Follicle-stimulating hormone (FSH) is a glycoprotein, which transmits its signal via a 75-kDa FSH receptor, FSH-R (Simoni et al.,1997). FSH is a key hormone of spermatogenesis (Sofikitis et al.,2008) and its essential role has been demonstrated by the use of knock-out mouse models for the FSH β-subunit and the FSH-receptor (Sairam and Krishnamurthy,2001; Danilovich and Sairam,2005). Spermatogenesis is dependent upon the presence of the somatic Sertoli cells that produce essential elements required for germ cell proliferation and differentiation (Griswold,1998). Most of the effects of FSH are cyclically regulated such as receptor mRNA (Heckert and Griswold,2002), receptor binding (Hermann et al.,2008), and the second messenger pathway (Welsh and Ireland,1992). It is well recognized that the main target of the hormone is the supporting Sertoli cells (Heckert and Griswold,2002), since these cells exhibit membranous FSH receptors (Reichert and Dattatreyamurty,1989) and their proliferation rate is FSH-dependent (Orth,1984,1986), although other studies reported that germ cells, mostly spermatogonia, can also present FSH receptors (Orth and Christensen,1978; Bacetti et al.,1998).
In the understanding of the role of FSH, endocytosis of the complex FSH/FSH-R is a key event. To date, findings on FSH endocytosis are sparse and the accurate kinetics of FSH internalization from the plasma membrane to the cytoplasm has not been clearly characterized. By using immunohistochemistry and in situ hybridization, the presence of IR-FSH and FSH-R mRNA has been detected within Sertoli cells (Böckers et al.,1994). Using pulse chase of radioactive labelled 131I-FSH, another group demonstrated that the final destination of the hormone was the degradation (Shimizu and Kawashima,1989), even if a recent work using C-terminal deletion of the FSH-R reported that most of the complex FSH/FSH-R could be recycled back to the plasma membrane (Krishnamurty et al.,2003). However, how long the final degradation process of the hormone took was not determined. By using colloidal gold particles coupled with transferrin, we previously examined the dynamic of this protein from the plasma membrane toward the endocytic pathway in both Sertoli and germ cells (Segretain et al.,1992; Segretain,1993). From these studies, it was obvious that the gold particles associated with the protein are first located on the plasma membrane before being in vesicles to gain endosomes afterwards. Thus, this methodological approach appears useful for further analyzing the dynamic of FSH internalization within Sertoli and germ cells.
By high-resolution immunofluorescence microscopy and time-lapse ultrastructural analysis of gold particles coupled to FSH experiments, associated with analysis at the ultrastructural level, we have precisely examined the fine details of the endocytic route of FSH in rat Sertoli and germ cells according to testis development. The present results demonstrate that the complex FSH/FSH-R, revealed by gold particles, accumulated for a consequent period of time in the endosomal compartment before final lysosomal degradation. Age differences in the time course of FSH routing were also evident in both cell types.
The presence of the FSH-R was first analyzed by immunofluorescence with a specific antibody directed against FSH-R in Sertoli and germ cells in immature (8 days) and mature (90 days) rats. FSH-R appeared as small dots in the basal region of the seminiferous epithelium in vimentin-positive cells identified as Sertoli cells as also in vimentin-negative basal and c-kit-positive germ cells, identified as spermatogonia (Fig. 1A). To validate the use of isolated cells in the following experiments, the presence of FSH-R was also examined in isolated Sertoli and germ cell fractions. High-resolution deconvolution microscopy revealed that the immunosignal for FSH-R was present on plasma membrane as well as inside the cytoplasm of Sertoli cells (Fig. 1B). Similar examination, in c-kit-positive cells, identified as spermatogonia, demonstrated that the FSH-R was localized mostly in the perinuclear region due to the small amount of cytoplasm (Fig. 1B). Altogether these observations demonstrate that FSH-R is present in Sertoli cells and in basally located germ cells within seminiferous tubules. Appropriate controls reveal no FSH-R labelling in vimentin and c-Kit-positive cells.
All the following investigations were realized with FSH coupled with gold particles in isolated seminiferous tubules and in vitro in isolated cells from adult seminiferous tubules and from prepubertal rats to follow the cellular pathway of the hormone (Figs. 1C, 2). The use of gold particles of various sizes coupled to FSH, reveals that 15-nm particles were more easily detectable in both cell types, but were less numerous in cytoplasmic membranous structures, whereas by using 5-nm gold particles, more grains were accumulated into the endo-lysosomal pathway of the cells examinated (data not shown). In order to better visualize gold-associated proteins, only experiments with large particles were presented here.
When seminiferous tubules were cultured in the presence of FSH-gold, gold particles were noticed at the base of the tubules and in the interstitial spaces between basal cells populating the tubules (Fig. 1C). In isolated cells, FSH-gold particles were detected within a few minutes on plasma membranes of Sertoli cells and spermatogonia, which are identified by their nuclear shape and by the cytoplasmic organelle organization: presence of rare endoplasmic reticulum cisternae with a globular shape in spermatogonia as compared to numerous elongated cisternae in Sertoli cells (Fig. 2). Quantitative analysis of 100 labelled Sertoli or germ cells performed in 4 different animals for each developmental age reveals that at 8 days, Sertoli cells and germ cells, mostly spermatogonia at this age, represented, respectively, 56.1 ± 3.7% and 58.5 ± 10.6% of the total cell count. Quantification of cells that were labelled with FSH-gold allowed us to determine the ratio of cells expressing FSH-R according to the cell type and the age of animals. Indeed, although 56.1 ± 3.7% of Sertoli cells and 58.5 ± 10.6% of total germ cell were labelled by FSH-gold in 8-day-old rats, more cells expressed FSH-R in the adult testis (70.7 ± 14.0% for Sertoli cells and 68.2 ± 9.4% for spermatogonia). Moreover, only 28.1 ± 4.4% of the total early spermatocytes cells, identified by their larger size and their well-organized cell organelles, expressed FSH-R (Fig. 2). These observations clearly demonstrate that most Sertoli and germ cells were labelled and that the FSH-R was differently expressed between immature and mature cells. The number of gold particles by cells was also, respectively, evaluated in Sertoli and germ cells of immature and mature testes. Examination of the total χ2 value for Sertoli cells (χ2 = 0.03) and for germ cells (χ2 = 0.04) between 8- and 90-day-old rats indicates that the observed gold-labelling distributions in the developmental groups were not significantly different. No statistically significant difference was observed between Sertoli cells and germ cells FSH-gold labelling (χ2 = 0.15).
As shown in Figure 3A, numerous FSH-gold particles were present on Sertoli cell plasma membranes 5 min after chase in both ex vivo tissue culture system and in vitro conditions. Gold particles accumulated also in membranous invaginations or pits corresponding to the beginning of endocytosis (Fig. 3A, inset). Careful examination reveals a clathrin-like coat on the cytoplasmic face of the membranous pit charged with FSH. Later, from 10 to 30 min after FSH exposure, gold particles were observed inside early endosomal structures (Fig. 3A, bottom left panel). Finally, after 60 and 120 min, numerous FSH-gold particles accumulated within dense multivesicular bodies and in membranous structures, containing membranous residues or with a dark homogenous content identified as lysosomes (Fig. 3A, bottom right panel). FSH-gold particles were undetectable in controls performed 5 min after chase in the presence of an excess of native FSH (Fig. 2, insets; Fig. 3A, Ctl1) and gold particles alone were unable to bind to plasma membrane (Fig. 3A, Ctl2). At 120 min, the number of FSH-gold particles was 9.6 ± 4.2 (n = 7) in early endosomes, undetectable in the recycling compartment (n = 6), 9.8 ± 3.5 in late endosomes (n = 4), and 189 ± 56 (n = 4) in lysosomes. The kinetic of apparition of the coupled hormone from plasma membrane within early endosomes and finally within lysosomes is analyzed in Figure 3B. If the general aspect of the kinetic appeared quite similar in 8- and 90-day-old Sertoli cells, the analyses of the time curves for membranous and endosomal fractions between young and adult rats revealed the existence of a time-dependent difference in the kinetics of the early events. Indeed, FSH-gold particles' internalization occurred more rapidly in the neonatal Sertoli cells than in mature Sertoli cells as evidenced by the loss of FSH-gold particles at the plasma membrane within 30 min and their concomitant redistribution in early endosomes. This observation was confirmed by analysis of the relative labelling index (RLI) that demonstrates a preferential localization of FSH-gold particles in plasma membrane of immature Sertoli cells at 5 min (RLI = 2.27, P < 0.01) and until 10 min (RLI = 1.88, P < 0.01) and in adult Sertoli cells at 5 min (RLI = 2.21, P < 0.01), 10 min (RLI = 1.51, P < 0.05), and until 30 min (RLI = 1.46, P < 0.05).
In spermatogonia, identified by their morphological criteria (large round nucleus and few cell organelles in a roundish small cytoplasm), FSH-gold particles were present on plasma membranes 10 min after FSH exposure (Fig. 4A). Some particles were noticed within typical plasma membrane invaginations (Fig. 4A, inset). At 30 min, FSH was found within endosomal clear structures (Fig. 4A bottom left panel). At longer time intervals (60 and 120 min), the FSH-gold particles accumulated within multivesicular elements representing future lysosomes (Fig. 4A, bottom right panel). Quantitative study of gold particles on spermatogonia reveals a classical pathway of internalization of FSH starting from plasma membrane and moving from invaginations or pits to form vesicles. Later, the gold labelling was sequentially detected in endosomes, multivesicular bodies, and lysosomes. If in 8- or 90-day-old rats, the complete kinetic of FSH endocytosis appeared quite similar to that observed for Sertoli cells, the time-course of the first internalization process was, however, reduced in 90-day-old rats (compare 8 days to 90 days in Fig. 4B). Statistical analysis demonstrates that FSH-gold particles preferentially labelled the plasma membrane of germ cells of 8-day-old rats at 5 min (RLI = 2.27, P < 0.05) until 10 min (1.70, P < 0.05), whereas this compartmentalization occurred after 30 min in the mature cells (at 5 min, RLI = 2.27, P < 0.01; at 10 min, RLI = 2.17, P < 0.01; at 30 min RLI = 1.13, P < 0.05). Compartmentalization of FSH-gold particles in function of age and of cell types was subsequently analyzed. In Sertoli cells, no significant difference was detected between 8- and 90-day-old rats (for degrees of freedom, χ2 = 4.95). In contrast, a significant difference between compartmentalization of FSH-gold and age of rat was evidenced significant for germ cells (for 4 degrees of freedom, χ2 = 37.98; P < 0.01). In 8 day-old germ cells FSH-gold labelled preferentially early endosomes (for 2 degrees of freedom, χ2 = 8.06; P < 0.05), whereas the plasma membrane was preferentially labelled in 90-day-old rat germ cells (for 2 degrees of freedom, χ2 = 8.98; P < 0.05).
In order to complete the study of the FSH pathway, analysis of FSH/FSH-R endocytosis was studied on early spermatocytes exposed to FSH-gold. Ex vivo tissue culture system and in vitro observations demonstrated that spermatocyte plasma membranes were rapidly decorated by gold particles. At longer time intervals, the gold labelling was detected within endosomal structures as also in pale multivesicular bodies. At 1 and 2 hr post-FSH exposure, the gold particles accumulated into dense multivesicular bodies and in lysosomes (data not shown).
Since recent data reported that the early endosomal compartment could be involved in numerous signalling pathways (Sadowski et al.,2009; Gould and Lippincott-Schwartz,2009), we precisely analyzed in the present study the localization of the FSH/FSH-R complexes into this endosomal structure. Interestingly, the current findings clearly evidenced that FSH coupled to gold particles was detectable in the early endosomes from 10 to 60 min (Fig. 5A–C) and present at endosomal membrane invaginations until at least 30 min. Moreover, as expected, the FSH-R colocalized within this structure with Rab5, the small GTPase protein regulating the endosomal organization and function (Fig. 5D). Similar data were obtained in germ cells (data not shown).
The careful examination of particles at the end of the FSH-gold internalization time-course reveals that gold particles accumulated in mature Sertoli cells within two types of lysosomes, identified by microscopic criteria (i.e., presence of dense bodies frequently containing membranous residues), which differed, however, by their structural aspect as determined by stereoscopic examination. Although these lysosomes were morphologically different, they exhibited similarities in both their electron-dense aspect and the inside quantities of FSH-gold particles. Analysis of 300-EM images revealed that round lysosomes formed about 80% of the total population of lysosomes. Three-dimensional stereoscopic examinations allowed us to visualize the presence of clusters of dense bodies, of lysosomal appearance, in the basal region of Sertoli cells. Around the centrally located dense multivesicular body, numerous lysosomes took place, several with a round shape (arrow in Fig. 6A) intermingled with elongated ones (open arrow in Fig. 6A). Two to three hours after FSH-gold particle exposure, gold particles were present in round lysosomes containing many grains (Fig. 6B), and more dispersed in more elongated lysosomes (Fig. 6C). Immunolabelling with anti-lamp-2 antibody and deconvolution microscopy examination confirmed the lysosomal nature of these elongated structures (Fig 6D–F).
By developing pulse-chase experiments of FSH coupled with gold particles, associated with fine ultrastructural quantitative analysis, we have analyzed the dynamic of FSH endocytosis in Sertoli and germ cells during development in an ex vivo tissue culture system and in vitro. The present study confirms the presence of FSH-R in both Sertoli and germ cells, as previously reported (Bacetti et al.,1998) and demonstrates for the first time that FSH accumulated within membranous cytoplasmic vesicles of an endosomal nature, during a consequent time before its final lysosomal accumulation in Sertoli cells, spermatogonia, and early spermatocytes. Our data also evidence differences in the celerity of the FSH endocytic process between neonatal and adult testicular cells.
If the structure of the FSH receptors and their role in spermatogenesis are well known (reviewed in Huhtaniemi,2000; Meduri et al.,2008), its cellular and intracellular distribution is questionable. Previous data, using antibodies directed against the receptor peptide complex, described its presence at the plasma membrane of Sertoli cells (Dattatreyamurty and Reichert,1993; Beau et al.,1998). In the present study, by developing deconvolution microscopy, we have confirmed the location of FSH receptors on Sertoli cell plasma membranes but also demonstrated their presence within the cytoplasms of these cells and in germ cells such as spermatogonia and spermatocytes.
By coupling gold particles to proteins of interest such as transferrin (Segretain et al.,1992), hCG (Ghinea et al.,1994), TSH (Baratti-Elbaz et al.,1999), hSBP (Gérard et al.,1991), and HRP (Morales and Hermo,1983), the biologically active pathways of these proteins were previously characterized in the testis. In the current study, we have successfully applied this methodological approach to FSH. Our data clearly demonstrate that FSH endocytosis followed an endo-lysosomal pathway in the ex vivo seminiferous tubules and in isolated Sertoli and germ cells of young (8-day-old post-partum) and adult rats. The present findings show that FSH was firstly present on Sertoli cell, spermatogonia, and spermatocyte plasma membranes, and, secondly, internalized by coated pits associated with clathrin, as described for hCG in Leydig cells (Ghinea et al.,1994) and for FSH in testicular endothelial cells (Vu Hai et al.,2004). In a previous study, FSH was detected in vesicular cytoplasmic components but the nature of these structures was not clearly identified (Bacetti et al.,1998). The present findings reveal that these structures could be associated with early endosomes as suggested by morphological criteria (pale content with small associated membranous tails). This possibility is also supported by the current dual immunolocalization experiments showing that the FSH-R colocalized with Rab5, a specific marker of early endosomes, and is in agreement with our earlier data in germ cells demonstrating that cationic ferritin, another membranous marker, is internalized into endosomal structures of a similar nature (Segretain,1989). Since it has recently been reported that signalling from internalized receptors persists in an endosomal compartment (reviewed in Sadowski et al.,2009; Gould and Lippincott-Schwartz,2009), it is quite possible that the presence of FSH/FSH-R within signalling endosomes ensures the temporal prolongation of signal transduction after FSH internalization and/or active signalling molecules to specific locations within the cell. In agreement with such a hypothesis is the observation that the complex FSH/FSH-R was located until 30 min at the endosome membrane or in endosomal membrane invaginations, and maintained for at least 60 min within this cellular compartment as determined by quantitative analyses. From these data, it could be suggested that the FSH/FSH-R complex still could be efficient during the first steps of the internalization process, whereas later, the acidic compartment of the endosome slowly leads to a dissociation of the complex in order to recycle the receptor and to degrade FSH. The possibility that these processes differ according to the age of the animals (immature vs. mature) and to the cell types considered (Sertoli or germ cells) could be hypothesized.
The present work also demonstrates that FSH-gold particles mainly accumulated within lysosome-like structures in Sertoli cell, spermatogonia, and early spermatocytes suggesting that Sertoli and germ cells used the same cellular compartments for FSH internalization. However, the present findings and the lack of detectable FSH-gold particles within the recycling endosomal compartment differ from the results of Krishnamurthy et al. (2003) who demonstrated that most of the FSH/FSH-R complex could be recycled to the plasma membrane. The reasons for this discrepancy are presently unknown.
Differences in the experimental models, transitory transfected cell lines that do not express FSH-R used in the latter study versus whole seminiferous tubules that express FSH-R in the present study, could explain these differences. It could be also due to methodological procedures: thickness of the sections and the size of the gold particles used. Indeed, the recycling endosomes are structured by a thin network of small anastomosed tubules (Segretain,1993), which are not easy to label. The development of new morphological ultrastructural techniques, such as tomography, could allow us to answer this question.
Quantitative analysis of our data reveals for the first time that the kinetic of the FSH-gold endocytic process in Sertoli cells was different between 8- and 90-day-old rats. If the route of the FSH/FSH-R complex was easy to determine by means of immunofluorescence and ultrastructural examinations, only the quantification of FSH-gold particles within well-identified structures at the ultrastructural level in pulse-chase experiments showed the difference in the dynamic of FSH endocytosis. Indeed, if the nature of cell organelles involved in endocytosis in both cell types was similar in young and adult rats, the analysis of the plots demonstrates that the time-course period, during which particles were present over the plasma membranes, was longer in the mature animals. It is also noteworthy that the time-course for gold particle accumulation within the early endosomal fractions was also longer in the older animals. The reasons for such differences are presently unknown and need further investigations. It is well recognized that although Sertoli cells respond to FSH stimulation, the nature of the response itself varies depending on the developmental status of the animal (Griswold,1993). In addition, recent studies in FSH-R knockout animals revealed that FSH-R-mediated actions are mainly required for early post-natal testicular development (reviewed in Sairam and Krishnamurthy,2001). Thus, it is possible that the rapid internalization of the FSH/FSH-R complex and the rapid turnover of the FSH-R in 8-day immature Sertoli cells magnify the effect of FSH on Sertoli cell proliferation and differentiation. It may also be hypothesized that FSH endocytosis was differently controlled during neonatal and adult development. In support of this hypothesis are the recent findings that testosterone and TGF-β2, which are differently expressed during rat testis development, control the kinetics of internalization of the blood testis barrier proteins in Sertoli cells (Yan et al.,2008). Lastly, this situation could rather result from a general cell mechanism between immature and mature cells. Such a possibility is in agreement with a previous study that described organelle morphological variations during testicular development in rat (Levy et al.,1999). Indeed, immature undifferentiated Sertoli cells exhibit a reduced number of cell organelles as compared to mature differentiated cells (Gondos and Berndston,1993) including endosomes (unpublished observations). Another explanation could be that the nature of the early endocytic compartment could be more elaborated in mature differentiated cells. In agreement with this hypothesis is the demonstration of two distinct populations of endosomes: a dynamic population that is highly mobile on microtubules and matures rapidly toward late endosomes and a static population that matures much more slowly (Lakadamyali et al.,2006).
The present findings also show that FSH coupled to gold particles was found during the final endocytic stages in Sertoli cell lysosomes. Previous data using two-dimensional electron microscopy analyzed the number of lysosomes in Sertoli cells but were unable to determine the fine three-dimensional structure of these organelles (Kerr,1988; Ueno and Mori,1990). By developing an ultrastructural stereoscopic approach and immunofluorescence analysis with an antibody directed against Lamp-2, a specific marker of lysosomes, we demonstrate here for the first time that mature Sertoli cells exhibited two distinct types of lysosomes characterized by their round or elongated shapes. It is noteworthy that the presence of such elongated lysosomes also has been observed in macrophages as common structures (Racoosin and Swanson,1993). Even if lysosomes have been associated for several years only with a degradation function, recent studies proposed a new role for these organelles in cell signalling in different physiological processes such as cell-to-cell communication and molecule recycling (reviewed in Lakkaraju and Rodriguez-Boulan,2008). Characterization of the protein content of lysosomes could have major implications for understanding the dynamics and the function of the lysosomes (reviewed in Lubke et al.,2009). These recent data conjugated with our current observation, suggest that these two types of lysosomes could be involved in two different functions that remain to be determined.
In conclusion, by using deconvolution microscopy and fine ultrastructural examination associated with gold particle quantification in pulse-chase experiments, we have detected FSH-R not only in Sertoli cells but also in spermatogonia and early spermatocytes of young and adult rats. If the role of FSH-R in Sertoli cells is well documented, their presence in germ cells remains to be elucidated. Our findings also demonstrated that the kinetic of FSH endocytosis in neonatal and mature rat Sertoli and germ cells follows the same endocytic pathway in both cell types, but that minute differences occur in the initiation of the dynamic endocytic process between immature and mature animals. The presence of the FSH/FSH-R within early endosomes for a consequent time period could be in favour of a potential function of this complex as a new signalling pathway during endocytosis. Our data also demonstrated the presence of two populations of lysosomes both containing FSH within mature Sertoli cells. Altogether, these studies could argue for a role of FSH during endocytosis of the complex FSH/FSH-R in Sertoli and germ cell control.
Male Sprague-Dawley rats of 8 days (n = 20) and 90 days (n = 10) were obtained from Laboratoire de Transgenèse (CEA Saclay). Animals were housed with 12L:12D cycles at 25°C and given water and food ad libitum. Rats were killed by CO2 asphyxiation and testes were collected for further experiments. All procedures on animals were conducted according to the standard ethical guidelines approved by the animal care committee of the Institut National de la Santé et de la Recherche Médicale (INSERM).
Hormone and Chemicals
Recombinant FSH (Merck Serono International S.A., Geneva, Switzerland) was used as a source of FSH. Tertachloroauric acid, trisodium citrate, and polyethylene glycol were purchased from Merck (Darmstadt, FRG). Antibody against FSH receptor was from Affinity Bioreagents (Ozyme, St-Quentin en Yvelines, France). Mouse anti-vimentin antibody was from Dakocytomation (Trappes, France), mouse anti-c-kit (sc-13508) from Santa Cruz Biotechnology Inc. (Santa Cruz, CA), and anti-Lamp-2 antibody (Cd107b/lamp-2) from Pharmigen (Beckton Dickinson, San Diego, CA). Goat anti-rabbit FITC conjugated and goat anti-mouse TRITC-conjugated antibodies were from Jackson Immunoresearch (Interchim, Montluçon, France).
FSH-Gold Particle Preparation
Tetrachloroauric acid (HAuCl4, 4 H2O from Merck) was used to prepare colloidal gold particles of 5- and 15-nm in diameter prepared according to Friends method and mixed with FSH as previously described (Segretain et al.,1992).
Tissue and Cell Preparations
Isolated seminiferous tubules without basal lamina were obtained by microdissecting tubules from testis (Rannikko et al.,1996) and incubated with collagenase in PBS (0.2 M, pH 7.4) for 10 min as previously described (Segretain,1989). Tubules maintained at 34°C were incubated with FSH-gold (initial solution 4 μg/ml, diluted at 1/100 before use). Controls were performed in the presence of an excess of native FSH (40 μg/ml) added before or simultaneously with the gold-labelled hormone. The binding of gold particles alone was also tested to verify that it is the FSH and not the gold particles that are selected and taken up by the cells. The pulse-chase method was used in the present experiments. Briefly, 5-min pulse incubation with FSH-gold particles was followed by chase with unlabelled FSH. At 5, 10, 30, 60, and 120 min post-incubation with unlabelled FSH to allow the observation of FSH endocytosis, isolated seminiferous tubules were fixed with glutaraldehyde (2.5%) in PBS (0.1 M, pH 7.4), washed overnight in PBS buffer, and post-fixed in ferrocyanide reduced osmium (Karnovsky,1971). Tissues were then dehydrated in alcohol and embedded in Epon 812. Thin sections were directly examined in the electron microscope (Philips CM10, from UNIC, Institut Alfred Fessard, CNRS, Gif sur Yvette, France). Isolated Sertoli and germ-cell fractions from 8- and 90-day-old rats were prepared according to Segretain et al. (1992). FSH-gold suspension was deposited on cells for time intervals similar as for isolated seminiferous tubules. Cells were fixed by glutaraldehyde (2.5%) in PBS and processed for electron microscopy as previously described (Segretain et al.,1992). Briefly, after glutaraldehyde fixation, a 1-hr post-fixation with ferrocyanide-reduced osmium was realized before epon embedding as for seminiferous tubules. Stereoscopic ultrastructural examinations were performed with the goniometric stage of the Philips CM10 electron microscope, tilting the specimen from −7° to +7° from 0° position, as previously documented (Segretain et al.,1981).
For analyzing FSH endocytosis, morphological criteria have been used to determine early endosomal compartments (membranous structures with a clear content containing few vesicles and showing remnant small tubules attached and with outside numerous clear vesicles around), pale and dense multivesicular bodies corresponding to the early steps of lysosomal pathway, and lysosomes characterized by dense bodies frequently containing membranous residues (Morales and Clermont,1986; Saraste and Goud,2007).
Immunofluorescence and Deconvolution Microscopy
Cross-section of testis, seminiferous tubule fragments, and isolated Sertoli and germ cells from 8- and 90-day-old rats were fixed in −20°C methanol for 6 min and permeabilized for 45 min with saponin. Cells and tissues were then incubated for 2 hr with rabbit anti FSH-R (1/100), mouse anti-vimentin (1/100), mouse anti-c-kit (1/200), anti-Rab5 (1/100, gift from M. Zerial, Max Planck Institute, Dresden, Germany), or anti-Lamp-2 (1/200) antibodies. Secondary antibodies marked with FITC or TRITC were added for 1 hr at room temperature as previously described (Decrouy et al.2004). Cells and tissues were mounted in medium with DAPI (Biovalley) to label nuclei. In controls, primary antibody was omitted. Three-dimensional high-resolution microscopy analysis was performed with a wide-field deconvolution microscope Nikon TE2000E (SCM, Université Paris Descartes) equipped with phase-contrast illumination mode (DIC Nikon, Tokyo, Japan), fluorescent-specific excitation and emission filter (Chroma Technology Corp., Bellows Falls, VT) in fast prior filter-well, and forced-air-cooled CCD camera (Coolsnap HQ2, Roper Scientific SAS, Evry, France). Acquired image stacks were deconvoluted using Autoquant image package algorithms.
Gold-particle distribution in Sertoli and germ cells, according to the developmental ages and the cellular compartments, have been compared by chi-squared (χ2) analysis to test if the observed distribution differs significantly from random and, if it is non-random, to identify compartments that are preferentially labelled as previously described (Mayhew et al.,2002). Statistically, analysis of FSH-gold particle compartmentalization during the time-course of endocytosis has been performed by calculating the Relative Labelling Index (RLI) using the line intersection method (Mayhew et al., 2008). Briefly, images randomly selected were acquired. Then, by using a plug-in of the image J software, horizontal lines were superimposed on the images and the number of gold particles was quantified on plasma membrane and into endosomal and lysosomal compartments, the only structures on which gold particles were detected. At the same time, the number of intersections between the superimposed lines and the membranes of these compartments was quantified. Between-group comparison was performed according to the method of Mayhew et al. after comparing observed and expected distributions of gold particles by means of a Chi-squared analysis (Mayhew et al.,2002,2009).
Jérome Gilleron was funded by the French Ministry of Research and Technology and Diane Carette was a fellowship recipient of the Agence Nationale de la Recherche (ANR, no. A05066AS) and ARC. This work was supported by the Institut National de la Santé et de la Recherche Médicale (INSERM) and funded by the Conseil régional d'Ile de France (Sesame grant E.1774). The authors thank Françoise Carpentier for technical assistance, Sylvain Theissier for animal care, and Laure Gilleron for reading the manuscript.