Endocrine Disruptors and Intersex Fish: Consequences of Anthropogenic Activity

Abstract: Endocrine disruptors are abundant in aquatic ecosystems worldwide. Numerous field studies have observed the incidence of intersex male fish in fresh water systems contaminated with sewage effluent, which likely contains estrogenic endocrine disruptors. Such intersex fish are characterized by the presence of female oocytes in testes, the expression of a female-specific protein, vitellogenin, and in severely feminized fish, ovarian ducts have been observed. Studies performed in the laboratory provide convincing evidence that EDCs in effluent do in fact cause intersexual qualities in male fish, and that the field studies are not merely correlational. I explore these field and laboratory studies, and show that both types of biological investigation were necessary to reveal the effects of EDC exposure on wild living fish populations. Additionally, I explore possible effects of EDCs on human males.

Endocrine disruptors (EDCs) are exogenous chemicals that interfere with animal endocrine systems that dictate hormone synthesis and distribution in the body (Mills & Chichester, 2005). Myriad substances are conjectured to alter endocrinology by mimicking endogenous hormones. That is, EDCs can bind to hormone receptors, inappropriately agonizing or antagonizing an organism’s physiology (Mills & Chichester, 2005). They have also been shown to change hormone synthesis, hormone metabolism, and hormone receptor expression. (Sonnenschein & Soto, 1998). It is thought that EDCs perturb sexual development and reproduction in vertebrate organisms, including humans, by interfering with endogenous sex steroid pathways (Mills & Chichester, 2005). EDCs have additionally been linked to increased rates of gonadal cancers, and precocious onset of puberty in human females (Kristof, 2013). EDCs have been researched in many vertebrate animals in both their natural habitat (the field) and the laboratory (Mills & Chichester, 2005). Such efforts will hopefully result in an understanding of EDCs’ consequences on wildlife and human health, as well as on the environment and ecosystems at large.

Aquatic environments are especially susceptible to EDCs, which enter fresh water, estuarine, and marine environments via treated sewage disposal, termed effluent (Van Der Kraack et al., 2001). Many EDC candidates, such as 17β -estradiol, estrone, estriol, and 17a-ethynylestradiol, are present in sewage effluent across the globe (Baronti et al., 2000). These estrogenic compounds can likely bind, at low concentrations, to estrogen receptors with an affinity that nearly matches that of endogenous estrogens (Mills & Chichester, 2005). Alkylphenols, biodegradation products of common detergents, are also present in effluent and similarly bind estrogen receptors, albeit with less potency (Korner et al., 2000). Pesticides with estrogenic capacities, such as atrazine, are persistent in aquatic environments and resistant to degradation (Hayes et al., 2011).

Fish are uniquely vulnerable to EDCs: they can be exposed via respiration, osmoregulation, dermal contact, and ingestion of contaminated food, among others (Mills & Chichester, 2005). There is evidence that EDC exposure reduces fertility in free-living fish, thus rendering the survival of such populations at risk. There is widespread concern that these consequences may translate to other vertebrate classes, including humans (Mills & Chichester, 2005). As such, researchers have investigated EDC exposure in fish populations to acquire a better understanding of their possible deleterious effects. To examine the existing evidence, I explore EDC field studies in combination with more controlled laboratory research. Under this model, the physiological relevance of EDCs’ effects in the wild can be coupled to knowledge obtained from manipulated laboratory experiments in a synergistic relationship that results in a rich understanding of the issue (Calisi & Bentley, 2009). I highlight the complementarity between the two methods of biological research: here, fieldwork has established a correlation between EDCs in effluent and repercussions in fish, while further laboratory investigations have confirmed that estrogenic compounds can indeed elicit these observed changes. Finally, I ask if humans are also at risk for EDC exposure, and explore the possible consequences of EDCs on ecosystem perturbation and the environment.

The HPG Axis in Fish

EDCs can exert adverse effects on the hypothalamic-pituitary-gonadal axis (HPG axis), a highly conserved hormone pathway that facilitates development and maturation of the reproductive system in virtually all vertebrates (Ankley & Johnson, 2004). Many EDCs that perturb the HPG axis enter aquatic ecosystems and thus affect fish reproduction (Ankley & Johnson, 2004).

The main tissues of the HPG axis include the hypothalamus, the pituitary gland, and the gonads — the ovaries in females and testes in males. Both the hypothalamus and pituitary are located in the brain. In fish, hypothalamic neurons directly innervate the intercellular space of the pituitary, releasing gonadotropin-releasing hormone (GnRH). GnRH triggers the synthesis and secretion of two further gonadotropins, FSH and LH-like, which diffuse into the bloodstream until reaching their target organs, the gonads. FSH triggers the maturation of oocytes and sperm cells, whereas LH induces sperm release or ovulation, as well as final maturation of gametes.

In fish ovaries, primary oogonia mature to oocytes via oogenesis. Germ cells in different stages of development are present concurrently in the ovaries. Ovarian thecal cells assist in the production of steroid hormones. Following secretion from the pituitary, LH acts on thecal cells and induces testosterone synthesis. This steroid subsequently diffuses into the granulosa, where FSH stimulates the conversion of testosterone to 17β- Estradiol (E2) by activating the enzyme aromatase. Among other functions, 17β- Estradiol is transported to the liver, where it elicits the synthesis and release of vitellogenin (vtg) into the blood. Vtg is transported to the oocytes, and serves as the precursor for egg-yolk proteins that ultimately provide nutrients to developing embryos.

In fish testes, germ cells proceed from primary spermatogonia to mature sperm. Sperm of similar developmental stages group together in what are termed testicular cysts. Sertoli cells, referred to as “nurse” cells, supply nutrients to maturing sperm, while Leydig cells provide the site for steroid synthesis. Specifically, testosterone (T) and 11-ketotestosterone (11-KT) are synthesized in such cells. Although male fish do not normally produce vtg, the receptors and genetic machinery for its synthesis have been evolutionarily conserved (Ankley & Johnson, 2004). Therefore, if male fish encounter EDCs capable of activating these receptors, they too synthesize vtg. Vitellogenin expression is therefore an accepted biomarker of estrogenic exposure in male fish (Ankley & Johnson, 2004).

Intersex Fish are found in Rivers Contaminated with EDCs

Much research is performed in a laboratory setting. This paradigm controls for as many variables as possible, facilitating easy interpretation of data. However, animals should also be investigated in their natural environments to acquire meaningful results. Recent research has shown that the same experiment, performed in both the field and laboratory, can yield differing results (Calisi & Bentley, 2009). Nevertheless, research in the field presents obstacles, as there are countless factors that are uncontrollable in field studies. To acquire the most valid results, both field and laboratory work should therefore be embraced. Here, the link between EDCs and intersex fish was initially revealed in the field. Performing controlled laboratory work further supported the hypothesis that EDCs cause intersexuality in fish. The following section presents the important EDC field studies performed across the globe.

Field Studies Report Relationship between EDCs and Intersex Fish

Research conducted by Jobling et al. (1998) provided strong evidence that estrogen contaminated effluents cause intersexuality in wild-living roach, a species of cyprinid fish. Jobling and colleagues initially provided important background information. Specifically, EDCs known to interact with the estrogen receptor are present in effluent, the treated sewage released into U.K. river systems. The authors collected 60–100 roach from eight different U.K. rivers, at locations both upstream and downstream of effluent entry, as well as five reference sites (lakes and canals devoid of effluent) to collect control fish. Upon macroscopic investigation, all fish appeared phenotypically male or female. However, after histological examination, the authors found that a significant proportion of genetic males were in fact intersex. That is, their gonads contained varying degrees of oocytes.

Intersex fish were found at all sample sites. However, there occurred a higher incidence at sites downstream of effluent entry, suggesting that the effluent was indeed responsible for the observed intersexuality. At two downstream river sites, the incidence of intersexuality reached 100%, while incidence ranged from 11.7% to 44.4% at upstream sites. The authors devised an intersexuality index system to rate intersexuality severity, which ranged from 2–7 in the collected fish. Intersex fish with a rating of two exhibited a relatively low number of oocytes in their testes, while intersex fish at the higher end of the scale possessed at least 50% ovarian tissue, no sperm duct, and the presence of an ovarian cavity instead (Jobling et al., 1998). Interestingly, oocytes of male testes were similar in appearance and size to oocytes of genetically female ovaries. The observed (male) oocytes were typically primary oocytes, although a small number of intersex fish collected from downstream sites contained secondary oocytes; this was especially true where the prevalence of intersex fish was high (Jobling, 1998).

To further support their hypothesis, the authors measured vtg expression, a well-established biomarker of estrogen exposure. Indeed, the authors found that intersex fish possessed relatively higher levels of this protein as compared to control fish, and further, that expression level was positively correlated with the degree of intersexuality (Jobling, 1998). These data provide additional evidence that estrogen exposure from the effluent is linked to intersexuality in wild fish populations. The authors additionally noted a positive association between effluent concentration in a particular river site and incidence of intersexuality. Importantly, downstream samples were collected at least a few kilometers from effluent entry, implying that EDCs in the effluent wreak significant effects over long distances. In this study, the authors collected a large sample size, and provided convincing data that EDCs present in effluent are correlated to the observed prevalence of intersex fish.

Aerle et al. (2001) performed a similar field study in which they established incidence of intersexuality in a second species of cyprinid fish, the gudgeon. The authors collected over 400 fish from two United Kingdom rivers contaminated with sewage effluents, and collected control fish from non-contaminated lakes. Upon histological examination, a proportion of males at every testing site exhibited testicular oocytes, indicating intersexuality. The highest incidence of intersexuality at a single site reached 15%. At one site, no intersex fish were observed; however, the sex ratio was highly biased toward females, with only 3% of sampled fish as male, as determined by histological analysis. Thus, genetic males possibly experienced complete sex-reversal during development, explaining the overwhelming predominance of phenotypic females.

In this study, most intersex fish possessed few testicular oocytes. However, at one downstream site, the degree of intersexuality was very severe, for 6 of 15 intersex fish exhibited ~75% ovarian tissue (Aerle et al., 2001). Although Aerle and colleagues observed lower incidences of intersexuality as compared to Jobling et al. (1998), it is possible that sampling error skewed their results. That is, EDCs often accumulate in soft sand and dirt at the bottom of rivers. Conversely, due to the utilized fishing methods, most fish were caught near rocky bottoms. The sample may have therefore been misrepresentative of the gudgeon population, and possibly, there exists a higher incidence of intersexuality in these fish. Aerle et al. (2001) provided further evidence that EDCs in sewage effluents wreak deleterious effects on male fish, and the authors showed strong evidence that such EDCs may even result in complete sex reversal of genetic males to phenotypic females.

Field studies in the United States have also demonstrated the presence of intersex fish in EDC-contaminated freshwaters. In a massive field study performed from 1995–2004, Hinck and colleagues illuminated the occurrence of intersex black bass in U.S. rivers. Fish of 16 species were collected at 111 river sites, from all regions of continental United States, which contained varying degrees of contamination. The authors found intersex fish in four of the sixteen fish species, and overall observed 97 intersex fish out of 3110 (Hinck et al., 2009). All but one appeared to be genetic males with differing degrees of ovotestes.

Importantly, the authors observed a high incidence of intersexuality in smallmouth and largemouth bass in the southeastern United States. Of the largemouth bass collected, 18% of males displayed intersex gonads, while 33% of smallmouth bass males were intersex. Concentrations of 17β-Estradiol were significantly greater in intersex fish, relative to normal males, suggesting that estrogen exposure contributes to the observed intersexuality. Interestingly, intersexuality was more prevalent in younger fish (1–3 years of age), which implies that EDCs may exert their effects early in the fish life cycle. This study established that the intersex condition is widespread in small- and largemouth bass in southeastern United States. It provided further proof that intersexuality in fish, triggered by EDC exposure, is prevalent in many regions of the world.

Intersex fish have additionally been observed in Mediterranean seas. Metrio et al. (2003) found a high incidence of intersex swordfish in two different locations of the Mediterranean Sea. The researchers collected 162 swordfish and performed histological analyses on the gonads and liver of the fish. Immunohistochemistry was performed to examine levels of vtg expression. Forty of the 162 swordfish, approximately 25%, were deemed intersex, as determined by the presence of testicular oocytes (Metrio et al., 2003). The abnormal oocytes were located mostly within the lumen of the seminiferous tubules. Additionally, concentric layers of fibroblast-like cells called encystations were observed in some intersex males. These structures are hypothesized to be oocytes undergoing atresia or reabsorption. Cells in the liver of both male and intersex swordfish exhibited vtg expression, as revealed by immunohistochemistry, suggestive of estrogenic EDC exposure. In summary, the relatively high incidence of intersex fish in combination with aberrant vtg expression in both intersex and male fish suggests that Mediterranean swordfish have likely been exposed to estrogenic EDCs. If intersexuality leads to reduced fertility, as the authors suggest, and overfishing continues, the survival of the species could be at risk. As swordfish are a top predator, such an outcome could cause massive disruption in aquatic food chains. The consumption of fish might also lead to the bioaccumulation of EDCs in humans, thus posing a risk to our health.

An Italian field study collected Barbel fish, a cyprinid species, from two sites in the Po River (Vigano et al., 2001). One site was upstream of a polluted tributary, while the second was downstream. Barbel were examined histologically to determine the proportion of intersex fish. In the upstream region, no intersex fish were observed, while eight of the sixteen fish from the downstream site exhibited varying degrees of intersexual gonads.

The intersex gonads appeared more like ovaries than testes, as the reproductive organs contained ovarian cavities. Five of the eight intersexual fish possessed largely ovarian tissue, as the degree of testicular tissue ranged from 1% to 10%. The remaining intersexual fish exhibited greater degrees of testicular tissue: 30%, 50%, and 70%, respectively. Here, the male germ cells were largely immature spermatogonia, with few spermatocytes and spermatids. As Barbel display no sexual dimorphism, and genetic probes to identify sex are unavailable, the authors could not conclude whether or not intersexuality was a result of genetic male feminization of genetic female masculinization. Initially, the authors propose that the overall ovarian quality of the gonads suggests that the fish were genetic females with varying degrees of masculinization. However, studies have shown that genetic males exposed to estrogenic compounds early in development can exhibit complete sex reversal. Such reversal can redifferentiate germ cells, as well as induce development of female reproductive ducts, such as the oviduct. Thus, it is possible that the intersex fish were indeed genetic males but had developed into phenotypic females. The authors have determined a high incidence of intersex fish in another species and region of the world, providing further testimony that EDCs are a global issue.

In French rivers, scientists have likewise observed intersex fish. Minier et al. (2001) conducted a large field study, sampling cyprinid fish from three contaminated rivers in France. This was an important study, as the occurrence of hermaphroditic fish in French waters was previously unknown. Three rivers with known roach populations were selected and fish were sampled via electrofishing, a common scientific method that electrically stuns fish before they are caught. Fish were collected both upstream and downstream of effluent entry. A dam separated the upstream and downstream locations, such that movement between fish populations was restricted.

The authors observed intersex in all three sampled rivers. In general, intersex fish displayed a small number of scattered primary oocytes, never exceeding a count of twenty per testicle. This is in contrast to the severely intersex fish observed in the Jobling study, where testicles contained hundreds to thousands of oocytes. Interestingly, the intersex testes lacked spermatocytes, or mature sperm. Thus, functional spermatogenesis was likely not occurring in the collected cyrpinids. In the first river, the Bresle, 9% of male fish sampled showed intersex conditions, while upstream fish were devoid of such abnormalities (Minier et al., 2000). Additionally, at the Bethune River, 21% of the downstream males possessed ovotestes, as compared to the normal upstream males (Minier et al., 2000). This river possessed relatively low levels of effluent contamination compared to other sampled sites, and thus the incidence of intersex fish was surprisingly high, suggesting that even in minute concentrations, EDCs are highly potent.

The authors only collected three male fish at the third French river. One of these three males exhibited testicular oocytes. The authors imply that this site likely contained a high incidence of intersex fish; however, the sample size was far too small to establish a firm conclusion. This field study established the prevalence of intersex fish in French rivers, which was previously unknown, revealing that fish in yet another region of the world are affected by EDCs in river effluents. Importantly, the population equivalent, or concentration of effluent, was substantially lower at French sites as compared to the sites sampled by Jobling in the U.K. Relatively low levels of EDCs can therefore induce hermaphroditism in wild fish. Further, the low population equivalent likely explains the low severity of intersex characteristics in these fish. Only few and scattered testicular oocytes were observed here, as opposed to the presence of oviducts observed in sampled U.K. fish. Again, this study has established incidence of intersex fish in yet another global region.

Purdom et al. (1994) established an association between effluent EDCs and vitellogenesis, the synthesis of vtg, in male fish. Naturally, ovarian estrogens trigger vtg synthesis in the liver. Specifically, 17β-Estradiol induces its production. Purdom hypothesized that if effluent triggers vtg expression in male fish, then such waste likely contains estrogenic compounds.

The authors purchased adult rainbow trout, placed them in steel-cages, and situated the caged fish in three sites downstream of known effluent entry, where hermaphroditic fish had previously been observed. Caged fish were additionally positioned at two control sites with tap water. Both male and female trout exhibited increased vtg expression, as compared to reference fish, for which the levels of plasma vtg could not be detected by radioimmunoassay.

Following positive results from their first field experiment, the authors undertook a second field study. They placed caged trout in different contaminated river sites, and exposed them to effluents for three-weeks. Controls were held in tap water. At fifteen experimental sites, male fish exhibited vtg expression, which increased as time progressed. That is, blood samples were obtained at weeks 1, 2, and 3 and analyzed for circulating vtg levels. At one week, the average male vtg expression reached 33 μg/ml, while at week two mean plasma vtg averaged at 192 μg/ml. Finally, after three weeks of effluent exposure, male fish plasma vtg peaked at 373 μg/ml (Purdom et al., 1994). The control males held captive in tap water, on the other hand, exhibited vtg concentrations that were hardly detectable via radioimmunoassay: the levels for each of the three weeks were 0.1 μg/ml, 0.06 μg/ml, and 0.04 μg/ml, respectively (Purdom et al., 1994). These data suggest that estrogenic compounds are indeed present in Sewage Treatment Works effluents, and act to feminize exposed male fish, which reached levels of vtg expression typical of sexually mature female trout.

In the second part of their study, Purdom et al. (1994) investigated the effects of the contraceptive pill and its metabolites on male fish. Contraceptive pill metabolites, excreted in urine, are speculated to contaminate effluent and feminize male fish. Thus, Purdom and colleagues administered 17α-ethynlestradiol, the primary estrogen of the birth control pill, along with the two main metabolites of the pill, to male trout via intramuscular injection. Sham injected fish served as controls. Over six days, the authors analyzed vtg expression as determined by radioimmunoassay, and observed a logarithmic increase in male vtg expression, which reached levels typical of sexually mature females. The authors further compared 17α-ethynlestradiol to 17β-ethynlestradiol, which naturally induces vtg synthesis in female fish. The former elicited vtg expression to a much higher extent as compared to the latter, suggesting that only minor contamination of effluents with this compound can result in major physiological changes in exposed male fish. The authors contest that their results confirm the presence of estrogenic substances in sewage effluent. However, they admit that, due to the nature of field work, their studies could not be replicated or reproduced, and thus further work must be performed to draw any firm conclusions. The authors provided persuasive evidence that EDCs in effluent have potent effects on the physiology of aquatic organisms. However, it should be noted that the authors did not use wild-living trout — the fish utilized in the study were purchased from fish farms, and thus were not native to the rivers in which they were studied. Future research should investigate free-living fish populations to acquire more meaningful results.

Skewed Sex Ratios are Correlated with Effluent EDCs

Many studies have observed skewed sex ratios in wild fish downstream of effluent entry, suggesting that EDCs induce complete sex reversal in addition to intersexuality. Woodling et al. (2006) collected white suckers from two Colorado rivers, both upstream and downstream of effluent entry. In the spring of 2002, the authors collected 21 fish downstream of wastewater treatment plant (WWTP) effluent. Only one fish was male, revealing that this fish population showed strong sex ratio bias toward females (Woodling et al., 2006). Furthermore, pathologies were observed in the male’s testes, which were heavily scarred and contained relatively little spermatogenetic activity, as compared to healthy control males. Seven females possessed asynchronous ovarian development, while the ovaries of another female exhibited deformed oocytes that were pre-vitellogenic. These data suggest that EDCs in effluent adversely affect the female reproductive tract as well. Alternatively, the females may be genetic males that were exposed to EDCs during the labile period, marked by extreme sensitivity to low levels of sex steroid hormones. Upstream of effluent entry, the authors collected five male and seven female white suckers as reference fish. No intersex fish were observed upstream, and all fish exhibited healthy gonadal tissues. As such, the data suggest that contaminants from treated sewage are disrupting reproductive anatomy in white suckers of Boulder creek.

At their second field site, the authors collected 20 white suckers from the South Platte River at locations both upstream and downstream of WWTP effluent. Again, the sex ratio was heavily biased toward female fish, as the authors collected sixteen females and four intersex fish (Woodling et al., 2006). Two intersex fish exhibited regressed male gonads with scattered oocytes. The others possessed gonadal tissue that was 30–50% oocytes, while the rest appeared to be testicular tissue proceeding through various stages of spermatogenesis. At the upstream reference site, no intersex fish were found, and all collected fish displayed healthy gonadal tissues. These data suggest that contaminants in WWTP effluent are responsible for the biased sex ratios, and that such EDCs could be triggering complete sex reversal in genetic males at early stages of development. Importantly, the authors point out that sampling error was likely not responsible for the observed sex ratios, as high ratios of males were obtained at upstream reference sites. However, caution must be exercised in interpreting these results, as the sample sizes were small. Further studies should acquire larger sample sizes, and explore the possibility of complete sex-reversal in early development by EDCs.

EDCs are Linked to Reduced Fertility

The link between EDCs and morphological abnormalities of the male reproductive tract has been established. However, it is largely unknown whether such malformations result in decreased fertility. Jobling et al. (2002) investigated this question to reveal whether or not ecological consequences will result from EDC exposure.

The authors collected roach from five river sites 3–10 km downstream of effluent entry, and additionally collected upstream reference fish. Roach spawn in response to a change in photoperiod in combination with increased water temperature. The authors exposed roach to spawning conditions, and additionally injected the fish with carp pituitary extract, an inducer of spawning. Following injection of CPE (carp pituitary extract), the number of fish to spermiate, or, release milt (fish semen) was observed. The intersex fish spermiated less than reference males, but the results were not statistically significant. The volume of milt was significantly lower in feminized males as compared to reference males, suggesting that semen quantity is reduced in intersex fish. Intersex severity was negatively correlated with milt volume in a statistically significant manner. Of the most severely intersex fish, 33.3% were totally incapable of releasing milt (Jobling et al., 2002).

Although there was no difference in sperm density between groups, sperm from intersex males exhibited reduced motility. Between 35 to 50 seconds post spermiation, only 25% of intersex sperm displayed motility, compared to 42% motile sperm from reference male milt (Jobling et al., 2002). Additionally, motile sperm from intersex males swam at a significantly slower rate.

In a second experiment, the authors observed a difference in sperm density between the two groups, as well as reduced motility in intersex sperm. In severely feminized roach, the motile sperm swam approximately 50% slower than motile sperm from reference milt (Jobling et al., 2002). This evidence suggests that EDC exposure reduces gamete quality and likely reduces fertility in exposed populations.

Finally, the authors crossed fish from the different groups (reference vs. exposed) and observed fertilization success, determined by development of the fertilized egg past hatching. When reference males mated with reference females, the success rate was 93%. Conversely, when sperm from intersex males was crossed with eggs from exposed females, the fertilization success rate was reduced to only 68%; the difference between the two groups was statistically significant. The authors furthermore found a negative correlation between intersexuality index and fertilization success. Only 22.3% of severely intersex fish successfully fertilized female eggs (Jobling et al., 2002).

Taken together, this study provides strong evidence that EDC exposure results in reduced quality of gametes: intersex sperm exhibited less motility, slower motility, and decreased semen density. EDC exposure furthermore decreases reproductive capabilities in exposed roach. Although many studies have observed testicular oocytes, this study was the first to actually link EDC exposure to reduced fertility. If fertility is reduced, EDC exposure may render wild fish populations at risk for extinction.

EDCs Affect Intersexuality, Fertility, and Sex Ratios in Manipulated Laboratory Settings

The field studies discussed above found telling correlations between EDCs and intersex fish. However, controlled laboratory work can help to better establish causation, confirming that EDCs are responsible for ovotestes, skewed sex ratios, and reduced reproductive capacities. Indeed, results from field research have motivated many scientists to investigate EDCs in the laboratory. I now review important laboratory research, which supports the notion that EDCs do in fact cause the abnormalities observed in the field.

Koger et al. (1999) exposed medaka fish, at different developmental stages, to biologically relevant concentrations of estradiol for 24 h. The developmental stages included late embryos, 1-day, 7-day, and 21-day old larvae. Five months post-exposure, fish were analyzed. Estradiol-exposed fish fertilized fewer eggs than unexposed controls, although both groups initially laid eggs in similar quantities. This suggests that exposure suppresses reproductive function.

Exposure elicited biased sex ratios in late embryos, 1- and 7-day post hatch groups, relative to the 21-day old group. Newly hatched fish therefore exhibit enhanced sensitivity to estrogen exposure. The embryonic group displayed a ratio of 15% male: 85% female, while the 1-day post hatch group was 13% male: 87% female, and 10-day post hatch was 24% male: 76% female. Sex was determined by the presence or absence of the dorsal fin, a secondary sex characteristic typical of male medaka. The authors fail to explain the cause of sex-ratio bias. Sex-reversal of genetic males to phenotypic females or decreased viability of exposed male offspring are two possible explanations. As many field studies speculated that EDCs induce complete phenotypic sex reversal, further laboratory research should work to confirm or reject this hypothesis.

Estradiol treatment induced intersex in all experimental groups, specified by oocytes in otherwise functional testes. Additionally, offspring of estrogen-exposed fish showed decreased viability in comparison to controls, as fewer embryos survived to one week post hatching. While field studies provide only correlations between estrogenic EDCs and their effects, this laboratory study demonstrates that estrogens are in fact capable of directly yielding biased sex-ratios and intersex fish.

Papoulias et al. (2000) utilized medaka to understand when and how EDCs exert their action. In many vertebrates, bipotent primordial gonads differentiate into ovaries or testes via sex steroid induction. The authors speculated that exposure to ethinyl estradiol, a synthetic estrogen, before sexual differentiation, could induce complete sex reversal. In medaka, males are XY, while female sex chromosomes are XX. Importantly, the Y chromosome carries a gene that assists in the deposition of an orange pigment, xanthathin. Males therefore exhibit a characteristic orange-red color following hatching, while females are white, making genetic sex determination possible.

To examine the effects of estrogenic EDCs on male medaka, ethinyl estradiol (single injection) was administered to medaka embryos. Following hatching, the authors found that exposed genetic males exhibited complete phenotypic sex-reversal, as the gonads developed into ovaries with concomitant oviducts. Furthermore, on the day of hatching, progenitor germ cells began dividing. This is characteristic of oogonia but not spermatogonia. By day-29 post injection, full oviducts had formed in genetic males, with separate openings for urinary and genital tracts. Taken together, the results of this paper suggest that a single nano-dose of a synthetic estrogen can induce complete sex reversal in genetic male teleost fish. Possibly, EDCs in effluent contaminate embryos and trigger sex-reversal prior to sexual differentiation. It is plausible that biased sex ratios observed in field studies and in the Koger laboratory study resulted from genetic males that were converted to phenotypic females via early estrogen exposure.

Atrazine Induces Aberrant Sex Steroid Ratios in Male Fish

Atrazine, an herbicide, has been linked to intersexuality in male vertebrates, from frogs, to fish, and even humans (Hayes et al., 2011). Thus, atrazine is likely present in sewage effluents and may contribute to ovotestes and skewed sex ratios in fish. Atrazine is resistant to degradation in freshwater and can persist in the aquatic environment for one year (Hayes et al., 2011). Spano et al. (2003) found that atrazine at nominal concentrations induces abnormalities in the sex steroid hormone pathways. The authors investigated the effects of atrazine on vtg synthesis, E2, and T concentrations. They additionally investigated the ratio of E2, the biologically active estrogen in female fish, to T.

The authors observed no change in gonadal T or E2 levels. However, atrazine treatment decreased circulating T in male goldfish, with a concomitant increase in plasma E2. These results support the hypothesis that atrazine exerts its effects via aromatase, the enzyme that converts T to E2. Possibly, atrazine enhances aromatase expression or activity to convert circulating T to E2. This explains why normal levels of gonadal T, along with an increase in plasma T, were observed in male goldfish. Plasma E2 exceeded plasma T, thus increasing the E2: T ratio, and further supporting the theory that atrazine converted gonadal T to E2.

In this study, the authors failed to detect increased vtg synthesis or expression, which is commonly observed in free-living exposed fish. The authors propose that the nominal atrazine concentrations administered in the laboratory did not reach the threshold necessary for induction of vtg synthesis. The data presented in this paper suggest that atrazine feminizes the sex steroid profile of male goldfish, and could contribute to the incidence of intersex fish and biased sex ratios in wild fish populations observed in the aforementioned field studies.

Toward a Biological Mechanism: Aromatase and Atrazine

Atrazine, discussed above, is associated with increased risk for both ovarian and breast cancer in humans. It is estimated that 29–34 million kg of atrazine (ATR) are utilized annually by the agricultural industry. Similarly, its presence has been confirmed in drinking water, ground water, and soil samples. Hollaway et al. (2008) extrapolates from the Spano study, further confirming that atrazine exerts its action through the enzyme aromatase, which catalyzes the final step in the conversion of testosterone to estrogen.

Human ovarian cells were obtained from women undergoing IVF, and treated for 24 hours with increasing concentrations of atrazine. The treatment elicited a dose-dependent increase in aromatase activity, as determined by a water formation assay. Atrazine treatment did not induce cell proliferation, indicating that increased aromatase activity was not simply caused by the presence of more cells. At both 0.1 and 1.0 μl atrazine, a 2- fold increase in aromatase activity was observed, and these results were statistically significant (p < 0.05). The authors concluded that ATR-induced aromatase activity likely increases local estrogen and thus causes the feminization observed in male vertebrates. They also noted that the concentration of atrazine necessary to induce aromatase activity is higher than the levels of atrazine that could possibly be present in human tissues. This work successfully links observations in the field to a specific biological mechanism in the laboratory, highlighting the complementarity of both forms of research. Future work is required to confirm that a 2-fold increase in aromatase activity is biologically meaningful. Specifically, the authors do not provide information as to the typical fold increase in female aromatase when biologically activated. If aromatase is normally induced 50-fold, for example, then a 2-fold increase in activity is rendered meaningless.

EDCs Attenuate Reproductive Behaviors in a Laboratory Setting

The extent to which EDCs affect mating behavior remains unknown. At Trent University in Ontario, Gray and colleagues tested the effects of octylphenol, an established EDC, on the reproductive behaviors of medaka. Octylphenol (OP) is a surfactant, used commonly in industrial, agricultural, and domestic products and processes. Its degradation products are hydrophobic, resistant to destruction, and are often found in freshwater nearby wastewater outfalls, making them candidates for the abnormalities observed in the field studies discussed above.

Octylphenol induces intersex gonads in male fish, and the authors hypothesized that octylphenol may reduce reproductive behaviors, in turn decreasing reproductive capacities of male fish. The authors devised four treatment groups, each with fifty male fish, that received increasing, nominal amounts of 99% 4-tert-octylphenol. The fish were exposed from 1 day to 6 months post-hatching, so as to expose the fish from sexual differentiation to sexual maturity. Following the exposure, 65 reproductive trials were conducted. Three sexually mature, unexposed females were placed in breeding enclosures with one exposed male. The fish were videotaped, and the footage was observed blindly. Three behaviors, each confirmed as indicative of sexual interest, were quantified: approaches, circles, and copulations. Additionally, eggs laid by unexposed females following copulation were transferred to embryo-rearing solution and examined to determine the effects of EDC exposure on the health of the second generation of fish.

OP exposure was negatively correlated with courtship behaviors. As OP exposure increased, quick circles performed by the males decreased, suggesting that octylphenol results in reduced expression of this reproductive behavior. Similarly, male medaka from the 25 μg/L and 50 μg/L groups fertilized fewer eggs as compared to the control males. The 50 μg/L group copulated less than control males, again suggestive of reduced reproductive behavior by OP exposure. Exposure induced ovotestes in two male fish, one of which was unable to fertilize any laid eggs. Finally, in three of the treatment groups, 10, 50, and 100 μg/L, the males were significantly less capable of fertilizing eggs, which suggests that sperm quality, sperm quantity, and reproductive behaviors are stifled by exposure to octylphenol. Interestingly, embryos from exposed males exhibited abnormalities, such as tube hearts, bent tails, and even premature embryo death. This implies that EDC exposure in the parent generation is detrimental to the offspring as well as the parent generation, thus increasing the concern for the propagation of populations exposed to EDCs.

The authors conclude that OP exposure causes reproductive failure in male medaka, which were less capable of inducing egg production in females and also failed to fertilize those eggs that were laid (Gray et al., 1999). It is possible that reductions in reproductive and courtship behaviors by male medaka is a result of lower plasma T, which is converted to E in the brain to trigger such behaviors. The authors importantly note that the low incidence of intersex fish was likely because exposure to the surfactant was terminated three months prior to sacrifice of the fish. Thus, the fish may have had adequate time to reactivate endogenous T, and therefore it is possible that ovotestes may have initially formed but ultimately regressed. In summary, this laboratory experiment suggests that surfactants present in sewage effluent may stifle reproductive behavior in fish populations, and this could in turn lead to a global decline in the species.

Are Human Males at Risk for EDC Exposure?

Many worry that human fertility will suffer in response to EDC exposure. Research has been performed to see if the changes observed in fish are also present in humans. In the latter half of the twentieth century, studies purported that human male sperm counts were declining. In 1995, Dr. Pierre Jouannet published an alarming longitudinal study. Jouannet and colleagues obtained semen samples from a sperm bank in a Paris hospital over the course of twenty years. The semen was acquired from fertile men who had previously fathered at least one child, and were healthy, unpaid volunteers. The samples were obtained via masturbation in the laboratory after the men had abstained from sexual activity for at least three days. Following, the sperm cells were counted with a hemocytometer. The motility index of sperm was calculated by counting the number of motile sperm divided by the observed sperm count. Sperm morphology was also observed under the microscope to determine if sperm abnormalities were present. The authors found that the volume of semen did not change over the twenty-year study. However, they reported that the mean concentration of sperm declined by about two percent each year, from 89 X 106/ millimeter in 1973, to 60 X 106/ millimeter in 1992 (Jouannet et al., 1995). The authors purported that the p value was less than 0.001, and thus their results were found to be highly significant. Additionally, the authors reported that motility decreased to a significant extent as time progressed, by a value of 0.6% per year. The proportion of morphologically normal sperm cells also decreased at a rate of 0.5% per year. The authors concluded that sperm counts have indeed declined over the past two decades, and that such reduction is independent of the age of the males from which the samples were obtained. Finally, the authors suggest that reductions in sperm likely result from estrogenic components present in the environment.

This study elicited critical responses from scientists internationally, who claimed that this research lacked sound statistical analyses. Scientists further claimed that the study did not utilize a random sample of people from the population (Kolata, 1995). In a New York Times article, Dr. Richard Sherins of the Genetics and I.V.F. Institute in Fairfax, Virginia, is quoted: “Who knows what subject selection there was for a sperm bank? Twenty years ago, men might have been chosen because they had particularly high sperm counts.” Furthermore, Dr. Sherins explained that sperm count and quality are highly variable, changing significantly from week to week, and thus the samples obtained may not provide accurate portrayals of present human male sperm profiles. Finally, Dr. Sherins claimed that poor sperm counts and abnormal sperm morphology do not necessarily indicate reduced fertility, and thus this study should not provoke alarm (Kolata, 1995). Conversely, Dr. Earl Gray from the United States Environmental Protection Agency warned that the results of this study should be taken seriously, and warrant further research. In 2011, another New York Times article provided an updated status on the issue (Kolata, 2011). The article cited a recent longitudinal study conducted in Denmark, where the authors spent fifteen years obtaining sperm from 5000 eighteen-year-old Danish military men. The authors observed no decline in sperm counts whatsoever. According to the New York Times article, scientists claimed that the results of this study utilized sound statistical analyses, and that the techniques for analyzing sperm were superior to the techniques used in the earlier study.

Swan et al. (2003) conducted a massive study to determine whether or not sperm quantity and quality differed between fertile men in urban vs. rural geographical locations. The authors obtained semen samples from the partners of 512 pregnant women at prenatal clinics in four American cities: Columbia, MO, which represented the rural location, along with L.A., New York City, and Minneapolis. Sperm concentration was significantly reduced in semen samples obtained from Columbia, MO (58.7 X 106 /ml), as compared to samples acquired from urban areas (102.9 X 106 /ml NYC, 98.6 X 106 /ml L.A., 80.8 X 106 /ml Minneapolis). Additionally, semen samples from Columbia, MO possessed lower numbers of motile sperm relative to the urban sperm samples, with values of 113 X 106 /ml, compared to 196-, 201, and 162 X 106 /ml. These results were also statistically significant. On the other hand, morphology and semen volume did not differ across geographical regions. The authors controlled for age, and all donators remained abstinent for 2–5 days before providing their samples.

Men from rural areas are likely exposed to more agricultural pesticides as compared to men in urban centers. As such, it is possible that the observed differences in sperm quantity and quality are due to pesticide exposure, many of which are likely endocrine disruptors. This study, as compared to the longitudinal sperm study discussed above, provides convincing evidence that EDCs have the capacity to affect the fertility of human males as well as lower vertebrates. However, the results must be analyzed with caution, and further studies must address whether or not reduced sperm quantity and motility will actually have a significant deleterious effect of human male fertility.

EDCs May Bias Human Sex Ratios

There is concern that EDCs affect human sex ratios, as observed in exposed fish populations. A longitudinal study performed in the United States examined the nation’s sex ratios from 1969–1995 to investigate this concern (Marcus et al., 1998). The authors obtained birth certificates, 3–4 million per year, from the National Center for Health Sciences and the CDC. The birth data contained information on sex, birth order, race, age, geography, and parental age. In 1969, the authors observed that the sex ratio was 105.3 Males: 100 Females. It was found that this ratio declined by 1995 to 104.9 Males: 100 Females, albeit only in births from Caucasian mothers. Additionally, this pattern was observed across all geographic regions under investigation in the study. Conversely, the authors found that the sex ratio increased among black newborns, again across geographical regions. As such, the authors conclude that an environmental contaminant is not responsible for the observed changes in sex ratios. Similarly, the results do not indicate that a drastic change in sex ratio has occurred, and thus further research is required to determine if concern over this issue is valid.

Mocarelli et al. (1996) investigated sex ratios in Seveso, Italy, where an accident in 1976 released kilograms of dioxin into the environment. Dioxin is a toxic human-made substance, and has been shown to affect hormone activity in animals. The authors examined the more contaminated “A” zone in the town in comparison to the less contaminated “B” zone, as determined by dioxin soil concentrations in the respective regions. According to the authors, of the 74 people born subsequent to the accident until December of 1984, there was a drastic bias in the sex ratio: 26 males were born in comparison to 48 females. Results from the B zone sex ratios were not provided. An excess of female births was found to be associated with parents from the A zone, suggesting that dioxin exposure in the parents resulted in biased sex ratios.

Conclusion

Endocrine disruptors are of great concern to the health and reproduction of vertebrate species. It is possible that the observed abnormalities in fish populations will translate to humans. Such consequences to human fertility remain largely unknown and debated. However, from numerous field studies with concomitant laboratory work, it is valid to conclude that fish populations are being highly affected by estrogenic EDCs present in effluent all over the globe. In addition to morphological abnormalities, EDC exposure has been shown to decrease fertility in free-living and laboratory-reared fish populations. In combination with the overfishing that already occurs, EDC exposure could eventually lead to the demise of many fish species; this could disrupt major ecological food chains in aquatic ecosystems. Further, many people rely on fishing for their livelihood, and many people also rely on fish as a source of protein. Thus, a decline of fish species would affect the job security and health of millions of people. Hopefully, the proliferation of research dedicated to EDCs will result in an effort to remove these harmful chemicals from our environment.

References

Aerle et al. (2001). Sexual disruption in a second species of wild cyprinid fish (the gudgeon, Gobio gobio) in United Kingdom freshwaters. Environ Toxicol Chem. 20(12), 2841–2847.

Ankley, G.T., Johnson, R.D. (2004). Small fish models for identifying and assessing the effects of endocrine-disrupting chemicals. ILAR J. 45(4), 469–483.

Baronti et al. (2000). Monitoring natural and synthetic estrogens at activated sludge sewage treatment plants and in a receiving river water. Environ Sci Technol. 34, 5059–69.

Calisi, R., Bentley, G. (2009). Lab and field experiments: Are they the same animal? Hormones and Behavior, 56(1), 1–10.

Gray, M., Metcalfe, C.D. (1996). Induction of testis-ova in Japanese Medaka (Oryzias latipes) exposed to p-nonylphenol. Env. Tox. Chem. 16 (5), 1082–1086.

Gray et al. (1999). Reproducive success and behavior of Japanese Medaka (Oryzias latipes) exposed to 4-tert-octylphenol. Env. Tox. Chem. 18 (11), 2587–2594.

Hayes et al. (2011). Demasculinization and feminization of male gonads by atrazine: consistent effects across vertebrate classes. J Steroid Biochem Mol Biol. 127 (1–2), 64–73.

Hinck et al. (2009). Widespread occurrence of intersex in black basses (Micropterus spp.) from U.S. rivers, 1995–2004. Aquat Toxicol. 95(1), 60–70.

Holloway et al. (2008). Atrazine-induced changes in aromatase activity in estrogen sensitive target tissues. J Appl. Toxicol. 28(3):260–70.

Jobling et al. (2002). Wild intersex roach (Rutilus rutilus) have reduced fertility. Biol Reprod. 67(2), 515–524.

Jobling et al. (1998). Widespread sexual disruption in wild fish. Environ. Sci. Technol. 32 (17), 2498–2506.

Jounnet, P., Auger, J., Kuntsmann, J.M. (1995). Decline in semen quality among fertile men in Paris during the past 20 years. New Eng. Journ. Med. 332 (5), 281–285.

Koger et al. (2000). Determining the sensitive developmental stages of intersex induction in medaka (Oryzias latipes) exposed to 17 beta-estradiol or testosterone. Mar Environ Res. 50, 1–6.

Nicholas Kristof, “Warnings from a Flabby Mouse,” The New York Times, January 19, 2013, accessed November 9, 2014. http://www.nytimes.com/2013/01/20/opinion/sunday/kristof-warnings-from-a-flabby-mouse.html

Gina Kolata, “In update on sperm, data show no decline,” The New York Times, June 6, 2011, accessed November 9, 2014. http://www.nytimes.com/2011/06/07/health/research/07sperm.html?_r=1&pagewanted=print

Gina Kolata, “Study finds sperm counts are declining,” The New York Times, February 2, 1995, accessed November 9, 2014. http://www.nytimes.com/1995/02/02/us/study-finds-sperm-counts-are-declining.html

Korner et al. (2000). Input/ output balance of estrogenic active compounds in a major municipal sewage plant in Germany. Chemoscphere 40, 1131–42.

Marcus et al. (1998). Changing sex ratio in the United States, 1969–1995. Fertil Steril. 70(2):270–273.

Metrio et al. (2003). Evidence of a high percentage of intersex in the Mediterranean swordfish (Xiphias gladius L.). Mar Pollut Bull. 46 (3), 358–361.

Mills, L.J., Chichester, C. (2005). Review of evidence: are endocrine-disrupting chemicals in the aquatic environment impacting fish populations? Sci Total Environ. 343(1–3), 1–34.

Minier et al. (2000). An investigation of the incidence of intersex fish in Seine-Maritime and Sussex region. Analusis 28(9), 801–806.

Mocarelli P, Needham LL, Marocchi A, Patterson DG, Jr, Brambilla P, Gerthoux PM, Meazza L, Carreri V. (1991) Serum concentrations of 2,3,7,8-tetrachlorodibenzo-p-dioxin and test results from selected residents of Seveso, Italy. J Toxicol Environ Health 32(4):357–366.

Purdom et al. (1994). Estrogenic effects of effluents from sewage treatment works. Chem and Ecol. 8(4), 275–285.

Spano et al. (2004). Effects of atrazine on sex steroid dynamics, plasma vitellogenin concentration and gonad development in adult goldfish (Carassius auratus). Aquat Toxicol. 66(4), 369–379.

Sonnenschein, C., Soto, A.M. (1998). An updated review of environmental estrogen and androgen mimics and antagonists. J Steroid Biochem Mol Biol 65, 143–50.

Swan et al. (2003). Geographic differences in semen quality of fertile U.S. males. Environ Health Perspect. 111)4): 414–420.

Van Der Kraack et al. (2001). Endocrine toxicants and reproductive success in fish. Hum Ecol Risk Assess. 7, 1017–25.

Vigano et al. (2001). First observation if intersex cyprinids in the Po River (Italy). Sci Total Environ. 269, 189–194.

Woodling et al. (2006). Intersex and other reproductive disruption of fish in wastewater effluent dominated Colorado streams. Comparative Biochem and Physiol. 144, 10–15.