Morphological characterization of spermatogenesis and spermatogonial stem cells in Larimichthys crocea, a seasonal breeding teleost^†

Abstract

Seasonal spermatogenesis in fish is a complex and highly regulated process in which spermatogonial stem cells (SSCs) undergo a series of cellular changes to differentiate into mature sperm. In this study, we systematically described testicular development and identified thirteen different germ cell types throughout the reproductive cycle in large yellow croaker (Larimichthys crocea), a commercially important marine cultured fish in East Asia. Using a set of specific antibodies (VASA, PCNA, DMC1, NANOS2 and GSDF), we developed a high-throughput immunohistochemistry method to identify different types of spermatogenic cells, with a particular focus on distinguishing spermatogonial subtypes. VASA was strongly expressed in all four types of spermatogonia (As, Apr, Adiff, and B) and decreased progressively during spermatogenesis. DMC1 exhibited distinct expression patterns in different spermatocytes subtypes, and GSDF was highly expressed in somatic cells surrounding type A spermatogonia. Particularly, NANOS2 was highly specific to As and Apr spermatogonia, supporting their role as SSC candidates. By morphological observation and co-staining of VASA and PCNA, we found that As spermatogonia exhibited dynamic development characteristics during the annual reproductive cycle. These findings provide a valuable tool for reproductive studies and potential applications in surrogate reproduction through SSCs transplantation in teleost fish.

Graphical Abstract

Open in new tabDownload slide

Introduction

Spermatogenesis is a complex and highly coordinated process by which diploid spermatogonia undergo a series of cellular changes, eventually differentiating into millions of spermatozoa (SZ) [1–3]. This process is driven by spermatogonial stem cells (SSCs), which possess the dual ability to self-renewal and differentiation [4–6]. Fish, as the most diverse group of vertebrates, display unique and varied patterns of spermatogenesis [7–9]. In most teleost species, spermatogenesis is both seasonal and cyclic, characterized by distinct morphological and cellular changes throughout the reproductive cycle [1]. Although there are extensive morphological examinations on fish spermatogenesis in the literature [10–14], comprehensive studies that describe the entire cytological process, particularly focusing on the molecular and cellular events involved in SSCs development in seasonal breeding fish, remain limited.

In teleost, undifferentiated type A spermatogonia (1-cell) are well-established as the SSCs pool [1, 15]. In the germ cell transplantation, it is considered that only SSCs can efficiently colonize and differentiate into mature gametes in the host fish [16, 17]. Therefore, characterizing different spermatogenic cells, especially the cellular and molecular events of SSCs is critical. However, it is difficult to identify the various generations of spermatogonia based solely on morphological observations, largely due to the lack of a unified classification system and specific cellular markers. Several markers, such as VASA, PCNA (proliferating cell nuclear antigen), DMC1 (DNA meiotic recombinase 1), NANOS2 (nanos homolog 2), and GSDF (gonadal soma derived factor), have been used to identify fish spermatogenesis. The Vasa gene, also known as Ddx4, is an ATP-dependent RNA helicase in the DEAD (Asp-Glu-Ala-Asp)-box protein family. Numerous studies have shown that Vasa is exclusively expressed in germline cells in fish [18, 19]. NANOS2, a type of NANOS, has been regarded a specific marker for germline stem cells in several fish species, such as Danio rerio [20], Oryzias latipes [21], and Oncorhynchus mykiss [22]. The cell mitotic proliferation marker PCNA is an essential component of the DNA replication machinery [23, 24], while the expression of DMC1, an essential element of the synaptonemal complex, indicates meiotic differentiation of spermatogenic cell [25]. Gsdf, a member of the TGF-β super-family, is often used as a Sertoli cells marker in teleost [26]. Although these markers have been selectively used in various studies, an integrated and comprehensive analysis of these markers by immunohistochemistry during spermatogenesis in seasonal breeding fish remains to be documented.

The large yellow croaker (L. crocea, family of Sciaenidae) is an important commercial, seasonally breeding marine fish cultured in East Asian [27]. In China, the production of L. crocea ranks first in marine fish aquaculture, with output exceeding 280 997 tons in 2023 [28, 29]. Understanding the fundamental processes of spermatogenesis, particularly the development and regulation of SSCs in this species, is of great importance not only for enhancing its reproductive research, but also for germ cell-based breeding such as surrogate reproduction. This study aims to systematically describe testicular development and spermatogenesis throughout the reproductive cycle of L. crocea. By employing a set of specific antibodies (VASA, PCNA, DMC1, NANOS2, and GSDF), we developed a high-throughput immunohistochemistry method to identify different spermatogenic cells, especially the SSCs. Our findings provide valuable insights into the cellular and molecular events governing spermatogenesis in seasonal breeding fish, which can facilitate the development of advanced breeding techniques for L. crocea.

Materials and methods

Samples

All animal experiments conducted in this study were approved by the Institutional Animal Care and Use Committee of Zhejiang Marine Fisheries Research Institute and were in accordance with the animal ethics guideline (2023). All the fish used in the study were obtained from the research station of Zhejiang Marine Fisheries Research Institute (Xixuan Island). Throughout the testicular developmental process, we randomly collected 115 males from 120 days post-hatching (dph), with approximately 11 individuals sampled at each time point (120, 150, 180, 210, 240, 270, 300, 330, and 360 dph, 1.5 and 2 years-old). The developmental stages of each testis were examined and identified from stage I to stage VI according to the definitions reported previously [17, 30].

Individual fish were euthanized using tricaine methane sulfonate (MS-222; MedChemExpress, Shanghai, China). Testicular tissues were quickly dissected and then fixed in 2.5% glutaraldehyde fixative (Biosharp, Anhui, China), Bouin’s fixative solution (Phygene, Fujian, China) and 4% paraformaldehyde fixative (PFA; Biosharp, Anhui, China) for histological and immunohistochemical analysis.

Histological analysis

The testicular tissues were fixed in Bouin’s fixative solution for 24 h, dehydrated by ethanol with gradient concentration, embedded in paraffin and cut into 3–5 μm-thick sections. The sections were stained with hematoxylin and eosin (H&E), Alcian blue and nuclear fast red (Alcian Blue; Beyotime, Shanghai, China), Masson’s trichrome (Masson trichrome; Beyotime), and Periodic Acid-Schiff staining (PAS; Beyotime). All stained sections were examined and photographed using an Axio Imager A2 microscope and Axiocam 506 digital camera (Zeiss, Oberkochen, Germany). Germ cell nuclear morphology and diameter were determined using Image-Pro Plus 6 software, with n ≥ 30 cells measured per cell type. Additionally, the area composition of seminiferous lobules (SL), including spermatogonia (SG), leptotene/zygotene stage spermatocytes (L/Z), pachytene/diplotene stage spermatocytes (P/D), secondary spermatocytes (SS), spermatids (ST), SZ, and lobular lumen (LO) were measured using Image-Pro Plus 6 software, with n ≥ 5 SL measured per stage.

Transmission electron microscopy assay

The testicular tissues were minced and fixed in 2.5% glutaraldehyde fixative (Biosharp, Anhui, China) at 4°C for 4 h. Subsequently, the tissues were washed with1× PBS for three times (10 min each) and postfixed by 2% osmium tetroxide (OsO₄) in 1× PBS over 4 h at 4°C. Then the samples were dehydrated in ascending acetone concentrations (30, 50, 70, 90, 95, and 100%) for 15 min each and embedded in resin (Epon 812, Germany). Ultrathin sections (0.07–0.08 μm) were sliced using a UCT ultramicrotome (Leica Microsystems, Wetzlar, Germany). The sections were stained with 3% aqueous uranyl acetate and 1% lead citrate and examined using a transmission electron microscope Zeiss Libra 120 (Zeiss). All the ultrastructure of different spermatogenic cell types was examined by transmission electron microscopy assay (TEM). According to the published papers [12, 13, 30], all the germ cell types were distinguished by the morphology.

Immunohistochemical analysis

The testicular tissues were fixed in PFA overnight at 4°C, dehydrated in graded ethanol, and embedded in paraffin wax. Paraffin sections (5 μm) were deparaffinized and rehydrated by ethanol with gradient concentrations and immersed in 1× universal powerful antigen retrieval solution (Biosharp) for 30 min at a sub-boiling temperature (95°C) for antigen retrieval. After blocking for 2.5 h in blocking solution (10% goat serum, diluted with 1× PBS) at room temperature, the sections were incubated with primary antibodies overnight at 4°C, followed by washing five times with 1× PBS (10 min each), incubation with secondary antibodies for 1 h at 37°C, and further washing with 1× PBS. For immunofluorescence, the primary antibodies used in this study included anti-VASA (1:1,000; Abcam, Shanghai, China), anti-PNCA (1:1,000; Abcam), anti-DMC1 (1:100; ABclonal). Primary antibodies were detected using goat anti-rabbit IgG H&L (Alexa Fluor® 488) or goat anti-mouse IgG H&L (Alexa Fluor® 488). For immunohistochemistry, the primary antibodies used in this study included anti-NANOS2 (1:200; Sangong Biotech, Shanghai, China) and anti-GSDF (1:200). Primary antibodies were detected using HRP-conjugated goat anti-rabbit IgG (Sangong Biotech). Chromogenic staining was performed using an enhanced DAB chromogenic kit (20×, Solarbio). Subsequently, nuclei were stained with hematoxylin, followed by two 5-min washes in absolute ethanol and two 10-min washes in xylene. For immunofluorescence co-staining, the following antibody combinations were used: VASA and PCNA, VASA and NANOS2, and NANOS2 and GSDF. The first antibody was conjugated with Alexa Fluor® 488, and the second conjugated with 568. The detailed information of the antibodies is provided in Supplementary Table 1. Nuclei were stained with DAPI (10 μg/mL, Biosharp) at room temperature for 5 min, and then washed with 1 × PBS (three times, 5 min each time). All sections were examined and photographed using an Axio Imager A2 microscope and Axiocam 506 digital camera (Zeiss, Oberkochen, Germany).

Fluorescence and immunohistochemical intensity analysis

To measure the fluorescence intensity of VASA, and the immunohistochemical intensity of NANOS2 and GSDF, 20 images were acquired. Intensity measurements were performed using ImageJ2 software. At least three images or regions from three independent testicular samples were analyzed. Within each image or region, the relative intensity of VASA was measured for each spermatogenic cell type, and the relative intensity of NANOS2 and GSDF was measured for three type A SG subtypes.

Statistical analysis

All data are presented as mean ± standard deviation (SD) and analyzed using GraphPad Prism 10.0 software. The cell and nuclear diameters, fluorescence intensity of VASA, and immunohistochemical intensity of NANOS2 and GSDF were analyzed using one-way ANOVA followed by Tukey’s multiple comparisons test. Two-way ANOVA by the Sidak’s multiple comparisons test were performed for single comparisons between two groups in Figure 2C.

Results

Cellular characteristics of spermatogenesis in L. crocea

Spermatogenic cells in L. crocea were divided into thirteen types based on their morphological criteria: four types of SG, five types of spermatocytes, three types of ST and one type of SZ. The detailed morphological changes, including cell and nuclear diameters are shown in Figure 1, Figure 2A, and Supplementary Figure 1.

The four types of SG were identified as type A single (As), type A paired (Apr), type A differentiated (Adiff) and type B (B). The As SG was completely enveloped as a single cell by Sertoli cells and is the largest germ cells (diameter, 13.63 ± 2.15 μm), with a large nucleus (diameter, 8.33 ± 1.53 μm) containing small dense granules nuage (Ng) and mitochondria (Mi) in the cytoplasm. The Apr SG was paired (2-cell) within the cyst and was smaller (10.65 ± 1.86 μm in cell and 6.72 ± 1.36 μm in nuclear diameter) than the As. The nuage and Mi in Apr SG were denser and smaller. The Adiff SG was grouped (4, 8-cell) within a cyst. The nucleus of Adiff SG was smaller (diameter, 5.85 ± 0.89 μm), and the nucleoli was obvious. In type B SG, the number of germ cells inside the cyst increased (n > 16). The nucleus was round (diameter, 4.52 ± 0.42 μm) and dense, and the chromatin (Cm) reached maximum density.

Spermatocytes in different phases of meiosis were identified by nuclear characteristics, including size, Cm condensation, and meiosis of the chromosome. L/Z spermatocytes had round nucleus (diameter, 4.13 ± 0.38 μm) compared to type B SG, showing clear Cm with synaptonemal complex (Sy). The nucleus of pachytene (P) spermatocytes was denser (diameter, 3.51 ± 0.34 μm) and contained heterochromatin (Hc) as bold lines extending from the periphery to the center of the nucleus. Diplotene (D) spermatocyte (3.30 ± 0.23 μm in nuclei diameter) was typically found together with metaphase (M) figures. In this cell type, the Cm reached the maximum degree of condensation. M spermatocytes were rare because they quickly entered the secondary meiosis. The nuclear of SS was round and smaller with a dense Cm (diameter, 2.40 ± 0.19 μm).

Spermiogenesis was characterized by a striking reduction in cellular volume and formation of flagellum. Three types of ST were divided: spermatid I (ST1), spermatid II (ST2), and spermatid III (ST3). The nucleus of ST1 was smaller (diameter, 1.78 ± 0.13 μm) than that of SS. In the ST2, nucleus became condensed (diameter, 1.68 ± 0.13 μm) with Mi in the cytoplasm. ST3 was present in the LO, with flagellum (F) formed and nucleus recessed (diameter, 1.59 ± 0.13 μm). SZ were mostly found in the LO and were characterized by their elongated oval nuclei and single flagellum. The nuclear diameter of the SZ (1.33 ± 0.14 μm) was significantly reduced. The midpiece was short and contained spherical and ovoid Mi. The flagellum was enveloped by the flagellar plasma membrane, which extended into the cytoplasm through cytoplasmic canals around the midpiece.

Phases of testicular development during annual reproductive cycle

The testis of L. crocea is lobular, which is typical for the cystic type of spermatogenesis in teleost. The testicular lobule, delineated by a basement membrane, houses the germinal epithelium where germ cells in each cyst develop synchronously. The testicular morphology stained with H&E, Alcian Blue, Masson trichome and PAS, is shown in Supplementary Figs. 2 and 3. The testes were categorized into six stages based on the main germ cell type: Spermatogonia stage (I), early spermatogenesis stage (II), late spermatogenesis stage (III), early spermiation stage (IV), late spermiation stage (V), and regression stage (VI). Additionally, the area composition of SL, including SG, L/Z, P/D, SS, ST, SZ, and LO was measured and shown in Figure 2B.

Spermatogonia stage (I). As shown in Supplementary Figure 2a-d and a’-d’, two types of SG were observed in this stage: As and Apr. Meanwhile, a large number of somatic cells (SC) were present. The interstitial tissue (It), stained blue or green, was located in the dorsal domain of testis in this stage (Supplemental Figure 2b and c).

Early spermatogenesis stage (II). Early spermatocytes (L/Z and P/D) were observed in this stage (Supplemental Figure 2e-h and e’-h’). The interstitial tissue extended inward from the dorsal to ventral regions. Meanwhile, the SL separated by interstitial tissue were present (Supplemental Figure 2f’ and g’). A small LO appeared in the center of the SL (Supplemental Figure 2f and g). The area proportion of the LO (4.81%) and spermatocyte (total for 12.5%, L/Z for 4.85%, P/D for 6.31% and SS for 1.35%) within the SL increased (Figure 2B). Type B SG were observed within a cyst (Supplementary Figure 2g’). Besides, a small number of ST were observed (Supplementary Figure 2h’).

Late spermatogenesis stage (III). As shown in Supplementary Figure 2i-l and i’-l’, various types of germ cell cysts and some single SG were distributed along the germinal epithelium of the testicular lobules. Blood vessels were frequently observed between the lobules. The area of the LO within testicular lobules increased to 6.00%. Spermatogenic cells at this stage were mainly composed of SG (area accounts for 51.43%), primary spermatocytes (25.35%), SS (3.12%), and ST (14.10%) (Figure 2B). The interstitial tissue continued to extend and separated SL from the gonadal edge (Supplementary Figure 2j-k and j’-k’).

Early spermiation stage (IV). The main characteristic of this stage was the appearance of numerous ST within the LO (Supplementary Figure 3a-d and a’-d’). Spermatogenic cells were mainly composed of SG (26.60%), L/Z (3.97%), P/D (29.18%), SS (8.19%) and ST (22.32%) (Figure 2B).

Late spermiation stage (V). In the maturation and spawning-capable stage, the LO was filled with released SZ (Supplementary Figure 3e-h and e’-h’). Meanwhile, a small number of SG (area accounts for 16.72%), spermatocytes (14.01%), and ST (4.78%) were also observed (Figure 2B). The combined area of SZ and lobule lumen accounted for 64.49% of the total seminiferous lobule area (Figure 2B).

Regression stage (VI). As shown in Supplementary Figure 3i-l and i’-l’, the germ cells were mainly composed of SG (71.00%). The size of testicular lobule decreased. The area of primary spermatocytes accounted for 14.43%, with L/Z and P/D contributing 4.03% and 10.40%, respectively (Figure 2B). Meanwhile, the interstitial tissue between spermatogenic cysts became thicken (Supplementary Figure 3j’ and k’).

Furthermore, we quantified the percentage of different developmental stages at each sampling time (Supplementary Figure 4). Testicular development exhibited a distinct annual development pattern. We also measured the cell and nuclear diameters of As and Apr SG across stages I–VI (Supplementary Figure 5). Notably, the cell and nuclear diameters of As and Apr SG gradually decreased from stage I to V, then increased at stage VI. Further analysis of the size differences between As and Apr SG at stages I and VI (Figure 2C, D) revealed that the diameters of both cell types were significantly larger at stage I compared to stage VI, indicating morphological characteristics during testicular development.

Identification of various spermatogenic cells using a set of specific antibodies

Progressively decreased expression of VASA during spermatogenesis

Immunofluorescence analysis using an anti-VASA antibody revealed that the presence of VASA signals in the testes at six stages (from I to VI) (Supplementary Figure 6). In stages I and II, VASA was obviously expressed in all spermatogenic cells, including type A and B SG and early L/Z spermatocytes (Supplementary Figure 6b, c, f, g). During stages III, IV and V, VASA was highly expressed in SG at the edge of the testis (Supplementary Figure 6j, k, n, o, p, r, s, t). In stage VI, VASA was abundant in the SG surrounding the SL (Supplementary Figure 6v, w, x).

As shown in Figure 3, different types of spermatogenic cell were visualized by the anti-VASA antibody, and the fluorescence intensity is shown in Supplementary Figure 7A. VASA was highly expressed in the cytoplasm of type A SG, including As, Apr and Adiff SG (Figure 3a-c, a’-c’). However, VASA signals were significantly decreased in type B SG (Figure 3d, d’). As type B SG differentiated into L/Z and P spermatocytes, VASA expression decreased dramatically (Figure 3e, f, e’, f’). Both D and SS showed weak expression of VASA (Figure 3g, h, g’, h’), while ST (Figure 3i-k, i’-k’) and SZ (Figure 3l, l’) showed no expression of VASA. In conclusion, the expression level of VASA decreases along spermatogenesis in L. crocea.

High expression of PCNA in mitosis and meiosis of spermatogenesis

In stage I, PCNA was highly expressed in the nuclei of SG (Supplementary Figure 8a-d). In stage II, PCNA was highly expressed in numerous SG and early spermatocytes including L/Z and P spermatocytes (Supplementary Figure 8e-h). In stages III and IV, PCNA was prominently detected in a large number of spermatocytes forming cysts around the SL (Supplementary Figure 8i-p). In stage V, the PCNA protein was expressed in some spermatocyte cysts surrounding the SL, but no signals were detected in ST or SZ (Supplementary Figure 8q-t). In stage VI, PCNA signals increased and were detected in spermatocytes around the SL (Supplementary Figure 8u-x).

As shown in Figure 4, different types of spermatogenic cell were visualized using the anti-PCNA antibody. PCNA was highly expressed in the nuclei of SG, including As, Apr, Adiff and type B SG (Figure 4a-d, a’-d’). However, as spermatocytes developed, PCNA expression decreased. Both P and D spermatocytes showed weak expression of PCNA (Figure 4f, f’, g, g’), while SS (Figure 4h, h’), ST (Figure 4i-k, i’-k’) and SZ (Figure 4l, l’), showed no expression of PCNA.

High expression of DMC1 in meiotic spermatocytes

In stages I and II, no DMC1 signals were observed in spermatogenic cells, although weak signals were detected in a few SG located at the periphery of testis (Supplementary Figure 9a-h). In stages III, IV and V, DMC1 signals became pronounced, with signals notably concentrated in the spermatocyte cysts within the SL (Supplementary Figure 9i -t). In stage VI, DMC1 signals were no longer detected (Supplementary Figure 9u-x).

As shown in Figure 5, different spermatogenic cell types were visualized using the anti-DMC1 antibody. Although weak DMC1 signals were present in the cytoplasm of As and Apr SG (Figure 5a-b, a’-b’), no signals were observed in Adiff and type B SG (Figure 5c-d, c’-d’). As SG differentiated into spermatocytes, DMC1 exhibited distinct expression patterns. In L/Z spermatocytes, DMC1 showed an obvious polar expression in one side of the nucleus (Figure 5e, e’). In P spermatocytes, the signals expanded across the nucleus in a striped pattern until occupying the whole nucleus (Figure 5f, f’). In D spermatocytes, the signals were dramatically reduced (Figure 5g, g’). No DMC1 signals were detected in SS, ST, and SZ (Figure 5h-l, h’-l’).

High expression of NANOS2 in type A spermatogonia

From stage I to IV, NANOS2 signals were prominently expressed in type A SG, including As, Apr and Adiff SG (Figure 6a-d, a’-d’). In stage V, NANOS2 expression was less pronounced, with only scattered signals in type A SG (Figure 6e, e’). However, in stage VI, NANOS2 signals increased significantly and were widely expressed in SG surrounding the SL (Figure 6f, f’).

Throughout spermatogenesis, NANOS2 signals were exclusively detected in the cytoplasm of type A SG, including As, Apr and Adiff SG. The signal intensity is shown in Supplementary Figure 7B. The strongest NANOS2 expression was observed in As SG, with gradual decrease in Apr and Adiff SG (Figure 6g-i, Supplementary Figure 7B). No signals were detected in type B SG, spermatocytes, ST, and SZ (Figure 6j-r).

High expression of GSDF in somatic cells surrounding type A spermatogonia

In stages I, II and VI, GSDF signals were strongly expressed in the SC surrounding germ cells (Figure 7a-b, a’-b’, f, f’). In stage III and IV, GSDF signals were also observed in the SC surrounding germ cell cysts (Figure 7c-d, c’-d’). In stage V, GSDF expression decreased, with only scattered signals detected in SC surrounding type A SG (Figure 7e, e’).

Throughout spermatogenesis, GSDF exhibited strong expression in SC surrounding type A SG, including As, Apr and Adiff SG (Figure 7g-i). The signal intensity is shown in Supplementary Figure 7C. The GSDF signal was most prominent in SC surrounding As SG and gradually diminished in those surrounding Apr and Adiff SG. As SG differentiated into spermatocytes, weaker GSDF signals were observed in SC surrounding the germ cell cysts (Figure 7j-n).

Characterization of spermatogonial subtypes using co-staining methods

Characterization of different spermatogenic cells by co-staining of VASA and PCNA

By co-staining with VASA and PCNA, the mitotic activities of different spermatogenic cell types were assessed. The subtypes of spermatogenic cells were identified based on the criteria of nuclear morphology and VASA expression, as previously described. As to the expression of PCNA in SG cells, some As SG showed high levels of PCNA expression in nucleus (Figure 8A a-a3), while others showed low level of expression (Figure 8A b-b3), suggesting that the former As SG were at mitotic cycle and the latter were at quiescent state. Notably, As SG at mitotic cycle were mainly detected in stage I, whereas quiescent As SG were predominantly observed in stage VI (Supplementary Figure 10). In Apr, Adiff and type B SG, PCNA was consistently highly expressed (Figure 8A c-c3, d-d3, e-e3). Although L/Z spermatocytes also showed high levels of PCNA expression (Figure 8A f-f3), both P and D spermatocytes showed weak expression of PCNA (Figure 8A g-g3 and h-h3), and SS showed no detectable PCNA expression (Figure 8A i-i3).

Characterization of different subtypes of spermatogonia by co-staining of VASA and NANOS2

We performed co-staining of VASA and NANOS2 to characterize different subtypes of SG (Figure 8B). As shown in Figure 8B a-a3, both VASA and NANOS2 were strongly expressed in the cytoplasm of As SG. In Apr and Adiff SG, the levels of VASA protein remained high, while NANOS2 expression was reduced (Figure 8B b-b3 and c-c3). In type B SG, the expression of both VASA and NANOS2 were extremely low (Figure 8B d-d3).

Characterization of different subtypes of type A spermatogonia by co-staining of GSDF and NANOS2

We also performed co-staining of GSDF and NANOS2 to characterize different subtypes of type A SG (Figure 8C). As shown in Figure 8C a-a3, NANOS2 protein was highly expressed in the cytoplasm of As SG, while GSDF was expressed in the surrounding SC. As the As SG developed into Apr and Adiff SG, GSDF expression weakened in SC around the germ cell cysts, and NANOS2 expression became less pronounced in these more differentiated SG (Figure 8C b-b3, c-c3). Finally, we proposed new criteria for identifying different spermatogenic cell types, especially the subtypes of SG, based on the expression patterns of VASA, NANOS2, PCNA, DMC1 and GSDF as well as detailed cell morphology (Figure 9).

Discussion

Implications for understanding seasonal spermatogenesis and SSC regulation

Spermatogenesis is a complex, highly regulated process that involves the proliferation and differentiation of SG into mature sperm [1, 2, 30]. Fish, particularly seasonal breeders, exhibit unique patterns of spermatogenesis with distinct cellular and molecular characteristics [9, 31–33]. Even though morphological examination has been vastly studied in many fish species, there is still debate on the identification of the true SSCs. The objective of our study was to take advantage of the organization of spermatogenesis in the L. crocea to identify the SSCs on morphological analyses and immunohistochemical characterization in a seasonal breeding fish species.

In this study, we firstly systematically described the cytological characteristics of spermatogenesis and identified different types of germ cells based on microstructure and ultrastructural observations. Additionally, we utilized specific antibodies to identify and characterize spermatogenic cells, especially distinguish distinct spermatogonial subtypes. Furthermore, we analyzed the specific localization of distinct spermatogonial subtypes at different stages of the reproductive cycle. To our knowledge, this is the first study to comprehensively characterize SG l subtypes across different testicular development stages in seasonal breeding fish. These findings provide new insights into SSCs in vertebrates, enhancing our understanding of SSC behavior in seasonal breeders. Moreover, our results lay a foundation for future studies on germ cell-based breeding techniques, such as surrogate reproduction, which holds significant potential for the conservation and commercial breeding of marine species.

Characterization of spermatogenic cells and testicular development stages

To characterize different spermatogenic cells, histological analysis has been widely used to visualize spermatogenesis in many fishes, such as D. rerio [10], Acrossocheilus fasciatus [11], Acipenser naccarii [34], Thymallus thymallus [13], Pampus argenteus [30], and Scophthalmus maximus [35]. In this study, the germ cells were classified into 13 types based on cell and nuclear morphological characteristics: four types of SG (As, Apr, Adiff, and B), five types of spermatocytes (L/Z, P, D, M and SS), three types of ST (ST1, ST2, and ST3) and SZ. Generally, in teleost, although the process of spermatogenesis is conserved [1], the types of SG is quite variable. To standardize nomenclature and facilitate comparisons with laboratory rodents (mice, rats and hamsters), SG have been morphologically classified into type A undifferentiated, Adiff, and type B SG in well studied fish species, such as D. rerio [36], S. maximus [37], Paralicthys olivaceus [38], and Sebastes schlegelii [39]. In this study, based on their morphology, single (As) and paired (Apr) type A SG were further distinguished. Compared with other fish, the nuclear size from Adiff to B was relatively similar, indicating that the process of spermatogenesis is conserved in teleosts [1]. However, the nuclei size of As and Apr was bigger than that observed in continuous reproductive species like D. rerio [36] and other seasonal breeding fish species, such as P. olivaceus [38], S. schlegelii [39], and P. argenteus [30]. This suggests that the size of SG may be genetically determined. The morphology of different germ cell types is generally consistent with that described in most examined teleost species [40].

Our study also categorized testicular development into six stages based on cellular compositions, which aligns with the seasonal and cyclic characteristics of spermatogenesis observed in other fish species [30, 33, 41]. In experimental fish models, such as D. rerio, spermatogenesis does not exhibit apparent seasonal variations [36], with different germ cell types present simultaneously and continuously in the testes. In contrast, in seasonal breeding teleosts, there are no significant differences in histological testis development, with the testicular development categorized into six stages based on the primary cellular compositions [30, 33, 37]. In this study, we comparatively evaluated the proportional area of spermatogenic cysts in six testicular developmental stages. Our findings are consistent with single-cell transcriptomic studies in this species [42], further supporting the histological observations and the classification of testicular development in seasonal breeding teleost.

Furthermore, we comparatively analyzed the cellular morphology of As and Apr SG at different testicular development stages. It was worth noting that the cell and nuclear diameters of both As and Apr SG in stage I were significantly larger than those in stage VI, suggesting distinct morphological characteristics of SG at different stages of testicular development. To our knowledge, this is the first study to identify different morphological traits in type A SG (1, 2-cell) across testicular development stages in seasonal breeding fish, providing new insights into SSCs in vertebrates. This observation may reflect the proliferative status of SG at different stages, where stage I likely representing a phase of heightened proliferation and cellular activity, leading to larger cell sizes.

Identification and characterization of spermatogenic cell subtypes using specific antibodies

The Vasa gene was the first specific marker identified for primordial germ cells in D. rerio [43]. Subsequently, Vasa homologs have been identified in many teleost species, including O. latipes [44] and O. mykiss [45]. In D. rerio, Vasa is specifically expressed in SG. In this study, however, high VASA expression was observed in SG, with a progressive decrease throughout spermatogenesis. This pattern aligns with findings in other fish species, such as Nibea albiflora [31] and Paralichthys dentatus [46]. The decreased VASA expression in differentiated germ cells suggests its role in confined to the early stages of germ cell development.

PCNA, an essential component of the DNA replication machinery, is a well-established marker for mitotic proliferation [24]. As a cell cycle-dependent protein, PCNA serves as a molecular marker to distinguish various phases of the cell cycle in proliferating germ cells [47]. Our immunohistochemistry results revealed that PCNA was present in the nuclei of germ cells from the SG l stages through to the spermatocyte stages of leptotene-zygotene and pachytene during prophase I. This expression pattern of PCNA during spermatogenesis is conserved across vertebrates [36].

Dmc1 (disrupted meiotic cDNA) is a conserved and functionally specific gene, which was found encoding a protein required for homologous chromosome synapsis during the process of meiosis from yeast to mammals [48]. In teleost, Dmc1 homologs have been identified and are found specially expressed in meiotic spermatocytes [49]. In this study, we found that DMC1 was specifically expressed in meiotic spermatocytes, including L/Z, P and D, indicating that DMC1 was a reliable marker for identifying meiotic cells. The polarization of DMC1 on one side of the L/Z nucleus could be explained by its role in DNA repair. In addition, we found weak expression of DMC1 in the cytoplasm of As SG, which was an unexpected observation. This weak expression could be the result of non-specific signal detected by the antibody, which was directed against the C-terminal region of zebrafish DMC1.

NANOS, an RNA binding protein, is confirmed to be essential in germ cell development in vertebrates. NANOS2, a type of NANOS, has been regarded as germline stem cells, GSCs-specific marker in fish, including D. rerio [20, 50, 51] and O. latipes [21]. Studies in D. rerio have demonstrated that the NANOS2 is required for the maintenance and survival of GSCs [20, 50, 51]. Our results showed that NANOS2 was specifically expressed in isolated (As) or two associated (Apr) SG in immature and maturing testis. This expression pattern aligns with the findings within seasonal breeding teleost O. mykiss [22], suggesting both the As and Apr SG are likely SSCs or possess the potential to achieve stemness under certain condition in L. crocea. In contrast, studies of the fish Oreochromis niloticus and P. olivaceus showed expression of NANOS2 from isolated A SG to cysts of eight differentiated cells [5, 21, 22, 52]. It is possible that the SSC generations are genetically determined in teleost.

GSDF, Gonadal soma-derived factor, a member of the TGF-β super-family, is often used in researching sex differentiation in teleosts [53, 54]. In O. niloticus [55], Halichoeres trimaculatus [56] and P. olivaceus [57], Gsdf had much higher expression in testes than in ovaries. In addition, GSDF has been considered as a key regulator of SG proliferation in some teleost species, such as O. mykiss [26], D. rerio [58] and O. niloticus [55]. Previous studies have showed that the location of Gsdf mRNA and protein specifically expressed in the SC surrounding SG in L. crocea [42] and P. olivaceus [59]. In this study, we also found that GSDF was specifically localized in Sertoli cells surrounding type A SG, underscoring its importance in maintaining the spermatogonial niche.

Interestingly, two subtypes of As SG were identified based on PCNA expression: mitotically active (As-1) and quiescent (As-2). Meanwhile, we also found that the As-1 SG were predominantly detected in testes in stage I, while As-2 SG were mainly found in testes at stage VI, indicating that As SG exhibit distinct proliferation rate in different developmental stages. In S. schlegelii, single cells sequencing revealed that undifferentiated SG, which had low expression of proliferation genes, were abundant in regressed testes [60]. This suggests that As-2 cells might serve as progenitors of SG. However, in O. mykiss, the transplantation efficiency and colony-forming ability of SG from regressed testes were lower compared with those in testes at the stage of spermatogenesis when using the spermatogonial transplantation method [17]. These findings highlight the complexity of SSCs stemness. This dynamic proliferative capacity likely plays a key role in determining the transplantation success and functional integrity of SG, particularly in seasonal breeding species. Moreover, we also detected NANOS2 in all the As and Apr SG found along the seminiferous tubule wall in degenerated testis. We propose that those NANOS2-positive SG are the putative SSCs, remaining in the tubule when spermatogenesis is arrested. Nevertheless, more solid conclusion should be concluded through SSC transplantation approach in the future.

Conclusions

In conclusion, this study systematically investigated testicular development and the complete cytological development process of male germ cells in the seasonal breeding fish L. crocea. We developed a simple and effective method to identify different spermatogenic cells, with a particular focus on distinguishing spermatogonial subtypes using a set of specific antibodies. Our findings indicate that the As and Apr SG (1, 2-cell) are potential SSCs, characterized by high levels of NANOS2 expression. Furthermore, As and Apr SG exhibit different morphological and molecular characteristics in different testicular development stages. These results provide a valuable framework for studying spermatogenesis and suggest potential application for surrogate reproduction through SG l transplantation in teleost fish.

Author contributions

YY, HY, and DX (Designed the research), YN, YY, and LM (Performed the experiment and wrote the manuscript), WH (Collected the samples), and YY and DX (Revised the manuscript).

Conflict of Interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.

Ethics approval and consent to participate

The study was approved by the Institutional Animal Care and Use Committee of Zhejiang Marine Fisheries Research Institute.

Consent for publication

Not applicable.

References

Author notes