered to express a reporter construct containing vitellogenin Estrogen Responsive Element (ERE) with a reporter gene that encodes firefly luciferase (Pons et al., 1990). One weakness of this assay is its potentially low specificity, depending on how the criteria for a positive response are defined. According to EPA, a “positive response” is when luminescence is above 10% of the positive control response at 1nM 17β-estradiol. Another limitation of ER transactivation assays is the growing appreciation for the role of non-genomic effects of estrogen. There is also an apparent lack of concordance between predictions from in vitro experiments and effects in vivo. A prime example has been in “pro-estrogen,” compounds that are more potent after being metabolized. A study by Conley et al. of estrogenic derivative methoxychlor, a chemical compound used is aerial spray and pesticides to control mosquitoes and other insects, found that it is 368.3-fold more potent in vivo than would have been predicted by the ER transactivation assay (Conley et al., 2016).

Another example is the uterotrophic assay which uses either sexually mature ovariectomized female rats or sexually immature intact female rats to determine the effect of exposure to estrogen agonists and antagonists on the weight of the uterus, uterine epithelial cell height, mammary gland cell proliferation and bone integrity (Anderson et al., 1972; Westerlind et al., 1998). This assay is expensive and examines only long-term effects, but it highlights the complexity of hormonal effects on target tissues in an intact organism. Another example is the effect of isoflavones when compared to estrogens on bone integrity measured in vivo, which has little or no correlation with in vitro activity (Gallo et al., 2005). These in vivo assays may be critical in examining long-term effects of trace compounds in the environment.

As indicated above, analytical approaches based on target chemical analyses are insufficient to depict meaningful environmental contaminations. Mechanism-based bioassays were recently introduced and shown to be powerful bioanalytical tools that assess contamination of environmental matrices. These are increasingly being used to detect estrogenic and other nuclear receptors in complex mixture of EDCs (Table I). One such bioassay is nuclear translocation of ER in response to ligand binding. Because ER normally shuttles between the nucleus and cytoplasm, a chimera of ER Ligand Binding Domain with the GFP-labeled remaining molecule derived from GR is a powerful high throughput screening tool to detect ER nuclear translocation irrespective of ligands’ chemical structure (Jones et al., 2020; Stavreva et al., 2012a; Stavreva et al., 2012c).

Elucidating the mechanisms of ER activation in different organisms by using combinations of model ligands may aid in understanding the effects of environmental samples on assays. The potential for synergistic or additive effects of EDCs on ERs and other nuclear receptors, as described in the GR section of this review, should also be considered. EDCs can also impact ER signaling indirectly through interactions with the aryl hydrocarbon receptor (AhR), G protein-coupled estrogen receptor (GPER), GR and PR (Figure 1), (Legler et al., 1999; Soto et al., 1994; Soto et al., 1995; Zhang et al., 2004). These data indicate that a variety of EDCs present at sub-effective doses can act simultaneously to evoke large scale ER-regulated responses.

Figure 1:

Sex hormones activate cellular receptors in the large family of nuclear receptors (NRs), some of which reside in the cytoplasm (such as AR and GR) while others are predominantly nuclear. Irrespective to their location, binding to a ligand (hormone or an EDC mimicking a hormone) induces a rapid and transient exclusive localization in the nucleus. Once in the nucleus, these NRs become ligand-regulated TFs that bind their respective DNA sequences and interact with variety of chromatin remodelers to influence chromatin state to open or close sites needed for their own actions as well as facilitating the binding of other TFs. Thus, activated NRs can assist each other in binding to DNA sites by a mechanism called “dynamic-assisted loading” and thus substantially modify the final transcriptional outcome (Goldstein and Hager, 2018; Madsen et al., 2014; Miranda et al., 2013b; Swinstead et al., 2016b).

H1 and H2 – Hormone 1and Hormone 2

NR1 and NR2 – Nuclear Receptors 1 and 2. NR1 depicts receptors which reside in the cytoplasm and translocate to the nucleus upon binding to a ligand (GR, AR, and aryl hydrocarbon receptor, AhR are exclusively cytoplasmic in the absence of the ligand). However, any NR can function as an initiating TF and assist the loading of other NRs to their usual or a novel regulatory site.

HSP70 and HSP90 – components of heat shock protein complexes that bind AR, AhR and GR to maintain them in the cytoplasm by conformation that hides their nuclear localization sites. Once a ligand binds to these receptors, the chaperone complex dissociates, allowing H/NR complex to translocate to the nucleus where they become TFs and bind DNA.

Chromatin remodelers complex – interact with TFs and reposition nucleosomes to allow TFs to access specific sites on DNA. Chromatin remodelers contain multiple proteins that are subdivided into four major families of: SWI/SNF, ISWI, CHD, and INO80 (Clapier and Cairns, 2009).

PROGESTINS

The term ‘progestogens’ refers to natural and synthetic molecules having progestational activity which prepares the uterus for pregnancy. Synthetic compounds designed to bind progesterone receptors (PRs, also known as NR3C3: PR-A and PR-B) constitute classes of molecules having different pharmacological properties and modes of action (Schumacher et al., 2007; Stanczyk et al., 2013).

Naturally occurring progesterone (P4) is a 21-carbon sex steroid hormone which was first isolated and synthesized in the mid-1930s by Adolf Butenandt. For his work on sex hormones he was awarded Nobel Prize in Chemistry in 1939 (Butenandt and Westphal, 1974). Soon after its discovery, progesterone and its synthetic analogs, named progestins, began to be used for prevention of premature birth, progestational insufficiency, assisted reproduction, threatened and habitual abortion, premenstrual syndrome, dysmenorrhea, endometriosis, osteoporosis, prevention of endometrial hyperplasia with menopausal hormone therapy, and many other conditions (Henderson, 2018; Piette, 2018; Swyer, 1960).

P4 is synthetized by the corpus luteum of the ovary early in the menstrual cycle and in the placenta during early implantation and pregnancy. If pregnancy does not occur, P4 levels fall promoting menstruation. In pregnancy, progesterone has multiple functions including converting the endometrium to the secretory stage in preparation of implantation during which it cooperates with thyroid hormone (Kakita-Kobayashi et al., 2020). This interaction decreases the maternal immune response thereby permitting fetal implantation and regulating the utero-placental circulation. In addition, decreased progesterone levels following delivery trigger milk production (Di Renzo et al., 2016). P4 is also produced by neurons and glia where it can be metabolized to neuroactive metabolites, the most important of which is allopregnanolone which regulates GABA receptor function and may have mood regulatory and neuroprotective effects (Boero et al., 2020). Small amounts of P4 secreted by the adrenal glands, Leydig cells of the testes and adipose tissue have different functions depending on the tissue type and developmental stage.

Progestins exert their actions by binding to the ligand binding domain (LBD) of PR-A and PR-B, members of the nuclear receptor superfamily of ligand-dependent transcription factors (Figure 1). The human PR-A and PR-B are encoded by a single gene on chromosome 11 (11q22-q23). The two proteins, PR-B (116 kDa), and the shorter PR-A (94 kDa), are controlled by the proximal PR-A promoter region and the shorter form is initiated from the second AUG (492 bases upstream) translational start site (Leonhardt et al., 2003; Wen et al., 1994). Although the two forms are co-expressed, their functional differences are conferred by the N-terminal segment which contains a third activation domain which is unique to PR-B (Weber et al., 1991). Ligand binding to their LBD leads to allosteric activation of the receptors. However, conformational differences between activation domains and their DNA-binding Domain (DBD) depend on the sequence and architecture of the DNA response element which acts as an allosteric ligand to modify DBD conformation allowing cooperative binding of cofactors and interaction with distant regulatory regions (Grimm et al., 2016).

PRs regulate expression of many genes involved in development, differentiation and proliferation of target tissues, circadian rhythm, and has been implicated in hormonally dependent human cancers (Brisken, 2013; Grimm et al., 2016; Stingl, 2011; Zhao and Fent, 2016). In addition, an imbalanced ratio of PR-A/PR-B expression has been implicated in endometriosis and documented in early stages of breast cancer (Mote et al., 2002; Mousazadeh et al., 2019). Thus, biological actions of progestins depend on tissue/cell-type, developmental-stage, availability of co-regulatory proteins, accessibility of target promoters and enhancers and cell-specific protein signaling pathways which regulate PRs posttranslational modifications (Abdel-Hafiz and Horwitz, 2014; Boonyaratanakornkit et al., 2001; Hagan and Lange, 2014; Hill et al., 2012; Hilton et al., 2015). For example, PRs can function as tumor suppressors, it is documented to block growth of endometrial cancer cells. In contrast, PRs promote growth of uterine leiomyoma (Patel et al., 2015).

Because progesterone regulates female menstrual cycle, pregnancy and embryogenesis, the natural hormone is highly expressed during pregnancy and peaks in each female menstrual cycle. It is excreted by humans, livestock and other vertebrates. In addition, synthetic progestins are key ingredients in contraceptives used for anovulatory actions in combination with synthetic estrogens. Progestins are used in hormone replacement therapy and treatment of endometriosis and several cancers. Over 20 known synthetic steroids have progesterone-like activity some of which also have androgenic, estrogenic and glucocorticoid activities (Fent, 2015; Sitruk-Ware and Nath, 2010). As a result of human physiology and farming considerable amounts of natural hormone and its synthetic analogs are shed into the environment.

In contrast to estrogens, ecotoxicological interest in progestins became apparent only recently and risks associated with their exposure are still not fully understood. Recently negative effects of progestins on aquatic organisms have been documented. In fish, for example, synthetic progestins in picomolar concentrations disturb sex development and reproduction and induce female masculinization (Liang et al., 2015a; Liang et al., 2015b; Runnalls et al., 2013; Zeilinger et al., 2009). Exposure to synthetic progestins also result in transcriptional changes in brain and ovaries of zebra fish adults (Zucchi et al., 2013) and embryos (Zucchi et al., 2012). Progestin concentrations in wastewater and surface water range between a few to tens ng/L (reviewed in (Fent, 2015). Considerably higher concentrations are reported in 1999–2000 by the US Geological Survey from 30 states in US rivers (Kolpin et al., 2002).

In France (Besse and Garric, 2009) and Switzerland (Fent, 2015), progesterone and its synthetic analogs are consumed in larger amounts than are other sex hormones: estrogens and androgens. In China and the US, androgenic progestins are routinely applied as growth promoters in livestock and are discharged into the environment from wastewater. Consequently, high concentrations of progestins are found in livestock manure, soil and runoff from farms. The largest contamination comes from sows and cows; up to 11,900 ng/L of progesterone were found in animal farm waste and 350 ng/L were detected in the associated runoffs (Liu et al., 2012a). Clearly, the waste produced by farm animals is a major source of progestins in waters adjacent to farms and cattle feeding operations (Liu et al., 2012b; Stavreva et al., 2021). In addition, other steroid hormones, including progestins, have been found in paper mill effluents which accumulate in river sediments (Chatterjee et al., 2007; Jenkins et al., 2003). These may arise from processing pine pulp or microbial degradation of abundant phytosteroids present in the waste from processing pine trees (Brockmeier et al., 2014).

Only recently have analytical methods for detection of progestins been fully developed and less than half of the ~20 synthetic progestins have been documented in aquatic systems (Table I). Furthermore, methods to detect antagonists in the environment are lacking (Leusch et al., 2017; Viswanath et al., 2008) and need further development. The progestins Norethindrone and Levonorgestrel in combination with other steroids were identified by liquid or gas chromatography-tandem mass spectrometry (LC-MS/MS and GC-MS/MS) in wastewater treatment plant effluents at concentrations up to 50 ng/L (Fernandez et al., 2007; Vulliet et al., 2007; Vulliet et al., 2008). Medroxyprogesterone, well-known for its association with cancer of female organs, androgenicity and spermatogenesis in the progeny (Liang et al., 1983; Shi et al., 2019) was detected in municipal wastewater plants effluents and surface water samples at the levels up to 15 ng/L and 1 ng/L, respectively (Kolodziej et al., 2003).

This is an unexpected outcome because progestins are presumed to be degraded in wastewater plants, mostly by biodegradation, with removal efficiency of over 90% (Chang et al., 2011; Muz et al., 2014). Inefficient removal or progestins has been reported in Europe (Sauer et al., 2018) and high levels of norethindrone (up to 188 ng/L) were detected in effluent from a wastewater plant in Malaysia (Al-Odaini et al., 2010). Furthermore, due to the hydrophobic nature of progestins and other steroid hormones, they accumulate in aquatic organisms (Fick et al., 2010; Kroupova et al., 2014; Kumar et al., 2015). For example, Levonorgestrel, was measured by liquid and gas chromatography-mass spectrometry and found to accumulate in rainbow trout plasma at levels as high as 8.5–12 ng/mL (from the effluent wastewater with 1 ng/L) (Fick et al., 2010). Deleterious effects of progestins specifically on male aquatic organisms and their reproduction are well documented (Kumar et al., 2015; Raghavan et al., 2018). Bioaccumulation of progestins and other steroid hormones in children and adults by seafood consumption from aquacultures has also been documented (Liu et al., 2015).

A summary of different approaches to measure progestins and PR activities is presented in Table I. Direct comparison between assays is difficult because different assays use different reference compounds and endpoints. Receptor binding and yeast reporter gene assays are usually normalized to P4, although the reporter assays use synthetic progestogens, such as levonorgestrel or promegestone, because P4 is rapidly degraded in these conditions. Despite this, the emerging concept is that contamination with progestins alone or in combination with other steroid hormones is widely and globally distributed in the aquatic environment due to human activities.

GLUCOCORTICOIDS and THEIR INTERACTIONS WITH OTHER NR BY DYNAMIC ASSISTED LOADING MECHANISM

Glucocorticoids (cortisol in humans and corticosterone in rodents) are natural steroid derivatives and their many non-steroidal agonists (De Bosscher, 2010), are widely used as anti-inflammatory agents in treating immune diseases such as asthma, rheumatoid arthritis, cancer and other human and animal conditions (Barnes and Adcock, 2009). Detection of glucocorticoids in the environment began in the last decade (Chang et al., 2009; Schriks et al., 2010) and glucocorticoid receptor activity in US water samples was first reported in 2012 (Stavreva et al., 2012b). However, extensive use of steroid hormonal therapy in humans, pets, and livestock resulted in increased detection of corticosteroids and corticosteroid antagonists in the environment (Barnes, 2011; Barnes and Adcock, 2009; Stavreva et al., 2021). Corticosteroid activity disruptors are metals, metalloids, pesticides, bisphenol analogues, flame retardants and other chemicals used in agriculture and industry and have become environmental pollutants (Martin et al., 2003). Disturbances of glucocorticoid physiology in humans and animals are linked to birth defects, mood and cognitive disorders, immune, allergic, metabolic, cardiovascular diseases and several cancers (Odermatt and Gumy, 2008; Reynolds et al., 1975)

Biological activity of glucocorticoids is mediated by glucocorticoid and, at higher concentrations, mineralocorticoid receptors (GR and MR, respectively). Both are widely expressed in eukaryotes and involved in many physiological processes. GR is the canonical ligand-dependent TF, whose biology has been addressed in detail in many studies and by multiple approaches. The complexity of GR physiology in mammals resides, in part, in combination of its diurnal, also called circadian (Migeon et al., 1956) in addition to the ultradian (pulsatile ~1hour) release. Cortisol secretion tightly regulates transcriptional responses which are confined to the duration of the hormonal pulse (Levine et al., 2013; Stavreva et al., 2019; Stavreva et al., 2009). Mammalian cells exhibit many pulsatile circuits, and it is likely that multiple pulsatile systems coexist and interact in individual cells. Ultradian GR actions induce cyclic transcriptional regulation, which is controlled by the chaperone-mediated GR recycling back to the cytoplasm and the rapid exchange of GR at its binding elements, GREs on DNA. Responses to the pulsatile activation of GR by the natural hormones are significantly different from those observed after constant exposure to even low nanomolar concentrations of a potent glucocorticoid like Dexamethasone (Stavreva et al., 2009). Prolonged exposure likely results in altered feedback on hypothalamic-pituitary-adrenal (HPA) axis and negative effect on delicate cyclic secretion of natural hormone. Thus, humans and other organisms exposed to uncontrolled GR agonists present as EDCs are likely to have altered circadian systems, immune responses and other key physiological processes. For example, increased susceptibility to infection as a result of a weakened immune system is associated with fish kills in the Shenandoah and Potomac rivers watershed (Ripley et al., 2008); some of these sites show direct GR activation from these rivers (Stavreva et al., 2012b).

Recent studies to probe genome-wide changes of open chromatin sites (by DNase I hypersensitivity or ATAC-seq assays, combined with RNA- and ChIP-seq), revealed unexpected dynamic effects of GR activation on transcriptional machinery (John et al., 2011; Johnson et al., 2018; Stavreva et al., 2015). GR interactions with chromatin remodeling complexes leads to wide chromatin landscape remodeling which facilitates binding of other NRs, such as activated ER, a process labeled “dynamic assisted loading” (Voss et al., 2011). This process is depicted on the lower part of Figure 1. Eukaryotes have four major families of chromatin remodelers: SWI/SNF, ISWI, CHD, and INO80 (Clapier and Cairns, 2009) which play distinct roles based on how they select and affect target nucleosomes, the basic structural units of chromatin. GR activity is particularly linked to the ATP-dependent action of SNF/SWI complex (Morris et al., 2014; Muratcioglu et al., 2015). The resulting distinct chromatin landscapes changes in eukaryotic cells determine how genetic information is organized, replicated, transcribed, and repaired (Yen et al., 2012). Rapid remodeling of chromatin landscape is not unique to GR. In fact, assisted loading relationship between TFs has been described in several systems and involves all sex hormones (Goldstein et al., 2017; Madsen et al., 2014; Miranda et al., 2013a; Nagaich et al., 2004).

Single-cell imaging experiments revealed that many TFs are only transiently bound to their response elements and repeatedly cycle on and off following activation (Bosisio et al., 2006; McNally et al., 2000; Sharp et al., 2006). For all TFs tracked by single molecule/single cell labeling, the dwell times in the nuclear environment are in the range of ses conds (Chen et al., 2014; Izeddin et al., 2014; Morisaki et al., 2014; Swinstead et al., 2016a; Swinstead et al., 2016b). Recently improved single cell/single molecule methods demonstrate broad distribution of DNA binding affinities for variety of NRs and other TFs (Garcia et al., 2021). This is an important aspect of transcription factor (TF) biology which is critical for short-lived effects and for facilitated the binding of other TFs for proper regulation of gene expression. Thus, unwanted exposure to glucocorticoids with other hormonal mixtures interferes with normal physiology in a complex manner. Because the dynamic assisted loading mechanism applies to other nuclear receptors, the combination of EDCs can affect circadian cycle and many vital functions in humans and other life forms.

The interaction of glucocorticoids with estrogens is not limited to basic biology studied in a single cell. Their interactions in brain function, plasticity, stress, and memory have been widely documented (McEwen et al., 2016; Taxier et al., 2020). Whether these activities also involve dynamic assisted loading mechanism remains to be determined.

CONCLUSIONS

We have reviewed existing data regarding environmental contamination, detection methods and potential biological effects exerted by sex hormones. Many chemicals have adverse effects on animals and humans in the range of environmental exposures which indicate that low doses found in the environment cannot be ignored. Furthermore, water is only one route of exposure, mostly pertaining to the wildlife and fish. The final effects in mammals are from combined exposure to EDCs from various sources including seafood, salt and fresh water. Fortunately, most currently used mammalian and yeast cell-based assays for EDCs detect low doses comparable to those observed for naturally occurring hormones. Thorough validation and awareness of the advantages and limitations of current chemical libraries and in vitro assays for predicting in vivo responses are critical for their successful implementation. The difficulty in using chemical libraries is that contaminants are present in small amounts in water and other sources and are insufficient for laborious chemical identification. In addition, the effects of in vitro assays may not correlate with in vivo activities. To overcome some of the limitations of existing chemical libraries, a substantial international study collected over 55,000 compounds from all existing chemical structures including simulated ToxCast™ metabolites and, using sophisticated bioinformatic analysis, derived multiple models of AR agonist and antagonist activities (Mansouri et al., 2020). Earlier large-scale international efforts also compiled over 32,000 compounds to derive potential models for ER (Mansouri et al., 2016). The consensus activity models provided high averaged predictive accuracy of ~80% for the evaluation set. The results of screened EPA DSSTox database of ∼875,000 chemicals are publicly available for future prediction of activities for novel chemical structures. These remarkable efforts demonstrate the potential in performing global studies to identify EDCs.

One paradoxical effect of mammalian metabolism has not achieved sufficient attention. In mammals, detoxification mechanisms convert potentially toxic organic compounds into inactive, water-soluble metabolites in preparation for excretion. However, some organic metabolites are biologically active, and some are more potent than their initial source. For example, testosterone, is converted to a more soluble metabolite, dihydroxy testosterone (DHT) by 5α-reductase in prostate, skin and to a lesser degree, in other tissues. DHT is 100-fold more active than is testosterone on target tissues. Another example is DDT, an androgen receptor antagonist, and its more soluble, stable and more potent metabolite, DDE. This underscores the importance of in vivo testing.

Another largely unexplored issue is the combined effect of multiple hormones detected in the same sample. Presence of sex hormones in combination with glucocorticoids, thyroid, aryl hydrocarbon and other hormonal agonists and antagonists is well documented. These environmental mixtures are likely to have yet unexplored effects on target tissues due to their interactions with multiple receptors and intranuclear modulators and chromatin remodelers including the “assisted loading” mechanism which modify chromatin conformation in response to mixed hormonal content to modulate gene expression (Figure 1). Some of these epigenetic changes may be long-lasting and possibly inheritable.

The hazardous effects of sex hormones in the environment largely rely on epidemiological data in humans and other life forms. Although sometimes difficult to interpret, many studies reveal substantial long-lasting metabolic, reproductive and behavioral consequences of EDC exposure. Low economic status and ethnic populations are more likely to be exposed to these hazards which results in poorer health outcomes.

Lack of consensus in the scientific community on quantitative methods for detection and “safe” levels of sex hormones and other EDCs in the environment is a major obstacle for development of a rational policy for efficient EDC monitoring and establishing safety limits to protect wildlife and human health. Contamination of water, air and soil with sex EDCs threatens the integrity of human, aquatic and other ecosystems, and poses serious concerns for human and animal health. Several avenues have been proposed to coordinate research programs within federal agencies, state/local governments, academia, industry, trade association, and NGOs. For example, the National Children’s Study was envisioned as a longitudinal, observational study to examine the effects of a broad range of environmental and biological factors on children’s health and development by following 100,000 children from the womb to age 21. Unfortunately, due to financial constraints this study was closed in 2014 after recruiting 5,000 children (http://www.nationalchildrensstudy.gov). A coordinated program encompassing governmental and public organizations and industry leaders with scientists would enable a science-based approach to understand the impact of EDCs and pollution on human health. Several short and long-range studies in many countries are currently underway, but some valuable information will be scattered among websites and scientific publications. Linking environmental pollution to large scale programs, such as NIH-led All-of-Us (https://allofus.nih.gov/), which provides in-depth socio-economic, health and genomic information, could be an appropriate starting platform.

This is the time to organize a high-level meeting that brings together all critical players with the twin goal of sharing information and considering options to address the critical problems of EDCs detection, monitoring and regulatory policies with particular emphasis on relation to human disease. Large scale international studies of chemical libraries described above (Mansouri et al., 2016; Mansouri et al., 2020) show that these efforts are now feasible. Virtual platforms, popular since 2020 during the COVID-19 pandemic, can facilitate such approaches. Only then the scientific community could provide national and international agencies an informed opinion on monitoring and controlling specific aspects of sex hormones and other environmental contaminants.

Highlights MCE-D-21–00213.

• Current information on endocrine disruptors is scattered in many publications

• Many studies report presence of multiple disruptors in the same sample

• We include compounds mimicking estrogen, androgen, progesterone and glucocorticoids

• We review the evolution and tabulated methodologies for their detection

• The dynamic assisted loading mechanism highlights the complexity of hormonal actions

• We discuss the difficulties and propose solutions to control the contaminants