Deciphering dynamic interactions between spermatozoa and the ovarian microenvironment through integrated multi-omics approaches in viviparous Sebastes schlegelii Free

The reproductive mode of an animal is undoubtedly closely related to its habitats and is a manifestation of the long-term adaptation of a species to its environment (Bush et al., 2016; Garant et al., 2007; Lodé, 2012). It impacts the health, survival and species continuity of living organisms. Common reproductive modes include oviparity with fertilization inside or outside the mother's body, ovoviviparity and viviparity. (Schindler and Hamlett, 2010). The reproductive process of viviparity is similar to that of ovoviviparity, except that the nutrients required for embryonic development are transferred from the maternal body through special nutrient transfer organs (Hamlett and Hysell, 1998; Uribe et al., 2014; Kwan et al., 2015), and the yolk content of oviparous species is usually higher than that of viviparous species, such that the yolk alone provides sufficient nutrition for embryonic development, whereas viviparous species are dependent on maternal nutrient supplies besides yolk resources (Lodé, 2012). In species with internal fertilization, asynchronous spermatogenesis and oogenesis require the storage of spermatozoa in the female genital tract over a certain period of time. Thus, females of some species have evolved organs specifically for storing spermatozoa, such as the spermatozoa storage tubes in birds and frogs (Holt, 2011), albumin-secreting glands in reptiles (Pearse and Avise, 2001), oviduct, uterine glands and uterine adnexa in therian and marsupials (Birkhead and Moller, 1993; Hood and Smith, 1989), and the gonoduct, a caudal portion of the ovary in teleosts such as black rockfish (Sebastes schlegelii) (Potter and Kramer, 2000; Storrie et al., 2008).

During the storage process, spermatozoa can be further modified and activated by substances inside the female genital tract (Saint-Dizier et al., 2020). In mammals, spermatozoa undergo capacitation in the oviduct (Martinez et al., 2020), and interact with dynamic and region-specific maternal components, including soluble proteins, extracellular vesicles and epithelial cells inside the female tract, for a few days before fertilization (Saint-Dizier et al., 2020). This interaction changes the state of spermatozoa and helps them to resist the maternal immune response (Rodriguez-Martinez, 2007). However, the molecular mechanism underlying the interaction between spermatozoa and the ovarian microenvironment remains largely elusive, although several substances are likely essential for the final maturation and activation of spermatozoa, such as N-acetyl-glucosamine or N-acetyl-galactosamine (Bergmann et al., 2012), β-amino acids, bicarbonate ion, progesterone and oviductins (Boatman, 1997). In fruit flies, the molecular ‘hand-off’ from males to females was demonstrated by semi-quantitative proteomics using sex-specific isotopic labeling, which show that female-derived proteins constitute one-fifth of the post-mating spermatozoa proteome (McCullough et al., 2022).

There are more than 30,000 fish species, among which over 500 species are ovoviviparous or viviparous (Smith and Wootton, 2014). Black rockfish, a viviparous teleost, displays long-term spermatozoa storage in the ovary (Niu et al., 2021). The male and female fish usually mate in late November, and the spermatozoa are stored in the ovary before fertilization with mature oocytes in April (He et al., 2019). Thus, the black rockfish is an appropriate model for exploring the interaction between spermatozoa and the ovarian microenvironment in terms of the long period of storage between mating and fertilization. In this study, transcriptomic, proteomic and metabolomic changes in the spermatozoa and ovaries during spermatozoa storage were investigated. The most significant changes were detected in spermatozoa shortly after their entry into the ovary. During spermatozoa storage, ovarian complement-mediated immunoregulation was induced by spermatozoa invasion but repressed at the intermediate stage (2 months after mating). Cellular junction pathways and communication between ovarian cells and spermatozoa were enhanced at a late stage (3.5 months after mating) in the ovary. These results provided an overall view of the dynamic molecular features in spermatozoa and ovarian tissues during spermatozoa storage in black rockfish. They contribute to a better understanding of the mechanism underlying spermatozoa-ovarian microenvironment interaction in viviparous fish.

In order to explore the dynamic molecular features of ovary tissues and spermatozoa during the process of spermatozoa storage in the black rockfish, proteomic and metabonomic profiles of ovarian tissues from mating and non-mating female fish were analyzed, as well as transcriptomic profiles of spermatozoa (Fig. S1) isolated from ovaries at three different points of spermatozoa storage in ovary (Fig. S2). Proteomic data of ovary in the early spermatozoa storage (ESS) stage indicated that the number of differentially expressed proteins (DEPs) between mating and non-mating groups was low, with 127 proteins upregulated and 330 proteins downregulated in the mating group (Fig. S3 and Table S1). KEGG enrichment analysis of DEPs at ESS stage was performed to explore the molecular response of ovary tissues triggered by the spermatozoa invasion. Notably, complement and coagulation cascades, along with tryptophan metabolism, emerged as the most enriched pathways, harboring a substantial number of DEPs (Fig. 1A and Table S1). Pathway network analysis further revealed the key role of complement and coagulation cascades, exhibiting the highest connectivity, and connected with other immune-related pathways (Toll-like receptor signaling pathway) and cell-adhesion molecules (Fig. 1C). To further identify the response of the ovary induced by spermatozoa, the protein expression profile in ovarian tissues incubated with spermatozoa in vitro was analyzed. Results revealed a significant upregulation of certain complement-related genes after 48 h of incubation (Fig. 1B). Conversely, incubation of spermatozoa with female fish serum (FFS) in vitro significantly diminished spermatozoa viability, which was further enhanced after complement activation induced by Mg²⁺ but alleviated when complement was inactivated with EDTA or heat treatment (Fig. 1D). These findings suggested that spermatozoa invasion triggered complement activation of ovarian tissue, which negatively affected spermatozoa survival.

KEGG enrichment analysis of DEPs between mating and non-mating groups at intermediate spermatozoa storage (MSS) stage was conducted to identify ovarian molecular features after prolongation of spermatozoa storage. Interestingly, complement and coagulation cascades still represented the most significantly enriched pathway (Fig. 2A). However, the majority of DEPs identified in this pathway were downregulated at the MSS stage, in contrast to the pattern observed at the ESS stage (Fig. 2A). This suggested that complement activation initially induced by spermatozoa does not sustain during storage. It was hypothesized that certain proteins and metabolites might be involved in self-regulation by the ovary. Thus, metabolomics analysis was employed to further explore the regulation of metabolites on complement activity. A positive correlation between hippurate and complement-related proteins was found using the association analysis of DEPs and differential metabolites (DEMs) at the MSS stage (Fig. 2B and Table S3). Hippurate, a downstream product of tryptophan metabolism, was identified as a DEM at both ESS and MSS stages (Fig. 2C and Table S2). Tryptophan metabolism and the complement and coagulation cascade represented the most significant pathways in the KEGG analysis using DEPs at ESS stage (Fig. 1A). The impacts of tryptophan and hippurate were further assessed by RT-qPCR, which showed that the expression of complement-related genes was downregulated in cultured ovarian tissues in the presence of tryptophan and hippurate (Fig. 2D and Fig. S4). These results likely suggest that tryptophan plays a role in regulating complement depression, contributing to an immunosuppressive environment to protect spermatozoa physiology at the MSS stage.

The observations above indicated that the spermatozoa entry induced a potent complement response in the ovarian microenvironment, which in turn had a detrimental effect on spermatozoa survival. However, spermatozoa may have developed a defense strategy to adapt this immune response. To explore this possibility, transcriptomic sequencing was carried out using spermatozoa RNA from testis and ovary from mating groups at ESS, MSS and late spermatozoa storage (LSS) stages (Fig. S1). First, genes expressed in spermatozoa were classed into four categories based on their distribution: spermatozoa-specific genes (SSGs), spermatozoa-testis genes (STGs), spermatozoa-ovary genes (SOGs) and spermatozoa-testis-ovary genes (STOGs). At the pre-mating (PRM) stage, the numbers of spermatozoa genes present in the above groups were 19, 59, 3 and 3746, respectively (Fig. S5A). The numbers of SSGs, STGs, SOGs and STOGs were then determined in the spermatozoa transcriptome during spermatozoa storage. SSGs remained largely constant, indicating that our sampling efforts and the transcriptome analysis process were reliable. However, it appeared that when spermatozoa entered the ovary, the number of STGs was reduced. Next, tread analysis was performed using STOGs, which were the most prevalent, to examine spermatozoa transcript profiles at various storage stages (Fig. S5B). As these genes are likely involved in the entire process of spermatozoa adaption to the ovarian microenvironment, we focused on the genes that showed increased expression from PRM to LSS stages. KEGG pathway enrichment was carried out using only gene set 19 and was found to be enriched in genes involving longevity regulation (Fig. S5C). Mgst3 was enriched by KEGG analysis, and it encodes an enzyme involved in glutathione metabolism as well as in cellular redox processes and energy metabolism.

We next performed differential expression analysis (DEA) using all spermatozoon genes (with TPM>0). According to the transcriptomic data, there were 6789 DEGs between spermatozoa in PRM and ESS-M groups (Fig. S6 and Table S4), indicating that dynamic changes in transcriptional activity occur after spermatozoa entry into the ovary. To understand this dramatic fluctuation in spermatozoa transcripts and determine key genes in each stage, a total of 12 specific gene modules were identified using weighted gene co-expression network analysis (WGCNA) (Fig. S7). Among these modules, the genes in the MEblue module primarily represents spermatozoa genes specifically expressed at the ESS stage, encompassing 1950 specific genes (Fig. 3A,B). KEGG analysis revealed that these genes were significantly enriched in pattern recognition receptor-related pathways, such as NOD-like receptor and Toll-like receptor (Fig. 3C). Additionally, hub genes in the MEblue module were identified using the top 200 genes in the association network (Fig. S8) and validated by RT-qPCR. The results highlighted Toll-like receptor pathway-related genes TLR2, TLR3 and Myd88, and the NOD-like receptor pathway-related gene NLRP12, which exhibited the highest expression levels in spermatozoa isolated from the ovary at the ESS stage (Fig. 3D). Furthermore, genes in the MEblue module were also enriched in the C-type lectin pathway (Fig. 3C). Different leptins were used to label the glycosyl bonds in the spermatozoa, and the function of complement proteins on glycosyl alterations were further investigated. The levels of di-sialylated T-antigen, [GlcNAc]1-3, N-acetylglucosamine and N-acetyl-β-D-glucosamine were significantly downregulated in spermatozoa when complement proteins were inactivated by heat treatment of FFS (Fig. 3E and Fig. S9). These observations suggested that spermatozoa undergo an immunoreaction and surface glycan remodeling to cope with complement-mediated ovarian immunoregulation. When sperm glycosyl synthesis was inhibited by tunicamycin (TM), the mortality rate of spermatozoa during incubation with FFS was further increased (Fig. S10). On the premise that TM treatment alone did not affect sperm survival, this experiment suggested an important role for glycans in the defense of spermatozoa against complement.

DEPs were analyzed to explore ovarian molecular changes during sperm storage. KEGG enrichment analysis revealed significant enrichment of DEPs in the GO terms related to carbon metabolism and intercellular interaction and regulation, including ‘focal adhesion’, ‘ECM-receptor signaling pathway’, ‘tight junction’ and ‘bacterial invasion of epithelial cells’ (Fig. 4A,B). Correlation analysis of DEPs and DEMs highlighted that focal adhesion and tight junction were central, with CDC42 and ITGB1 being involved in multiple pathways (Fig. 4B and Fig. S11). Positive correlations were observed between FN1A and multiple differential metabolites. L-ergothioneine was positively correlated with MRLC2 and MYL9, whereas 4-methyl-5-thiazoleethanol showed negative correlation with CDC42b and ACTN4. These observations suggested that these proteins and metabolites might participate in regulating the intercellular state of the spermatozoa-ovarian microenvironment. Furthermore, genes in MEsalmon (Fig. 4C and Fig. S8), the WGCNA module primarily associated with spermatozoa at the LSS stage, were also associated with gap junction and focal adhesion pathways (Fig. 4D). During spermatozoa storage, there were no significant histological difference in ovarian cells between the mating and non-mating groups (Fig. S12). Spermatozoa were found to be randomly dispersed in the slits surrounding the outer germinal epithelium at the ESS stage. They were attracted to the surface of the epithelium by ciliary oscillations at the MSS stage, and gradually became embedded in the zona pellucida (ZP) layers at the LSS stage (Fig. S13). Thus, the signaling pathways described above might be involved in regulating the interactions between epithelial cells and spermatozoa.

In our previous study, zpb2a was specifically expressed in ovary of black rockfish, and expression of ZPB2a was significantly elevated with ovarian development. Furthermore, protein localization was gradually shifted from cytoplasm to zona pellucida (ZP) within the oocyte, which indicated that ZPB2a protein is a component of the zona pellucida of black rockfish egg. In addition, ZPB2a could bind to the posterior side of the spermatozoa head (Li et al., 2022). As glutathione S-transferase mu 3 (GSTM3) emerged as the exclusive protein predicted to interact with ZPB2a in the MESalmon module of WGCNA (Table S6), AlphaFold was used to predict the model of SsGSTM3-ZPB2a interaction to identify important domains (Table S8). Additionally, the findings from the comparison of the binding model across black rockfish, human and zebrafish, revealed that a majority of the amino acid residues potentially involved in the interaction are located within the structural domain GST_Mu and are conserved (Fig. S14A). Moreover, the predicted structural model indicated that the amino acid residues involved in binding were spatially close (Fig. S14B). To validate the physical interaction between ZPB2a and SsGSTM3, co-immunoprecipitation and bimolecular fluorescence complementation (BiFC) assays were performed (Fig. 5A and Fig. S15). Anti-GSTM3 antibody was proved to be specific to endogenous GSTM3 in black rockfish by western blot (Fig. S16). GSTM3, a member of the glutathione transferase gene family, serves a crucial role as an important antioxidant enzyme in spermatozoa survival (Riis Poulsen et al., 2018). In black rockfish, SsGSTM3 was predicted to localize in the cytoplasmic membrane and cytosol (Table S6). Furthermore, SsGSTM3-EGFP fusion protein was expressed in HEK 293T cells after transfection, and it was found to be localized in the cytoplasm and membrane (Fig. S17A); in the spermatozoa, SsGSTM3 protein was predominantly distributed in the spermatozoa head region (Fig. S17B). In addition, blocking the activity of SsGSTM3 using a specific antibody reduced spermatozoa survival, suggesting that ZPB2a upregulates spermatozoa survival through interactions with GSTM3 (Fig. 5B). GSTM3 also exhibited an inhibitory effect on Notch pathway in vitro, as determined by a dual-luciferase assay (Fig. 5C). In order to investigate the effect of the Notch signaling pathway on spermatozoa, the spermatozoa were treated with DAPT, an inhibitor of the Notch pathway. Genes associated with glycan and lipid metabolism, as well as apoptosis, were predominately downregulated upon suppression of the Notch signaling pathway in spermatozoa (Fig. 5D,E). Collectively, these findings suggested that the SsGSTM3-ZPB2a interaction facilitates spermatozoa survival by inhibiting the Notch signaling pathway.

The molecular landscape of spermatozoa-ovarian microenvironment interactions was elucidated from PRM to LSS stages (Fig. 6). At the ESS stage, spermatozoa invasion induced a strong ovarian immune response. In particular, the level of complement-related proteins (CD59, F13 and C7) was significantly elevated, and these proteins could form the membrane attack complex (MAC). To circumvent this immune attack, the expression of pattern recognition receptor genes was significantly elevated and surface glycans were remodeled in spermatozoa. As spermatozoa storage progressed towards the MSS stage, ovarian complement activation was downregulated by tryptophan and hippurate, resulting in a homeostatic ovarian microenvironment for spermatozoa storage at the MSS stage. In addition, cellular junction proteins were involved in spermatozoa-ovarian microenvironment interactions at the LSS stage. Levels of ovarian proteins associated with cell adhesion (CDC42, RAC and RHOG) and ECM-receptor interaction (collagen β1, laminin β1 and laminin α6) were significantly elevated, and spermatozoa genes related to gap junctions (tubb4, gnai3 and pdgfa) and focal adhesion (cav1, ppp1cb, col6a6 and capn2) were also upregulated in spermatozoa. Some of the LSS stage DEPs (AP2B1, DNM2, SLC1A3 and HSD17B14) were enriched in the synaptic vesicle cycle pathway (Fig. 5A), suggesting potential contributions of exosome vesicles to the connection between spermatozoa and epithelial cells. When spermatozoa reached the ZP layer, ZPB2a bound to SsGSTM3 at the surface of spermatozoa, suppressing the Notch signaling pathway as well as downregulating the expression of genes associated with glycan and lipid metabolisms, and apoptosis. These regulations contributed to the dormancy and survival of stored spermatozoa. It is likely that all these changes establish a conducive environment for spermatozoa approaching the oocyte and preparing for fertilization.

In animals with internal fertilization, spermatozoa are transported from testis to ovary during mating and are subsequently stored in the ovary until fertilization. During storage, spermatozoa and the ovarian microenvironment undergo dynamic molecular changes (Bloch Qazi et al., 2003). In our present study, extensive changes in the spermatozoa transcriptome at the ESS stage were found, suggesting that spermatozoa undergo rapid transcriptional regulation to adapt to the new surrounding environment. With the improvement of high-throughput sequencing techniques, accumulating evidence suggests the presence of partial self-transcriptional activity in spermatozoa, challenging the traditional notion that spermatozoa are unable to transcribe due to the replacement of histones by highly condensed protamine. In mammals, alterations in the RNA content of morphologically mature spermatozoa occur as they transition from the testis into the epididymis (Santiago et al., 2021). Moreover, it has been reported that histones in mammalian spermatozoa are not completely replaced by protamine, as a certain percentage (ranging from 1% to 15%) of histones remains present in transcriptionally active promoter regions (Yoshida et al., 2018). A significant effect of cryopreservation on spermatozoa transcriptome has been observed in many species, including black rockfish (Niu et al., 2022), suggesting that the transcription profile of mature spermatozoa could be influenced by the external environment. In the present study, it was discovered that the co-expressed gene module MEblue exhibited a positive correlation with ovarian effects on spermatozoa at the ESS stage. These genes were associated with pattern recognition receptors, which play a crucial role in innate immunity.

In mammals, innate and adaptive immune defenses are present in both ovaries and the reproductive tract to prevent infection by pathogenic microorganisms (Fichorova et al., 2005; Wu et al., 2004). The complement system serves as the first line of defense in innate immunity (Defendi et al., 2020). It has been shown that complement-related genes are expressed in the human ovary (D'Cruz et al., 1990; Clarke et al., 1984) and are upregulated by spermatozoa (Parish et al., 1967; Tauber et al., 1975). Our proteomic analysis at the ESS stage also unveiled an induction of ovarian complement activation by spermatozoa. Incubation experiments further showed that complement in the female serum reduced spermatozoa viability. This suggested that complement-mediated immunoregulation is not favorable for the preservation of spermatozoa in ovary. However, spermatozoa can be stored in the ovary for over 4 months in black rockfish (Liu et al., 2019), which implies that spermatozoa could overcome ovarian immune defense by developing appropriate strategies. In other words, the storage period in the ovary might act as a screening process for spermatozoa, where robust spermatozoa are selected for fertilization based on their ability to survive in the presence of increased complement levels. Our results support this idea, as the overall mortality of spermatozoa was increased by complement enhancement, yet at least 20% of viable spermatozoa remained in the ovary (Fig. 1). This mechanism is conserved during the evolution of species with internal fertilization, as it has been found in several other species (Roan et al., 2017; Han et al., 2019; Nogueira et al., 2011).

Members of the TLR family have been shown to exert regulatory activities on spermatozoa (Zhu et al., 2016; Fujita et al., 2011). At ESS stage, genes coding for pattern recognition receptors, including Toll-like receptors (TLR2, TLR3 and Myd88) and NOD-like receptor (NLRP12), were significantly upregulated in spermatozoa. There is evidence that TLR2, TLR3 and TLR9 are expressed in different regions of mouse spermatozoa (Saeidi et al., 2014), whereas TLR2, TLR5, TLR7, TLR15 and TLR21 are expressed in chick spermatozoa (Das et al., 2011). The motility of spermatozoa decreases when Toll-like receptor signaling is activated (Zhu et al., 2016). These observations support the regulation of Toll-like receptors on spermatozoa in black rockfish. The glycocalyx on the sperm surface is crucial for sperm survival, maintenance of physiological status and defense against immune responses in the female reproductive tract (Schröter et al., 1999). Knockout of galectin 1 resulted in reduced spermatozoa motility and defective membrane potential hyperpolarization (Vasen et al., 2015). The complex interactions between various glycan complexes have been reported to affect the binding of spermatozoa with eggs and the fertilization process (Umezu et al., 2020; Clark, 2014). In mammals, the surface of spermatozoa is composed of a dozen different monosaccharides, including sialic acid, galactose, GalNAc, glucose, GlcNAc, mannose, xylose, fucose, glucosamine, glucuronic acid and iduronic acid. Some of these monosaccharides could be further modified by sulphation and/or acetylation (Tecle and Gagneux, 2015). The mature glycocalyx allows spermatozoa to penetrate the cervical mucus (Gilks et al., 1989; Tollner et al., 2008a) and protects them from humoral and cellular immunity in the uterus, where they encounter female antibodies, complement and immune cells, such as macrophages and neutrophils. In addition, glycan-mediated interactions are involved in the attachment of spermatozoa to the oviductal epithelium (Tollner et al., 2008a,b). In this study, DEGs of spermatozoa were enriched in the glycan synthesis-related pathway at ESS stage and the remodeling of surface glycan was upregulated by complement, demonstrating that glycans participate in regulating spermatozoa survival from complement-mediated ovarian immune defense.

Ovarian complement is an important microenvironment factor for spermatozoa. At the MSS stage, proteomics analysis showed that ovarian complement was depressed, resulting in an immunosuppressive environment to protect spermatozoa physiology. The regulatory role of tryptophan on ovarian complement was further analyzed in this study. Tryptophan is an important amino acid that is widely involved in protein synthesis and metabolic reactions (Karahoda et al., 2020). It can be metabolized into 5-hydroxytryptamine, an inhibitory neurotransmitter related to the function of the nervous system and found at high concentrations in the mammalian cerebral cortex and synapses (Bley et al., 1994). In addition, tryptophan is involved in many physiological mechanisms of the neuroendocrine-immune network and participates in the immune function of macrophages and lymphocytes. The balance of tryptophan in the fish diet is essential for the stability of their immune system (Machado et al., 2019). Our metabolomic analysis showed no significant differences in tryptophan levels between mating and non-mating ovaries. However, the level of hippurate in the mating ovary was significantly higher at the ESS stage. Tryptophan can be metabolized by anthranilate to benzoate, which can be converted to hippurate in the liver (Machado et al., 2019). Hippurate and betaine have been identified in the metabolites of the intestinal immune system of pigs (Riis Poulsen et al., 2018). The level of hippurate has been considered as a valid biological marker of inflammatory bowel disease in humans (Takis et al., 2016), suggesting a correlation between hippurate and the balance of the immune system. Indeed, as shown in this study, complement-associated proteins were highly correlated with hippurate, whereas complement-mediated immunoregulation was negatively correlated with tryptophan and hippurate levels.

There is a predominant role of cell junction-related proteins at the LSS stage. In the ovary, DEPs were significantly enriched in focal adhesion and tight junction pathways. In spermatozoa, the co-expressed gene module MEsalmon was also associated with gap junction and focal adhesion. Throughout spermatogenesis, various pathways regulate the state of intercellular communication, such as tight junction, focal adhesion and regulation of actin cytoskeleton (Siu and Cheng, 2004). These pathways are mainly a component of the blood-testis barrier that protect spermatogenesis from external disturbances (She et al., 2021). Regulation of the actin cytoskeleton also plays roles in spermatid polarity formation, which is important to support spermatogenesis (Li et al., 2018). Focal adhesion kinase (FAK) is a non-receptor protein tyrosine kinase that shows phosphorylation-dependent localization in the spermatic cord epithelium of the adult rat and regulates Sertoli cell adhesion through its effect on the occludin-zonula occludin 1 complex (Lie et al., 2012). In addition, ANXA1, ANXA2, GSTM3 and HSP genes were also identified in the MEsalmon module. ANXA2 and HSP genes are considered to be the markers of extracellular vesicles, and are induced by spermatozoa and involved in spermatozoa-oviduct proteins binding in birds (Cordeiro et al., 2021). GSTM3, a member of the glutathione transferase family, has been implicated in spermatozoa survival in previous literature (Llavanera et al., 2020). GSTM3 is distributed on the plasma membrane of porcine spermatozoa and bound to the ZPB protein of the egg, thereby mediating the initial step of sperm-egg binding (Petit et al., 2013). All the signaling pathways and related genes mentioned above could participate in the spermatozoa-ovarian cell interaction, which facilitates a closer connection between spermatozoa and ovarian cells.

The mechanism of spermatozoa storage in viviparous fishes remains a relatively understudied area, although some studies in mammalian and avian species have been undertaken (Mahé et al., 2021; Wang et al., 2022; Matsuzaki and Sasanami, 2022). With the completion of the black rockfish genome in our previous report (He et al., 2019), a significant expansion of the ZP family was found when the black rockfish genome was compared with other teleost fishes, an interesting finding given that ZP layers are involved in spermatozoa storage (He et al., 2019). In addition, ZPB2a protein has a strong capacity to bind spermatozoa (Li et al., 2022). Building upon this, our current study reveals a true interaction between ZPB2a and SsGSTM3 on the surface of the spermatozoa head, and reveals a potential role in downregulating Notch signaling to enhance spermatozoa survival.

In addition, this study highlights the crucial role of glycan in black rockfish spermatozoa as a defense against complement, suggesting a conserved protective function for glycan in spermatozoa, aligning with previous reports demonstrating the importance of glycans for spermatozoa survival and storage in the female reproductive tract in mammals and insects (Tecle and Gagneux, 2015; Miller, 2015). In summary, a comprehensive characterization of the dynamic molecular features of spermatozoa and ovarian tissues was presented during spermatozoa storage. At the ESS stage, complement-mediated immunoregulation was triggered in the ovary by spermatozoa invasion, leading to an upregulation of pattern recognition receptor-related genes and a remodeling of surface glycans, enabling spermatozoa to withstand the ovarian immune defense. Conversely, at the MSS stage, ovarian complement-mediated immunoregulation was attenuated by tryptophan and hippurate, contributing to the establishment of an immunosuppressive microenvironment conductive to spermatozoa storage. At the LSS stage, cell junction proteins were found to be involved in spermatozoa-ovarian cell interactions, which may be beneficial for substance exchange. This investigation in black rockfish provides valuable insight into the intricate journey of spermatozoa within the ovary and provides clues to the mechanism behind spermatozoa survival in the female body.

A total of 200∼2.5-year-old females and 50 males were cultivated separately at Weihai Shenghang Ocean Science and Technolog (Shandong, China). Artificial insemination was performed in November, and female fish were divided into three groups: a PRM (pre-mating) group, a mating group (M) and a non-mating group (NM, control). Samples were collected at three time points: late December (ESS stage), mid-February (MSS stage) and late March (LSS stage). Spermatozoa isolated from the ovaries at each stage were processed for RNA extraction. Specifically, the ovaries were clipped to fully release the internal spermatozoa into PBS buffer. The sperm suspension was obtained by differential centrifugation combined with cell sieve filtration. First, the suspension was filtered through a cell sieve of 40 μm and centrifuged at 500 g at 4°C for 5 min, then centrifuged at 700 g at 4°C for 5 min to remove other cells, including blood cells and oocytes. After each centrifugation at 700 g, microscopic observation was carried out to gradually increase the proportion of spermatozoa in the supernatant; when the purity of the spermatozoa reached more than 95%, the spermatozoa were obtained by centrifugation at 4,500 g at 4°C for 10 min, and then treated with liquid nitrogen and stored at −80°C. Ovaries from the M and NM groups were immediately frozen in liquid nitrogen and stored at −80°C for extraction of proteins and metabolites. For proteomic analysis, six female fish of equal size were randomly selected for each of the mating and non-mating groups at each stage for sequencing and analysis, with two fish ovaries mixed together to create one biological replicate. For the metabolome, six female fish were selected for each of the mating and non-mating groups. For the transcriptome at each stage, the same six fish used for proteomic and metabolomic analyses were also selected for mating, with sperm collected in each of the two fish (as for proteomic analysis) mixed together to create one biological replicate.

All animal experiments were conducted in accordance with the guidelines of China Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing, Research, and Training (China, 1988). Every experimental process was approved by College of Marine Life, Ocean University of China (Qingdao, Shandong, China).

Proteins from ovary tissues were extracted as described previously (Wu et al., 2021). Subsequently, 100 μg of proteins were hydrolyzed, desalted and separated in liquid phase using a Shimadzu LC-20AD liquid phase system.

The number of identified peptides and proteins are shown in Table S1. Analysis of differential expression was performed using MSstats. Fold changes between different groups of more than 1.5 and P<0.05 were considered as differently expressed proteins (DEPs), which were further submitted to Blast2GO for enrichment analysis. Significantly enriched terms or pathways were determined by comparing significantly differentially expressed proteins with total identified proteins using the hypergeometric test. Enrichment analysis was based on Fisher's exact test and Benjamini-Hochberg to calculate P-values; pathways with P<0.05 were considered significant.

Ovarian tissues (about 25 mg) were incubated with 800 μl of extraction solution (methanol:acetonitrile:water, 2:2:1, v/v) pre-chilled at −20°C and 10 μl of internal standard metabolites. After centrifugation at 16,000 g at −20°C for 15 min, the supernatant was removed and concentrated using a cryo-vacuum concentrator. The samples were vortexed for 1 min and sonicated in a water bath at 4°C for 10 min. After centrifugation at 4°C for 15 min at 25,000 g, the supernatant was placed in a bottle and mixed with QC quality control samples to determine the reproducibility and stability of the LC-MS analysis process.

The formulations and procedures of elution were carried out as described previously (Wu et al., 2021). Primary and secondary mass spectrometry data acquisition was performed using a Q Exactive mass spectrometer (Thermo Fisher Scientific). LC-MS/MS data processing was performed using Compound Discoverer 3.0 (Thermo Fisher Scientific) software for peak extraction, peak alignment and compound identification. Differential metabolites (DEMs) were screened based on the following criteria: (1) VIP of the first two principal complements of the PLS-DA model ≥1; (2) fold-change ≥1.2 or ≤0.83; and (3) P-value<0.05. Statistical analysis, metabolite classification annotation and functional annotation were performed using the metaX (Wen et al., 2017).

Total RNAs from spermatozoa and from ovarian tissues were extracted using a RN29-Allprep Tissue Cell RNA/DNA Sorting Kit (Aidlab) and TRIzol reagent (Invitrogen), respectively, according to the manufacturers’ protocols. The quality of RNAs was analyzed by 1.5% agarose gel electrophoresis and the concentration was determined using a NanoDrop spectrophotometer. A total of 12 cDNA libraries were generated as described previously (He et al., 2019).

RNA-seq raw data were evaluated using FastQC (v0.11.9) and clean data were obtained after removal of the adaptors and low-quality bases using SOAPnuke (Li et al., 2008). Clean reads were then mapped to the black rockfish genome using HISAT2 (Kim et al., 2015). The statistics for pre-processing transcriptome data are shown in Table S4. Differentially expressed genes (DEGs) between different stages were identified using DEseq2 (Love et al., 2014). Genes with adjusted P<0.05 and fold-change>2 were considered as DEGs. GO terms and KEGG pathways were analyzed using DAVID (https://david.ncifcrf.gov/). The WGCNA package (v1.47) with R functions was used to explore the correlation patterns among DEGs across multiple samples and modules highly correlated to different spermatozoa storage stages were identified (Langfelder and Horvath, 2008). Identified genes from spermatozoa were classed into four categories: (1) sperm-specific genes with TPM value>5 in spermatozoa but <1 in testis and ovary; (2) sperm-testis genes with TPM value>5 in both spermatozoa and testis but <1 in ovary; (3) sperm-ovary genes with TPM value>5 in both spermatozoa and ovary but <1 in testis; (4) sperm-testis-ovary genes with TPM value >5 in spermatozoa, testis and ovary. Gene expression profile was grouped according to tread analysis using Short Time-series Expression Miner (STEM) software and visualized by OmicShare tools platform (www.omicshare.com/tools).

The ovaries at the ESS stage were washed six times in PBS buffer containing 5% penicillin-streptomycin-nystatin solution (Gibco). Subsequently, they were cut into small pieces (0.3 cm³) in F-15 medium containing 5% penicillin-streptomycin-nystatin solution. The ovarian pieces were placed onto 24-well plates and ∼1×10⁵ spermatozoa were added into each well. Other pieces without spermatozoa treatment were considered as the control group. After 24 h and 48 h, ovarian tissues were frozen in liquid nitrogen and stored at −80°C for RNA extraction.

To assess the effect of complement on the survival of spermatozoa, serum was prepared from the blood due to its high content in complement (Li et al., 2015). Blood was collected from the caudal vein of female individuals and allowed to coagulate for 12 h at 4°C. Serum was prepared by centrifugation at 3000 g for 10 min at 4°C and stored at −20°C. About 2×10⁵ spermatozoa in 400 μl Dulbecco's modified Eagle medium (DMEM) (Biological Industries) were mixed with an equal volume of the following solutions: (1) DMEM, (2) control female fish serum (FFS), (3) FFS containing 0.02 M MgCl₂ as a natural co-factor involved in the formation of stable C3 convertase to activate complement (Acevedo and Vesterberg, 2003), (4) FFS containing 0.04 M ethylenediaminetetraacetic acid (EDTA) disodium salt to chelate Mg²⁺ or (5) FFS treated at 56°C for 30 min to inactivate the complement. After incubation for 3 h at 10°C, the spermatozoa were washed twice with PBS and stained using Calcein-AM and propidium iodide (Meilunbio). Their viability was examined using a Nikon Eclipse Ti-U microscope (Nikon, Tokyo, Japan). At least 100 spermatozoa at each group were counted. To investigate the role of glycans on the spermatozoa surface in resisting complement, spermatozoa were pre-incubated with 10 μg/ml tunicamycin (TM) for 1 h, followed by incubation with FFS for 2 h. The specific incubation methods and the statistical survival assay of sperm were performed as described above.

To explore the regulatory effect of the Notch signaling pathway on spermatozoa, spermatozoa were incubated with 50 μM DAPT in DMEM medium at 10°C for 24 h. The spermatozoa without DAPT treatment were selected as the control group. Subsequently, spermatozoa were washed twice with PBS and collected for RNA extraction and qPCR.

Ovaries from mating group at the ESS stage were washed six times in PBS containing 5% penicillin-streptomycin-nystatin solution and were cut into small pieces, as above. The ovarian pieces were then put onto a pre-impregnated cellulose acetate film placed on an agarose block soaked in L-15 medium and other essential ingredients, as previously described (Kortner et al., 2009). After 24 h of incubation, tryptophan and hippurate were added to the L-15 medium at a concentration of 50 μM to examine their effects on the complement from ovarian follicles. After 72 h, the ovarian pieces were frozen in liquid nitrogen and stored at 80°C for RNA extraction. The detail culturation process of ovarian tissue blocks have been described in previous literature (Kortner et al., 2009).

This was performed on spermatozoa removed from male testis and incubated with FFS, and on spermatozoa incubated with inactivated FFS at 56°C for 0.5 h. Subsequently, ∼2×10⁷ spermatozoa were incubated with different lectins, including LEL, ABL, JAC and WGA (wheat germ agglutinin) at their standardized concentrations at 37°C for 15 min (Batra et al., 2020). A drop of spermatozoa were placed on a clean glass slide (SAS Medical, Beijing, China), and the fluorescence was examined using a laser confocal scanning microscope A1R (Nikon). Six random views were taken, and each view contains at least 100 spermatozoa. Fluorescence intensities were calculated using Image J software.

Ovarian tissues at PRM, ESS, MSS and LSS stages were fixed in 4% PFA overnight. They were dehydrated using graded ethanol. After xylene treatment and embedding in paraffin wax, ovarian tissues were sectioned at 5 μm, deparaffinized in xylene and stained using a Hematoxylin and Eosin (H&E) kit (Solarbio). Images were taken using a Nikon Eclipse Ti-U microscope.

Equal amounts of GSTM3 prokaryotic protein and ZPB2a prokaryotic protein (400 μg each) were mixed and incubated at room temperature for 1 h; part of the mixture was taken as the input sample and the remaining part was incubated at room temperature for 1 h with magnetic beads (Beaver, Suzhou, China) that were pre-incubated with GSTM3 antibody at room temperature for 1 h. Protein elution, SDS-PAGE electrophoresis and western blot were performed as in a previous study (Qu et al., 2022). Antibodies were as follows: anti-GSTM3 (bs-16341R, Bioss, Suzhou, China; 1:1000), anti-His (CW0286, CWBIO, Taizhou, China; 1:1000), anti-ZPB2a (from our lab, Li et al., 2022; 1:500), goat anti-rabbit IgG, HRP conjugated (CW0103, Taizhou, China; 1:2000) and goat anti-mouse IgG, HRP conjugated (CW0102, CWBIO, Taizhou, China; 1:2000). A bimolecular fluorescence complementation (BiFC) assay was performed as described in a previous study (Qu et al., 2022). Briefly, pBiFC-SsGSTM3-VC155 and pBiFC-SsZPB2a-VN173 were constructed and co-transfected into HEK 293T cells. Co-transfection using either pBiFC-SsGSTM3-VC155 and pBiFC-VN173 or pBiFC-SsZPB2a-VN173 and pBiFC-VC155 was considered as the negative group. To determine the subcellular localization of SsGSTM3 in the cell, GSTM3-N1 was transfected into HEK 293T cells with Lipofectamine 3000 (Thermo Fisher Scientific). pEGFP-N1 was selected as the control group. At 24 h after transfection, cells were fixed with 4% PFA at room temperature for 15 min. The cells were washed twice with PBS and then nuclei and membranes were stained with DAPI (5 μg/ml) and WGA (50 μg/ml) coupled with rhodamine, respectively, for 15 min. All photographs were acquired using the laser confocal scanning microscope A1R (Nikon).

Spermatozoa were washed twice with 1×phosphate-buffered saline (PBS) and fixed with 4% PFA at room temperature for 30 min. Spermatozoa suspensions were added dropwise to poly-lysine-coated slides and dried at 60°C. Spermatozoa were incubated in blocking buffer (3% BSA in PBS) at room temperature for 1 h, and incubated with anti-GSTM3 antibody (bs-16341R, Bioss, Suzhou, China) diluted in PBS at a ratio of 1:200 at room temperature for 2 h. A secondary fluorescein isothiocyanate AffiniPure goat anti-rabbit IgG antibody was used to display green fluorescence signal. Spermatozoa were stained with DAPI (4′,6-diamidino-2-phenylindole) to visualize nuclei and photographed on a fluorescence microscope.

The amino acid sequences of GSTM3 and ZPB2a of black rockfish were submitted to the website version of Alphafold (https://colab.research.google.com/github/sokrypton/ColabFold/blob/main/AlphaFold2.ipynb). Two sequences were separated by a colon (:), and the files in PDB format were downloaded, then submitted to the PDBePISA website (https://www.ebi.ac.uk/msd-srv/prot_int/pistart.html) for the prediction of protein interactions, and to obtain the amino acid residues and other information about the interactions between the SsGSTM3 and ZPB2a. The detail prediction information is shown in Table S8. The structural model was adjusted and displayed using PyMOL software.

A Strand cDNA Synthesis SuperMix with gDNA digester (Yeasen) according to the manufacturer's instruction. Gene expression analysis by qPCR was performed using the Light-Cycler 480 (Roche). Gene-specific primers are listed in Table S5 and their specificity was validated based on the single peak of melting curves generated after an amplification reaction. Rpl17 was selected as the reference gene, as described previously (Jin et al., 2021). The relative expression levels of the target genes were calculated using the 2^−ΔΔCt method. All data are mean±s.e.m. calculated from nine samples with three independent experimental repeats.

In order to explore whether SsGSTM3 regulates Wnt, Notch and TGF-β signaling pathway, pEGFP-N1-SsGSTM3 was co-transfected with TopFlash, RBP-JK-luc, p3TP-lux respectively. pRL-TK vector was considered as the internal reference vector. HEK 293T cells (SCC-120511) were purchased from Solarbio Life Science, and authenticated and tested for contamination before use. Approximately 3×10⁵ HEK 293T cells were seeded in 24-well plates, cultured in Dulbecco's modified Eagle medium (Biological Industries) containing 10% fetal bovine serum (Gibco) and 1% penicillin-streptomycin (Gibco) overnight under 5% CO₂ at 37°C, and transfected with the plasmids using Lipofectamine 3000 Reagent (Life Technologies) based on the manufacturer's protocol. Briefly, total 610 ng plasmids, including 300 ng luciferase reporter vector (TopFlash, RBP-JK-luc, p3TP-lux), 300 ng pEGFP-N1-GSTM3 and 10 ng pRL-TK vector, were co-transfected into HEK 293T cells. The transfection groups of pEGFP-N1 and non-expression vectors were treated as a control and mock group, respectively. After 48 h of transfection, HEK 293T cells were washed twice with PBS buffer and lysed with lysis solution for 30 min at room temperature. Relative luciferase values for each well were calculated using the Reporter Assay System (Promega) according to the manufacturer's instrument. At least three biological repeats in each group were performed and analyzed.

All data are presented as mean±s.e.m. (standard error of mean) of three independent experiments. Significant differences were statistically analyzed using a one-way two-tailed ANOVA followed by a Games-Howell test for at least three groups or an independent-samples t-test for two groups (SPSS 20.0, IBM) with P<0.05 indicating a significant level.

We sincerely appreciate Dr Shi Deli's suggestions and revisions of this manuscript.

The expression profiles of genes characterized in this project have been deposited in NCBI Sequence Read Archive (SRA) under accession number PRJNA857487.

The peer review history is available online at https://journals.biologists.com/dev/lookup/doi/10.1242/dev.202224.reviewer-comments.pdf

This article is part of the Special Issue ‘Uncovering developmental diversity’, edited by Cassandra Extavour, Liam Dolan and Karen Sears. See related articles at https://journals.biologists.com/dev/issue/151/20.