Metformin exposure affects human and mouse fetal testicular cells.

BACKGROUND: Metformin is a drug used in the treatment of diabetes and of some disorders related to insulin resistance, such as polycystic ovary syndrome. Gestational diabetes can cause complications for both mother and child, and some studies have shown a beneficial effect of metformin during pregnancy without an increase in perinatal complications. However, the effects on the gonads have not been properly studied. Here we investigated the effect of metformin administered during pregnancy on the development and function of the fetal testis. METHODS: A dual approach in vitro and in vivo using human and mouse models was chosen. Cultures of human and murine organotypic testes were made and in vivo embryonic testes were analysed after oral administration of metformin to pregnant mice. RESULTS: In human and mouse organotypic cultures in vitro, metformin decreased testosterone secretion and mRNA expression of the main factors involved in steroid production. In vitro, the lowest observed effect concentration (LOEC) on testosterone secretion was 50 µM in human, whereas it was 500 µM in mouse testis. Lactate secretion was increased in both human and mouse organotypic cultures with the same LOEC at 500 µM as observed in other cell culture models after metformin stimulation. In vivo administration of metformin to pregnant mice reduced the testicular size of the fetal and neonatal testes exposed to metformin during intrauterine life. Although the number of germ cells was not affected by the metformin treatment, the number of Sertoli cells, the nurse cells of germ cells, was slightly yet significantly reduced in both periods (fetal period: P = 0.007; neonatal period: P = 0.03). The Leydig cell population, which produces androgens, and the testosterone content were diminished only in the fetal period at 16 days post-coitum. CONCLUSIONS: This study showed a potentially harmful effect of metformin treatment on the development of the fetal testis and should encourage future human epidemiological studies.

Effect of mono-(2-ethylhexyl) phthalate on human and mouse fetal testis: In vitro and in vivo approaches.

The present study was conducted to determine whether exposure to the mono-(2-ethylhexyl) phthalate (MEHP) represents a genuine threat to male human reproductive function. To this aim, we investigated the effects on human male fetal germ cells of a 10⁻5 M exposure. This dose is slightly above the mean concentrations found in human fetal cord blood samples by biomonitoring studies. The in vitro experimental approach was further validated for phthalate toxicity assessment by comparing the effects of in vitro and in vivo exposure in mouse testes. Human fetal testes were recovered during the first trimester (7-12 weeks) of gestation and cultured in the presence or not of 10⁻5 M MEHP for three days. Apoptosis was quantified by measuring the percentage of Caspase-3 positive germ cells. The concentration of phthalate reaching the fetal gonads was determined by radioactivity measurements, after incubations with ¹4C-MEHP. A 10⁻5 M exposure significantly increased the rate of apoptosis in human male fetal germ cells. The intratesticular MEHP concentration measured corresponded to the concentration added in vitro to the culture medium. Furthermore, a comparable effect on germ cell apoptosis in mouse fetal testes was induced both in vitro and in vivo. This study suggests that this 10⁻5 M exposure is sufficient to induce changes to the in vivo development of the human fetal male germ cells.

Meiosis initiation in the human ovary requires intrinsic retinoic acid synthesis.

BACKGROUND: The initiation of meiosis is crucial to fertility. While extensive studies in rodents have enhanced our understanding of this process, studies in human fetal ovary are lacking. METHODS: We used RT-PCR and immunohistochemistry to investigate expression of meiotic factors in human fetal ovaries from 6 to 15 weeks post fertilization (wpf) and developed an organ culture model to study the initiation of human meiosis. RESULTS: We observed the first meiotic cells at 11 wpf, when STRA8, SPO11 and DMC1 are first expressed. In culture, meiosis initiation is observed in 10 and 11 wpf ovaries and meiosis is maintained by addition of fetal calf serum. Meiosis is stimulated, compared with control, by retinoic acid (RA) (P < 0.05). No major change occurred in mRNA for CYP26B1, the RA-degrading enzyme proposed to control the timing of meiosis in mice. We did, however, observe increased mRNA levels for ALDH1A1 in human ovary when meiosis began, and evidence for a requirement to synthesize RA and thus sustain meiosis. Indeed, ALDH inhibition by citral prevented the appearance of meiotic cells. Finally, 8 wpf ovaries (and earlier stages) were unable to initiate meiosis whatever the length of culture, even in the presence of RA and serum. However, when human germ cells from 8 wpf ovaries were placed in a mouse ovarian environment, some did initiate meiosis. CONCLUSIONS: Our data indicate that meiosis initiation in the human ovary relies partially on RA, but that the progression and regulation of this process appears to differ in many aspects from that described in mice.

[Development and regulations of testicular functions in the human foetus]. / Ontogenèse et régulations des fonctions testiculaires chez le foetus humain.

Two major functions are assumed by the testis: the production of male gametes (that is, spermatozoa) and the production of steroid hormones. Both two functions are established during fetal life and are essential to the adult fertility and the masculinization of the internal tract and genitalia. For many years, our laboratory has been interested in the ontogeny of those two functions in rodents and, since 2003, in collaboration with gynecology and obstetrics service of professor R. Frydman in Antoine-Béclère hospital, we have studied them in human. The first aim of this work was to improve the global knowledge of the human fetal testis development by using both our experimental data and the literature. Then, we focused on the different defects that can occur during the fetal testis development. Indeed, male reproductive abnormalities have been steadily increasing since the last decades and are thought to be related to the concomitant increase of the concentration of contaminants and particularly of endocrine disruptors in the environment. Thus, we decided to study the effect of endocrine disruptors on human fetal testis and, more particularly, the effect of phthalates, by using an organ culture system developed for human. In contrast to the data obtained in rat, mono (ethylhexyl)-phthalate (MEHP), an active metabolite of the most widespread phthalate in the environment, does not disturb the steroidogenic function. On the other hand, it has a negative effect on the male germ cells number. This study is the first experimental demonstration of a negative effect of phthalates directly on human fetal testis.

A new method for toxicity assays on human and mouse fetal testis.

Exposure to environmental pollutants (EP) is associated with a wide range of toxic effects, in particular in testis development. Uranium is a potential pollutant of nuclear industry and over the last few years, its environmental concentrations have increased. In animals, the current procedures for evaluating the potential developmental toxicity of uranium are based on in vivo studies. These methods do not allow to know the direct effects on testicular cells and are obviously excluded for human experiments. Consequently, we have developed an in vitro culture system of the whole testis. In the present study we characterized and validated this organ culture system in both mouse fetal testes and human fetal testes recovered during the first trimester (6-12 weeks) of gestation. We compared the histological aspect, the number of germ cells and the testosterone production, before and after culture. Testicular architecture and intercellular communications were preserved, and organ culture appears as a powerful method for studying the early development of testicular gametogenesis and steroidogenesis in both species. Thus by using this method we will be able to investigate the effects of uranium on mouse and human developing testis. The mouse model will allow us to determine the dose range of interest without restriction of material.

Retinoid-sensitive steps in steroidogenesis in fetal and neonatal rat testes: in vitro and in vivo studies.

Retinoic acid (RA) was recently shown to modify testosterone secretion of the fetal testis in vitro. We characterized this effect by culturing rat testes explanted at various ages, from Fetal Day 14.5 to Postnatal Day 3. In basal medium, RA inhibited, in a dose-dependent manner, both basal and acute LH-stimulated testosterone secretion by testes explanted on Fetal Days 14.5, 15.5, and 16.5. It had no effect on testes from older animals. The negative effect of RA did not result from a diminution in the number of Leydig cells but from a decrease in P450c17 mRNA levels and in LH-stimulated cAMP production. However, the RA-induced decrease in P450C17 mRNA levels was also observed with neonatal testes, suggesting that this enzymatic step is no longer rate limiting at this developmental stage. To study the physiological relevance of RA effects, we used fetuses and neonates issued from mothers fed a vitamin A-deficient (VAD) diet, resulting in a threefold decrease of plasma retinol concentration. On Fetal Day 18.5 and on Posnatal Day 3, testosterone secretion by the testis ex vivo was significantly increased in VAD animals. This shows that the endogenous retinol inhibits differentiation and/or function of fetal Leydig cells before Fetal Day 18.5 and is required for the normal regression of fetal Leydig cell function that occurs after Fetal Day 18.5. In conclusion, our results show that retinoids play a negative role on the steroidogenic activity during the differentiation of rat fetal Leydig cells.

Development of the foetal and neonatal testis.

The foetal testis originates from a proliferation of the mesonephric and the coelomic epithelia which are colonized by the primordial germ cells. In the foetal testis, the development and functions of the three main cell type precursors (Leydig, Sertoli and germ cells) do not depend upon gonadotropins. Numerous intra- and extra-testicular factors are candidates for the control of its development and functions. To study the potential involvement of these factors, we developed an organotypic culture system. In absence of any growth factors or hormone, this system allows a development of the three main cell types which mimics that observed in vivo. The effects of different regulators (gonadotropin-releasing hormone, follicle-stimulating hormone, transforming growth factor-beta, insulin-like growth factor-I, anti-Mullerian hormone, retinoic acid, oestrogens) were tested in this system. Whether or not some of the effects observed in vitro have a physiological relevance was evaluated using appropriate transgenic mice. It is concluded that the foetal testis cannot be considered as an adult mini-testis since it has a specific physiology which largely differs from that of the immature or adult testis.

Retinoid receptors involved in the effects of retinoic acid on rat testis development.

We have previously shown that retinoic acid (RA) is able to act on the development of Leydig, Sertoli, and germ cells in the testis in culture (Livera et al., Biol Reprod 2000; 62:1303-1314). To identify which receptors mediate these effects, we have now added selective agonists and antagonists of retinoic acid receptors (RARs) or retinoid X receptors (RXRs) in the same organotypic culture system. The RAR alpha agonist mimicked most of the effects of RA on the cultured fetal or neonatal testis, whereas the RAR beta, gamma, and pan RXR agonists did not. The RAR alpha agonist decreased the testosterone production, the number of gonocytes, and the cAMP response to FSH of fetal testis explanted at 14.5 days postconception (dpc). The RAR alpha agonist disorganized the cords of the 14.5-dpc cultured testis and increased the cord diameter in cultured 3-days-postpartum (dpp) testis in the same way as RA. All these RA effects could be reversed by an RAR alpha antagonist and were unchanged by an RAR beta/gamma antagonist. The RAR beta agonist, however, increased Sertoli cell proliferation in the 3-dpp testis in the same way as RA, and this effect was blocked by an RAR beta antagonist. The RAR gamma and the pan RXR agonists had no selective effect. These results suggest that all the effects of RA on development of the fetal and neonatal testis are mediated via RAR alpha, except for its effect on Sertoli cell proliferation, which involves RAR beta.

Effects of retinoids on the meiosis in the fetal rat ovary in culture.

We investigated the effect of retinoids on the entrance of female germ cells into meiotic prophase and their progression through it, using explants of rat ovaries from 14.5 days post coïtum (dpc) fetuses cultured with or without 10(-6) M retinoic acid (RA) or 10(-9) M retinoic acid receptor alpha (RARalpha) specific agonist. The percentages of oogonia and of oocytes at each meiotic stage in the ovary were evaluated at explantation (D0) and after 3 (D3), 5 (D5) and 9 (D9) days of culture and on equivalent stages in vivo (i. e. 17.5 and 23.5 dpc). The number of germ cells per ovary were counted at D0, D3 and D9. Newly explanted (D0) ovaries contained no germ cell in meiosis. In control medium some germ cells had spontaneously reached the stage leptotene and very few the zygotene on D3. The first pachytene were observed on D5 and the first diplotene on D9. This pattern mimicked that which occurs in vivo although with a slight delay. RA reduced the percentage of oogonia by more than half and increased the percentage of zygotene by more than 22-fold on D3, showing that it accelerated entrance into meiosis. This effect was also observed in response to RARalpha agonist. RA increased the percentage of zygotene and reduced the percentage of pachytene on D9, showing that it can also delay the zygotene/pachytene transition. Lastly, RA reduced the total number of germ cells present on D3 but not on D9. This may be the result of a double effect of RA on the number of germ cells: negative when the cells are in proliferation (D0 to D3) and positive when they entered in meiotic prophase (after D3). Thus, RA is a potential regulator of germ cells meiosis and number in the fetal ovary.

Multiple effects of retinoids on the development of Sertoli, germ, and Leydig cells of fetal and neonatal rat testis in culture.

We investigated the effect of retinoids on the development of Sertoli, germ, and Leydig cells using 3-day culture of testes from fetuses 14.5 and 18.5 days post-conception (dpc) and from neonates 3 days postpartum (dpp). Addition of 10(-6) M and 3.10(-8) M retinoic acid (RA) caused a dose-dependent disruption of the seminiferous cords in 14.5-day-old fetal testes, without any change in the 5-bromo-2'-deoxyuridine (BrdU) labeling index of the Sertoli cells. RA caused no disorganization of older testes, but it did cause hyperplasia of the Sertoli cells in 3-dpp testes. Fragmentation of the Sertoli cell DNA was not detected in control or RA-treated testes at any age studied. The cAMP produced in response to FSH was significantly decreased in RA-treated testes for all studied ages. Both 10(-6) M and 3.10(-8) M RA dramatically reduced the number of gonocytes per 14.5-dpc testis. This resulted from a high increase in apoptosis, which greatly exceeded the slight increase of mitosis. RA caused no change in the number of gonocytes in testes explanted on 18.5 dpc (the quiescent period), whereas it increased this number in testes explanted on 3 dpp (i.e., when gonocyte mitosis and apoptosis resume). Lastly, RA and retinol (RE) reduced both basal and acute LH-stimulated testosterone secretion by 14.5-dpc testis explants, without change in the number of 3beta-hydroxysteroid dehydrogenase-positive cells per testis. Retinoids had no effect on basal or LH-stimulated testosterone production by older testes. In conclusion, RE and RA are potential regulators of the development of the testis and act mainly negatively during fetal life and positively during the neonatal period on the parameters we have studied.

Expression and effect of insulin-like growth factor I on rat fetal Leydig cell function and differentiation.

Insulin like growth factor I (IGF-I) is believed to be a potent para/autocrine stimulator of Leydig cell function in adult testis. We investigated whether IGF-I is also an intratesticular regulator of fetal Leydig cell function by measuring its production in the fetal testis and its ability to affect testicular steroidogenesis during fetal development. Northern blot analysis revealed one major IGF-I transcript of 7-7.5 kb and two minor transcripts of 3.8 and 1.8 kb in 20.5 day fetal testis. IGF-I was detected by RIA in 16.5 fetal day testes, and the amounts of IGF-I secreted by 16.5 and 20.5 fetal day testes in vitro were much greater than the amounts contained in the testes, indicating active synthesis in culture. The secretion of IGF-I by the fetal testis in vitro was increased with testicular age and time in culture. It was not modified by gonadotropins or (Bu)2cAMP. Testosterone secretion by fetal testes explanted 13.5, 16.5, 18.5, and 20.5 days after conception and cultured in the presence or absence of 100 ng/ml LH for 3 days was not affected by the addition of 50 ng/ml IGF-I to the medium. In contrast, the addition of IGF-I to dispersed fetal testicular cells cultured for 3 days in the presence or absence of LH increased the number of Leydig cells identified by a positive cytochemical reaction for 3beta-hydroxysteroid dehydrogenase (3betaHSD). This was more pronounced with cells from 16.5- day-old fetuses (stage when the fetal Leydig cells are differentiating in vivo) than with 20.5-day-old fetuses cells (stage when the number and the function of fetal Leydig cells are stable or decreasing). It results from both an increased differentiation of mesenchymal cells in fetal Leydig cells and an increase in the mitotic index of the fetal Leydig cells, as inferred from the small increase in the percentage of bromodeoxyuridine/3betaHSD-positive cells. Both LH and IGF-I increased significantly testosterone production by day 16.5 cells. In the presence of LH, a high amount of testosterone was produced per 3betaHSD-positive cell; IGF-I further increased this production. This effect was not observed with day 20.5 cells. The amounts of testosterone produced per 3betaHSD-positive cell cultured in the presence of both LH and IGF-I were more than additive. Like IGF-I, insulin (50 ng/ml) increased testosterone secretion per 3betaHSD-positive cells in cultures of day 16.5 cells, but not in those of day 20.5, cells. Lastly, IGF-I also increased the steroidogenic activity of each Leydig cell in cultures containing (Bu)2cAMP, but its effects were weaker than those observed in the presence of LH. This suggests that IGF-I has sites of action both upstream and downstream cAMP generation. These results suggest that IGF-I acts as paracrine/autocrine factor in the differentiation and activity of fetal Leydig cells.

Effect of anti-Mullerian hormone on Sertoli and Leydig cell functions in fetal and immature rats.

Anti-Mullerian hormone (AMH) is mainly involved in the regression of Mullerian ducts in male fetuses, but it may have other functions linked to gonadal development. The present study examines the effect of AMH on steroidogenesis by Sertoli and Leydig cells in fetal and immature rats during the period where AMH is physiologically produced in the testis. The basal aromatase activity of Sertoli cells in primary culture was strongly stimulated (77-91%) by cAMP. AMH (35 nM) reduced cAMP-stimulated aromatase activity by 49-69% as early as fetal day 16 and until postnatal day 20. This effect was dose dependent and was seen after 48 h in culture. AMH also blocked the Sertoli cell aromatase activity stimulated by FSH, but LH did not stimulate this activity, confirming that the aromatase activity effectively resulted from Sertoli cells and not from contaminating Leydig cells. RT-PCR analysis showed that AMH reduced aromatase activity by decreasing the amount of aromatase messenger RNA. AMH also inhibited the LH-stimulated testosterone production by dispersed fetal Leydig cells in culture in a dose-dependent manner. The inhibitory effect of AMH did not depend on the fetal stage studied (16 or 20 days postconception) and resulted from a drop in the steroidogenic activity of each Leydig cell without affecting the number of 3beta-hydroxysteroid dehydrogenase-positive cells. These data provide the first evidence that AMH, like other members of the transforming growth factor-beta family, has an autocrine/paracrine effect on testicular steroidogenic function during the fetal and prepubertal periods.

Immunohistochemical localization of transforming growth factor-beta 1 in the fetal and neonatal rat testis.

The localization of transforming growth factor-beta 1 in the fetal and neonatal rat testis (from day 13.5 of fetal life to postnatal day 20) was investigated by an immunohistochemical staining method employing a polyclonal anti-TGF-beta 1 antibody that does not cross react with either TGF-beta 2 or TGF-beta 3. In testis and mesonephros tissue, immunostaining for TGF-beta 1 was undetectable on fetal day 13.5 and appeared exclusively in the primordial Sertoli cells on fetal day 14.5. Staining in Sertoli cells was still clearly observed on days 15.5 and 16.5 of fetal life and became faint from fetal day 18.5 onwards. In fetal Leydig cells, a positive reaction for TGF-beta 1 appeared on day 16.5 and became very intense during late fetal life. After birth, fetal-type Leydig cells, which were still observed on postnatal days 4 and 20, also exhibited a very strong immunostaining for TGF-beta 1, whereas adult-type Leydig cells, observed on day 20, showed a slight staining. No immunoreactivity for TGF-beta 1 was found in germ cells and peritubular cells on any day studied. In conclusion, TGF-beta 1 is present very early in the fetal rat testis and its prevailing localization shows age-related changes, which suggests that this factor plays an autocrine/paracrine role in the regulation of testicular function and differentiation, during early development.

Stimulatory effect of follicle-stimulating hormone on basal and luteinizing hormone-stimulated testosterone secretions by the fetal rat testis in vitro.

The in vitro effect of FSH on testosterone secretion by the fetal rat testis was studied. Testes were cultured in the presence or absence of either commercial human (h) FSH (Metrodine; 200 mIU/ml) or recombinant hFSH (200 mIU/ml) for 3 days and with 100 ng/ml ovine LH during the last 4 h of culture. To avoid a stimulatory effect by the 0.4% LH that contaminates Metrodine, the cultures were performed in the presence of a monoclonal anti-hLH beta antibody and with a concentration of Metrodine that had no short term stimulatory effect on testosterone production by the fetal testes in vitro. Metrodine treatment had a positive long term effect on both basal and LH-stimulated testosterone secretion by fetal testes explanted on days 18.5, 20.5, and 22.5 postconception, which was abolished by the addition of a monoclonal anti-hFSH beta antibody. LH-free recombinant FSH also augmented basal and LH-stimulated testosterone secretion of testes explanted on days 13.5, 14.5, and 18.5 postconception. The positive effect of recombinant hFSH appeared during the second day of treatment with day 14.5 and 18.5 testes and on the third day of treatment with day 13.5 testes. As it is widely accepted that FSH receptors are exclusively localized on Sertoli cells, these results suggest that on or before day 15.5 of fetal life, 1) Sertoli cells are able to respond to FSH, 2) Sertoli cells can produce factors that are able to act on Leydig cell function, and 3) Leydig cells are sensitive to FSH-induced Sertoli cell factors. In conclusion, this study points out a potential paracrine control of fetal Leydig cell function and/or differentiation by fetal Sertoli cells as soon as fetal Leydig cells differentiate.

Developmental changes in testosterone production by the rat testis in vitro during late fetal life.

The age-related evolution of the in vitro capacity of the rat testis to produce testosterone and to respond to luteinizing hormone (LH) during 3 and 24 h incubation was studied from day 18.5 to day 21.5 of fetal life. Basal testosterone production by testes from 18.5-day-old fetuses was significantly higher than production by testes from 20.5- and 21.5-day-old fetuses when secretion was expressed as either per testis or per microgram of testicular protein. When maximal LH-stimulated testosterone secretion was expressed on a per testis basis, it was significantly lower for day 18.5 testes than for day 20.5 and 21.5 testes. However, when it was expressed on the basis of testicular protein content, it decreased significantly between days 18.5 and 20.5. Basal and LH-stimulated secretions displayed the same time-related decrease throughout 24 h of incubation for the three ages studied. The dose-response curves for LH showed that the sensitivity was similar for day 18.5 and 20.5 testes (ED50 = 10 vs. 14 ng/mL, respectively). These results showed an age-related decrease in testicular steroidogenic capacity without a change in the coupling efficiency of LH receptor to testosterone production during late fetal life in rats.