Stem cell activity of type A spermatogonia is seasonally regulated in rainbow trout^†

Abstract

Spermatogonial stem cells (SSCs) support continuous production of sperm throughout the male's life. However, the biological characteristics of SSCs are poorly understood in animals exhibiting seasonal reproduction, even though most wild animals are seasonal breeders. During the spermiation season in rainbow trout, the lumen of the testes contains only spermatozoa and scattered type A spermatogonia (ASG) along the walls of the testicular lobules. These few remaining ASG, designated “residual ASG,” are the only germ cells capable of supporting the next spermatogenesis, suggesting that the residual ASG are true SSCs. However, whether residual ASG can behave as SSCs in any teleost species is unknown. In this study, we attempted to clarify the biological characteristics of SSCs associated with seasonal reproduction in rainbow trout using spermatogonial transplantation. We found that the stem cell activity was clearly regulated seasonally during the annual reproductive cycle. Although the residual ASG exhibited moderate transplantability and colony-forming ability at the beginning of the spermiation season, these parameters decreased dramatically later and remained low until the next spermatogenesis was initiated. Furthermore, no clear correlations were observed between these qualitative changes and previously described morphologic characteristics of ASG or plasma sex steroid levels. Our results suggest that the biological properties of SSC populations in rainbow trout are seasonally regulated by a novel mechanism.

Introduction

Spermatogenesis produces a large number of sperm during the breeding season throughout the male life span, and this process is heavily dependent on a robust stem cell system. Spermatogonial stem cells (SSCs) are the foundation of spermatogenesis and, like other adult stem cells, have a capacity for both self-renewal and the generation of progenitor cells committed to differentiation into spermatozoa [1, 2]. The balance between self-renewal and differentiation of SSCs is strictly regulated extrinsically by the surrounding microenvironment and intrinsically by various gene networks in order to prevent excessive self-renewal or differentiation, which can cause male infertility [3].

Studies on this critical stem cell population, particularly in vertebrates, have been hampered because the number of stem cells among the testicular cell population is very low, and methods to distinguish SSCs from their progenitor cells have only recently been developed. In 1994, Brinster and colleagues first reported the establishment of a spermatogonial transplantation assay in mice [4, 5]. Using this assay, SSCs could be functionally and quantitatively detected among testicular cells based on the direct measurement of colony formation in recipient testes [6, 7]. This technique has accelerated the accumulation of knowledge concerning the biology of SSCs, especially in rodents [8, 9].

Teleosts exhibit a broad and diverse array of reproductive strategies. Some species display apparent seasonal reproduction, and others show active gametogenesis throughout the year [10]. In salmonids, most Pacific salmon die soon after a single reproduction event [11, 12], whereas trout spawn multiple times throughout life [13]. In some species, adult individuals change their sex after a certain age or upon a specific social cue [14]. In teleost species, spermatogenesis is regulated in different ways depending on species-specific reproductive strategies, suggesting that studies of SSCs using species that display different reproductive strategies would provide numerous clues that would help elucidate the functional and regulatory aspects of vertebrate SSCs.

Spermatogonial transplantation systems have been established for some fish species [15, 16]. In rainbow trout, only type A spermatogonia (ASG) isolated from donor individuals were transplanted into the peritoneal cavity of the allogenic recipient hatchlings. As a result, donor ASG were incorporated, forming donor-derived colonies in the recipient gonads and consequently producing functional donor-derived gametes for several years [17]. Since by definition SSCs are capable of both self-renewal and the generation of progenitors committed to differentiation into spermatozoa, this result suggested that the incorporated donor ASG behaved as SSCs in the recipient gonads. Furthermore, this study also found that only a subset of the ASG population (five donor ASG from among 10 000 transplanted ASG) could colonize and resume gametogenesis in the recipient gonads. The above-mentioned results indicate that rainbow trout ASG constitute the most undifferentiated germ cells capable of behaving as SSCs in recipients. In addition, functional evidence suggests that ASG are heterogeneous with respect to stemness. This was further confirmed by side population analyses using rainbow trout ASG [18].

Detailed morphologic analyses of ASG have been conducted for several fish species. These studies revealed that various ASG morphologic characteristics, such as nuclear size, number of nucleoli, and amount of heterochromatin, vary across ASG populations. Thus, ASG can be subdivided into various types based on morphologic criteria [10, 19–21]. The results of analyses of the expression of putative SSC markers identified in rodents (nanos2 and gfrα1) also showed internal heterogeneity among ASG that are positive and negative for each marker [22–24]. However, what subpopulation(s) of heterogeneous ASG behave as stem cells in vivo remains unknown, as methods enabling the identification and isolation of live SSCs have not been established. Therefore, our knowledge concerning SSCs in teleost species remains limited. The spermatogonial transplantation assay is the only method currently available for functionally and quantitatively detecting the presence of SSCs among ASG based on their colony-forming ability.

Rainbow trout, which belong to the Salmonidae family, are seasonal breeders that reproduce multiple times throughout their life span. Spermatogenesis exhibits strict seasonality and proceeds almost synchronously in the entire testis. This spatiotemporal organization of spermatogenesis provides advantages for researchers who wish to analyze the physiologic properties of the testes at a specific spermatogenic stage, such as spermatogonial proliferation, onset of meiosis, or spermiogenesis. Furthermore, during the spermiation season, the lumen of the testicular lobules contains only spermatozoa, with a few scattered ASG remaining along the testicular lobule walls. These histologic observations suggest that these few remaining ASG, designated “residual ASG,” are maintained for spermatogenesis in the next reproductive cycle [11] and may actually be true SSCs. However, whether the residual ASG population exhibits greater stem cell activity than ASG populations in the testes during the other spermatogenic stages in any teleost species has not been determined.

Our long-term aim is to understand the regulation of SSCs in animals exhibiting seasonal reproduction, such as rainbow trout. A basic requirement for such a study is a thorough biological characterization of the ASG population throughout the reproductive cycle, including quantitative data regarding stem cell activity, which would enhance understanding of the biological regulation of ASG populations, including SSCs. In salmonids, studies of spermatogenesis have been performed primarily using specimens that have yet to undergo the first testicular maturation. In this situation, it is difficult to evaluate the seasonal regulation of ASG properties apart from the physiologic effects associated with testicular development and sexual maturation. Therefore, in order to focus solely on events corresponding to the annual reproductive cycle, we carried out biological analyses of ASG populations using adult male rainbow trout. In this study, we used a transplantation assay combined with detailed histologic analyses to reveal unexpected behaviors of ASG populations during the annual reproductive cycle in male rainbow trout.

Materials and methods

Fish

All rainbow trout (Oncorhynchus mykiss) used in this study were maintained at the Oizumi station of the Field Center of Tokyo University of Marine Science and Technology (Yamanashi, Japan) under a natural photoperiod and at a water temperature of 10.5°C. In this study, we used a transgenic rainbow trout strain (vasa-Gfp) that carries the green-fluorescent protein (Gfp) gene driven by vasa gene regulatory elements [25, 26]. In this transgenic strain, the germ cells exhibit green fluorescence, predominantly in ASG [27].

Sampling

In order to analyze the biological characteristics of trout ASG throughout the annual reproductive cycle, adult male rainbow trout, age 2 to 3 years, were sampled at 0, 1, 3, 5, 7, 9, 11, and 14 months after confirmation of the onset of spermiation (months postspermiation (mps)) (Figure 1). Spermiation, which occurred from July to September in the first year, was confirmed by the application of gentle abdominal pressure. Since the onset of sperm release in the next reproductive cycle was delayed for approximately 1 month compared with the first cycle, a spermatogenic stage of 14 mps corresponds to a stage of 1 mps in the second reproductive cycle. Sampled fish were anesthetized with 2-phenoxyethanol (Wako Pure Chemical Industries, Ltd) at a concentration of 400 ppm. The whole fish and isolated testes were weighed to calculate the gonadal somatic index (GSI = gonad weight/body weight × 100). The isolated testes were used for the histologic analyses and transplantation assay described below. Blood was also collected from the caudal vein using a heparinized syringe. The collected blood was centrifuged at 1000 × g for 10 min, and the plasma was stored at −80°C until used for hormone analyses.

Figure 1.

Diagram of experimental schedule. In order to analyze the biological properties of ASG during the annual reproductive cycle in rainbow trout, adult trout reaching first sexual maturity were sampled at 0, 1, 3, 5, 7, 9, 11, and 14 months after confirming the spermiation (mps). Zero month postspermiation indicates the period within the month after confirming the first sperm release (2+). One month postspermiation indicates the period from 1 to 2 months after confirming the first sperm release. Fourteen months postspermiation correspond to 1 mps in the next cycle (3+).

Histologic analyses of rainbow trout testes

Part of each isolated testis was fixed in Bouin's fixative at 4°C for 24 h prior to histologic analyses. The fixed specimens were dehydrated using an ethanol series and then embedded in paraffin, sectioned to a thickness of 4 μm, and stained with hematoxylin and eosin (HE). The developmental stage of the germ cells was determined based on the definitions reported previously [10, 19]. In order to determine the precise localization of ASG in testicular lobules, paraffin sections were subjected to immunofluorescent staining using the method described previously [28]. In the present study, mouse anti-GFP antibody (1:1000 dilution; product no. 1181446000; Roche Applied Science) was used for the detection of ASG in vasa-Gfp transgenic rainbow trout carrying ASG that exhibited strong GFP expression (Supplemental Table S1). In order to accurately determine the type of cells exhibiting positive fluorescence, the sections were subsequently stained with HE after observation of fluorescence. Merged images were created by combining the fluorescence and HE staining images using Adobe Photoshop 6.0 (Adobe Systems Inc.).

Isolation of type A spermatogonia from testicular cells

For spermatogonial transplantation, one pair of the isolated testes from a single adult fish was dissociated using the following method, and the resulting cells were used as donors. Three to seven adult fish were used as donors at each experimental time point. In each transplantation, immature testes containing only ASG were also collected from three prepubertal under-yearling fish, pooled and used for the subsequent steps as a positive control. Isolated testes were minced using scissors and rinsed by gentle pipetting with Eagle minimum essential medium (MEM (pH 7.8); Nissui Pharmaceutical Co., Ltd) supplemented with 5% fetal bovine serum (Invitrogen Co.) and 20 mM Hepes (Sigma-Aldrich, Co.) to remove free spermatozoa. The resulting testicular fragments were dissociated by enzymatic treatment in accordance with the method described previously [29]. The dissociated testicular cells were rinsed with the above-mentioned medium to eliminate enzymatic activity. After centrifugation at 200 × g for 10 min at 10°C, the resulting pellets were suspended in the medium and filtered through 42- and subsequently 20-μm pore-size nylon screens (Tokyo Screen Co., Ltd) to remove nondissociated cellular clumps. As a testicular cell suspension prepared from mature testes contains a large number of spermatozoa (Figure 2A), residual spermatozoa were removed using centrifugation in Percoll (Percoll Plus; GE Healthcare UK Ltd). ASG-rich fractions were recovered from between 25 and 45% isotonic Percoll solution (Figure 2B), and the fractionated cells were stored on ice until cell sorting. In order to avoid possible negative effects on the incorporation efficiency of ASG associated with the presence of testicular somatic cells among the donor testicular cells [30], highly enriched ASG were used for the transplantation assay. To do so, ASG were purified by GFP-dependent cell sorting using a Moflo XDP (Beckman Coulter) [27, 31]. ASG (3 × 10⁵ cells) were sorted into microtubes containing the MEM and kept on ice until transplantation (Figure 2C and D).

Figure 2.

ASG isolated from donor rainbow trout. Green fluorescent cells are ASG derived from vasa-Gfp transgenic rainbow trout. (A) Testicular cells of mature testis after enzymatic digestion contain a large number of spermatozoa (nonfluorescent cells) and a relative paucity of ASG. (B) Fractionated testicular cells after Percoll gradient are enriched in ASG but still contain somatic cells (GFP-negative cells). (C) Highly enriched ASG were isolated using a fluorescence-activated cell sorter (FACS). (D) ASG from prepubertal testes were also enriched by FACS and used as a control group at each transplantation assay. Scale bar indicates 10 μm.

Spermatogonial transplantation

ASG isolated from vasa-Gfp transgenic rainbow trout were used as donor cells for spermatogonial transplantation. Approximately 500 ASG were transplanted into the body cavity of each nontransgenic rainbow trout hatchling (25–35 days postfertilization) using the method developed previously [17]. We used 80–180 recipient hatchlings for each transplantation and repeated transplantations three to seven times at each experimental time point. At 20 and 60 days post-transplantation (dpt), the recipients were dissected, and colonization of donor ASG was assessed using a fluorescent microscope (model BX-51N-34FL; Olympus) equipped with a GFP filter (U-MWIB; Olympus). Incorporation efficiency was defined as the number of recipients carrying donor ASG in their gonads divided by the total number of analyzed recipients. The data were collected from at least 20 recipients in each experiment. Since we have often observed that the incorporation efficiency was affected by recipient batches (Sato and Yoshizaki, unpublished data), the incorporation efficiency obtained from the adult donors was expressed as a relative value compared to that of the control prepubertal group in order to compensate for variations in acceptance of donor cells by the recipient batches.

Additionally, in order to evaluate the colony-forming ability of incorporated donor ASG, the number of donor ASG colonized in each recipient gonad was counted at 20 and 60 dpt. The presence of recipients with large colonies derived from donor ASG in their gonads was also analyzed at 60 dpt. Because the colonized ASG tended to proliferate more actively in female recipients, “large colony” was defined separately according to sex. Colonies consisting of at least 15 and 30 donor ASG were defined as “large” in male and female recipients, respectively. The appearance rate of recipients with donor ASG-derived large colonies was defined as the number of recipients with large colonies of donor ASG divided by the total number of recipients with donor ASG in their gonads.

Morphological characterization of trout type A spermatogonia

ASG were subdivided in accordance with the criteria of Loir (1999) [19]. The largest type of ASG, with a nuclear diameter of approximately 10 μm and one prominent, dense body, were categorized as A1 cells. The other ASG, which had smaller nuclei than A1 cells and possessed several dense bodies, were categorized as A2 cells. The appearance ratio of A1 cells was defined as the number of A1 cells divided by the total number of ASG analyzed (at least 50 randomly selected ASG were analyzed for each donor). Nuclear size was measured using ImageJ software (National Institutes of Health).

Type A spermatogonia apoptosis and proliferation analyses

In order to detect the appearance of apoptotic or proliferating ASG, the sections were subjected to double-immunofluorescence staining: mouse anti-GFP antibody (1:1000 dilution) for detection of ASG in vasa-Gfp transgenic rainbow trout, together with rabbit anti-active caspase3 antibody (1:1000; product no. G7481; Promega Co.) for the detection of apoptotic cells, or rabbit anti-phospho-histone H3 (PH3) antibody (1:1000 dilution; product no. 06–570; Merck Millipore) for the detection of proliferating cells (Supplemental Table S1). After washing the primary antibodies, appropriate secondary antibodies (1:200 dilution; Alexa Fluor 488-conjugated goat anti-mouse IgG; product no. A-11001; Life Technologies and 1:200 dilution; Alexa Fluor 546-conjugated goat anti-rabbit IgG; product no. A-11035; Life Technologies) were applied to the sections (Supplemental Table S1). Fluorescent signals were observed under a fluorescent microscope (model BX-51N-34FL; Olympus) equipped with appropriate filters: U-MWIB or U-MNIBA (Olympus) for Alexa Fluor 488 and U-MWIG (Olympus) for Alexa Fluor 546. The percentages of apoptotic or proliferating ASG were calculated by determining the ratio of double-positive cells among 500–1000 randomly selected ASG (cells positive for anti-GFP antibody) from each donor. The types of cells with positive signals were verified by staining the sections with HE after the observation of fluorescence.

Estimation of the total number of type A spermatogonia in trout testes

In order to assess changes in the total number of ASG in each donor, the relative total number of ASG in donor testes was estimated. The middle portion of fixed donor testis was embedded in paraffin, sectioned (4 μm), and stained with HE. Images of randomly selected areas of these sections were captured using a digital camera (DP72; Olympus) configured to a light microscope (model BX-51N-34FL; Olympus), and all of the ASG in each captured image were counted. The number of ASG per gram of tissue was estimated under the assumption that the density of testicular tissues is 1 g/cm³. Subsequently, the estimated total number of ASG in each donor testis was calculated by multiplying the estimated number of ASG per gram of tissue by the weight of the donor gonads. Data are shown as relative values compared with the estimated total number of ASG in donor testes at 0 mps.

Hormone assays

Sex steroids were extracted from plasma samples three times using a 5× volume of diethyl ether. Levels of 11-ketotestosterone (11-KT), testosterone (T), and 17, 20β-dihydroxy-4-pregnen-3-one (DHP) were measured using a time-resolved fluoro-immunoassay according to the method [32], with a slight modification. 11KT-, T-, or DHP-oxime–bovine serum albumin (T-3-CMO-BSA and 11KT-3-CMO-BSA were purchased from Sigma-Aldrich, whereas DHP-3-CMO-BSA was prepared according to the method described previously [33]) was immobilized by physical adsorption to the wells of a microtiter plate (Thermo Fisher Scientific Inc.) as an antigen in the solid phase. A competitive assay was performed between the solid-phase antigen and an objective antigen in the plasma samples against the primary antibody (anti-Testosterone (3), anit-11KT-(3)-CMO-BSA, and anit-DHP-(3)-CMO-BSA, Cosmo-Bio Co., Ltd) at 4°C overnight. After washing the wells, a Europium (Eu)-labeled secondary antibody (Perkin Elmer, Inc.) was used to detect the antigen–antibody complex formed in the solid phase (Supplemental Table S1). After the antigen–antibody reaction, Eu³⁺ was dissociated from the secondary antibody using enhancement solution (Perkin Elmer, Inc.). The intensity of Eu fluorescence was measured using a time-resolved fluorometer (Infinite 200 pro; Tecan Group Ltd) operated with i-control software (Tecan Group Ltd). Plasma levels of 17β-estradiol were measured using an enzyme immunoassay kit (Cayman Chemical Co.) according to the manufacturer's instructions. Signal intensity was measured using a microplate absorbance reader (iMark; Bio-Rad Laboratories, Inc.).

Statistical analysis

All data are presented as the mean ± SEM. Statistical significance was determined using one-way ANOVA followed by Tukey multiple comparisons test, using a statistical significance level of P < 0.05. All analyses were carried out using GraphPad Prism software, version 5.0 (GraphPad Software Inc.).

Ethics

All experiments described herein were carried out in accordance with the Guide for the Care and Use of Laboratory Animals from Tokyo University of Marine Science and Technology.

Results

Histological changes of testes during the annual reproductive cycle in adult rainbow trout

In order to understand the structural changes that occur in rainbow trout testes during the annual reproductive cycle, we performed histologic analyses using adult fish. Germ cell stages were as follows (Figure 3): the largest isolated or paired germ cells with a prominent or sparsely fragmented nucleolus and clear cytoplasm were categorized as ASG (Figure 3A); germ cells that appeared smaller than ASG and formed clusters were categorized as type B spermatogonia (BSG) (Figure 3B). After the final mitosis, BSG differentiate into spermatocytes and spermatids through meiosis, finally transforming into spermatozoa, which were identified based on the nuclear morphologic characteristics reported by Schulz et al. [10] (Figure 3C–F). Further, the structural changes of testis were categorized into four stages, as follows:

1. Early spermiation stage (0, 1, and 14 mps): in the first year, at age 2 years, sperm release occurred from July to September. In the month after the first sperm release was confirmed (i.e. 0 mps; Figure 1), a large number of spermatozoa were released into the lumen of the testicular lobules, and some spermatogenic cysts could still be observed along the walls of the spermatogenic lobules (Figure 4A-I). Additionally, singly isolated or aligned ASG were dispersed along the walls of the lobules (Figure 4A-II). The incidence of spermatogenic cysts decreased during this period. At 1 mps, although a few spermatid cysts still remained, spermatozoa became dominant in the lobules (Figure 4B-I). Widely scattered residual ASG were also observed along the walls of the tubules (Figure 4B-II). At 14 mps, which corresponds to the 1 mps spermatogenic stage in the next reproductive cycle, histological feature closely resembled that at 1 mps (Figure 4B and I).

2. Late spermiation stage (3 and 5 mps): the testis was filled with a large number of spermatozoa (Figure 4C-I and D-I), and only residual ASG were observed along the walls of the lobules (Figure 4C-II and D-II).

3. Regression stage (7 mps): from March to April, sperm release ceased. Spermatozoa and ASG were the only germ cells found in the testes. In some of the testicular lobules, the residual spermatozoa were progressively phagocitized by somatic cells constituting the walls of the lobules and were completely removed by 11 mps (Figure 4E and F).

4. Spermatogenesis stage (9 and 11 mps): cysts containing differentiating germ cells were actively formed (Figure 4G-I and H-I). At 9 mps, although ASG and BSG were still prominent in the testes, a small number of spermatocytes and spermatozoa began to appear (Figure 4G). At 11 mps, differentiated germ cells (spermatocytes, spermatids, and spermatozoa) became dominant (Figure 4H). The formation of BSG cysts was observed until the onset of sperm release.

Figure 3.

Germ cells at various differentiation stages. (A) ASG: black arrowhead denotes a type A1 spermatogonium in which the nucleus is ≥10 μm with one prominent dense body; white arrowheads indicate type A2 spermatogonia with a smaller nucleus with fragmented dense bodies. (B) Type B spermatogonium. These cells are smaller than ASG and form clusters. (C and D) Spermatocyte in pachytene (C) and meiotic figure from the first meiotic phase (D). (E) Spermatid. (F) Spermatozoa. Scale bar indicates 10 μm.

Figure 4.

Seasonal change in testicular structure during the annual reproductive cycle in rainbow trout. Upper panels (I) show an overview of testicular structure, and lower panels (II) are higher magnifications focused on ASG expressing GFP. The lower panels were created by combining an image of fluorescent ASG stained with an anti-GFP antibody with an image of HE staining. (A and B) Early spermiation stage (0 and 1 mps). (C and D) Late spermiation stage (3 and 5 mps). (E and F) Regression stage (7 mps). Sperm release was stopped and the residual spermatozoa were phagocytized by somatic cells (indicated by black arrows) constituting the wall of lobules (F). In some of the lobules, the residual spermatozoa were completely removed and ASG are the only germ cells remaining in the lobules (E). (G and H) Spermatogenesis stage (9 and 11 mps). (I) Early spermiation stage in the next reproductive cycle (14 mps corresponds to 1 mps in the next cycle). Scale bars indicate 50 μm and 10 μm in upper and lower panel, respectively.

Seasonal changes of gonadal somatic index during the annual reproductive cycle in adult rainbow trout

The mean GSI changed in association with the spermatogenic stages (Figure 5). In early spermiation stage, the mean GSI was at its maximum value at 0 mps (5.76 ± 0.61) and then approximately halved at 1 mps (3.63 ± 0.24). At 3 mps, the GSI was slightly decreased but remained at a constant value during this late spermiation stage (3 mps: 2.86 ± 0.12; 5 mps: 2.81 ± 0.33). When the spawning season finished at 7 mps, a marked reduction in the mean GSI was noted (0.93 ± 0.14). During the spermatogenesis stage, the mean GSI increased rapidly due to active formation of cysts of differentiated germ cells (9 mps: 0.36 ± 0.04; 11 mps: 3.20 ± 1.05). At 14 mps, which corresponds to the 1 mps spermatogenic stage in the next reproductive cycle, the mean GSI was similar to that at 1 mps (3.18 ± 0.26).

Figure 5.

Change in the GSI (%) during the annual reproductive cycle in rainbow trout. Different lowercase letters denote mean values that differ significantly (P < 0.05).

Seasonal changes in the transplantability of type A spermatogonia isolated from adult rainbow trout

In the control prepubertal group, the incorporation efficiency (defined as the appearance rate of recipients receiving donor-derived ASG in their gonads among all analyzed recipients) showed no significant difference at either 20 or 60 dpt (Figure 6A and B). In contrast, the incorporation efficiency of ASG derived from adult testes changed dramatically during the annual reproductive cycle (Figure 6C and D). In the early spermiation stage, at 0 and 1 mps, the incorporation efficiency at 20 dpt was relatively high (0 mps: 0.52 ± 0.22; 1 mps: 0.87 ± 0.16) (Figure 6E). In contrast, in the late spermiation stage, the incorporation efficiency declined rapidly and remained at the lowest level until the end of the spermiation stage (3 mps: 0.09 ± 0.03; 5 mps: 0.03 ± 0.02). In the regression stages, the incorporation efficiency increased (7 mps: 0.37 ± 0.11). During the spermatogenesis stage, the incorporation efficiency recovered and remained high (9 mps: 0.66 ± 0.08; 11 mps: 0.31 ± 0.21). At 14 mps, the incorporation efficiency was similar to that at 1 mps (14 mps: 0.97 ± 0.13). A similar seasonal change in incorporation efficiency was observed at 60 dpt, although the incorporation rate of ASG isolated at 7 mps was low (7 mps: 0.02 ± 0.06) (Figure 6E and F).

Figure 6.

Transplantability of ASG populations throughout the year. (A and B) Efficiency of incorporation of donor ASG isolated from prepubertal testes (control group) at 20 dpt (A) and 60 dpt (B). (C and D) Efficiency of incorporation of donor ASG isolated from adult testes at 20 dpt (C) and 60 dpt (D). (E and F) Relative values of the efficiency of incorporation of donor ASG isolated from adult testes at 20 dpt (E) and 60 dpt (F). Average ± SEM from 3, 3, 4, 6, 5, 7, 4, and 5 independent transplantations for 0, 1, 3, 5, 7, 9, 11, and 14 mps are shown, respectively. Different lowercase letters denote mean values that differ significantly (P < 0.05).

Seasonal changes in colony-forming ability of type A spermatogonia throughout the annual reproductive cycle

In order to assess the colony-forming ability of ASG, the number of colonized donor ASG in recipient gonads was determined. At 20 dpt, several donor ASG were observed in the recipient gonads (Figure 7A). The mean number of incorporated donor ASG did not show large differences between groups that received ASG prepared from both adult donors and prepubertal donors (as controls) throughout the experimental period, other than 5 mps when incorporation of donor ASG was observed in only one recipient individual (Figure 7B).

Figure 7.

Colony-forming ability of donor ASG in recipient gonads. (A) Incorporated donor ASG (GFP-positive) in a recipient gonad at 20 dpt. (B) Number of colonized donor ASG in recipient gonads at 20 dpt. Left graph is a result obtained from an adult donor and right graph is a result obtained from a prepubertal donor. (C and D) Colonized-donor ASG (GFP-positive) in male (C) and female (D) recipient gonads at 60 dpt. (E and F) Number of colonized donor ASG in male (E) and female (F) recipient gonads at 60 dpt. (G and H) Appearance ratio of recipients with large colonies derived from donor ASG in male (G) and female (H) recipient gonads at 60 dpt. Colonies consisting of at least 15 and 30 donor ASG were defined as “large” in male and female recipients, respectively. (A, C, and D) Left is fluorescent view and right is bright view. (B, E, and F) Each dot indicates the number of incorporated or colonized donor ASG in one recipient. Scale bar indicates 50 μm.

The proliferation of donor ASG within recipient gonads was also analyzed at 60 dpt (Figure 7C and D). In the control prepubertal group, the number of donor ASG that had colonized recipient gonads remained essentially constant throughout the year (Figure 7E and F). Furthermore, recipients carrying large colonies derived from donor ASG were observed at a constant rate during the experiment period in both sexes of the control group (Figure 7G and H). On the other hand, the number of colonized donor ASG isolated from adult trout changed during the annual reproductive cycle. Using donor ASG isolated from testes at 0 and 1 mps (i.e. residual ASG in the early spermiation stage), donor ASG actively proliferated, especially in female recipients (Figure 7E and F), and formed large colonies in recipients of both sexes at 60 dpt (Figure 7G and H). However, between 3 and 7 mps (i.e. residual ASG in late spermiation and regression stages), the number of colonized donor ASG per recipient was small, and few large colonies from donor ASG were observed in recipient gonads, although we could not obtain a sufficient number of recipients carrying donor ASG during this season. At 9 and 11 mps, the number of colonized ASG increased, and the donor ASG formed large colonies in the recipient gonads, though no large colonies were found in male recipients when the donor ASG were isolated at 11 mps. In addition, the colony-forming ability of donor ASG isolated from the testes at 14 mps (i.e. 1 mps in the next reproductive cycle) was very similar to that at 1 mps.

Seasonal changes in morphologic characteristics of type A spermatogonia

In order to determine whether a correlation exists between the morphologic characteristics of ASG and their stem cell activity, we analyzed ASG morphology in each donor testis using previously defined trout ASG morphologic criteria. Loir (1999) [19] subdivided rainbow trout ASG into two types based on nuclear size and the number of dense bodies in the nucleus. Briefly, A1 cells are the largest ASG (nuclear diameter ≥ 10 μm), with one prominent dense body, whereas A2 cells possess a smaller nucleus with several dense bodies (Figure 2A). During the spermiation stage (0–5 mps), although ASG with one dense body and a clear cytoplasm were frequently observed in the testes, the nuclear size of the ASG was notably smaller compared to that of A1 cells (Figure 4A–D). Therefore, we could not find any ASG with a similar morphology to A1 cells in the testes during the spermiation season (Figure 8). On the other hand, in both the regression and spermatogenesis stages (7–11 mps), the nuclei of the ASG were larger; A1 cells were often observed in the testes at these stages (Figure 8).

Figure 8.

Appearance ratio of A1 cells [19] among donor ASG at each spermatogenic stage. Average ± SEM from 3, 3, 4, 6, 5, 7, 4, and 5 adult testes for 0, 1, 3, 5, 7, 9, 11, and 14 mps are shown, respectively.

Seasonal changes in the number of type A spermatogonia in adult testes during the annual reproductive cycle

In order to clarify whether the frequency of transplantable ASG among donor ASG changes seasonally in synchrony with the appearance rate of apoptotic or proliferating ASG in donor testes, adult testes were examined immunohistochemically using antibodies against active caspase3 (apoptosis marker) and PH3 (proliferation marker) (Figure 9A and B). Although the appearance rate of apoptotic ASG at 0, 7, and 9 mps tended to be higher than the rates at the other stages, no significant differences were found throughout the experimental period (Figure 9C). Further, the percentage of apoptotic ASG remained low (less than 1%) throughout the experimental period. In contrast, the appearance rate of proliferating ASG within the overall ASG population changed seasonally. Proliferating ASG were observed at a low frequency in the testes from 0–5 mps and at 14 mps (i.e. throughout the spermiation stage), but they were observed at a higher frequency between 7 and 11 mps (i.e. during the regression and spermatogenesis stages) (Figure 9D). Furthermore, we estimated the total number of ASG in donor testes using conventional histologic analyses. Although no significant differences in the estimated total number of ASG in donor testes were observed up to 9 mps, the number of ASG gradually increased beginning at 7 mps, reaching a maximum value at 11 mps (Figure 9E), which was maintained until 14 mps. These results suggest that the residual ASG do not actively proliferate during the spermiation season. In contrast, ASG actively proliferate from the regression to spermatogenesis stages, resulting in the replenishment of ASG in the testes.

Figure 9.

Change in the number of ASG during the annual reproductive cycle in adult rainbow trout. (A) Double-immunostaining analyses with anti-GFP (green: detection of ASG) and anti-active caspase3 (red: detection of apoptotic cells) antibodies. (B) Double-immunostaining analyses with anti-GFP (green; detection of ASG) and anti-phospho-histone H3 (red; detection of proliferating cells). (C) Frequency of apoptotic ASG within donor testes. Average ± SEM from 3, 3, 4, 4, 3, 5, 3, and 4 adult testes for 0, 1, 3, 5, 7, 9, 11, and 14 mps are shown, respectively. (D) Frequency of apoptotic or proliferating ASG within donor testes, respectively. Average ± SEM from 3, 3, 4, 6, 5, 7, 3, and 5 adult testes for 0, 1, 3, 5, 7, 9, 11, and 14 mps are shown, respectively. (E) Relative total number of ASG contained in the testes. Average ± SEM from 3, 3, 4, 6, 5, 7, 4, and 5 adult testes for 0, 1, 3, 5, 7, 9, 11, and 14 mps are shown, respectively. Scale bar indicates 10 μm. Different lowercase letters denote mean values that differ significantly (P < 0.05).

Relationship between the plasma level of sex steroid hormones and type A spermatogonia transplantability

Plasma levels of several sex steroids were analyzed in order to clarify the relationship between these hormones and ASG transplantability. The plasma level of 11-KT, which is the major androgen in fish, varied seasonally (Figure 10A). Plasma 11-KT was maintained at a high level during the spermiation season. When sperm release ceased at 7 mps, plasma 11-KT rapidly decreased to its lowest level. During the spermatogenesis stage, the level of 11-KT gradually increased. At 14 mps, the level of 11-KT was similar to that at 1 mps. Plasma T levels were the highest at the onset of spermiation (0 mps) and decreased gradually during the early spermiation stage until reaching a low level that was maintained through the late spermiation stage (Figure 10B). The lowest plasma T level was reached during the regression stage, and the level then gradually increased during the spermatogenesis stage. The plasma level of T at 14 mps was similar to that at 1 mps. Plasma levels of 17β-estradiol (E₂), which is one of the major fish estrogens, were also analyzed (Figure 10C). The highest level of plasma E₂ was observed at 0 mps, and the level subsequently declined until 3 mps. The plasma E₂ level remained low from 3 to 7 mps and then increased slightly during spermatogenesis. At 14 mps, the plasma E₂ level was similar to that at 1 mps. Plasma levels of DHP, a hormone involved in the onset of meiosis and spermiogenesis [10, 34], remained high throughout the spermiation stage but subsequently declined from the end of the spermiation stage to the onset of the regression stage (Figure 10D). The lowest DHP level was observed throughout the spermatogenesis stage.

Figure 10.

Changes in plasma levels of sex steroid hormones during the annual reproductive cycle in adult rainbow trout. Plasma levels of 11-KT (A), testosterone (B), 17β-estradiol (C), and 17, 20β-dihydroxy-4-pregnen-3-one (D). Average ± SEM from 3, 3, 4, 4, 5, 6, 4, and 5 adult donor individuals for 0, 1, 3, 5, 7, 9, 11, and 14 mps are shown, respectively. Different lowercase letters denote mean values that differ significantly (P < 0.05).

Discussion

Motivated by the lack of biological knowledge concerning the mechanisms that regulate SSCs in seasonal breeders, despite the fact that most vertebrates exhibit seasonal reproduction, we sought to thoroughly characterize the biological properties of ASG in adult rainbow trout testes. Based on function, rainbow trout ASG are defined as the most undifferentiated germ cells included among the SSCs. In this species, spermatogenesis shows strict seasonality and proceeds almost synchronously in the whole testis, although some ASG remain in the walls of the testicular lobules throughout the year. This structural characteristic of spermatogenesis afforded a unique opportunity to isolate and analyze ASG from the testes at various spermatogenic stages. Consequently, this study revealed that the stem cell activity of ASG is seasonally regulated during the annual reproductive cycle in rainbow trout.

Widely scattered ASG remained in the trout testes during the spermiation season (1–5 mps), and these cells were designated “residual ASG” [11]. Residual ASG are thought to provide for the production of spermatozoa in the next reproductive cycle [35], suggesting that residual ASG are either true stem cells or highly enriched in stem cells. Previously, we reported that donor ASG incorporated in recipient gonads are capable of both self-renewal and differentiating into functional gametes over a period of several years in the recipient gonads, indicating that transplantable ASG possess stemness [17]. Therefore, we predicted that the residual ASG population would exhibit greater transplantability compared with ASG in the testes during the other spermatogenic stages. In this study, however, it was revealed that the transplantability of residual ASG declines dramatically from the early (0–1 mps) to late (3–5 mps) spermiation stages and remains low until right before the onset of the next reproductive cycle (Figure 11). The results of spermatogonial transplantation assays using donor cells prepared from prepubertal testes that contained only ASG and no germ cells of other stages showed that the transplantability remained constant over the entire experimental period, suggesting that the ability of recipients to accept donor ASG into their gonads is not affected by season. Our data thus indicate that the stem cell activity of the residual ASG declines during the late spermiation stage (3–5 mps). This unexpected result suggests that a novel seasonal mechanism regulates the stem cell activity of ASG in rainbow trout.

Figure 11.

Schematic illustration of biological characteristics of ASG during the annual reproductive cycle in rainbow trout.

We postulated that the proliferation activity of incorporated donor ASG in recipient gonads might also be reflected by qualitative alterations in the SSCs and/or recipient environment. At 60 days after transplantation, we determined the number of colonized ASG in the recipient gonads. Donor ASG derived from prepubertal testes produced similar-sized colonies constantly throughout the experimental period, suggesting that the recipient environment remains constant. However, when the donor ASG were isolated from testes during the late spermiation and regression stages (3–7 mps), the colonized ASG showed lower proliferation activity and did not form large colonies in the recipient gonads (Figure 11). This result suggests that the proliferation ability of SSCs in recipient gonads is downregulated specifically from the late spermiation stage to the regression stage. Generally, adult stem cells maintain a slow cycling state as a mechanism to avoid the accumulation of DNA damage caused by extracellular stressors and to ensure life-long tissue renewal [36, 37]. As the residual ASG population in rainbow trout is an essential germ cell population for sustaining spermatogenesis in the following reproductive season, the residual ASG might remain dormant until the next spermatogenic cycle in order to avoid the accumulation of the DNA mutations associated with cell division. Although the colonized ASG isolated from regression-stage testes did not proliferate in the recipient gonads, they did actively proliferate in the testes (Figure 11), indicating that there might be a unique mechanism that continually activates dormant SSCs in the surrounding testicular environment during the regression stage.

Previously, it was reported that ASG in rainbow trout are morphologically heterogeneous, particularly with regard to nuclear morphology. Loir (1999) subdivided ASG into two types [19]: A1 cells and A2 cells. A1 cells are either single or paired and have a large nucleus with one prominent, dense body. ASG with a morphology similar to A1 cells have been observed in other fish species, and they were often designated SSCs and considered to be in the most undifferentiated state based on morphologic criteria [10, 20, 21, 23, 38]. Because the transplantable ASG have ability to behave as stem cells in the recipient gonads following transplantation, we initially postulated that only A1 cells would have the transplantability. If this hypothesis is correct, the appearance frequency of A1 cells among the overall ASG would correlate directly with the incorporation efficiency of donor ASG in recipient gonads. Contrary to our hypothesis, however, the appearance of A1 cells did not necessarily correspond with the incorporation efficiency of the donor ASG. In the early spermiation stage (0–1 mps), although no A1 cells were found in the testis, donor ASG exhibited a high transplantability and colony-forming ability (Figure 11), indicating that the above-mentioned morphologic criteria are unsuitable as an indicator for identifying SSCs. The appearance of A1 cells among the overall ASG population coincided with an upregulation of ASG proliferation in the testis (Figure 11), raising the possibility that the morphology of A1 cells mirrors that of proliferating ASG. Therefore, it is necessary to identify new morphologic and/or molecular markers capable of distinguishing SSCs from other ASG subtype(s). However, since the conventional morphologic criteria remain an important signature of the cellular state, more detailed analysis is required to determine the functional properties of each type of ASG.

The expression patterns of some of the germline stem cell markers identified in mammals have been analyzed in rainbow trout [23, 24, 28]. One such marker is nanos2, which was identified as a spermatogonial and oogonial stem cell marker in mice and medaka, respectively [39, 40]. Bellaiche et al. recently reported that nanos2 transcripts are expressed in a subset (less than 20%) of ASG in prepubertal testes and that all ASG remain in the spermiated testes, and they hypothesized that the ASG expressing nanos2 were SSCs [23]. Based on the morphologic description by Bellaiche et al. [23], we predicted that nanos2-positive ASG in spermiated testes would be similar to the residual ASG observed during the spermiation season (1–5 mps) in this study. If only nanos2-positive ASG exhibited the ability to behave as SSCs, the incorporation efficiency of residual ASG should be more than five times higher than that of ASG derived from prepubertal testes. However, the results in this study clearly demonstrated that the incorporation efficiency of residual ASG was nearly equal to or lower than that of ASG isolated from prepubertal testes, suggesting that nanos2 expression is not a suitable marker for SSCs in rainbow trout. In this study, however, the suitability of rainbow trout nanos2 as a molecular marker for SSCs in steady-state testes could not be determined. This would be an important task in future studies, as studies in mice showed that undifferentiated spermatogonia behave differently in steady-state testes and during the regeneration process following spermatogonial transplantation [41–43]. Furthermore, we previously reported that a rainbow trout GFRα1 homologue is expressed homogeneously in all ASG, including residual ASG [28], suggesting that this molecule is also unsuitable as a molecular marker for SSCs.

We next hypothesized that there would be a drastic reduction in the relative number of SSCs within the residual ASG populations from the early spermiation stage (1 mps) to late spermiation stage (3 mps), leading to a decrease in the transplantability of donor ASG. In order to test this hypothesis, it is necessary to determine the appearance rate of SSCs among donor ASG; however, there are no specific markers to identify SSCs prospectively. Therefore, we characterized the changes in the total number of ASG in donor testes using immunofluorescence and histologic analyses and examined the feasibility of assessing quantitative changes in SSCs within donor testes. As a result, no significant differences were observed in the appearance frequency of apoptotic ASG throughout the spermiation stage (0–5 mps) (Figure 11). Further, the percentage of apoptotic ASG were too low to decrease the total number of SSCs drastically. In addition, proliferation of the residual ASG was maintained at a low level during this period (Figure 11). These results demonstrate that the residual ASG population represents neither an increase in apoptosis nor a decrease in proliferation, suggesting that a significant change in the number of total ASG would not occur between 1 and 3 mps. This assumption is also supported by the estimated total number of ASG in donor testes; the total number of ASG did not change significantly during the spermiation stage (0–5 mps) (Figure 11). Thus, the decreased transplantability of residual ASG populations might be caused by qualitative changes in each SSC rather than a reduction in the number of SSCs. However, as it is possible that the frequency of SSCs would be extremely low among all ASG, we might not be able to detect a change in the relative number of SSCs based on the methods used in this study. Therefore, in order to further assess this hypothesis, more precise studies will be needed in the future. We found that the transplantability of donor ASG was restored between the regression stage (7 mps) and the onset of spermatogenesis (9 mps) in the next cycle (Figure 11). At that time, no significant difference was observed in the frequency of apoptotic ASG among all donor ASG, but the ASG actively proliferated (Figure 11), suggesting that SSCs preferentially proliferate to restore the transplantability of the overall ASG population.

Several previous studies in teleost species have suggested that an association exists between spermatogonial proliferation and sex steroid hormones [10, 35]. For example, E₂ stimulates the proliferation of SSCs [44], and 11-KT promotes spermatogonial proliferation and subsequent differentiation [45, 46]. Therefore, we predicted that some sex steroid hormones play roles in regulating the proliferation of SSCs. Our analyses showed that when both transplantability and proliferation of ASG in donor testes were upregulated in the regression stage, plasma levels of all of the sex steroid hormones examined were at their lowest levels (Figure 11), suggesting that these sex steroids are not involved in this phenomenon. Miura et al. [44] reported that even 10 pg/ml of E₂ can stimulate cultured eel SSCs to proliferate. As the plasma level of E₂ in the regression stage was approximately 80 pg/ml, it should have been high enough to stimulate the proliferation of SSCs. Even though we could not detect direct correlation of any of the plasma steroid levels and stem cell activity of the ASG, we cannot deny the possibility that local paracrine actions of the steroids are involved in regulation of ASG functions. Therefore, further studies regarding the local paracrine actions of the various steroids and the other growth factors will also be needed.

Previous studies in several teleost species revealed that ASG have the capacity to differentiate into both sperm and eggs in recipient gonads [17, 47]. The establishment of a spermatogonial transplantation system in teleost species opened new avenues for studying the biology of teleost SSCs and contributed enormously to the establishment of methods for the xeno-transplantation of germ cells isolated from valuable fish species or strains, such as endangered species and farmed fish with desirable traits [15]. However, if adult fish exhibiting seasonal reproduction are used as donors, the success rate of transplantation might be altered, depending on the season in which the donor ASG are isolated. Thus, in order to obtain surrogate broodstock that efficiently produce donor-derived gametes, it is important to consider the possibility of seasonality in the stem cell activity of ASG.

Our most important finding is that the qualitative characteristics of rainbow trout ASG, especially their transplantability and colony-forming ability, are seasonally regulated in association with spermatogenic activity during the annual reproductive cycle. We found that the stem cell activity of residual ASG, which were considered as true SSCs based on histological observations, decline dramatically and remain in a state of dormancy for a specific period during the annual reproductive cycle. The results suggest that there is a novel mechanism to regulate the stem cell activity of ASGs in rainbow trout. A previous study using hamsters, which exhibit seasonal breeding, also suggested that SSCs are seasonally regulated in a spermatogenic activity-dependent manner [48]. Therefore, the findings of the present study would provide valuable clues that help elucidate the common mechanisms behind the seasonal regulation of SSC biological properties in a variety of vertebrates.

Supplementary data

Supplementary data are available at BIOLRE online.

Supplementary Table S1. List of antibodies

Author notes

†

Grant support: Grant-in-Aid for Scientific Research on innovative Areas, “Mechanism regulating gamete formation in animals” from The Ministry of Education, Culture, Sports, Science and Technology (MEXT) (25114005) Ocean Resource Use Promotion Technology Development Program from MEXT.