The endocrine disrupting effects of phytoestrogens has been of emerging interest in both animal production and human health. The full spectrum of effects on reproduction continue to be elucidated. Initially of interest in human reproduction, the impact of phytoestrogens on reproduction is becoming increasingly relevant in livestock production because several key pasture species are estrogenic. The effects of estrogenic pasture on livestock reproduction are a considerable economic and animal welfare issue.

Several species of subterranean clover (“sub clovers”) were incorporated into Australian livestock grazing systems in the 19^th century (1, 2). These sub clovers were widely adopted due to their nutritive value as pasture legumes, their resilient characteristics such as the ability to remain dormant over many seasons, their lower requirement for fertilizer than other pasture species, and their persistence under grazing (1). As a result, sub clover pastures are well integrated into animal production systems in Australia. Many of those species are now recognized to have high levels of the isoflavone formononetin, which is metabolized to equol in the rumen. Equol causes a spectrum of moderate to severe functional and morphological changes to the ewe reproductive tract (3–5) that has been termed “clover disease”. Although it is suspected that clover disease continues to cause fertility and welfare issues in sheep, investigation into it has been dormant since the 1990’s.

The female has been the focus of livestock research into phytoestrogen effects on reproduction. Early investigations into the disruptive effects of estrogenic clover described masculinized genitalia, morphological changes to the cervix, and irregularities in the duration of the estrous cycle (3, 4, 6, 7). While there have been reports on the profound effects of phytoestrogens on male fertility in several species (8–10) and in vitro (11), there has been very little investigation into reproductive compromise in the ram.

The obvious solution to mitigate clover disease in sheep production is pasture renovation to eliminate the sub clover cultivars that are estrogenic. However, pasture renovation can take several years, and does not guarantee that estrogenic sub clover will not persist or re-emerge, given the resilient characteristics of this species (1).

Though the known presentation of clover disease has changed over time, with reduced severity compared to initial reports (4, 12, 13), an impact on sheep reproduction is likely still present (14). With the expansion of research into the impacts of other endocrine disruptors on reproduction (15, 16), new perspectives emerge on the potential mechanisms behind how estrogenic pasture influences on sheep reproduction. The first step toward prevention of the issue, therefore, is to define the severity, nature, and distribution of clover disease in commercial systems, to establish the mechanistic basis behind the phenomenon, and to increase awareness of the potential impacts of estrogenic sub clover within the industry. Accordingly, this review examines phytoestrogens as endocrine disruptors in sheep reproduction, with reference to known and postulated mechanisms of phytoestrogens in both the sexes.

Ovine clover disease was first detected and described in Western Australia in the 1940’s (17). Decades later, the disease was attributed to the high formononetin content of several cultivars of Trifolium subterraneum, a common pasture legume in sheep production systems in Australia (3, 4, 6, 7, 18). In the rumen, formononetin is converted to the estrogenic metabolite equol, causing masculinized genitalia, morphological changes to the cervix ( Figures 1A, B ), and irregularities in the duration of the estrous cycle (3, 4, 6, 7). These changes are accompanied by a reduction in the fertility rate, an increase in dystocia, an increase in the number of stillborn lambs, and an increase in the incidence of uterine prolapse after parturition (3, 17–19). Though these studies were critical in identifying the role of estrogenic sub clover in the compromise of sheep reproduction, the mechanisms behind the disease are not fully understood. Though the severe manifestations of clover disease appear to have been eradicated, anecdotal reports indicate that more subtle manifestations of sub-fertility have persisted when sheep graze on sub clover cultivars that are estrogenic.

Despite evidence that clover disease remains an issue in Australian sheep production systems (14), there has been little progress toward understanding the disease or developing solutions that will mitigate the impact on ovine fertility. Investigations in other species have revealed that phytoestrogens operate via diverse pathways, including competitive binding to nuclear estrogen receptors, that can result in both agonistic and antagonistic actions, as well as by modifying the epigenome. While these studies in other species confirm that exposure to phytoestrogens has detrimental effects on reproduction, particularly during some sensitive developmental periods, there remains a lack of comprehensive evidence on the causal pathways.

New additions continue to be made regularly to the spectrum of plant-derived estrogenic compounds that are known broadly as phytoestrogens. The predominant classes of phytoestrogens are isoflavones (daidzein, formononetin, genistein, glycitein), flavones (luteolin), flavonoids (quercetin, kaempferol), coumestans (coumestrol), stilbenes (resveratrol), and lignans (lariciresinol, matairesinol, pinoresinol, secoisolariciresinol) (20, 21). When they are consumed by an animal, many of these compounds can be further metabolized by the gut microflora into estrogenic metabolites. An example of particular relevance is the estrogenic by-product equol, resulting from the breakdown of formononetin in the mammalian digestive tract (22, 23).

The mechanism(s) that link phytoestrogens and reproduction remain controversial, though several key pathways have been proposed via which estrogenic compounds could operate. Perhaps the most discussed pathway is the interaction of phytoestrogens with estrogen receptors, albeit with a far lower binding affinity than the endogenous ligand, 17‐β‐estradiol (24, 25). In vertebrates, there are two main subtypes of nuclear estrogen receptor, ERα and ERβ (which are comprehensively reviewed by 26). The biological functions of the two receptor subtypes vary, as does their distribution between tissues, between the sexes, and between species (27–29). When they are activated by a ligand, the nuclear receptors ERα and ERβ are thought to act through either direct genomic modulation or indirect signaling. In the former, when ligand binds to the estrogen-receptor, the receptor dimerizes and binds to estrogen response elements in the promoter region of specific genes, thereby stimulating or repressing the transcription of that gene (30, 31). Indirect signaling leads to changes in transcription without a direct interaction with DNA (32, 33). Importantly, phytoestrogens not only competitively bind with ERα and ERβ, but can promote transcription from this interaction (34, 35).

Many of estrogens actions occur after stimulation of ERα. Stimulation of ERα is known to induce cellular proliferation, including the priming of the uterus via uterine cell production (36) and cancer progression, particularly in breast tissue, reproductive tissues, and bone (37–39). Estrogen- ERα binding is also able to induce estrogen-neuroendocrine feedback in the hypothalamus (40, 41), and increase transcription in most cell types (42). While ERβ is often co-expressed with ERα (26, 43), the stimulation of ERβ is thought to oppose the actions of ERα on gene expression (26, 31, 44–46). Thus, the outcome of estrogen stimulation for a particular cell will depend on the relative expression of ERα and ERβ. Both subtypes are involved in glucose (47–49) and lipid (50, 51) homeostasis in the brain, liver, pancreas, heart, and skeletal muscle (52). Most phytoestrogens have a higher binding affinity for ERβ than ERα (53), though that bias does not clearly translate to predictable physiological outcomes in vivo in mammalian systems. Notably, the expression of both receptor subtypes is increased by increased estrogen exposure (26), which may be a crucial mechanism behind the cumulative pathology resulting from long-term phytoestrogen exposure (3, 6). In the context of estrogenic pasture, an increase in receptor expression could underly the clinical manifestations, whereby long-term phytoestrogen exposure confers permanent infertility in the ewe, rather than the transient infertility that is observed with short-term exposure.

In addition to actions mediated by ERα and ERβ, phytoestrogens can interact with the G protein-coupled estrogen receptor (GPER). While ERα and ERβ are nuclear receptors, and therefore require their ligands to access the nucleus, GPER is a membrane associated receptor that is distributed across a multitude of tissues, at least in humans and rodents (54–57). When it binds ligand, GPER operates through non-genomic signaling, a comprehensive discussion of which can be found in a recent review by Luo and Liu (54). Of note, GPER is present in the reproductive tissues of both sexes, including breast tissue (58–60), the testis (61), prostate (62), ovary (63, 64), and endometrium (65, 66). Exogenous estrogenic compounds, including genistein, bisphenol A, and zearalenone, have a similar binding affinity for GPER as they do for the nuclear estrogen receptors (67). Of interest, some phytoestrogens, including equol and genistein, have been reported to have a greater binding affinity for GPER compared to 17β-estradiol (68). Thus, these extra-nuclear estrogen receptors may be a significant pathway through which exogenous estrogens can exert endocrine-disruptive effects through non-genomic, or indirect transcriptional pathways.

Phytoestrogens could also alter reproductive function via an indirect effect on genome transcription (69) that is often referred to as organizational effects. Modifications to the epigenome, such as DNA methylation, non-coding RNA’s, and histone modification, can influence the expression of the genome without altering the DNA sequence (70, 71). There is evidence that endocrine disrupting agents, and specifically exogenous estrogenic compounds, alter the epigenome in several tissues. For example, in neonatal rodents, coumestrol, genistein, and the estrogenic metabolite equol induce hypermethylation in several regions of the genome, including tissue-specific alterations in the uterus, kidney, and pancreas (72–76), as well as broadly increasing DNA methylation in the epigenome (73), in proto-oncogenes (72), and in dermal tissue (74). In addition to the endocrine disrupting actions of phytoestrogens, it is likely that in utero exposure to estrogenic compounds is detrimental to reproductive outcomes, given that gestation is an important period of epigenetic remodeling. Phytoestrogens can cross the blood-placenta barrier and infiltrate fetal tissues, including the brain (77, 78). To date, the phenotypic consequences of the epigenetic effects of phytoestrogen exposure have not been investigated.

It is therefore possible that epigenetic modifications that follow phytoestrogen exposure will have biologically relevant effects on the ovine reproductive system. Epigenetic changes that are induced by environmental factors, such as by assisted reproductive technologies and climate, are known to modify sperm function (79, 80), reproductive success in both male and female (81, 82), and disease onset (83) across a range of mammalian species. Moreover, these modifications can persist through generations (83, 84). Targeted and broad epigenetic modifications can influence the clinical severity of other reproductive pathologies, such as varicocele (85), polycystic ovary syndrome (86), and endometriosis (87). Theoretically, epigenetic changes in the reproductive system of sheep after exposure to phyto-estrogens could be at least partially responsible for the wide variation in clinical manifestations of clover disease, and could account for the persistence of flock sub-fertility even after sheep are removed from clover pasture.

Thus, the effects of phytoestrogens can be exerted via several mechanisms and pathways. An understanding of the systemic and direct mechanisms that lead from phytoestrogen exposure to reproductive pathology will be essential to find solutions. The developmental period during which an individual is exposed to phytoestrogens is likely critical in the extent of reproductive dysfunction. Developmentally sensitive periods include uterine life, neonatal life, prior to and during puberty, and times when estrogen has an essential biological role, such as during the estrous cycle.

Despite its categorization as a female oriented issue, mounting evidence suggests that the subfertility and infertility that is associated with clover disease is likely far more complex. Since a 1960’s report that wethers on estrogenic pasture exhibited mammary development, lactation, and pathology of the bulbourethral glands (198), there is little investigation into the impact of phytoestrogens on ram reproduction. Similarly, there is no information on how ram spermatozoa perform in the female tract after phytoestrogen exposure in either, or both, sexes. The profound effects of phytoestrogens on male fertility in other species ( Table 3 ) suggests strongly that phytoestrogen impact on ram reproductive function may be a critically overlooked component of clover disease. Of note, human males exposed to estrogenic compounds at relatively high dosages or duration have smaller seminiferous tubules and impaired spermatogenesis (199). Roosters and rats exposed to estrogenic compounds have lower testosterone production (200, 201), more oxidative stress, and reduced sperm function (5). In the boar, estrogenic compounds cause premature capacitation (202) and a stronger immune response to spermatozoa in the female tract (203).

We propose that exposure to phytoestrogens could compromise both the male and the female, as well as interactions between spermatozoa and the female tract, creating conditions within the female reproductive system that are hostile, more so than usual, to spermatozoa. The exposure to exogenous estrogens could drastically impair the ability of spermatozoa to traverse the female tract and successfully fertilize, which could partially account for the poor fertility in sheep that graze estrogenic pasture (4, 204). The addition of physiological concentrations of equol to ram spermatozoa in vitro negatively alters sperm function, with decreased sperm motility, increased ROS, increased membrane fluidity, and greater DNA fragmentation compared to non-exposed ram spermatozoa (11).

One obvious factor that will influence the interaction between spermatozoa and the ewe reproductive tract is the morphological changes to the female reproductive tract that are observed after phytoestrogen exposure (3, 6, 128). Prolonged grazing of estrogenic pasture has been found to induce a loss of crypts and folds in the cervix of ewes (4, 6, 13). The crypts of the ovine cervix represent “privileged pathways”, potential routes where conditions of cervical mucous flow, composition, and viscosity are conducive to the successful progression of spermatozoa toward the oocyte (205, 206). Thus, the morphological changes that are observed in female tract and in the sperm physiology after exposure to phytoestrogens could synergize to compromise the ability of spermatozoa to traverse the female reproductive tract.

The ability of spermatozoa to navigate the female reproductive tract is not merely a function of motility, but rather involves dynamic interactions between the male gamete and the tract environment. Particularly noteworthy in the context of reproduction after animals are exposed estrogenic compounds, is the role of ovarian steroids in the production and composition of cervical mucous, and in the immune response that is stimulated by the presence of the incoming spermatozoa (205, 207). Exogenous hormones, such as those that are used to synchronize the estrous cycle of ewes, are known to vastly alter the composition of cervical mucous thereby impairing the function of spermatozoa, leading to a reduction in fertility (205, 207–210). Infertility in ewes that graze sub clover species that are estrogenic could be partially due to changes in cervical mucous that are induced by phytoestrogens. During the follicular phase, the viscosity of the cervical mucous decreases in response to rising estrogen concentrations, lowering the barrier to sperm migration through the female reproductive tract. That change optimizes the conditions in preparation for ovulation and potential insemination (205, 207). Very early studies in sheep showed that the viscosity of mucus was lower after chronic exposure to estradiol and in ewes with clover disease (12, 211). However, the lower mucus viscosity in ewes with clover disease was also correlated with reduced fertility (212). This suggests that a factor other than viscosity, such as the biochemical composition of mucus following long term phytoestrogen exposure, is less favorable to sperm transport.

While exogenous estrogen is known to alter the female immune response and the composition of seminal plasma in the male (163), it is not known if phytoestrogens elicit similar effects. When seminal plasma enters the female tract immune cells are recruited and activated, releasing cytokines and chemokines that play a role in the sperm selection process in the female tract (213). In mammals, the magnitude of this immune response varies in response to endogenous estrogen (203), exogenous hormones (214), and specifically in sheep, between breeds (205). There are reports of the production of anti-sperm antibodies in the female reproductive tract after insemination, though a predominant focus of immune studies has been the known immunogenic properties of seminal plasma (215, 216). Because some of the constituents of seminal plasma have an immunomodulatory effect (215, 217), anything that influences the composition of the seminal plasma could affect the female immune response, with an impact on sperm survival. Greater numbers of leukocytes in cervical mucus increases sperm capacitation and the acrosome reaction (218). Because leukocytes produce ROS at far higher levels than do spermatozoa (219), the production of ROS might be the underlying mechanism for the promotion of these events. It will be worthwhile to investigate whether phytoestrogens alter the composition of seminal plasma, and if that then modulates the immune response that is mounted by the female after copulation.

In summary, the combination of loss of privileged pathways for sperm in the cervix, changes in the composition of cervical mucous, and changes in the immune response of the female to spermatozoa all represent possible mechanisms that could contribute to the fertility issues that occur in sheep when they graze sub clover ( Figure 1 ). Investigation is warranted into male-female gamete interactions after either sex is exposed to phytoestrogens.

There are two clear rationale for further research into clover disease; there are clear agricultural benefits to be gained, and this syndrome in sheep presents a unique opportunity to develop a model of the reproductive consequences of phytoestrogen exposure, particularly during critical developmental periods. The review of the known literature points logically to several potential sites of action, and mechanisms of action, of phytoestrogens on sheep reproduction. This includes estrogen receptor binding, genomic and epigenetic alterations and immune modulation, at sites including the reproductive tract, brain and gametes of both the male and female. It is likely that a systemic interplay between those factors is responsible for the spectrum of pathological changes that are observed in ovine clover disease, though that itself requires clarification. In our view, it is necessary not only to understand the consequences of exposure to phytoestrogens on gamete production in both sexes, but also to understand how that exposure alters the interaction of the gametes in the female reproductive tract. An interesting avenue of research would be to explore the physiological and genetic basis of individual susceptibility to phytoestrogens. The identification and strategic breeding of resilient animals is a logical start to address the issue, as well as the removal from breeding of persistently sub-fertile individuals and the early identification of susceptible flocks. Those strategies could be used in parallel with pasture renovation to eliminate estrogenic clover.

KP drafted, edited and wrote the final manuscript. FC conducted the meta-analysis. DB and JS reviewed, edited and contributed to the manuscript. All authors approved the final version of the manuscript.

KP is supported by the Lefroy Bequest, and DB was supported by a Meat and Livestock Australia grant (P.PSH.1138).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.