Estrogen, Efferent Ductules, and the Epididymis

Abstract

Estrogen's presence in the male reproductive system has been known for over 60 years, but its potential function in the epididymis remains an important area of investigation. Estrogen is synthesized by germ cells, producing a relatively high concentration in rete testis fluid. There are two estrogen receptors (ESR), the presence of which in the head of the epididymis is well documented and consistent between species; however, in other regions of the epididymis, their expression appears to be isotype, species, and cell specific. ESR1 is expressed constitutively in the epididymis; however, its presence is downregulated by high doses of estrogen, making the design of experiments complicated, as the phenotype of the Cyp19a1^−/− mouse does not resemble that of the Esr1^−/− mouse. Ligand-independent and DNA-binding Esr1 mutant models further demonstrate the complexity and importance of both signaling pathways in maintenance of efferent ductules and epididymis. Data now reveal the presence of not only classical nuclear receptors, but also cytoplasmic ESR and rapid responding membrane receptors; however, their importance in the epididymis remains undetermined. ESR1 regulates ion transport and water reabsorption in the efferent ducts and epididymis, and its regulation of other associated genes is continually being uncovered. In the male, some genes, such as Aqp9 and Slc9a3, contain both androgen and estrogen response elements and are dually regulated by these hormones. While estrogen pathways are a necessity for fertility in the male, future studies are needed to understand the interplay between androgens and estrogens in epididymal tissues, particularly in cell types that contain both receptors and their cofactors.

Estrogen, synthesized by germ cells, targets receptors in efferent ductules and epididymis, inducing fluid reabsorption through nuclear, cytoplasmic, and membrane receptors.

OVERVIEW OF EPIDIDYMAL MORPHOLOGY AND FUNCTION

The epididymis consists of one highly convoluted duct that links efferent ductules to the vas deferens. In most species, the epididymis can be divided into four major regions known, from proximal to distal, as the initial segment, caput, corpus, and cauda (Fig. 1) [1, 2]. In the human, the initial segment is not as extensive, but a very tall epithelium is present [3]. These regions are further subdivided by connective tissue septa into discrete intraregional zones with compartmentalized gene expressions [1]. Each zone has distinct and overlapping functional properties that stem from structural and molecular differences among the five basic epithelial cell types [4, 5]: principal, narrow, apical, clear, and basal cells (Fig. 2). In large mammals, the caput epididymis includes the coiled branches of efferent ductules [3, 6]. Ciliated cells are found only in the efferent ductal epithelium [7].

FIG. 1.

Schematic and histological representation of the male reproductive tract and excurrent ducts. Schematic (left) showing relative orientation of the efferent ducts and the proximal (IS and caput) and distal segments (corpus and cauda) of the epididymis. The cauda connects to the ejaculatory duct or vas deferens. Sagittal section (right) of the efferent ducts and epididymis depicting the convoluted nature of the duct as well as its complex and changing epithelium. eff duct, efferent ductules; IS and init segment, initial segment. Bar = 1.5 mm.

FIG. 2.

Schematic representation of epididymal cell types. Principal cells make up the majority of the epithelial layer throughout the epididymis. Clear cells are large endocytic cells present in the distal caput, corpus and cauda regions. Narrow and apical cells of the initial segment and intermediate zone have a unique morphology and function that overlaps somewhat with clear cells. Basal cells are located against the basement membrane and communicate to the luminal side of the epithelium via narrow body projections and interdigitations with the principal cells.

Principal cells comprise approximately 65%–80% of the total epithelial cell population, and synthesize essentially all proteins secreted into the epididymal lumen [8–11]. Morphology of the principal cell reveals a prominent, branched, microvillus, absorptive border, but the cell changes dramatically from a tall columnar structure in the initial segment to low cuboidal cells in the cauda [2, 12].

Narrow and apical cells of the initial segment and intermediate zone have a unique morphology and function that overlaps somewhat with clear cells [13]. Nuclei of these cells are oval to spherical, and reside in the apical region of the cell. The narrow cell cytoplasm tapers between principal cells as it touches the basement membrane, but its apical cytoplasm may bulge slightly into the lumen with numerous vacuoles, endocytic vesicles, lysosomes, and mitochondria. Carbonic anhydrase II (CAR2) is found only in the narrow cells, but lysosomal enzymes cathepsin D and beta-hexosaminidase A are present in both narrow and apical cells [14]. Narrow and apical cells show subtle differences in their expressions of specific components of the endocytic-lysosomal pathway. Both cell types appear to be responsible for H⁺ secretion and bicarbonate resorption [15–17].

Clear cells are large endocytic cells interspersed between principal cells within the caput, corpus, and cauda regions, and are characterized by an apical region with numerous coated pits, vesicles, endosomes, multivesicular bodies, lysosomes, as well as lipid droplets [5, 9, 18–21]. The endocytic activity of clear cells is greater than that of any other cell type in the epididymis, and is particularly active in the cauda [8, 19]. These cells are responsible for the uptake of a number of different proteins excreted by the epididymal epithelium, as well as the contents of the cytoplasmic droplet [8, 19, 22, 23]. Clear cells also secrete protons and express NHE3 (Slc9a3) and V-ATPase (Atp6v1e1) [24–26].

Basal cells account for 15%–20% of the epithelium throughout the epididymis [9, 13, 27], and are characterized by their location against the basement membrane. Although previously unrecognized, basal cells are now known to have narrow projections that contact the luminal side of the epithelium [25]. Furthermore, these cells share extensive interdigitations with the plasma membrane of adjacent principal cells. It is believed that basal cells can endocytose factors derived from the blood or principal cells [28, 29], and may help to regulate principal and clear cell functions [25, 30, 31]. Basal cells express angiotensin II type 2 receptor, which, when activated by angiotensin II, increases proton secretion by adjacent clear cells [25].

The principle function of the epididymis is to provide a luminal environment that transforms spermatozoa info fully mature cells. In addition to sperm “maturation,” the epididymis also plays an important role in sperm transport, protection, and storage [5, 10, 32, 33]. The formation of this luminal environment is the result of net secretory and absorptive processes of the epithelium, which continually changes along the duct [5, 34, 35]. These changes include net water, Na⁺, Cl⁻, and HCO₃⁻ reabsorption, K⁺ secretion, and luminal acidification [36]. In general, the secretions function to protect, stabilize, or modify the sperm surface, with the end product being spermatozoa that are viable, motile, and able to fertilize an egg [5, 33, 37, 38].

ESTROGEN AND ITS RECEPTORS IN THE MALE REPRODUCTIVE TRACT

Aromatase

It has been known since the 1930s that developmental exposure to estrogenic compounds can induce malformations and abnormal functioning of the male reproductive tract [39–42]. However, the role that estrogen plays in male reproduction took several decades to appreciate. We now know that estrogen is produced in significant quantities in testes (i.e., rat rete testis fluid concentration is 248 ± 95 pg/ml) [43], and is present in the semen of several species [43–46]. This is due to the presence of cytochrome P450 aromatase (CYP19A1), an enzyme that catalyzes the irreversible conversion of androgens into estrogens [47, 48]. In the immature rodent testis, CYP19A1 activity is high within Sertoli cells, but becomes more prominent in the adult Leydig cells, which actively synthesize estradiol (E2) at a much higher rate than that seen in adult Sertoli cells [41, 49]. Furthermore, the presence of Cyp19a1 transcripts and protein has been demonstrated in spermatogenic cells of several species, including human, where it is localized to the Golgi of round spermatids and throughout the cytoplasm in later-stage cells, where it actively synthesizes estrogens [47, 49–53]. Spermatozoa also show intense immunostaining for CYP19A1 in the cytoplasmic droplet (Fig. 3), and actively synthesize estrogens within the lumen of the epididymis [48]; thus, epididymal sperm serve as a unique endocrine source of estrogens that target estrogen receptors (ESRs) found in efferent ductule and epididymal epithelia [54]. Interestingly, studies that have investigated CYP19A1 presence in male germ cells have found that it is more abundant in sperm that are motile [53, 55, 56]. Other studies also suggest that CYP19A1 may be present in the mouse caput epididymal epithelium and interstitium (Fig. 3), as well as in rat, human, and monkey [57–59].

FIG. 3.

CYP19A1 (P450Aromatase) in mouse spermatozoa. Luminal sperm in the intermediate zone of the caput epididymis show positive immunostaining for CYP19A1 [51] in the cytoplasmic droplet (cd). It is noteworthy that cytoplasm of the apical cell (Ap) shows low intensity for the protein and the interstitial connective tissue area (In) shows intense staining. Bar = 25 μm.

Estrogen Sulfotransferase

Estrogen sulfotransferase and sulfatase are also found in the epididymal epithelium [60–63]. The sulfotransferase may serve to help protect the epithelium from excess estrogen [64] arriving from efferent ducts and CYP19A1 activity in spermatozoa [47, 52, 65, 66]. The sulfatase, which is capable of increasing free estrogen [62], may have a role in the regulation of estrogen action in epididymis through interplay between CYP19A1, estrogen sulfation, and free estrogens. Within the epididymal lumen, estrogen sulfotransferase may play a role in stabilizing the acrosomal membrane through its ability to sulfate membrane cholesterol [61, 67].

Estrogen Receptors

In order to mediate their biological effects, estrogens typically interact with ESR1 and ESR2 (also known as ERα and ERβ), both belonging to the nuclear receptor (NR) family of transcription factors. Like other members of the NR family, ESRs contain conserved, structurally and functionally distinct domains. The DNA-binding domain (DBD) is nearly 99% conserved, and is involved in DNA recognition and binding, whereas ligand binding occurs in the C-terminal ligand-binding domain (LBD). The N-terminal domain is not as highly conserved, and represents the most variable domain both in sequence and in length [68]. Transcriptional activation is facilitated by two distinct activation functions (AFs), the constitutively active AF-1 located at the N terminus of the receptor, and the ligand-dependent AF-2 that resides in the C-terminal LBD. Both AF domains recruit a range of coregulatory protein complexes to the DNA-bound receptor.

There are several lines of evidence suggesting that estrogens and ESRs can signal through multiple distinct pathways [69]. The classical DNA-binding pathway involves ligand-bound receptors that bind directly to estrogen response elements (ERE) in the promoters of target genes [70], or can interact with other transcription factor complexes, including AP-1-responsive elements, to influence transcription of genes whose promoters do not harbor EREs [71, 72]. Alternatively, there are nongenomic rapid effects of estrogen action that are not completely understood [73, 74], but include activation through other signaling pathways, such as growth factors, which stimulate kinases that subsequently activate ESRs or associated coregulators in the absence of ligand [75].

We have known since the 1970s that the epididymal epithelium binds ³H-estradiol [76–78]. ESRs are localized in specific cells of the testis, efferent ducts, and epididymis, with considerable variability between species [see reviews in references 41, 47, 54, 79–83]. In the testes of nearly all species examined, Leydig cells express Esr1 [47, 54]. Recent studies, using cell cultures and alternative methods of fixation, have shown that Sertoli cells also express Esr1, as well as the G protein-coupled receptor (GPR) 30, and respond to estrogen activation [84]. If these more recent discoveries are confirmed, they will help to explain the reported testicular degeneration that occurs in the aromatase knockout mouse [85].

The first evidence of ESR expression in the efferent ducts and initial segment of the epididymis is on Day 16 of gestation in the mouse [86], suggesting a role for estrogen during development. The epididymis exhibits variable staining for ESR1, depending upon species and epithelial cell type studied. Figure 4 illustrates immunostaining for ESR1 in the marmoset, hamster, and mouse. The one region of the male reproductive tract that consistently shows intense immunostaining for ESR1 across all species is the epithelium of efferent ducts [54, 81, 87–95], where mRNA expression is 3.5-fold higher than uterus [88], the traditional standard for ESR expression. However, the ciliated cell of the efferent ducts, in some species, such as the marmoset, exhibit reduced immunostaining (Fig. 4F), which is surprising, as ciliated cells are known for being estrogen responsive in the female [96] and male [97].

FIG. 4.

ESR1 immunostaining [89, 95] in monkey, hamster, and mouse efferent ductules and epididymis. A–D) Marmoset monkey. Intense nuclear staining is found in proximal and distal efferent ductal nonciliated epithelial cells. Ciliated cells are negative to slightly positive. Caput epididymidis is negative, but cauda epithelium shows low-intensity staining in the cytoplasm of all cells. E–H) Golden hamster. Intense nuclear staining is found in proximal and distal efferent ductal nonciliated and ciliated epithelial cells. The cytoplasm also appears to be slightly positive. Intraepithelial lymphocytes are negative. Caput epididymis is negative except for the nuclei of occasional basal cells. Cauda epithelium is negative. I–M) Mouse. Intense nuclear staining is found in proximal and distal efferent ductal nonciliated and ciliated epithelial cells. The cytoplasm also appears to be slightly positive. In the initial segment epididymis, apical, and narrow cells are strongly positive, while principal and basal cells are slight positive to negative. All cells of the caput epididymal epithelium are positive, but the apical cell nuclei are more intensely stained. Some nuclei of the peritubular smooth muscle are also positive. In cauda epithelium, clear cells are strongly positive, while principal cells are slightly positive, as well as the cytoplasm. Nuclei of connective tissue and smooth muscle cells are also positive. A, apical cell nuclei; B, basal cells; C, ciliated epithelial cells; Cl, clear cells; L, lymphocytes; N (A–H), nonciliated epithelial cells; N (I–M), narrow cells; P, principal cells. Bar = 25 μm.

In contrast to androgen receptor (AR), which is localized nearly ubiquitously throughout the epididymis [95, 98], nuclear ESR1 is lacking throughout the epididymal epithelium in several species, although present in mouse, cat, and monkey [54, 81, 89, 95, 99]. In the marmoset, there appeared to be low-intensity staining in cytoplasm of cauda epididymal epithelial cells (Fig. 4D). Others have shown ESR1 staining in cytoplasm [84, 94, 100], but the significance remains controversial. Interpretation of ESR1 expression data throughout the epididymis has been somewhat confusing. Studies using autoradiography, E2 binding assays, and RT-PCR all indicate that ESR mRNA and protein are present in epididymal tissues [41, 77, 78, 81, 101], with the highest concentration of cytoplasmic receptor in the cauda epididymis of rabbits [102]. Immunohistochemical results have varied with the use of different antibodies and tissue processing techniques [54]. Some studies show the epididymal epithelia as only slightly positive, while others demonstrate strong ESR1 staining in principal cells and other cell types in a region-specific manner. In hamster epididymis, ESR1 staining was absent in all epithelial cells, except for a very low intensity within nuclei of the basal cells (Fig. 4K). In the mouse, ESR1 shows intense expression [95] throughout efferent ductules and epididymis (Fig. 4). In mouse epididymis, narrow, apical, and clear cells show more intense staining for ESR1 than do the principal and basal cells.

Similar to AR, ESR2 is widely expressed throughout the male reproductive tract, and is found in nearly every cell type of the testis, efferent ductules, and epididymis [89, 95]. In the rat, the pattern of ESR2 staining is opposite to that of ESR1, as ESR2 shows increased staining intensity from distal to proximal efferent ductules [103]. Despite its nearly ubiquitous expression, a defined role for ESR2 in the male reproductive tract remains to be elucidated, although one study showed an increased number of spermatogonia in the ESR2 knockout mouse [104].

ESTROGEN REGULATION OF EFFERENT DUCTULES AND EPIDIDYMIS

Androgens are the primary hormones that regulate epididymal function, with DHT playing the most active role in the caput [5, 105, 106]. However, efferent ductules and initial segment epididymis show selective regulation that is dependent upon luminal factors from the testis. Unlike the caput through cauda regions of the epididymis, androgen resupplementation following rete testis ligation or castration does not rescue epithelial morphology of the efferent duct and initial segment regions [5, 7, 107–110]. Therefore, other factors, such as estrogens, have been postulated to help regulate their function [5, 9, 41, 111, 112].

Early experiments with estrogen were either inconclusive regarding its ability to play a major role in epididymal function [78, 113, 114], or the experimental design (high dosages) prevented us from drawing any definitive conclusions. In one of the first experiments to suggest that estrogen could influence epididymal function, Meistrich et al. [115] reported a decrease in sperm transit times with exposure to E2; however, the dosage was very high. Nevertheless, estrogen was hypothesized to “… act directly on the epididymis. …” More recent studies have shown that estrogen does indeed regulate epididymal contractility by upregulating the calcium-sensitizing RhoA/ROCK pathway in epididymal smooth muscle [57], which maintains epididymal sensitivity to oxytocin and endothelin-1 [116–118].

Estrogen treatment following uni- or bilateral castration does not provide an ideal experimental model for the determination of estrogen's role in epididymal function, because ESR1 is downregulated to nearly undetectable levels with high dosages of estrogen [119]. Estrogen concentrations following systemic treatment can reach pharmacological levels [111], which would preclude alterations in ductal function being ascribed to direct effects. Others have shown increased fluid reabsorption in efferent ducts with testosterone, but inhibition with E2 [120]; however, in those studies, it was likely that ESR1 was downregulated. It is noteworthy that E2 treatment following bilateral castration increased efferent duct/epididymal weight, but not as much as did testosterone. Surprisingly, the combination of testosterone and E2 restored the weight to control levels, which were greater than testosterone alone [119].

A recent study reported that some genes expressed in the efferent ductules contain both ERE and androgen response element (ARE) in the promoter region, suggesting that a balance of estrogen and androgens may be required in this unique epithelium [70]. Compared with epididymal epithelium, the efferent ducts have a much higher expression of ESR1 [88], and thus would likely have more significant direct estrogen and estrogen/androgen dual regulation. For example, the sodium/hydrogen exchanger 3 (Slc9a3) message is reduced nearly 6-fold in the Esr1^−/− and ICI 182 780 (ICI) antiestrogen-treated mice [121, 122], but, in the castrated male, testosterone increases the mRNA [111]. Snyder et al. [111] treated males with estrogen, but found no response in Slc9a3; however, the dosage they used would have removed most of the ESR1 present in efferent ducts.

Aqp (aquaporin) is another example of a gene that contains both AREs and EREs, and is regulated dually by estrogen and androgen (Fig. 5). AQP9 is found in specific epithelial cells of efferent ductules and epididymis [123, 124] and shows differential responses to estrogens and androgens [125–127]. AQP9 staining is reduced significantly in efferent ducts of Esr1^−/− mice and ICI-treated rats [127, 128]. However, antiestrogen or antiandrogen treatments, as well as castration or ductal ligation, selectively reduce AQP9 in the epididymis with ICI having no effect in initial segment [127], while castration and ligation reduced immunostaining in the initial segment and in clear cells throughout the epididymis, but not in principal cells elsewhere [123, 127, 129, 130]. In efferent ducts, DHT and E2 or testosterone and E2 together stimulated AQP9 expression after castration, but testosterone alone was ineffective [127]. In cauda epididymis, testosterone restored AQP9 after castration [130]. In initial segment epididymis, DHT [127] and 5-alpha-androstane-3-beta-17-beta-diol, a metabolite of DHT that has higher affinity for ESR2, were effective [129]. The complexity in the results of these studies is likely due to several factors: 1) the efferent ductules express high levels of ESR1 and ESR2 constitutively, along with AR [88, 95, 119]; 2) the initial segment epididymis has the highest expression of 5α-reductase and is most sensitive to DHT [131], while the cauda would respond to testosterone [130]; 3) ICI treatment reduced AQP9 prior to the loss of microvilli [127]. AQP9 is located on the microvilli [123, 124], and can be lost if treatment alters epithelial cell morphology. Therefore, the regulation of genes in specific tissues of the epididymis will depend upon the expression of 5α-reductase, the presence or absence of ESR1 and AR, as well as their unique DNA response elements (Fig. 5).

FIG. 5.

Summary diagram of potential estrogen action in epithelia of efferent ductules and the epididymis. At least 4 potential pathways are considered. 1) testosterone (T) can enter the cell or be converted to E₂ by aromatase (Arom) found in luminal sperm [54]. Testosterone binds AR and translocates into the nucleus, where it binds to AREs, on the promoter regions of genes with or without EREs. 2) E₂ will either enter the cell, as did testosterone or bind the membrane ESR (mESR1). It remains controversial whether E₂ binds GPR30 in the membrane [144] or collaborates with mESR1 to mediate epidermal growth factor receptor (EGFR) activation (nonclassical) of kinases and phosphorylation [180]. It is well known that E₂ binds ESR1 and translocates into the nucleus for classical mediation of transcription through EREs and recruitment of numerous cofactor proteins (C1–3). It is unknown how AR and ESR1 compete for these cofactors, or what happens when the steroid balance is altered in a cell expressing both receptors. 3) The ESR1 can also be activated through phosphorylation and mediate transcription through the ERE. 4) It is well documented in other tissues that mESR1 binds E₂, resulting in very rapid cell signaling [84, 175, 178]. This rapid steroid activity through the membrane receptor involves caveolin-1, G proteins, and the phosphorylation/dephosphorylation cascades, which mediate transcription either through the ESR1/ERE or other transcription factors (TF).

ESTROGEN RECEPTOR NULL MALES

Much of what we have learned about the role of estrogen and ESR1 in the male reproductive tract has been derived from the analysis of Esr1 knockout mice [97, 122, 128, 132–140] and treatment of wild-type mice with the pure antiestrogen ICI [84, 90, 99, 103, 127, 136, 141–145]. Animals lacking a functional Esr1 gene are infertile, and sperm recovered from the cauda epididymis exhibit a lower percent motility, beat less vigorously, and are ineffective at in vitro fertilization [132]. Furthermore, the concentration of sperm in the cauda epididymis decreases as these mice age. These effects on sperm appear to be due to defects in the luminal environment provided by the efferent ducts and epididymis, and not to a direct effect on spermatogenesis [146]. This is evidenced by transplantation studies in which Esr1^−/− germ cells were capable of normal fertilization when transplanted into a wild-type testis and allowed to transverse a normal reproductive tract [147].

Testes of Esr1^−/− mice show normal histology until puberty, at which time they begin to increase in weight, show seminiferous tubular degeneration, and atrophy [132, 136, 141]. It has since been determined that this testicular degeneration results from an inhibition of normal fluid reabsorption in the efferent ducts, leading to an accumulation of fluid in the lumen [79, 97]. This is illustrated by an increase in the luminal diameter of the rete testis and efferent ductules. Efferent ductule epithelium appears undifferentiated, short in height, and lacking microvilli and components of the endocytic apparatus [97, 135–138]. As a consequence of fluid accumulation, there is a transient increase in testis weight in Esr1^−/− males, followed by a steady decrease and subsequent testicular atrophy. Long-term atrophy of the testes due to backpressure of accumulating luminal fluid is a well-recognized pathogenesis found after exposure to toxicants and ductal ligation [135, 148].

In Esr1^−/− mice, efferent ductule epithelium exhibits significant reduction in the expression of several proteins that are important in fluid/ion equilibrium, including SLC9A3, CAR2, and two water channels, AQP1 and AQP9 [122, 128]. Analysis of mRNA levels shows that Esr1 directly influences Slc9a3 transcription, whereas the lowered expression of CAR2 and AQP1 is likely secondary to the morphological defects of the efferent duct epithelium, flattening of the epithelium, loss of microvilli, and a reduced endocytotic apparatus [97, 136, 137]. AQP9 protein is reduced prior to the loss of microvilli after ICI treatment [127]; therefore, it is likely that estrogen regulates fluid reabsorptive pathway genes independent of its regulation of epithelial morphology.

Epididymal morphology in the Esr1^−/− mouse also exhibits significant abnormality [97], with hyperplasia of the narrow cells and accumulation of vacuoles and periodic acid-Schiff reaction-positive granules in narrow, apical, and clear cells, which is indicative of abnormal ion transport and excessive endocytosis (Fig. 6). These changes are associated with recent discoveries that the epididymal luminal fluid in Esr1^−/− mice is more alkaline and hypo-osmotic relative to wild-type, which is at least partly explained by decreased expression of Slc9a3, Car2, and Slc4a4 in the proximal portion of the Esr1^−/− epididymis [121, 146]. While these Esr1^−/− molecular deficiencies are similar in both the efferent ducts and epididymis, the epididymal defect is relatively cell and function specific, and not representative of global epididymal dysfunction. The alkaline, hypo-osmotic luminal fluid in Esr1^−/− males is associated with an increased frequency of damaged sperm membranes and abnormal sperm morphology, both of which likely contribute to the infertility of these mutant mice.

FIG. 6.

Epididymal epithelium from the Esr1 knockout mouse. A) Nuclei of narrow cells (N) are protruding abnormally into the lumen of the initial segment epididymis. B) An apical cell contains multiple nuclei (1, 2) and vacuoles (v) in its apical cytoplasm. C) An apical cell in the intermediate zone of the caput epididymis contains large, periodic acid-Schiff⁺ (PAS⁺) granules or lysosomes (L). n, nucleus. D) The apical cell contains PAS⁺ lysosomal granules in the basal cytoplasm, and the apical cytoplasm with large vacuoles (v) protrudes into the lumen. Bar = 20 μm.

In marked contrast to the Esr1^−/− males, Esr2^−/− mice (known as ERβKO) are fertile [149] and have reproductive tracts that are grossly and histologically normal, although there has been one study showing Leydig cell hyperplasia and an increase in the number of spermatogonia in animals lacking Esr2 [104]. Furthermore, the double Esr1, Esr2 knockout (known as ERαβKO) have characteristics that are identical to the Esr1^−/− mice [133, 150]. Thus, further investigation is required to more accurately decipher the role of ESR2 in the male reproductive tract.

EFFECTS OF ANTIESTROGENS ON EPIDIDYMIS

Antiestrogen effects on male reproduction are somewhat confusing, as there are considerable differences between species and between treatment compounds. ICI blocks both ESR1 and ESR2, and, in the adult male, produces effects similar to the Esr1^−/− mouse [41, 136, 141, 143, 151], although there are considerable differences between species. Immunostaining of ESR1 is decreased by ICI, but there is no effect on ESR2 and AR [90]. Thus, ICI effects were due primarily to ESR1 blockage. In contrast, Tamoxifen appears to function as an ESR agonist in the male reproductive tract [118, 152–156], similar to its action in select female tissues [157].

ICI has shown the most consistent antiestrogen activity in the male, having inhibitory effects on the fluid reabsorption pathway in most species tested, with dilation of efferent ductule lumens and changes in the expression of specific genes [99, 135, 136, 141, 143, 158, 159]. ICI inhibition of fluid resorption is rapid, but its effect on efferent ductal epithelium is delayed until about 8 days in the rat [141], suggesting that multiple gene pathways are involved. It remains undetermined why ICI treatment does not result in backpressure atrophy of the testis in mice and monkey [99, 141].

Similar to what has been discovered in the Esr1^−/− mouse, ICI alters the expression of numerous genes in efferent ducts and epididymis, particularly those associated with endocytosis and ion/water transport, but also numerous other pathways. For example, ICI decreases expression of Aqp1, 4, 9, Wnt4, Nptx1, Ren1, Lect1, Cryba4, Cyp4f4, Bin2a, Slc30a2, Gckr, Prom2, Ceacam1, Ctsd, Slc9a3, Krt19 (keratin), and Car2 [99, 103, 135, 136, 143, 151, 160, 161], and increases expression of Slc26a3, Slc34a2, Cftr, Atp1a1, Mmp7, Spp1, Glycam1, Ctnnb1, Ctsc, Adam7, Umod, Sftpd, Crisp1, Slc34a2, Slc38a5, Npw, and Pemt [151, 158, 161].

In rats, ICI induced a transient increase in lysosomal bodies and microvillus height in efferent ducts prior to their reduction [103]. In monkey, ICI increased ESR1 expression in caput epididymidis, while efferent ductules exhibited dilation that is typical in Esr1^−/− mice [99]. Thus, ESR activity in the epididymis appears to be more complicated than first surmised. For example, AQP9 is modulated by both estrogen and DHT in efferent ducts, but in initial segment epididymis, primarily DHT and testosterone control AQP9 [119]. In another study, DHT deficiency reduced Esr1 and Esr2 expressions in epididymal epithelial cells and moved the receptors from nucleus to cytoplasm [162]. Therefore, a balance between estrogen and androgen activity, and potential effects on the distribution of their receptors in specific regions of the male tract, will likely determine the significance of both classes of hormones (Fig. 5).

AROMATASE NULL MALE

The aromatase-null mouse model (Cyp19a1^−/−) was created to remove endogenous sources of estrogens. In one study, male Cyp19a1^−/− mice (12–14 wk) were fertile [163], but in another the males were infertile due to failure to mount [164]. However, in both studies, testicular morphology appeared to be normal in young males, but age-dependent disruption of spermatogenesis was observed, with degenerated and apoptotic round spermatids and a significant reduction in testis weight [140, 165]. In addition, these animals displayed hyperplasia and hypertrophy of Leydig cells.

It is noteworthy that the male Cyp19a1^−/− phenotype does not mimic that of the Esr1^−/− mouse, and their age-dependent testicular degeneration occurs without efferent ductule abnormalities that were observed in the Esr1^−/−. This is likely explained by several observations [47], including: 1) Esr1 is expressed in the Cyp19a1^−/− male reproductive tract [119, 140], and could be activated in a ligand-independent manner [166–170]; and 2) Esr1 ligands other than estrogen are active in the tissues of Cyp19a1^−/− males [129].

MODELS OF NONCLASSICAL ESTROGEN SIGNALING

Other novel animal models are contributing to our understanding of estrogen action in the male reproductive tract. These include mice that have mutations either in the Esr1 DBD [139, 171] or in the LBD [172, 173]. In the ENERKI male, ligand-independent Esr1 is capable of restoring activity in the efferent ducts and preventing fluid accumulation, partially rescuing the infertile phenotype of the Esr1^−/−. However, ENERKI animals are subfertile, likely owing to increased apoptosis of spermatocytes and subsequent reduced sperm numbers [173]. Similarly, despite the absence of DNA binding in NERKI mice, these animals rescue the classical Esr1^−/− phenotype, resulting in normal sperm counts and sperm motility, despite persistent reduction in AQP1, and possibly SLC9A3, expression in the efferent ducts [139]. However, as these NERKI animals age, they begin to exhibit testicular pathology and abnormal histology [174]. Regardless, studies of the ENERKI and NERKI mice demonstrate that both estrogen-independent as well as estrogen-dependent signaling pathways function in the male reproductive tract [173].

Recently, several studies have shown that the GPR30 or GPER, an integral membrane protein, is capable of mediating rapid effects of estrogen through a nonclassical pathway [142, 144, 175–179], and may cooperate with a membrane ESR1 to mediate phosphorylation pathways via epidermal growth factor receptors [180]. However, the presence and activity of GPR30 in epididymal tissues remains to be uncovered, although another GPR, the downregulation of which results in efferent ductule effects similar to those in the Esr1^−/− mouse, has been localized to efferent ducts and initial segment epididymis [181, 182].

CONCLUSION

The epididymis is highly responsive to androgens, but estrogen has a predominant role in efferent ductules and initial segment epididymis. Estrogen function in the remainder of the epididymis remains uncertain, and a high degree of species variability often complicates definitive conclusions. Abundant sources of estrogen are available for epididymal target cells, and ESRs are selectively expressed along the reproductive tract, with efferent ductules expressing ESR1 in high concentrations. However, based upon recent studies, it will be important to re-examine ESR1 and GPR30 expressions in specific cell types of the epididymal epithelium using improved methods and new antibodies. There remains a lack of knowledge regarding NR cofactors in the male reproductive system and their combined activities when present in cells containing both androgen and ESRs (Fig. 5). Such studies are necessary to better explain the consequences of exposure to environmental endocrine disruptors, as well as to provide potential targets for the development of a nonandrogen male contraceptive.