Estrogens in male germ cells

Abstract

The biosynthesis of steroids and the production of spermatozoa are two major functions of the mammalian testis, which are tightly controlled by gonadotropins and numerous locally produced factors.  Among these are the estrogens that are produced within the seminiferous epithelium via the irreversible transformation of androgens (C19) into estrogens (C18) by aromatase.  We have recently reported that male germ cells are the new source of estrogens in the testis.  For instance, estrogen receptors (ER) are found mainly in spermatids that give rise spermatozoa.  Moreover, it is noteworthy that GPR30 (a transmembrane ER) induces rapid responses after estradiol binding, which, in turn, modulates cyclins and proapoptotic factors (e.g., BAX) to affect germ cell cycle progression and apoptosis.  In summary, at least in the animal species that were studied thus far, germ cells are the major source and the target of estrogens, affecting normal male gonadal development and spermatogenesis, in particular spermiogenesis.  These findings have also shed new light on the possible adverse effects of endocrine disruptors having estrogenic activities that can cause abnormal development of the male genital tract.

Male Germ Cells as a New Source of Estrogens in the Testis More than 80 years have past since the pioneering work of Zondek1 who demonstrated that the gonad in stallions produces estrogens. This was followed by a study by Dorrington et al. reporting that the rat testis is able to synthesize estrogens from androgens. In the testis, as in other tissues, cytochrome P450 aromatase is the enzyme that irreversibly transforms androgens into estrogens. In fetal and immature animals, Sertoli cells are the main source of estrogens,5–7 and the expression of aromatase in Leydig cells is very low.3,4 Regardless, the cellular sites of aromatisation change during development, and in the adult mammalian testis, aromatase is localized mostly in Leydig cells.4 During the last decade, there has been growing evidence that within the seminiferous tubules, germ cells, at least in the species studied so far, also participate in the production of testicular estrogens.8 For example, Nitta et al.9 have shown that adult mouse germ cells express a functional aromatase and that estradiol output is equivalent to that of Leydig cells. Similar to Leydig cells, aromatase expression is relatively low in adult rat Sertoli cells,10 and Sertoli cell aromatase is negatively controlled by germ cells.11 In addition, we have demonstrated that adult rat germ cells are an important source of estrogens and that the level of aromatase contributed by germ cells represents ∼60% of that of the whole testis.8 Moreover, numerous aromatase transcripts were found to be unable to encode a functional protein.12 Interestingly, aromatase activity is much higher in late haploid germ cells than in less differentiated germ cells,8 illustrating these quiescent cells have the ability to “regulate” their fate and development during spermatogenesis. Very recently, we have also reported the presence of aromatase mRNA in gonocytes and preleptotene spermatocytes (Bois et al. submitted data), illustrating that all testicular cells express aromatase except peritubular myoid cells.13 Additionally, studies using genetic models have demonstrated the significance of germ cell derived estrogens on male reproductive function. For instance, male mice deficient in aromatase (ArKO) develop normally and are able to breed and to produce litters; however starting at the age of 5 months and onwards, ArKO males present with a failure of spermatogenesis and by the age of one year, they are all infertile. Specifically, spermatogenesis is abnormal in these ArKO mice with a blockage of germ cell maturation at the spermatid stage, leading to a 50% decrease in the number of round and elongated spermatids and their apoptosis.22

Male Germ Cells: A Target for Estrogens

In order to exert a biological effect, estrogens must bind to estrogen receptors (ER), which in turn modulates the transcription of estrogen-responsive target genes (genomic effect). Alternatively, estrogens can activate different signaling pathways by interacting with membrane-bound binding proteins (non-genomic effect). Two types of nuclear ERs are known to exist: ERα (ESR1) and ERβ (ESR2). The distribution of these ERs has been described in the male rat gonad, but there is still debate about their localization.8 Likewise, we have recently reported the presence of transcripts for both ERs in the adult rat testis14 and showed the presence of ERs proteins in purified germ cells.15,16 It has been speculated that round spermatids may be one of the target cell types of estrogens since high levels of ERβ have been demonstrated in these cells.17

Additional evidence came from observations of patients with either congenital estrogen deficiency or estrogen resistance, as well as from animal models such as mice deficient in aromatase (ArKO) or in ERs (ERKO). It should be noted that impaired sperm motility and germ cells arrest (at the spermatid stage) were observed in both aromatase deficient men and mice.18,19 Thus, these observations illustrate a putative link between the lack of estrogen activity and infertility in men, illustrating the important role played by estrogens on male reproductive function. Conversely, exposure of adult male rats to a high phytoestrogen diet disrupts spermatogenesis via an increase in germ cell apoptosis,20 and overexpression of aromatase in mice leads to male infertility in all or in 50% of animals when it takes place either in fetal life or at puberty, respectively.21 In short, there is a delicate balance between androgens and estrogens to maintain normal testicular physiology and reproductive function with aromatase being an important molecule in these cellular events (Fig. 1).

In several mammalian species the synthesis and the role of estrogens have been recently explored. In the black bear (Ursus americanus), the presence of aromatase has been reported at the beginning of testicular recrudescence in Sertoli cells and then in round and elongating spermatids during the mating season.23 In the Siberian hamster, estrogens are able to induce initiation of spermatogenesis, independently of FSH in the photo-regressed adult male.24 In roe deer (Capreolus capreolus), estrogens were shown to be implicated in sperm production and in spermatozoal maturation.25 In the wild male ground squirrel (Citellus dauricus Brandt), a positive immunoreactivity for aromatase has been detected in Leydig and Sertoli cells and all types of spermatogenic cells but only during the breeding season. These authors concluded that estrogens play an important role in spermatogenesis, testicular recrudescence and regression.26 Similar observations have been published in the bank vole by the group of Bilinska.27,28

Moreover, treatment of adult monkeys with an aromatase inhibitor have suggested the involvement of estrogens in spermatid differentiation.29,30 The study conducted by the group of Balasinor in rats suggested that the process of spermatid elongation from steps 8 to 19 is androgen-dependent, whereas differentiation of round spermatids from steps 1 to 6 is regulated by estrogen.31 The observation of abnormal acrosome development in the ArKO mouse suggests that acrosome biogenesis may be an estrogen-dependent process.22 This hypothesis is in agreement with the observation that a high level of aromatase is detected in the Golgi complex of the developing spermatid,9 as well as by the presence of ERs in rat spermatids.14,16 Estrogens can also induce rapid effects via non-genomic pathways, that is, they bind onto membrane receptors to activate different signaling pathways via the production of second messengers (e.g., Ca²⁺, cAMP, NO) and the activation of tyrosine kinase receptors (e.g., EGFR, IGF-1R) and protein/lipid kinases (e.g., PI3kinase, Akt, members of the MAP kinase family or Src kinases, PKA, PKC).32,33

Recent data have illustrated that the seven transmembrane receptor associated to a G protein (GPR30) is involved in the proliferation-inducing effects of estrogens in a GC1 cell line34 and that GRP30 is present in human and rodent testicular cells.8 GPR30 is notably involved in the induction of expression of apoptotic markers (e.g., BAX, FAS and FASL) by estradiol in rat pachytene spermatocytes and round spermatids.15,16 Studies conducted in rodents and primates have shown that spermatogenesis is under the control of estrogens at different levels. For example, in the immature bank vole, exposure to a low dose of 17β-estradiol was shown to accelerate the onset of spermatogenesis, which was blocked by the injection of the antiestrogen, ICI 182,780 (Fig. 1).28 Other studies have demonstrated an improvement in the recrudescence of spermatogenesis in estradiol-treated rodents.24,35 Neonatal administration of estrogen to rats induced an increase in the number of spermatogonia at day 16 post-coitus.36 Similarly, studies from Culty's group have shown that rat gonocyte multiplication is partly under estradiol control,37,38 except that endogenous estradiol can inhibit gonocyte development in mice.39 There is also evidence for the direct involvement of estrogens in preventing germ cell apoptosis as demonstrated by the effects of estradiol in human adult seminiferous tubule cultures.40

Conversely, exposure of adult male rats to a high phytoestrogen diet disrupts spermatogenesis by increasing germ cell apoptosis.20 A significant decrease of the numbers of round and elongated spermatids, but not spermatogonia and spermatocytes, in ArKO mice clearly demonstrates a role of estrogens in regulating survival factor(s) in spermatids.22

It is also well known that estrogens can trigger pachytene cells to undergo apoptosis before further differentiation.45 Thus, estrogens are important to maintain the proper Sertoli:germ cell ratio in the seminiferous epithelium. Additionally, estradiol also stimulates spermatozoal motility in golden hamsters41 and humans.42,43 In addition, Fraser and collaborators have reported a positive effect of estrogenic xenobiotic genestein on capacitation in human and mouse spermatozoa.44

In contrast to androgen receptors, which are localized mostly in somatic cells, estrogen receptors are found in most testicular cells, including germ cells. The ERα gene is highly expressed in the adult testis, in particular at stages VII to XIV. In addition, the steady-state mRNA levels of both ERs are higher in purified round spermatids than in pachytene spermatocytes, suggesting a putative role of estrogens in the haploid steps of spermatogenesis, namely spermiogenesis.14

As already demonstrated in ArKO mice22 and from data obtained from the ERαKO, estrogens do play an important role in germ cell development.8 Nevertheless, it has been claimed that GPR30 is not involved in estrogenic responses in reproductive organs in mice.46 However, these authors generated Gpr30 deficient mice and showed that mutant males and females are both fertile; it is noteworthy that data were not provided for any functional aspects of spermatogenesis. Conversely, in zebrafish the identification of a functionally-active membrane ER has been reported with a high homology when compared to mammalian GPER, which was expressed in spermatogonia, spermatocytes and Sertoli cells.47 Moreover, it has been reported that in immature rat Sertoli cells, besides ERs, a functional GPER was expressed and involved in anti-apoptotic control.48

Estrogens, Human Fertility and Future Perspectives

Idiopathic infertility is a worldwide health issue, and no efficient treatment is available to date. While the role of estrogens in spermatogenesis is still a matter of debate, but from the observations of reduced sperm number and motility in men genetically deficient in aromatase,18,19 which coupled with our data and those of literature as briefly discussed above, these findings suggest a role for aromatase/estrogens not only during the development and maintenance of spermatogenesis but also in spermatozoa maturation via spermiogenesis since not only aromatase transcript but also a biologically active protein namely Gpr30 as well as ERs have been demonstrated.49 In humans, the main source of estrogens is the Leydig cell.4,50 However, adult Sertoli cells are also able of synthesizing estradiol in vitro,51 and estrogens are found in ejaculated spermatozoa.52 Aromatase mRNA was also reported in immature germ cells and in ejaculated spermatozoa,53 and estrogen seems to be related to the quality (such as motility) of sperm cells.54,55

In short, it is increasingly clear that estrogens, ERs and aromatase are found and/or produced by immature germ cells including spermatocytes and spermatids, as well as ejaculated spermatozoa in humans and many other species. Recent studies have demonstrated their involvement in male gamete maturation, in particular spermiogenesis, and the final steps of sperm maturation including capacitation and/or acrosome reaction. More importantly, ERs are found throughout the spermatozoon and localized intensely within mitochondria, an organelle found in the mid-piece of spermatozoa that provides the energy source for flagellar movement in the female reproductive tract. In addition, aromatase and ER mRNA transcripts could be relevant markers for assessing infertility in men and may serve also as diagnostic tools for clinicians to evaluate the sperm quality.49 In short, this is an emerging concept that post-meiotic haploid spermatids play an unexpected role in regulating their development during spermiogenesis as well as spermatozoal function via their ability to produce estrogens locally in the seminiferous epithelium.

Figures and Tables

Figure 1 A schematic illustrating estrogen targeting in a testicular cell via genomic and non-genomic pathways involving estrogen receptors (ERs) and GRP30, respectively. G₁, agonist of GPR30; G15, antagonist of GPR30; PPT, agonist of ESR1; DPN, antagonist of ESR2; TF, transcription factors; ER, estrogen receptors; ICI, selective estrogen inhibitor (ICI 182780).

Acknowledgments

The authors are indebted to their students (J. Levallet, D. Silandre, C. Bois and F. Mohammadi-Dubois). In addition, the financial support of the French Ministry of Education and Research, INRA, to our laboratory for the past decade is gratefully acknowledged.