Introduction

The “endocrine disruptor” concept, the interest for specific studies on their endocrine activity and the link between human health and environmental exposure flashed for the first time at the European Workshop on the Impact of Endocrine Disruptors on Human Health and Wildlife (Weybridge, UK) in 1996. Nowadays, the term endocrine-disrupting chemical (EDC) is used to define a structurally diverse class of synthetic and natural compounds that possess the ability to interact with the endocrine system. These anthropogenic chemicals are continuously released into the environment through many sources, mainly of anthropogenic origin, e.g., wastewater treatment plants, terrestrial runoff and precipitation, and finally targeting humans and wildlife health (Illuminati et al. 2010, 2017; Annibaldi et al. 2015; Afifah et al. 2017; Ribeiro et al. 2017). EDCs are diverse and include herbicides and pesticides, plastic contaminants, heavy metals, biocides, heat stabilizers, chemical catalysts, pharmaceuticals and dietary components, flame retardants, and others. Given their physicochemical differences and distinct biological effects, it is not surprising that EDCs interact with the endocrine system through a wide range of mechanisms of action (Henley and Korach 2006). They interfere with the neuroendocrine and endocrine functions involved in embryo development and reproduction (Godfrey et al. 2017). Once released in the environment, they exert their effects by mimicking, antagonizing, or altering endogenous steroid levels (androgens or estradiol) by changing rates of their synthesis or metabolism and/or expression or action at receptor targets. EDCs, in fact, share structural homologies with natural hormones, being able to bind their cognate receptors and triggering their same biological response. Considering the reproductive pathway, the main gonadal steroids include estrogens, androgens, and progesterone, which are derivatives of cholesterol. In their biosynthesis, testosterone derives from androstenedione, while estrogens are synthesized by the aromatization of either testosterone (T) into estrogen (17β-estradiol) or androstenedione into estrone, until final conversion into estradiol. Two specific aromatase enzymes, responsible for catalyzing these reactions are encoded in fish: CYP19a mainly expressed in the gonads and CYP19b mostly found in the brain (Zhang et al. 2014c). Estrogens are known to be involved in vitellogenesis regulation, oogenesis, gonadotropin regulation, testicular development, and many other reproductive processes (Yilmaz et al. 2015). Androgens, including testosterone, working via the androgen receptor (AR), play a critical role in males’ development and secondary sex characteristics (Schulz et al. 2010). Despite most of the EDCs have environmental concentrations far below the range established by environmental agencies, it should be considered that an environmental health problem regarding EDCs is related to their bioaccumulation, chronic exposures, and eventual occurrence of synergistic/antagonistic interaction among them. Furthermore, EDCs are known to act through non-monotonic dose-response curves, and therefore it has been argued that effects observed at high doses cannot be used for the extrapolations into the low dose range. Endocrine disruption has been demonstrated to occur in wildlife, particularly in aquatic species or in species that are connected to the aquatic food chains. Dietary uptake due to consumption of polluted food represent an important source of contamination (Maradonna et al. 2014, 2015; Traversi et al. 2014; Ruhí et al. 2016; Carnevali et al. 2017). Given the widespread environmental distribution of the pollutants, the questions being addressed today, are not whether endocrine disruption occurs in wildlife but rather in which concentration it occurs, what mechanisms are involved, and whether disruption of the endocrine system will lead to ecologically relevant effects.

The aim of this review is to summarize the effects induced by the exposure to EDCs on reproduction, focusing on gametogenesis, giving a general overview of all aspects dealing with this issue. This will provide the knowledge to build an integrated model, using morphological, biochemical, and molecular results, to be used in further studies in an environmentally relevant contest. Results shown will provide evidence to sustain that many of the molecular changes are associated with ecologically relevant effects in terms of their impact on population and further generations.

Effects of EDC exposure on ovarian physiology

Most of studies described so far documented the ability of EDCs to affect ovarian morphology, oocyte maturation and steroidogenesis, altering the normal functioning of the endocrine and reproductive systems by mimicking or inhibiting endogenous hormone action, modulating the production of endogenous hormones or altering hormone receptor populations (Sonnenschein and Soto 1998). A major mechanism of endocrine disruption is the EDC binding to hormone receptors. Other mechanisms, besides receptor-mediated events, may include mechanisms of inhibition or stimulation of hormone metabolism or alterations in serum hormone-binding proteins.

EDCs affects ovarian morphology: oocyte growth, maturation, and fertility

Gobious rarus exposure to bisphenol A (BPA) treatments (14 and 35 days) induced ovarian morphological alterations consisting on the presence of a relatively higher proportion of pre-mature oocytes and the occurrence of several atretic follicles (Zhang et al. 2014a). Moreover, in the same species, the exposure to 2,4-Dichloro-6-nitrophenol, a chloride-based herbicide, induced detrimental effects both in male and female fish. In female gonads, an increase in the number of deformed follicles along with the presence of degenerating vitellogenic oocytes were observed, suggesting a defect in follicle recruitment (Chen et al. 2016). Effects on male are described in the section below. When catfish, Clarias gariepinus, were exposed, from hatching to 50 dph, to ethinylestradiol (EE2) (50 ng/l) and diethylstilbestrol (DES) (10 ng/l), and development was monitored until adulthood, morphological deformities, such as stunted growth, spinal curvature, and yolk-sac fluid retention were observed. In addition, histological studies revealed rudimentary or malformed ovaries with precocious oocyte development as well as follicular atresia (Sridevi et al. 2015). In zebrafish female, Danio rerio, the chronic exposure to 25 ng/l EE2 and to 0.02, 0.2, 2, 20, and 40 μg/l Di-(2-ethylhexyl)-phthalate (DEHP), severely impaired ovulation and embryo production. Ovarian observation suggested the ability of both EE2 and DEHP to modulate vitellogenesis inducing a reduction of previtellogenic oocytes associated with an increase of vitellogenic ones. After the exposure, fish were crossed and the number of embryos obtained was about 1% of the embryos produced by the control, showing a clear impairment of fecundity (Carnevali et al. 2010). Recently, in the same species, the effects of the chronic exposure to five different concentrations (0.42, 4.2, 42, 420, and 4200 μg/l) of di-isononyl phthalate (DiNP), a substance which was introduced to replace DEHP, were analyzed focusing on a number of reproductive parameters, including follicles number and size frequency and macromolecular composition of vitellogenic oocytes. Histological analysis revealed a significant reduction in the number of vitellogenic oocytes in gonad of fish exposed to the lower DiNP concentrations (0.42, 4.2, and 42 μg/l) as well as a reduction in the number of mature oocytes in those of fish exposed to higher doses (420 and 4200 μg/l) (Fig. 1). In vitellogenic oocyte sampled from exposed fish, a decrease of lipids, phosphates, and proteins components was measured by FT-IR, suggesting that DiNP might affect vitellogenin uptake and induce changes of lipid composition. Furthermore, results demonstrated that DiNP adversely affects female reproductive physiology acting in a non-monotonic fashion. In the same experimental model, exposure from 2 h post fertilization (hpf) to sexual maturity, to the flame retardant tris (1,3-dichloro-2-propyl) phosphate (TDCPP) promotes oocyte maturation as documented by an increase in the percentage of late vitellogenic/mature oocyte (Wang et al. 2015). Moreover, in the F1 generation, high malformation rates were observed demonstrating that parental exposure to TDCPP causes developmental toxicity in offsprings (Wang et al. 2015). In marine medaka, Oryzias latipes, ovary, the exposure to 8.5 μg/l of 3,3′-diindolylmethane (DIM), induced an increase of VTG storage, but a reduction of cathepsin enzyme activities, suggesting a blockage of VTG uptake into growing oocytes. This reduction occurred together with lower levels of eggshell proteins leading to an inhibition of primary oocyte maturation and contributing to a decreased fecundity (Chen et al. 2017). In trout (Salmo trutta f. fario), a 3-month exposure to 5 μg/l BPA, starting at the beginning of the spawning season, affected the percentage of ovulated eggs and shifted ovulation after 3 weeks, without affecting the quality of the eggs (Lahnsteiner et al. 2005). Occurrence of atretic oocytes characterized by disorganization of the ooplasm was documented in carp treated with 2 mg/l Perfluorooctanoic acid (PFOA) for 56 days (Giari et al. 2016). In a 21-day reproduction study on fathead minnow exposed to the pharmaceutical dutasteride (100 μg/l), several ovarian histopathological conditions including presence of atretic follicles and mitotically dividing oogonia, proliferation of interstitial connective tissue containing different types of somatic cell, and macrophage aggregates were detected. Accumulation of adipose tissue and interstitial proteinaceous fluid was also observed (Margiotta-Casaluci et al. 2013). In tilapia, Oreochromis niloticus, the action of diuron, a substituted urea herbicide with a documented anti-androgenic action, and three of its metabolites were evaluated. Pollutant exposure induced an increase of gonadosomatic index and the ovary presented a higher number of vitellogenic oocytes and a decrease of germinative cells, suggesting that diuron metabolites have estrogenic action and accelerate this species ovarian development (Boscolo Pereira et al. 2016).

Fig. 1
figure 1

Histological analysis of ovaries from control fish and fish exposed to five DINP concentration. a Percentage of previtellogenic, vitellogenic, and mature oocytes in the ovary of control (C) and fish exposed to DINP; b representative ovarian section showing previtellogenic (PreV), vitellogenic (Vit), and mature (M) oocytes

Full size image

EDC effects on molecular markers linked to steroidogenesis, ovarian growth, and maturation

Exposure of G. rarus to different BPA concentrations differently affected the expression of steroidogenic enzymes star, cyp11a1, cyp17a1, hsd3b, hsd11b2, and cyp19 expression at lower (5 and 15 μg/l) concentrations; BPA induced the increase of ovarian steroidogenic gene transcription and sexual steroid receptor genes, while at higher concentration (50 μg/l), the increase of nr5a1s. Variation of the steroidogenic gene expression was probably linked to changes in er and ar transcripts levels able to affect both the steroid hormone signaling and nr5a1s expression, suggesting a possible epigenetic regulation (Zhang et al. 2014a). Moreover, in the same fish, alteration of gdf9 and bmp15 mRNA expression and their levels may play a key role in the weight gain and in the abnormal ovarian development, above described (Zhang et al. 2014b). In catfish, transcript levels and activity of aromatase, the rate-limiting enzyme of estrogen biosynthesis were significantly altered by EE2 and DES exposure (Sridevi et al. 2015). More specifically, upon exposure to EE2, at 50 and 150 dph, the transcript levels of star and P450scc exhibited significant increase. Regarding cyp17, 3b-hsd, and 17b-hsd1, a gradual increase from 50 to 350 dph following EE2 treatment was observed. Conversely, exposure to DES decreased cyp17 transcript levels, while 3b-hsd and 17b-hsd1 transcripts were unaltered. Neither EE2 nor DES exposures affected ovarian aromatase (cyp19a1a) levels (Sridevi et al. 2015).

The decrease of fecundity observed in zebrafish chronically exposed to EE2 or DEHP, was associated with the increase of BMP15 levels and the reduction of lhr, mprb, and ptgs2 expression, the final triggers of ovulation. By an in vitro maturation assay, the inhibitory effect of DEHP on germinal vesicle breakdown was further confirmed (Carnevali et al. 2010). In the same species, exposure to DiNP resulted in modulation of genes involved in steroidogenesis in a dose-related, but non-monotonic manner. Greater differences were observed for star and cyp11a1 transcript levels among the different tested concentrations (0.42, 420, and 420 μg/l). Only the lowest dose affected fshr transcript. Esr1 and esr2b levels were affected only in the 420 μg/l trial. Exposure to DiNP reduced esr2a transcript level in a non-monotonic manner, with great differences observed among the lowest (0.42 mg/l) and the higher concentrations (420 μg/l; 4200 μg/l). Exposure to DiNP did not change pgrmc1 transcript level in any of the treatment groups, but in the 420 and 4200 μg/l exposures, a reduction of lhcgr and pgrmc2 transcript levels was measured. Conversely, bmp15 transcript level was significantly increased only in fish exposed to the 0.42 μg/l concentration. The results described suggested that low DINP concentrations mainly interfere with steroidogenesis and oocyte growth while higher concentrations impair oocyte maturation (Santangeli et al. 2017b), suggesting that there is still the need to find a safer DEHP substitute (Forner-Piquer et al. 2017).

In addition to the estrogenic effects, EDCs may also exert and anti-androgenic action as evidenced in fathead minnow (Pimephales promelas) treated with the Dutasteride, a pharmaceutical product acting as a 5α-reductase (5αR) inhibitor, the enzyme that converts testosterone into dihydrotestosterone (DHT). Dutasteride impairs reproductive functions in fish, consistent with an anti-androgenic mode of action (Margiotta-Casaluci et al. 2013).

In vitro exposure of trout post-vitellogenic oocytes to prochloraz, an imidazole fungicide, widely used in agriculture, affected oocyte maturation process. It is well-known that prochloraz antagonizes both androgen and estrogen receptors and agonizes the aryl hydrocarbon receptor. Prochloraz (10−5, 10−6, or 10−7 M), in combination or not with LH, was able to alter the GVBD in Rainbow trout post-vitellogenic follicles. Prochloraz administered alone is able to trigger oocyte maturation stimulating the transcription of a set of genes responsible for pre-ovulatory follicular recruitment and differentiation, such as igf1, igf2, connexin 43, and hsbd20. Some of them were also triggered by LH, which seems to synergically cooperate with prochloraz in inducing GVBD (Rime et al. 2010).

Most of the studies described so far documented the ability of many xenobiotics to affect ovarian morphology, oocyte maturation, and steroidogenesis following the classical genomic pathway (Blair et al. 2000). More recently, interest has moved to the investigation of the non-genomic mechanisms activated by xenobiotics (Thomas and Doughty 2004). The key role of a novel seven pass-transmembrane receptor, GPR30, in mediating several rapid estrogenic responses was demonstrated. In zebrafish, it was demonstrated that low BPA concentrations (10–200 nM) and three alkylphenols, e.g. tetrabromobisphenol A, 4-nonylphenol, and tetrachlorobisphenol A (5–100 nM) disrupted oocyte maturation by a non-genomic mechanism of action involving the activation of the Gper/Egfr/Mapk3/1 pathway (Fitzgerald et al. 2015).

Effects of EDC exposure on male gonadal physiology and functionality

The main determinant of male fertility includes the evaluation of sperm motility once released into the environment, spermatozoa velocity, and motility duration (Linhart et al. 2008). In vertebrates, androgens regulate testicular functioning, spermatogenesis, secondary sexual characteristics and behavior, and their effects are mediated by androgen receptors (Martyniuk and Denslow 2012). EDC exposures caused testes growth inhibition and maturation delay, changes of male sex cell types, low semen quality, occurrence of ova-testis, and changes in estrogen and testosterone levels in several teleost models (Kime 1999). The adverse impacts of ECDs occur at lower concentrations as duration of exposure increases (Kumar 2013). Although technical and biological differences among studies make it difficult to classify toxicity of EDCs, it seems that heavy metals have greater potential to impair spermatozoa functioning and exhibit higher toxicity than organic EDCs do (Popek et al. 2006).

Effects of EDCs on testis morphology and spermatogenesis

In marine medaka, Oryzias melastima, exposure to relatively low doses (0 and 8.5 μg/l) of 3, 3″ diindolylmethane (DIM), an antifouling agent, led to anomalous production of vitellogenin and eggshell proteins in the testis, clearly highlighting its estrogenic potency (Chen et al. 2017).

In zebrafish, the effects of a chronic exposure to environmentally relevant concentrations of DEHP (0.2 and 20 μg/l) were analyzed, highlighting DEHP ability to impair reproduction inducing a mitotic arrest during spermatogenesis and decreasing embryo production (up to 90%). These changes were associated with the increase of spermatozoa DNA fragmentation (Corradetti et al. 2013). Similarly, zebrafish exposure to a mixture of environmentally relevant concentrations of triclocarban and inorganic mercury, induced severe histological lesions in the testis, which size and mature sperm abundance was decreased if compared to those of fish exposed to the compounds individually (Wang et al. 2016). Similarly, the exposure to DES, flutamide (FLU) and their combination induced an impairment of spermatogenesis, decreasing sperm concentration, and affecting both meiotic and apoptotic processes (Yin et al. 2017).

A possible delay of spermatogenesis was hypothesized in male carp, Cyprinus carpio, exposed to 2 mg/l PFOA for 56 days, where spermatogonia and spermatocytes were the dominant germ cells. Moreover, an increased amount of interstitial tissue and proliferation of Sertoli cells was observed in the testis (Giari et al. 2016). In the same species, a 14-day exposure to graded-BPA concentrations, from 1 to 1000 μg/l, caused severe alterations of the lobular structure and spermatogenic cysts, with free spermatozoa often degenerating into the lumen (Mandich et al. 2007). In goldfish exposed to BPA, an increase of hepatic vitellogenin was induced, but gonadosomatic index (GSI), hepatosomatic index (HSI), and E2 levels were not affected, suggesting that BPA might act on testicular spermatogenesis causing an alteration of sperm maturation (Hatef et al. 2012a, c).

Spermatozoa production, morphology motility, and velocity

Morphological sperm alteration, including damage to the flagellum and its beating, can be revealed using a dark-field microscope equipped with stroboscopic illumination, or a phase contrast microscope with high-speed video camera (Alavi et al. 2012a, b). Computer-assisted sperm analysis (CASA) systems, which are based on video recording approaches, are employed for analysis of sperm motility and velocity (Alavi et al. 2012a). Valuable parameters of sperm velocity are curvilinear velocity (VCL), straight line velocity (VSL), and the angular path velocity (VAP), being especially this last parameter, differently affected by ECDs.

Exposure to cadmium, mercury, BPA, zinc, and tributyltin decreased VCL and VSL of spermatozoa in different fish species. As reported by Hatef et al. (2013), specific spermatozoa responses can be supposed when different species are exposed to a given EDCs. In African catfish—Clarias gariepinus, burbot—Lota lota, European chub—Leuciscus cephalus, a decrease of sperm motility was observed immediately after copper or cadmium exposure of spermatozoa, while no evident effects were induced in brown trout, S. trutta f. fario, spermatozoa exposed to the same metals (Lahnsteiner et al. 2004). Different results were also evidenced when the effects of exposure to mercury, zinc, lead, nickel, cyclohexane, and 2,4-diclorophenol were compared among seabass (Abascal et al. 2007), African catfish, burbot, and European chub spermatozoa motility (Lahnsteiner et al. 2004). Exposure to cadmium negatively affect sperm quality also in D. rerio, including motility, time of motility, curvilinear velocity, average path velocity, and straight line velocity, which alteration may reduce the fertility rate of these animals (Acosta et al. 2016). No effects on motility were observed in medaka, exposed to scalar concentration of nonylphenol (Hara et al. 2007). In croaker, many estrogenic and non-estrogenic pollutants, at environmentally relevant concentrations, blocked the ability of 17,20β,21-trihydroxy-4-pregnen-3-one to induce sperm motility, a process physiologically mediated by a rapid, non-genomic steroid action (Thomas and Doughty 2004).

These evidences suggest that despite the recognized mode of action of the EDCs, the variability in the responses may be related to differences in spermatozoa and seminal plasma physiology and biochemistry. Sperm velocity was decreased in Perca fluviatilis sperm incubated for 3 h with mercury. A trial was set up to gain information on the effects of mercury chloride (HgCl2) on both sperm physiology and structure, to identify sites of action of HgCl2 and to investigate the mechanism of action of HgCl2 on spermatozoa. Results demonstrated that HgCl2 acts on different sites of sperm, such as the plasma membrane, the axoneme, and the midpiece. HgCl2 exposure decreased stored ATP and increased morphological alterations (Hatef et al. 2011). In the same species, in vitro treatment of sperm with BPA, decreased both motility and velocity and its capacity to damage the flagella structure changing its shape from the straight position to the C-shaped one was demonstrated (Hatef et al. 2010). In sea bass, a 5-min incubation with 1 mg/l mercury or an instant exposure to 100 mg/l, significantly affected sperm velocity, while no effect was induced by exposure to lead or copper (Abascal et al. 2007). Velocity was also affected in common carp, after a 24-h incubation with TBT (Rurangwa et al. 2002) and to graded concentration of mercury (Chyb et al. 2001a,b) and zinc (Chyb et al. 2000). Impairment of spermatogenesis was observed in trout, S. trutta f. fario, exposed to environmentally relevant concentrations of BPA (1.75, 2.4, and 5 μg/l) during pre-spawning and spawning periods. Results showed that at the beginning of spawning, the two lowest BPA concentrations, reduced sperm density, motility rate, and swimming velocity. In the middle of spawning season, BPA reduced swimming velocity and only the 2.4 μg/l concentration, reduced motility rate. The highest concentration completely impaired spermatogenesis and only one out of eight fish gave low-quality semen (Lahnsteiner et al. 2005). In goldfish, a 20- or 30-day exposure to BPA decreased sperm motility at 15-, 30-, 60-, and 90-s post-activation; a shorter exposure, 10 days, affected sperm velocity at 30-, 60-, and 90-s post-activation. These alterations were possibly caused by variation of hormonal levels, as described below (Hatef et al. 2012a). In the same species, using a phase contrast microscopy, a shorter flagella length was measured in spermatozoa in presence of high mercury concentration (Van Look and Kime 2003).

In rainbow trout, seminal plasma was used as incubation medium for monitoring short-time exposure to mercury and cadmium ions on sperm. Although mercury exposure affected sperm motility immediately after dilution with milt as well as at 4 h of exposure, no differences were found respect to control sperm after a 24-h treatment, suggesting that in trout, seminal plasma has a protective role against the toxic effects of mercury ions on motility (Dietrich et al. 2010).

In Gobius niger, exposed for 48 h to 1 or 10 mg/l Cd acetate, hsp70, mtt and casp3 mRNA levels measured in the testis, significantly increased, highlighting the toxicity of this metal on testis physiology (Migliarini et al. 2005). Similarly, in starlet was recently demonstrated that exposure of sperm to NP and DES can induce reactive oxygen species in spermatozoa, leading to an impairment of sperm quality, e.g. decrease the percentage of intact sperm cells (Shaliutina et al. 2017).

Recently, the effect of seven heavy metals on the motility parameter of zebrafish sperm was tested in order to develop an in vitro toxicological test system as an alternative to live animal testing. The endpoints analyzed were progressive motility, VCL, and linearity. Results demonstrated that progressive motility was the most sensitive of the three investigated parameters, suggesting its use as an accurate and fast bioindicator of heavy metal load (Kollár et al. 2018).

Effects of EDCs on steroidogenesis and hormone synthesis

The alteration of testis morphology described in zebrafish exposed to triclocarban and inorganic mercury was associated to variation of the transcription of a set of steroidogenic genes, such as cyp19a, 3β-HSD, cyp17, and 17β-HSD (Wang et al. 2016). In G. rarus, the exposure to BPA evidenced non-monotonic effects on the mRNA expression of StAR, cyp11a1, 3β-HSD, cyp17a1, and cyp19a1a resulting the medium concentration, 15 μg/l, to be the most detrimental. More specifically, the increase of the expression of several mRNAs codifying for testicular steroidogenic enzymes suggested an increase of androgens, which effect could be counteracted by the non-significant downregulation of cyp17a1 mRNA expression (Liu et al. 2014a). The correlation analysis at mRNA level demonstrates that the BPA-mediated actions on testicular steroidogenesis might involve sex steroid hormone receptor signaling, gonadotropin/gonadotropin receptor pathway, and transcription factors, such as the nuclear receptor subfamily 5, group A (Nr5a), and the fork head box protein L2 (Foxl2) (Wang et al. 2012; Zhang et al. 2013). Similarly, the exposure to 25 ng/l EE2, modulated the expression of the same steroidogenic genes StAR, cyp11a1, 3β-HSD, cyp17a1, and cyp19a1a affected by BPA. The authors hypothesized that this inhibition might occur in response to a negative feedback of EE2 on pituitary follicle stimulating hormone (FSH) or could be mediated by a direct EE2 effect on the testis (Liu et al. 2012).

The endocrine disruptive activity of DEHP was documented in goldfish, where a 30-day exposure to this contaminant reduced the levels of 11- KT and LH, suggesting that the decreased sperm quality could be caused by the contaminant effects on testicular hormone levels. However, DEHP does not affect GnRH and Kiss-1/Gpr54 system. In addition, DEHP does not exhibit estrogenic activity and does not act through sex steroid receptor (Golshan et al. 2015). In a similar trial, goldfish were exposed to three nominal vinclozolin (VZ) concentrations (100, 400, and 800 μg/l). Dose-dependent VZ effects were observed, resulting in an impairment of sperm quality via disruption of steroidogenesis. Aside from VZ effects mediated by a competitive binding to AR, potential effects of VZ by direct inhibition of 11-KT biosynthesis were also demonstrated (Hatef et al. 2012b). In the same species, a chronic exposure to BPA (0.6, 4.5 and 11 μg/l) significantly decreased testosterone (T) and 11-KT levels and impaired sperm motility and velocity (Hatef et al. 2012a).

In zebrafish exposed to DES and FLU, the impairment of spermatogenesis was associated to a reduction of plasmatic 11-KT and a deregulation of aldh1a2, cyp26a1, nonos1, sycp3, dmc1, bax, and bcl2 genes codifying for signals regulating meiosis and apoptosis and dmrt1, sf1, cyp11b2, cyp17a1, ar, fshr, and lhr genes orchestrating gonadal steroidogenesis (Yin et al. 2017).

Di-n-butyl phthalate exposure significantly increased plasma testosterone concentrations and significantly reduced spiggin levels in male three-spined sticklebacks, Gasterosteus aculeatus, exposed to 35 μg/l of DBP after 22 days. Thus, DBP appears to be able to act as an endocrine disruptor at environmentally relevant concentrations in sexually mature three-spined sticklebacks, suggesting that, similarly to mammals, DBP acts as anti-androgenic also in fish (Aoki et al. 2011).

In newborn guppies, Poecilia reticulata, exposed for 3 months to a sub-lethal dose of nonylphenol (NP), an anomalous transcription of hepatic vitellogenin correlated with significant reduction of the gonadosomatic index (GSI) was described. Males did not approach females and typical posturing and sigmoid courtship behaviors were impaired. Normal courtship behavior had a 3-month delay, suggesting that NP has an estrogenic potency sufficient to temporary disrupt reproduction (Cardinali et al. 2004).

Effects of EDCs on sex reversal

In fish, EDCs exposures during critical developmental stages can alter sex phenotypes (Fenske and Segner 2004; Scholz and Klüver 2009), impair gonad development (Zhu et al. 2016), and disrupt the progress of the reproductive process (Fenske et al. 2005; Schäfers et al. 2007). In teleosts, indeed, the tight interplay between genetic and environmental mechanisms of sex determination is responsible for the final gonadal differentiation, making teleosts excellent experimental models to study this process (Devlin and Nagahama 2002; Guiguen et al. 2010). Estrogens play a crucial role in differentiation of fish ovaries, resulting susceptible to several estrogenic mediators and endocrine disruptors. In a study exposing medaka (Oryzias latipes) embryos to a potent nonsteroidal estrogen, the diethylstilbestrol (DES), molecular and cellular mechanisms involved in ovarian differentiation were analyzed (Paul-Prasanth et al. 2011). A short-term exposure, from 0 to 8 dpf, of XY fish to DES, induced a reduction of the expression of genes codifying for the male-dominant somatic cell markers, such as GSDF, SCP3, and 42SP43, while a longer period, 28 days, caused a strong reduction of their gene expression, and the gonads developed as ovaries clearly highlighting the estrogenic properties of this pollutant.

In general, the presence of oocytes in testis of gonochoristic species is considered a pathological condition, and in the roach, Rutilus rutilus, a higher prevalence of intersex has generally been observed downstream of major wastewater treatment plants in English rivers (Jobling et al. 1998, 2002). Detection of intersex individuals and upregulation of aromatases, vitellogenins, and zona radiata proteins in juvenile and male fish have been also described in a wide array of mullet species worldwide, clearly evidencing a xenoestrogenic contamination. Much effort has been focused on the identification of new molecular markers for a precocious identification of feminization responses and intersex conditions in fish populations. The ovarian RNA commonly shows a strong expression of 5S rRNA in oocytes, supporting the use of such RNA as molecular marker of oocyte presence in the testis to assess xenoestrogenicity in field conditions (Ortiz-Zarragoitia et al. 2014). Intersex in the form of testicular oocytes was described in largemouth bass (Micropterus salmoides) collected over a 5-year period survey in surface waters on the Delmarva Peninsula, USA, a region dominated by agricultural land use and poultry production (Yonkos et al. 2014). Disruption by EDCs can be transitory or permanent, mainly depending on the nature, the concentration and the windows of exposure to the chemical. However, despite an increase of vtg mRNA transcripts and protein synthesis was documented in G. niger exposed for 48 h to alkylphenols, no histological changes were observed in the testis (Maradonna et al. 2004). Nevertheless, a number of studies described the ability of zebrafish to recover from estrogenic exposures (Larsen et al. 2009; Baumann et al. 2014a, b). An exposure to the estrogen-antagonist fadrozole, alone or in combination with EE2, permanently disrupts zebrafish sexual development, inducing masculinization and severe pathological testicular alterations. Gonad histopathology revealed interstitial proteinaceous fluid deposits, ovarian atresia, and presumably degenerative mineralization. In addition, the gonadal changes induced by EE2 alone seem to be partially reversible after a recovery period (Luzio et al. 2016). BPA exposure induced an intersex phenotype in carp: in testes of males exposed to 1000 μg/l BPA, in addition to several inflammatory cells into the residual lobules, few previtellogenic oocytes were scattered within the testicular tissue (Mandich et al. 2007).

Effects of EDCs on epigenetic process

The scientific community started to talk about epigenetics in 1942 when Conrad Waddington described it as “…the interaction of genes with their environment which brings the phenotype into being” (Waddington 1942). Epigenetic modifications are now defined as reversible and heritable chromatin chemical modifications resulting in adjustment of its activity without changes in the underlying DNA sequence. Epigenetic plays an important role in many cellular processes from differentiation, growth, metabolism, and regulation of gene expression by silencing or enhancing specific genes. In addition, it was shown that epigenetic mechanisms enable the communication between gene transcription and environment (Esteller 2007). Many epigenetic modifications include DNA methylation and histone modifications (Labbé et al. 2017). Recently, the implication of microRNA deregulation in response to EDCs has been noted in many vertebrate species (Cameron et al. 2016). In this context, data obtained in D. rerio and in Carassius auratus demonstrated that the exposure to fluoxetine, except from a deregulation of the hormone-dependent processes, causes changes in miRNA expression, potentially representing a new group of biomarkers of the exposure to toxicants. Noteworthy would be the increasing knowledge on the transcriptional variation of miRNA after chronic or low-level EDCs exposure, providing an epigenetic fingerprint of environmental exposure. The detrimental effects of EDCs on epigenetic mechanisms were evidenced a long time ago; for example, the hexabromocyclododecane (HBCD) and the 17-β oestradiol (E2) were evaluated on global DNA methylation levels in the gonads of the three-spine stickleback (Gasterosteus aculeatus). Results demonstrated an increase of global genomic hypermethylation in both ovaries and testis although the increase seen in the female gonads was not statistically significant (Aniagu et al. 2008). In addition, recent studies demonstrated the ability of EDCs to induce epigenetic transgenerational inheritance (Skinner 2014).

Many compounds, including dioxins, phthalates, and polychlorinated biphenyls can directly affect the germline. Oocytes and spermatozoa are specialized cells which possess the ability to fuse and to generate an embryo. At puberty, this embryo will develop into a mature organism producing gametes. At this point, epigenetic modifications are major actors of cell differentiation and reprogramming processes, both during gametogenesis and embryo development (Labbé et al. 2017). In this context, it was recently demonstrated that reproductive physiology was significantly impaired in zebrafish females chronically exposed to 5 ul/l BPA. In mature follicles, a downregulation of signals involved in oocyte growth and maturation, associated with a promotion of apoptosis was reported (Santangeli et al. 2016). This likely occurred through changes in the chromatin structure mediated by histone modifications, indicating that the negative effects of BPA on the female reproductive system may be due to its upstream ability to deregulate epigenetic mechanisms. In the zebrafish ovary, BPA interferes with histone modification, leading to the downregulation of lhcgr mRNA levels, potentially affecting also global methylation and interfering with the dnmt expression (Santangeli et al. 2016, 2017a) (Fig. 2). In rare minnow, G. rarus, it was demonstrated that a 7-day exposure to 15 μg/l BPA significantly upregulates the expression of gonadal cyp19a1a mRNA levels, while a longer exposure, 35 days, resulted in a downregulation of aromatase mRNA. Only in the ovary, an inverse correlation was demonstrated between cyp19a1 DNA methylation status and its mRNA levels, suggesting that in the ovary, DNA methylation could be responsible for the transcriptional control of cyp191a1 (Liu et al. 2014b). In both gonads, BPA could change the DNA methylation in the 5′ flanking region of cyp17a1 and cyp11a1 at specific CpG loci. While a direct correlation was found between CpG methylation at S4 loci and cyp11a1 gonadal mRNA levels, an inverse correlation was found between CpG methylation at P4 and P6 loci and cyp17a1 mRNA levels (Zhang et al. 2017). In the testis, BPA caused a hypermethylation of global DNA, inducing a decrease of ten-eleven translocation proteins (TETs) after a 7 day exposure to 15 ul/l. The exposure to a higher (225 μg/l) BPA concentration significantly increased the global DNA methylation due to DNMTs upregulation, which was due to an increased de novo synthesis of glutathione (Yuan et al. 2016). On the contrary, in zebrafish ovary, exposure to 1 mg/l BPA induced a decrease of the expression of dnmt1, resulting in a downregulation of the global methylation (Laing et al. 2016). These contrasting results suggest that the epigenetic effects induced by the single pollutants can be concentration- and time-dependent and vary among vertebrates. In a field survey, juvenile silver eels, Anguilla anguilla, were sampled in two locations in the southwest of France, presenting a different level of contamination. Results evidenced that the DNA methylation levels of the genes encoding for the aromatase and fshr were higher in contaminated fish than in fish from the clean site, suggesting that pollution could be responsible for an increase in the methylation level of gonadal genes involved in oocyte growth and differentiation. Chronic pollution experienced by animals throughout their life can affect their reproductive capacities and possibly, their offspring, inheriting such epigenetic marks (Pierron et al. 2014).

Fig. 2
figure 2

Effects of BPA on histone modification. a Histograms bar represents the enrichment of H3K4me3 and H3K27me—trimethylation of lysine 4 (K4) and 27 (K27) in the amino terminal of histone 3 (H3). Results are analyzed with the method of percentage of input. b Table represents the gene expression level of several genes involved in oogenesis process. In red is pointed out the lhcgr gene expression which downregulation was correlated with the specific profile of the abovementioned histone modifications

Full size image

SOS environment: evidence from the wild

Laboratory trials are essential to understand and focus on the effects of the single pollutants on fish physiology. In the wild, fish are exposed to many xenobiotics, potentially xenoestrogens, whereby understanding the impacts of mixtures could be of great importance. In this contest, zebrafish were exposed to a mixture of the same contaminants widespread in the Douro River (Portugal). Results clearly demonstrated a reduction of germ cell volume in both male and female gonads, but gametogenesis was significantly disrupted only in males. In testis, the presence of abundant interstitial tissues and the presence of a vitellogenin-derived proteinaceous fluid was evidenced (Silva et al. 2012). Hypotheses have been formulated to explain the deleterious effect exerted by these environmental xenoestrogen mixes in males; a direct action on testis leading to an inhibition of androgen production or a different-level interference with the hormonal cascade that regulates maturation, finally inhibiting spermatogenesis, have been hypothesized.

A major source of contamination for wild species is represented by wastewaters treatment plant (WWTP) discharges in the ecosystem. Fish inhabiting highly industrialized or the downstream areas, present severe reproductive problems from impaired gametogenesis to intersex gonads. A field study conducted in two different sites of the Bay of Biscay clearly evidenced the presence of intersex gonads in several mullet species associated with vtga gene transcription in the liver of both intersex and male fish. In females, transcription levels of cyp19a1a and gtf3a suggested an alteration of gametogenesis (Valencia et al. 2017). In a USA field survey, male largemouth bass, Micropterus salmoides, were sampled from two locations in Lake Mead, a site influenced by treated municipal wastewater effluent and urban runoff (Las Vegas Bay), and a reference site (Overton Arm). GSI and sperm motility did not differ between sites, but sperm count was lower by nearly 50% in fish from Las Vegas Bay. A non-linear positive association between ketotestosterone (KT) and GSI was identified (Goodbred et al. 2015). In conclusion, the higher concentration of contaminant body burdens coupled with the reduced levels of KT and sperm count in fish from Las Vegas Bay suggest that male reproductive condition was influenced by contaminant exposures. Plasma KT and GSI were associated in a non-linear fashion. This observation provides a framework to understand why GSI was similar between male bass from both sites despite their large difference in plasma KT. The non-linear model also suggested the existence of post-gonadal growth functions of KT at high concentrations (Goodbred et al. 2015). In another monitoring study, adult females Hoplias malabaricus were sampled at two locations in São Paulo State (Brazil), one reference site, Ponte Nova (PN) reservoir, and the polluted Billings (BIL) reservoir. The GSI, including the predominance of vitellogenic oocytes, was higher in spring and summer in both locations, but the oocyte recruitment dynamics were different. In winter, females from BIL presented vitellogenic oocytes and high 11-ketotestosterone levels suggesting precocity in the vitellogenic phase respect to the females from PN. In animals from PN, high deposition of lipids occurred in the ovaries. However, plasma estradiol levels did not vary throughout the annual cycle. In animals from BIL, plasma estradiol levels peaked during the summer, but the ovarian lipid content remained unchanged throughout the year. Fish of both sites regularly reproduce (Gomes et al. 2015). Summarizing, results show that in the polluted area, H. malabaricus developed an efficient reproductive plasticity with a longer reproductive period and an adjustment in hormonal profiles. Recently, 21 PCBs and 5 organochlorines were detected on the order of ng/g in anchovy, Engraulis encrasicolus, caught during a field survey in the western Adriatic Sea. Female sex-specific reproductive biomarkers, vitellogenin, vitellogenin receptor, and genes encoding for the zona radiata proteins were found transcribed also in male tissues; in addition, intersex was histologically identified in the 13% of the testis (Miccoli et al. 2017). These findings contribute to build a novel scenario; to date, in fact, variations of the European anchovy stocks have been linked to uncontrolled fisheries and climatic changes, without considering the negative effects of environmental pollution on fertility and reproduction.

Omics technologies: new paths forward

In the last decades, the development of omics technologies (transcriptomics, proteomics, and metabolomics) has provided the scientific community with new integrative approaches to explore new aspects of biology. These technologies enable researchers to assess thousands of genes, proteins, or metabolites in a single sample offering the potential to investigate responses to chemical stressor by identifying molecular-level evidence indicative of a specific mechanism of action and to improve our understanding of molecular toxicity pathways (van Ravenzwaay et al. 2012; Ellinger-Ziegelbauer and Ahr 2014; Baker and Hardiman 2014). Moreover, they offer great potential to study EDCs. While the studies reviewed in the previous sections provide essential knowledge about gametes adverse outcomes triggered by EDCS (i.e., gametes quality and intersex gonads), they only partially investigate the molecular mechanisms eliciting the observed effect.

Although, many studies have already applied omics approaches to investigate the molecular mechanisms involved in ovary and testis maturation (Tingaud-Sequeira et al. 2009; Groh et al. 2011; Martyniuk et al. 2013; Xu et al. 2016), only few of them have focused on elucidating the factors involved in the control of egg (Kohn et al. 2015; Yilmaz et al. 2017; Żarski et al. 2017) and sperm (Li et al. 2017) quality in fish and even less are those investigating how EDCs affect fish gametes quality. Santos and collaborators (Santos et al. 2007) investigated the gonadal transcriptome responses following exposure to EE2 in both male and female zebrafish and identified genes involved in the regulation of cell cycle, ubiquitin system, and glutathione peroxidase to be affected and associated with the changes observed in gametes quality. Gao et al. (Gao et al. 2017) identified EE2 to disrupt oocyte development and spermatogenesis in adult rare minnow (Gobiocypris rarus) by applying gonadal transcriptome analysis. These studies highlight the ability of omics technologies to unravel mechanisms of toxicity of EDCs underlying gametes disruption.

Conclusion

Environmental EDCs have a negative impact on both male and female gametogenesis. These impacts not only influence the number of gametes but also their quality on a genetic and epigenetic level, concealing a potential transgenerational effect. The physiopathology leading to altered fertility is based on hormonal disturbances. The appearance of intersex gonads, gamete quality, and epigenetic alterations, probably work in combination, to explain this negative impact. The main concern is the difficulty to identify the individual role of specific pollutants in the environment, since fish in the wild are exposed to several pollutants simultaneously. Understanding the mechanisms of EDCs action is a priority due to the growing number of reports describing the negative effects of such compounds on aquatic organism reproduction. This is a highly challenging task, because most of the time, these pollutants occur in water bodies at very low concentrations. The evidence obtained in the diverse experimental models could be of great importance to formulate recommendations for water quality and to set the maximum level of pollutants allowed into the environment. Studies need to be carried out, and the necessary policy changes should be done in a definite time-frame in order to have a proper assessment of endocrine disruptors. The damage that has already been done needs to be assessed and dealt with. Regulatory decision and research on EDCs should be based on principles of endocrinology. We envisage that application of omics technologies for the investigation of EDCs effects on fish gametes have the potential to address these challenges and provide essential knowledge to be integrated into regulatory decisions.