SEARCH

SEARCH BY CITATION

Keywords:

  • tilapia;
  • spermatogenesis;
  • primary culture;
  • testis;
  • spermatogonia;
  • endocrine control;
  • flow cytometry

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Spermatogenesis in vertebrates is controlled by endocrine and paracrine factors and involves the communication between somatic and germ line cells. To elucidate some of the relevant factors in the complicated molecular control processes, we established an in vitro test system using primary cultures of tilapia (Oreochromis niloticus) testis cells. The cultures were enriched for germ line cells and Sertoli cells and largely depleted of spermatozoa. By staining the cells with propidium iodide and carboxyfluorescein succinimidyl ester (CFSE), different cell populations could be identified cytologically and, in addition, quantified by flow cytometry. Cells that had gone through one or more divisions could be identified unequivocally based on their CFSE staining intensity. In parallel cultures maintained for up to 16 days in the presence of 11-ketotestosterone (KT), insulin-like growth factor I (IGF), and/or human chorionic gonadotropin (hCG) the initiation of meiotic and mitotic divisions was monitored. Although KT was important for the initiation of meiosis, spermatogonial mitotic divisions between 10 days and 16 days of culture were promoted by IGF and/or hCG in the presence of KT. These results illustrate the potential of the established in vitro test system for the analysis of the molecular control mechanisms of spermatogenesis. Developmental Dynamics 233:1238–1247, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Spermatogenesis in vertebrates is an attractive model system to analyze developmental processes such as, for example, the first steps from primordial germ cells to spermatogonia (or oogonia), the regulation of meiosis and the postmeiotic steps of cytological differentiation (spermiogenesis). The control mechanisms of these and other relevant differentiation processes appear to involve the communication between the somatic cells and the germ line cells. The principal differentiation processes that characterize spermatogenesis appear to be similar in all vertebrate species (Pudney,1995), but differences in detail are apparent both concerning the cytological organization of the testis and the genetic control.

The structure of fish testes is comparatively simple and is characterized by spermatogenic cysts containing synchronously differentiating germ line cells associated with somatic Sertoli cells (Loir,1999; Miura et al.,2002). The fully differentiated testis of adult male tilapia contains spermatogenic cysts at all stages of spermatogenesis, including type A and type B spermatogonia, spermatocytes in different stages of meiosis, and spermatids in the final stages of spermiogenesis. Finally, spermatozoa differentiate that are capable of fertilizing ovulated eggs. Spermatozoa typically represent the largest fraction of any cell type. Similar to the mammalian testis, only the somatic Sertoli cells are directly associated with the germ line cells. However, in contrast to the mammalian situation, all germ line cells within one cyst pass through all stages of spermatogenesis in developmental synchrony, whereas the individual cysts differ in their stage of development and with respect to the phase in the meiotic or mitotic cell cycle. The steroid hormone-producing Leydig cells are located outside the cysts in the stroma as in the mammalian testis.

The study of spermatogenesis in fish has some experimental advantages compared with mammals. In particular, spermatogenesis can be studied conveniently in primary cell cultures in vitro, thus facilitating the analysis of the molecular control mechanisms. Using primary cultures, sperm production in vitro has been reported in different fish species, including eel (Miura et al.,1991), zebrafish (Sakai,2002), and medaka (Saiki et al.,1997). In the latter species, quantitative data on spermatogonial cell proliferation and spermatocyte differentiation were obtained (Song and Gutzeit,2003). The medaka is the only organisms in which the differentiation of spermatogonia to mature and fertile sperm in vitro has been demonstrated. This process may occur in the absence of Sertoli cells (Saiki et al.,1997) or even in the absence of any somatic cells, as elegantly shown recently by Hong et al. (2004). Originating from a primary culture, these authors produced a permanent spermatogonial cell culture that could be induced to differentiate mature spermatozoa. However, with such cultures, the molecular control of spermatogenesis, which clearly involves somatic cells, cannot be analyzed.

In this communication, we carried out the experiments in another species, tilapia (Oreochromis niloticus), which is economically an important fish in world aquaculture (Harvey et al.,2002). This species is in many ways ideally suited for studies using primary cultures, in particular for experiments that require more cells than can be provided by the small testes of other model teleosts, the medaka and the zebrafish. The tilapia is one of the best-characterized species for the analysis of gonad differentiation in teleosts (Nakamura et al.,1998), genetic data are available, and a genome project is under way (Kocher et al.,1998). Furthermore, a range of modern genetic techniques have already been applied in this species by aquaculture industries (Hulata,2001), and, using transgenic technology, growth-enhanced lines of tilapia have been produced (Maclean et al.,2002).

Primary cultures of gonad cells offer several experimental advantages. The cells of interest can be studied in vitro under the chosen experimental conditions, and in many cases, the cells remain viable and retain their cell type-specific properties (Loir,1999; Song and Gutzeit,2003). Primary cultures from different organs of fish can successfully be prepared and maintained for several weeks (Ma and Collodi,1999). The viability of the primary cells is high; furthermore, some essential cellular functions such as cell proliferation and spermatocyte differentiation are maintained in vitro (Saiki et al.,1997; Song and Gutzeit,2003).

It was our aim in this study to establish a test system with primary testicular cells (germ cells and somatic cells) and to analyze the effects of some selected hormones and growth factors on cell proliferation and differentiation. Because the starting culture contained several cell types, in particular spermatogonia, spermatocytes, and Sertoli cells, a high-resolution analytical tool was essential that could be used to distinguish the cell types and to monitor their response to the test compounds separately. The parallel analysis by confocal microscopy and by flow cytometry after labeling the cells with carboxyfluorescein succinimidyl (CFSE) and propidium iodide (PI) turned out to serve this purpose optimally. Using this technique, some aspects of cell proliferation using human leukemia cells (Tokalov and Gutzeit,2003) and of cell differentiation with medaka testis cells have been studied successfully before (Song and Gutzeit,2003).

Although numerous aspects of fish spermatogenesis have been addressed by many authors, a clear picture of the molecular control mechanisms has not emerged. Progress has been hampered by technical problems, the complexity of cellular interactions, and endocrine/paracrine control processes. In experiments with live animals, these variables are particularly difficult to control, and in vitro cell cultures often give only partial answers because of the difficulty to establish conditions that resemble in vivo conditions. These problems in mind, we have initiated this study and developed methods to analyze the cytological effects of known signalling molecules on proliferation and differentiation of the germ line cells.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Cytological Characterization of Primary Testis Cells

The primary cultures were first (day 0) analyzed by laser scanning microscopy (LSM) to characterize the different populations of germ line cells and somatic cells. For fluorescent labeling of the cells, fluorochromes were chosen that helped in the cytological identification of the different cell types and, at the same time, allowed analysis of the cells by flow cytometry (see below). The nuclei were labeled with PI (red fluorescence), and the cytoplasm appeared in green fluorescence after labeling with CFSE (Fig. 1A). The latter compound in the diacetate form is colorless and diffuses into the cells where it becomes fluorescent when acetate groups are removed by intercellular esterases. The amino-reactive compound forms stable dye–protein adducts. The double labeling allows separation of the respective cell types based on the size of the cells and the nuclei, the nuclear/cytoplasmic ratio, the presence (or activity) of intracellular esterases and the intracellular distribution of CFSE.

thumbnail image

Figure 1. A–H: Primary testis cell cultures of tilapia were characterized by confocal microscopy (A,H) and by flow cytometry (B–G). To distinguish the different cell types, the cells were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE, green fluorescence) and propidium iodide (PI, red fluorescence). The different cell sizes, the nuclear/cytoplasmic ratio, and the labeling pattern allow us to distinguish each cell type analyzed: spermatogonia A (SgA), spermatogonia B (SgB), spermatocytes (Sc), spermatids (St), spermatozoa (Sz), Sertoli cells (S), and erythrocytes (Er). Longer exposure times allowed the visualization of the CFSE-labeled flagella of spermatozoa (Sz*). B–E: The cell types were characterized by single-parameter (PI, B) or by two-parameter analysis (PI/CFSE, C; and PI/SSC, E). To correlate the cell populations defined by flow cytometry with the microscopic images, the CFSE fluorescence of the major cell types was quantified by area morphometry. D: Significant differences (P < 0.05) with respect to the fluorescence measured in Sertoli cells (set arbitrarily to 100) are indicated (asterisks). E: Sertoli cells were further characterized by their phagocytic activity using fluorescent beads (B) as markers. The two-parameter analysis (Fluorescence/FSC) shows that most Sertoli cells have taken up one fluorescent bead (S+B) and that only a few cells have taken up two beads (S+2xB). G: Germ line cells and Sertoli cells typically aggregated in the cultures as shown for a 2-day culture of 11-ketotestosterone-treated cells. The presence of Sertoli cells in such aggregates could be demonstrated when the cells were prepared for analysis by flow cytometry (disaggregation, fixation, PI staining). H: When viewed with the laser scanning microscope, the identity of the cells (Sertoli cells, spermatogonia, and spermatocytes) could be determined. These data were consistent with the data obtained by flow cytometry of the same preparations. FSC, reflecting cell size (forward scatter); SSC, reflecting number and size of granules in cytoplasm (side scatter).

Download figure to PowerPoint

Based on the different staining properties, spermatogonia type A (SgA, also referred to as early or primary spermatogonia) can easily be distinguished from type B spermatogonia (SgB) in the microscope (Fig. 1A). Type A spermatogonia are larger and possess more cytoplasm, prominent CFSE stained nucleoli, and a nucleus with a granular appearance in contrast to the type B spermatogonia. As a result, the green fluorescence prevails in primary spermatogonia when viewed in the fluorescence microscope. The quantitative analysis by flow cytometry shows, as expected, no difference in PI staining (reflecting the amount of DNA in each cell, Fig. 1B), but the CFSE signal is stronger in type A than in type B spermatogonia, which resulted in an elongated distribution of diploid cells (Fig. 1C). Because of intermediate stages in the population of primary cells, a separation of both spermatogonial cell types by gating was not possible.

Occasionally two (or more) connected cells were observed (Fig. 1A, 2xSgB), thus reflecting the mitotic activity of these cells and the incomplete cytokinesis that is typical for spermatogonial divisions (Loir,1999). Spermatocytes (Fig. 1A, Sc) are characterized by large nuclei. With each meiotic division, the cells and the nuclei become smaller and meiosis is completed with the formation of haploid spermatids (Fig. 1A, St). During the following process of spermiogenesis functional spermatozoa differentiate (Fig. 1A, Sz) and the CFSE label becomes focused mostly at a single position. Upon longer exposure, a weak fluorescent signal can be visualized along the flagellum (Fig. 1A, Sz*) and the localized CFSE signal can be detected at the “neck” region of spermatozoa. The yellow color is due to the superposition of the nuclear red signal and the cytoplasmic CFSE signal. The Sertoli cells are characterized by their large size, elongated contour, and well-defined and indented nuclear envelope. The cytoplasm contained dark granules of various sizes. Because in freshly prepared primary cell cultures erythrocytes are often present, it was important to distinguish these cells from the germ line cells. The ovoid shape and the low level of CFSE fluorescence characterizes these cells and allows their identification (Fig. 1A, Er).

The cells were then cultured at a density of 5 × 104 cells/cm2 in 25-ml flasks (10 ml, ∼5 × 105 cells/ml) in humidified atmosphere with 5% CO2 (vol/vol) at 25°C overnight. After this incubation, fibroblasts and interstitial cells adhered firmly to the bottom of the flasks, whereas spermatogonia and Sertoli cells did not (Miura et al.,1996). Hence, from the somatic cells types, only Sertoli cells and erythrocytes were generated in sufficient numbers to allow identification by flow cytometry.

Defining Cell Populations by Flow Cytometry and Quantitative Area Morphometry

Spermatogenesis is characterized by meiosis and the successive reduction of the DNA content ranging from 4C in spermatocytes at the G2 stage to 1C in spermatozoa. The quantification of the DNA content by PI staining and flow cytometry allows separation of the respective cell populations of germ cells. The distribution of these cells according to PI staining is shown in Figure 1B. Clusters of cells with 2× or 4× the diploid DNA content were apparent in the flow cytometric analysis (Fig. 1B, labeled 2xSgB, 4xSgB). The analysis by LSM revealed the identity of these populations: small numbers of spermatogonia Typ B (SgB) adhered tightly to each other and were presumably connected by intercellular bridges, a feature that is characteristic for these cells.

With respect to the somatic cells, erythrocytes with diploid nuclei and Sertoli cells were identified in the primary cultures. These cell types can be distinguished from the germ line cells by two-parameter analysis, including the CFSE content of the respective cells (Fig. 1C) or side scatter (SSC), a parameter varying according the number and size of granules in the cytoplasm (Fig. 1E). Sertoli cells (S) had the highest level of CFSE fluorescence (100 ± 10; P < 0.05) and could be defined as a distinct cell population (Fig. 1C). Compared with Sertoli cells, the level of CFSE fluorescence in germ line cells was low (Sg, 18 ± 3; Sc, 33 ± 4; St, 5 ± 3; Sz, 2 ± 1; Fig. 1C). Erythrocytes (Er) were characterized by a low level of CFSE fluorescence (3 ± 2, Fig. 1C) but relatively high level of SSC (Fig. 1E). This cell type was strongly reduced in the primary cultures, because the small pieces of testis prepared before collagenase treatment were thoroughly washed and erythrocytes were flushed out of the tissue.

The identity of the cell populations described above and characterized by flow cytometry was verified by LSM analysis. The typical cytological features of the respective cell types and the quantification of the CFSE content by area morphometry (Fig. 1D) allowed correlation of the cellular parameters obtained by flow cytometry and LSM and clear identification of the respective cell types. As expected, the measurements in the microscope showed that the Sertoli cells (S) had the highest level of CFSE fluorescence (set to 100 ± 15 per cell). Compared with Sertoli cells, the level of CFSE fluorescence in germ line cells was significantly (P < 0.05) lower (SgA, 20 ± 2; SgB, 10.4 ± 0.3; Sc, 13.6 ± 0.9; St, 3.6 ± 0.2).

Another independent test for Sertoli cells in the primary culture is based on their phagocytotic activity. Fluorescent beads were added to a primary culture, and the uptake of the beads by Sertoli cells was quantified after a 1-hr incubation using flow cytometry (Fig. 1F). Most Sertoli cells had taken up one bead during that time (Fig. 1F, S+B, 95 ± 3%), and only a few cells had taken up two beads (S+2xB, 5 ± 3%).

Enrichment of Spermatogonia, Spermatocytes, and Sertoli Cells

Because it was our aim to analyze spermatogonial proliferation in the presence of Sertoli cells, it was necessary to enrich these cell types from primary testis cell cultures and to remove the large number of spermatids and spermatozoa in mature testes. To this end, we fractionated the primary cells by a single-step Percoll gradient centrifugation. In Figure 2, three different samples are compared: the cells released from the testis during dissection and washing steps before collagenase digestion (Fig. 2A,D), primary cells before gradient centrifugation (Fig. 2B,E), and the cell population after the centrifugation (Fig. 2C,F). The cell populations of interest were defined in two-dimensional plots using FSC (cell size) and SSC (granular structure of the cytoplasm) as parameters (Fig. 2A–C) and FSC in combination with PI fluorescence (Fig. 2D–F). Distinct cell populations were gated and designated as I–III. The cells released from dissected testes were represented a homogeneous population of the small (FSC, 12 ± 2), haploid cells with low SSC (12 ± 3) as evident in two-dimensional plots (Fig. 2A,D, gated area I). In the microscope, the cells were easily identified as spermatozoa. The primary cells prepared by enzymatic separation of testis cells are composed of three major populations of cells in FSC/SSC plots (Fig. 2B), which were separately analyzed by gating (I–III, Fig. 2A): area I contains spermatozoa (as in Fig. 2A), area II is characterized by larger (FSC, 24 ± 5, P < 0.05) haploid spermatids (see Fig. 2E) with the same SSC (8 ± 3, P > 0.05), and the third population (area III) of the largest cells (FSC, 50 ± 10; P < 0.05) with the highest level of SSC (35 ± 7; P < 0.05) contains both 2C and 4C cells (Fig. 2B,E) and consists of spermatogonia, spermatocytes, and Sertoli cells. Because somatic cells do not proliferate in the mature testis (Steinberg, 1971; Orth,1982) and the portion of spermatogonia cells undergoing mitotic divisions is very small (Bellve and Feig,1984), the 4C cells represented primary spermatocytes at prophase I of the meiotic division. Of the total number of cells in the culture, 70 ± 10% were present in area I (Fig. 2B,E), 20 ± 6% in area II, and 10 ± 5% in area III.

thumbnail image

Figure 2. The enrichment of diploid cells from primary cultures was monitored by flow cytometry. A–F: Three distinct cell populations could be separated by plotting the parameters FSC/SSC (A–C) of unfixed cells and FSC/PI fluorescence (D–F) of fixed cells stained with PI. The cell populations labeled I–III contained spermatozoa (I) whose identity was verified in analyzing a nearly pure preparation of mature spermatozoa (A,D), haploid spermatids (II), and all diploid cells in different phases of the cell cycle (III, cells with 2C, 4C DNA content are marked). FSC, reflecting cell size (forward scatter); SSC, reflecting number and size of granules in cytoplasm (side scatter); PI, propidium iodide.

Download figure to PowerPoint

Of three different Percoll concentrations tested (12, 20, and 28%), the concentration of 20% gave the best results, because the cell population of interest in area III (Fig. 2C,F) increased to 55 ± 5% (∼5.5-fold enrichment, P < 0.05). Under these conditions, most haploid could be removed while the diploid cells of interest were mostly retained. The analysis of the enriched cell population by LSM showed that this cell population consisted of spermatogonia and Sertoli cells (2C), spermatocytes (4C), and spermatids (1C).

Factors Controlling Mitotic and Meiotic Divisions

The population of primary cells as defined above (Fig. 2C) were labeled with CFSE and split into five parallel cultures: a control culture with no additional test compounds and cultures containing KT, KT+IGF, KT+ hCG, or KT+IGF+ hCG. The starting population (Fig. 3A, day 0) consisted mostly of spermatogonia (predominantly type B at the G0+1 phase) and primary spermatocytes (4C DNA content). These cell populations defined in the boxed area were further analyzed for a period of 16 days (see below). At the beginning of the experiment, a small number of haploid spermatids but no spermatozoa was present in the culture. Also, a population of Sertoli cells with strong CFSE fluorescence was consistently identified by flow cytometry. Apparently, the Sertoli cells did not proliferate under our culture conditions, which is in keeping with the observation of other authors (Steinberg, 1971; Orth,1982). They consistently occupied the same position in the dot plot histograms during the course of the entire experiment, and no changes concerning their cell cycle distribution were observed (Fig. 3A, 0–16 days). When the different cultures were viewed in the microscope, it was apparent that, in all KT-containing cultures (but not in the controls), the cells formed aggregates of germ line cells and Sertoli cells (Fig. 1G). In these aggregates, Sertoli cells could not be distinguished from the numerous germ line cells. However, when the clusters were processed for flow cytometry and stained with PI, spermatogonia, spermatocytes, as well as a few Sertoli cells could consistently be identified. Such a preparation with separated cells was also analyzed by LSM, and the respective cell types could easily be identified (Fig. 1H).

thumbnail image

Figure 3. Changes in cell division in a mixed population of primary cells (as defined in Fig. 2C) were monitored over a period of 16 days. At the beginning of the experiment (day 0), cells were labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE) and split into five parallel cultures: a culture with no additional test compounds (control) and cultures containing 11-ketotestosterone (KT), KT+insulin-like growth factor I (KT+IGF), KT+human chorionic gonadotropin (KT+hCG) or KT+IGF+hCG. After 0, 2, 4, 8, 10, 12, and 16 days the cells were fixed, labeled with propidium iodide (PI) and analyzed by flow cytometry. A: Examples of the major findings are presented in two-dimensional plots PI/CFSE. Spermatogonia (Sg), spermatocytes (Sc), spermatids (St), and Sertoli cells (S) were labeled with CFSE (0 day, control). In the presence of KT, spermatocytes (Sc, DNA content 4C) differentiated into spermatids (St*) with reduced CFSE content. The fraction of apoptotic cells (Ap) were quantified (boxed area, 2 days, KT). In the presence of KT, the spermatogonia (Sg) re-entered the cell cycle but arrested at the G2+M phase of the cell cycle (10 days, KT). In the presence of IGF, the cell cycle arrest was alleviated and Sg began to divide mitotically as shown by the appearance of a new generation (Sg*) with reduced CFSE content (16 days, KT+IGF). B: Quantitative data of the cell cycle analysis in the gated field (Fig. 3A, 0 day, control) are presented. The fraction of cells in the respective cell cycle phases represents the arithmetic mean ± SD of six separate experiments. For each time point, significant differences (P < 0.05) to the control are indicated (asterisks).

Download figure to PowerPoint

After 2, 4, 8, 10, 12, and 16 days of culture, the same number of cells were collected, fixed, labeled with PI, and analyzed by flow cytometry and LSM. The combination of both techniques allowed us to monitor the dynamics of cell proliferation and, at the same time, to identify the respective cell types and to interpret the data obtained by flow cytometry. In Figure 3B, the changes in the cell cycle distribution of the populations included in the boxed area (Fig. 3A, 0 day) are quantified. In control cultures, the distribution of cells remained nearly constant during the 16-day culture period (Fig. 3B), except for a small reduction of G2+M cell percentage at the second day of culture and a concomitant increase in S-phase cells. The same tendencies on the second day of culture were observed under all conditions tested. In KT-containing samples, the reduction of G2+M phase amounted to 8 ± 3% (P < 0.05) after 2 days of culture. The analysis by flow cytometry gave an explanation for the observed effect: spermatocytes (4C, Fig. 3A, 0 day) developed during 2 days of the incubation into spermatids, as shown by the reduced number of spermatocytes compared with the 0-day control and the formation of new spermatid population with low CFSE content (labeled St*, Fig. 3A, 2 days), which later differentiated into spermatozoa (labeled Sz*, Fig. 3A, 10 days). Spermatogenesis was apparently maintained in vitro between 2 and 16 days of culture as indicated by the appearance of newly formed haploid cells during the entire culture period. Because the cultures were depleted of St* and especially Sz* by the regular medium changes and spermatozoa do not survive for long under the culture conditions (Song and Gutzeit,2003), the meiotic divisions as well as spermiogenesis must have been maintained during the entire 16-day culture.

Significant changes in the cell cycle distribution were found between 2 and 10 days of culture (Fig. 3B). During this period, the fraction of cells in G2+M phase increased continuously and reached 47 ± 5% (KT), 58 ± 4% (KT+IGF), 52 ± 6% (KT+hCG), and 58 ± 5% (KT+IGF+hCG) after 10 days of culture (Fig. 3B, 10 days), while in control cultures, 35 ± 4% of the cells were in the G2/M phase. The analysis by flow cytometry showed that this effect is due to cell cycle arrest at the G2+M phase (Fig. 3A, 10 days). The percentage of apoptotic cells (Ap) was determined by flow cytometry based on their reduced DNA content (boxed area, Fig. 3A, 2 days). On the average, 7 ± 4% of the primary cells, by that criterion, were apoptotic in cultures maintained for 2 to 10 days.

During the final phase of the experiment (between day 10 and 16), significant differences between the different cultures were observed. The fraction of cells in the G2+M phase significantly decreased to 15 ± 3% (KT+IGF; P < 0.05), 28 ± 4% (KT+hCG; P > 0.05), and 20 ± 3% (KT+IGF+hCG, P < 0.05) with reference to KT-treated (54 ± 4%) cultures (16 days, Fig. 3B). These changes were accompanied by a concomitant increasing percentage of G0+1. At the same time, a new population of diploid spermatogonial cells with half CFSE fluorescence (Sg*, compare with Sg) appeared when analyzed by flow cytometry in two-dimensional plots (Fig. 3A, 16 days). Apparently, in the presence of KT, both IGF and hCG were able to induce mitosis of gonial cells either alone or in combination. However, the kinetics of induction differed between the two compounds: IGF-treated cells initiated mitosis approximately 2 days earlier than hCG-treated cultures (Fig. 3B), and this finding is also apparent in the cell cycle analysis. The IGF-containing culture initiated the second round of mitotic division on day 16 (Sg* cells in S-phase: 36 ± 3% on day 16 compared with 27 ± 2% on day 10; P < 0.05). At the same time, the spermatogonia of the delayed hCG-treated cells were still in the G0+1 phase (Sg* cells in S-phase on day 16 reduced to 21 ± 2% in comparison to 28 ± 2% on day 10; P < 0.05). In the presence of both mitogens, the IGF-typical kinetics was observed (Fig. 3A, 16 days).

The number of Sertoli cells was not determined in our study. However, this cell type consistently occupied the same position in the dot plot histograms, and no changes concerning the cell cycle distribution of Sertoli cells were observed (Fig. 3A, 0–16 days).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The analysis of developmental processes in primary cultures poses a technical challenge that has prevented rapid progress in the past. However, an in vitro approach has many practical advantages and the development of cell culture systems in which defined cellular functions are maintained and can be analyzed with biochemical and genetic techniques is most rewarding. Spermatogenesis is a developmental process that is amenable to in vitro studies both in mammals (Woodruff et al.,1992; Beumer et al.,2000; Jeyeraj et al., 2002) and in fish (Miura et al.,1991; Loir,1999; Sakai,2002; Song and Gutzeit,2003), and this work has demonstrated the potential of this approach.

Recently, medaka spermatogonia cultures were shown to differentiate in vitro and to develop into mature spermatozoa (Hong et al.,2004). However, there is ample evidence from mammalian and fish spermatogenesis that the Sertoli cells are critically involved in sex determination and the control of spermatogenesis and oogenesis alike (Pudney,1995; McLaren,1998). The molecular interactions between the somatic cells and the germ line cells in fish have not been analyzed in detail but the endocrine and paracrine mechanisms controlling spermatogenesis in vertebrates are largely conserved (Nagahama,1994). Apparently, spermatogonia are capable of following their differentiation program autonomously and the somatic cells, in particular the Sertoli cells, mainly control the dynamics of the process and provide the link to the regulatory circuits of the body. The differentiation of spermatogonia in isolation, therefore, may be regarded as a default program that is executed in the absence of controlling influences. Because cellular interactions and paracrine signalling can only be studied in cultures with a realistic representation of the relevant cell types, we used primary cultures that were depleted of haploid cells (spermatids and spermatozoa), the somatic cells (Leydig, myoid cells, macrophages), fibroblasts, and blood cells. The problem with such heterogeneous cultures is that analytical tools are required which allow identification of the different cell types and to analyze the response of individual cells in the mixed culture. We have shown here that flow cytometry is the method of choice to meet these demands.

The major cell types represented in newly prepared primary cultures were spermatogonia, spermatocytes, and Sertoli cells as shown by the defined cytological properties. When these cells were labeled with CFSE and PI, they occupied distinct positions in two-dimensional plots (Fig. 4, grey fields). In addition, CFSE labeling allows monitoring of meiotic and mitotic divisions of the respective cell types by the loss of CFSE fluorescence (Fig. 4, asterisks). Based on the observed effects of the tested substances on the cell germ line cells and taking into account data from the literature (Nagahama,1994; Miura et al.,1996; De Gendt et al.,2004), the likely target cells for hCG and KT are the Sertoli cells and for IGF the spermatogonia (Fig. 4, dashed arrows).

thumbnail image

Figure 4. Simplified graphic illustration of cell division and differentiation that was observed during the 16 day cultures of primary cells (data shown in Fig. 3). Together with additional data taken from other authors (see text for details), a tentative scheme for the molecular control and the cellular targets of the respective hormones and growth factors is shown. The fields shown in gray color represent the cell populations at the beginning of the culture, i.e., Sertoli cells (S), spermatogonia (Sg), spermatocytes (primary, Sc1; and secondary, Sc2), and spermatids (St). Black arrows illustrate the mitotic (solid) and meiotic (dashed) divisions observed in the cultures. The position of daughter cells (marked with stars) appeared in culture after successful mitotic or meiotic divisions and were identified by their reduced carboxyfluorescein diacetate succinimidyl ester (CFSE) content. Cells that had undergone apoptosis are located in the indicated area. The dotted arrows show the likely cellular targets of the analyzed compounds 11-ketotestosterone (KT), insulin-like growth factor I (IGF), and human chorionic gonadotropin (hCG).

Download figure to PowerPoint

In mammals, it has been known for a long time that testosterone is an essential hormone for the initiation and maintenance of spermatogenesis (Steinberger,1971). In fish, the major androgenic steroid is KT (Borg,1994) and this hormone seems to play a similarly important role as testosterone in mammals. KT is thought to be synthesized by the Leydig cells and to act on the Sertoli cells (Nagahama,1994; Miura et al.,1996). KT induced complete spermatogenesis in testicular explants from immature Japanese eel (Miura et al.,1991), and our observations confirmed this finding. Whereas in control cultures some spermatocytes appeared to follow their differentiation program and formed spermatozoa (Song and Gutzeit,2003), the number of newly differentiated haploid cells was much higher in all cultures that contained KT. Because androgens are known to act on Sertoli cells, the promotion of meiosis and spermatogenesis is presumably a Sertoli cell-mediated function. The observed KT-mediated aggregation of primary cells supports this interpretation, and other authors have made similar observation in primary cell cultures of the eel (Miura et al.,1996) and zebrafish (Sakai,2002) and emphasized the importance of cell contacts between germ line cells and Sertoli cells.

Apart from the stimulation of meiosis and the generation of spermatozoa, KT appears to have a strong effect on cell proliferation when compared with control cultures lacking KT. The spermatogonia progressed through the cell cycle and finally arrested in stage G2+M (Fig. 3B). Of interest, the arrest can be relieved by IGF or hCG. IGF is known to induce DNA synthesis in a broad array of cell types (Pestell et al.,1999) and plays an essential supporting role in allowing germ cells to proceed through spermatogenesis (Nader et al.,1999). IGF may be produced by different cell types, and the action may be endocrine (produced in the liver and transported by the blood stream), paracrine (somatic cells to germ cells), or autocrine (germ cells to germ cells). All of these pathways appear to be present in mammals (Lejeune et al.,1998). The relevance of IGF for fish spermatogenesis is evident from the work of Loir and Le Gac (1994), who showed that IGF stimulates germ cell proliferation in rainbow trout upon binding to the IGF receptor. In primary testis of cell cultures of the Japanese eel, IGF stimulated all stages of spermatogenesis in the presence of KT (Nader et al.,1999). Our results and the supporting evidence from other authors (Loir and Le Gac,1994; Nader et al.,1999) suggest that 11-KT is necessary for the induction of spermatogenesis, whereas IGF is necessary for the continuation of the process.

It is well established that gonadotropins play a major role in the endocrine regulation of spermatogenesis in vertebrates (Schulz et al.,2001). However, it is not clear whether hCG acts on spermatogonia cell proliferation either through mediation of KT production by Leydig cells as evidenced for the Japanese eel (Nagahama,1994) or by acting on some other testicular cell types such as Sertoli cells (Loir,1999). In the presence of all the main testicular cell types, hCG was shown (Loir,1999) to moderately stimulate proliferation of spermatogonia of rainbow trout in vitro.

The induction of spermatogonia proliferation by the action of IGF and hCG occurs with different kinetics; hence, different molecular pathways are likely to be involved. Cells responded with an approximately 2-day delay after hCG treatment compared with IGF exposure (Fig. 3B). From our data and the findings of other authors it seems plausible that IGF acts directly on spermatogonia, whereas the effect of hCG is mediated through the somatic Sertoli cells (Fig. 4) and, hence, results in a delayed reaction.

Our observation that mitotic divisions of in vitro cultured testis cells of tilapia are initiated after 12 days is in keeping with the results of other authors. The Japanese eel 11-KT–induced spermatogenesis in testis fragments was monitored and the first mitotic division of spermatogonia, the appearance of zygotene spermatocytes, and the formation of spermatozoa was observed after 9, 18, and 21 days, respectively (Miura et al.,1991). Furthermore, the evidence from in vivo data suggests that the dynamics of spermatogonia divisions and of spermatogenesis is similar to the respective developmental processes in vivo (Khan et al.,1987; Miura et al.,1991,2002).

The analysis of developmental processes in vitro requires not only optimal culture conditions for primary cells but, above all, the study of molecular signalling between different cell types and requires analytical tools that allow monitoring and quantification of effects on a single-cell basis. We have shown in this study that the combination of microscopy and flow cytometry provide powerful tools for the analysis of heterogeneous primary cell cultures.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Animals

Male tilapia (Oreochromis niloticus) of 150–200 g in body weight were kindly provided by Dr. Baroiller (CIRAD-EMVT, SCRIBE, Campus de Beaulieu, France). They were kept in circulating fresh-water tanks (500 liters) at 20°C.

Cell Culture

Dissociated testicular cells were prepared from normal testes of adult fish according to Sakai (2002). Aseptically removed testes were minced into ∼2 mm3 pieces with sterile scissors, washed five times with phosphate buffered saline (PBS) to remove free spermatozoa and blood cells and placed in Leibovitz cell culture medium (L-15, Gibson, France) containing 500 U/ml of collagenase (Sigma, Germany) and 0.03% DNAse (wt/vol, Amersham, Braunschweig, Germany). The cells were incubated on a ST3 Shaker (ELMI, Germany) in a six-well plate (Nunc, Germany) and incubated for 2 hr at 25°C under constant shaking (100 rpm). Testes cell suspensions were diluted seven times with L-15 medium with bovine serum albumin (1%), successively filtered through three meshes of decreasing pore size (100, 50, and 25 μm, Partec, Germany) and centrifuged at 40 × g for 10 min (Hettich, Germany). Supernatant containing mature sperm was discarded, and the centrifugation pellet was resuspended in 1 ml of Dulbecco's modified Eagle (equal volumes of F12 and DMEM) medium. The cell suspension was centrifuged using a one-step Percoll (Fluka, Switzerland) gradient to increase the fraction of diploid spermatogonia and to reduce the fraction of haploid spermatids and spermatozoa. An isotonic solution containing 90% Percoll was obtained by mixing nine volumes of Percoll (D = 1.13 g/ml, Fluka, Germany) with 1 volume of 10-fold concentrated PBS (Gibco BRL, Scotland). This 90% Percoll stock solution was used to prepare 28% (D∼1.04 g/ml), 20% (D∼1.03 g/ml), and 12% (D∼1.02 g/ml) Percoll solutions in L-15 medium. The solution (13 ml) was pipetted into a 15-ml conical tube, and a layer of the cell suspension (1 ml) was carefully placed on top. After centrifugation (100 × g, 30 min), cells from the pellet were washed three times in PBS, resuspended in culture medium (F12/DMEM) supplemented with 10% heat-inactivated fetal calf serum (Biochrom, Germany), 100 IU/ml penicillin, and 100 μg/ml streptomycin (Gibco). The cells were then cultured at a density of 5 × 104 cells/cm2 in 25-ml flasks (10 ml, ∼5 × 105 cells/ml) in humidified atmosphere with 5% CO2 (vol/vol) at 25°C overnight. After this incubation, fibroblasts and interstitial cells adhered firmly to the bottom of the flasks, whereas spermatogonia and Sertoli cells did not (Miura et al.,1996). The suspended cells consisting of spermatogonia, spermatocytes, spermatids, and Sertoli cells were spun down (100 × g, 10 min) and cultured in fresh medium supplemented with 1 μg/ml of bovine pancreas insulin (Sigma, Germany) in five 25-ml flasks (10 ml, ∼5 × 105 cells/ml). One culture served as control, and the other four cultures contained 10 ng/ml of KT (Sigma, Germany), KT and 100 ng/ml of IGF (Roche, Germany), KT and 357 ng/ml of hCG (5 IU/ml, Sigma, Germany), or the combination of KT and IGF and hCG. The cell cultures were maintained for up to 16 days, and the medium was changed every second day.

Flow Cytometry

The different cell types in primary testis cell cultures were defined by four-parameter flow cytometry using next criteria: cell size (forward scatter, FSC), number and size of granules in cytoplasm (side scatter, SSC), DNA content (by staining with PI), and a level of fluorescence of carboxyfluorescein diacetate succinimidyl ester (CFSE) dye that stains the cell cytoplasm and, furthermore, can be used to identify proliferating cells by the decreasing fluorescent signal in successive cell generations (Lyons,1999). The technique has been applied to medaka primary testis cell cultures previously (Song and Gutzeit,2003). Briefly, the isolated tilapia cells were stained with 10 μM CFSE (Molecular Probes, Eugene, OR) in PBS for 10 min at 25 × C. The cells were washed with PBS, and culture medium was added. After different culture times (see Results section), cells were removed from the culture and prepared for the analysis by flow cytometry. The cells were washed with PBS and centrifuged at 100 × g for 10 min. The cell pellet was resuspended in 100 μl of PBS, fixed in 70% (vol/vol) ethanol by adding 1 ml of cold (−20°C) ethanol, and stored overnight at −20°C. The cells were spun down again, and the pellet was resuspended in 1.5 ml of PBS at room temperature. After centrifugation, the cell pellet was resuspended in 1 ml of DNA staining solution containing 50 μg of PI and 0.2 mg of RNase (both Sigma, Germany) and incubated for at least 45 min at room temperature in the dark.

The same number of cells (5 × 105) per sample were analyzed by flow cytometry (CyFlow, Partec, Germany). The excitation wavelength was 473 nm, and green fluorescence (520 nm for CFSE) and red fluorescence (>590 nm for PI) were recorded. In addition, the parameters FSC and SSC were determined. For each variable (exposure conditions, culture periods, and so on), a minimum of six samples were quantified. The flow cytometer was calibrated with 2.5-μm polyfluorescent beads AlignFlow (Molecular Probes, Eugene, OR) before each series of measurements. The fraction of cells present in different cell generations and their representation in the respective cell cycle phases, the mean level of measured fluorescence and the phagocyte activity was calculated using CyFlow software (Partec, Germany).

Latex Beads Incorporation Assay

Phagocytosis of latex beads by Sertoli cells was performed as described previously (Shiratsuchi et al.,1999) with some modifications. Primary testis cells were mixed with carboxylate-modified latex microspheres (fluorescence, 488/645 nm; diameter, 1.0 μm; TransFluoSpheres, Molecular Probes) and incubated for 1 hr at 25°C. The selective incorporation into Sertoli cells was observed by confocal microscopy and quantified by flow cytometry.

Confocal LSM

Parallel to the analysis by flow cytometry, the respective cell cultures were analyzed using a laser scanning microscope (LSM 510, Zeiss, Germany) equipped with Ar–Kr laser (30 mW) operating at 488 nm. Two channels were used for simultaneous data acquisition: green fluorescence was detected on channel 1 (505- to 550-nm bandpass filter) while red fluorescence was detected on channel 2 (585-nm long-pass filter). Cells were attached to polylysine-coated coverslips (∼105 cells per slip) after a brief wash with PBS. Fluorescence emission was recorded through a Plan-Nufluor 40×/1,3 oil objective (single cell visualization) and Plan-Neofluar 20×/0.5 objective (imaging cell clumps). Accumulated digital images (1,024 × 1,024 × 8 bits per channel) were converted to bitmap files for analysis by area morphometry. The total fluorescence of the cells (100 cells per sample) was quantified by area morphometry analysis using appropriate software (Optimas Co., Washington, DC).

Statistics

The experimental data are expressed as the mean ± SD of six independent experiments. The significance of the recorded effects was assessed by analysis of variance.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Nadja Bachmann for maintaining the fish stocks.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES