Data-Independent Acquisition-Based Quantitative Proteomics Analysis of Fertile Red Eggs and Spermatozoa in Hermatypic Coral Galaxea fascicularis: Revealing Key Proteins Related to Gamete Maturation and Fertilization

Abstract

Sexually propagated scleractinian corals are in high demand for coral reef restoration. However, for threatened reef-building corals, many of the molecular mechanisms related to their reproduction remain largely unknown, which forms a major bottleneck in the large-scale cultivation of sexually reproducing corals. In this study, we analyzed the proteomic signatures of red eggs and spermatozoa from the ecologically significant coral Galaxea fascicularis, using a data-independent acquisition mass spectrometry (DIA-MS) method. A total of 7741 and 7279 proteins from mature red eggs and spermatozoa were identified, respectively. Among these proteins, 596 proteins were spermatozoa-specific and 1056 were egg-specific. Additionally, a total of 4413 differentially abundant proteins (DAPs) were identified, among which 3121 proteins were up-regulated in red eggs and 1292 proteins were up-regulated in spermatozoa. Furthermore, anenrichment analyses showed that DAPs identified in red eggs were mainly involved in the progesterone-mediated oocyte maturation pathway and lectin pathway; and DAPs detected in spermatozoa were mainly involved in the insulin secretion pathway and metabolic pathways for the generation of energy. This result will contribute to the discovery of the intrinsic regulation pathway of gamete maturation and fertilization. Furthermore, at least 57 proteins associated with gamete maturation and reproduction were identified, including the red fluorescent protein (RFP), vitellogenin proteins (VG), the egg protein (EP), the testis-specific serine/threonine-protein kinase family (TSSKs), and the EF-hand Ca²⁺-binding protein family (EFHC1 and EFHC2). Particularly, the third yolk protein EUPHY was reported for the first time in G. fascicularis. In conclusion, this study unveiled groundbreaking molecular insights into coral sexual reproduction, paving the way for more effective conservation and sustainable development of coral reef ecosystems

1. Introduction

As the foundational organisms of coral reef ecosystems, reef-building corals play an important role in primary productivity and nutrient cycling within reef areas [1]. The sustainability of coral reef ecosystems relies on the high survival rates of hermatypic corals, extensive coral coverage, and a sufficient supply of coral recruits [2]. However, the combined impacts of climate change and human disturbances have led to frequent and severe mass-bleaching events with limited recovery periods for reef-building corals. Consequently, 30% of coral reefs are already severely damaged, and up to 60% could be lost by 2030 [3]. Thus, sexually propagated corals are in high demand for coral reef restoration.

Current reef restoration efforts primarily focus on the asexual reproduction of corals, known as clonal processes. Techniques such as direct transplantation, coral gardening, and the construction of artificial reefs have been widely implemented to increase coral coverage and rehabilitate degraded reefs, yielding promising restoration outcomes [4,5,6,7,8]. However, given the challenges posed by global warming, asexual propagation alone is insufficient to enhance coral genetic diversity and ecosystem resilience [2,9]. Innovative restoration methods involving the sexual reproduction of corals, such as gamete hybridization and coral seedling cultivation, offer promising avenues for developing ‘strong corals’ capable of withstanding significant environmental changes [9,10]. In addition, the sexual propagation of corals provides sufficient sources for coral reef ecological restoration and ornamental trade, thereby avoiding the harvesting of wild corals. However, the intrinsic reproduction mechanisms of corals are still unclear, which leads to current artificial coral breeding being primarily reliant on empirical knowledge and limits the mass production of corals in aquaculture. Consequently, to achieve the mass culture of sexually propagated reef-building corals, coral propagation methods need further refinement, underscoring the need for molecular studies on coral reproduction to support these approaches.

High-quality and well-developed eggs and spermatozoa are key for reproduction; abnormal gametes always mean unsuccessful fertilization. Compared to other cells, mature spermatozoa are transcriptionally and translationally inactive and always contain higher amounts of membrane proteins, which are important for sperm–egg recognition and fusion [11,12]. Eggs, abundant in proteins such as vitellogenin, are not only involved in oogenesis, but also could serve as crucial nutrient sources that directly influence early larval development. Therefore, proteomics-based techniques are particularly suitable for understanding the molecular functions of spermatozoa and eggs, and providing useful information for reproduction. For instance, using liquid chromatography–mass spectrometry (LC-MS), 162 and 110 proteins were identified in the mature female and male gonads of the greenlip abalone, respectively. Among these, 47 proteins were found to be involved in spermatozoa and oocyte structure, sperm motility, acrosome response, and fertilization [13]. Through the combined use of sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and LC-MS, 550 spermatozoa proteins of mussel (Mytilus galloprovincialis) were identified, revealing several key fertilization proteins with the highest mean nonsynonymous substitution rates, which may play crucial roles in speciation [14]. Similarly, using proteomics methods, 286 spermatozoa proteins were identified in the Hong Kong oyster (Crassostrea hongkongensis), including a “bindin” protein crucial for sperm–egg fusion [15]. Proteomic technology has also advanced significantly with the development of new methods for protein research. A data-independent acquisition mass spectrometry method (DIA-MS) provides higher sensitivity and protein coverage than the classic data-dependent acquisition (DDA) method, as demonstrated in a multi-laboratory evaluation [16]. Applying the DIA-MS method, 2109 proteins were discovered in the spermatozoa of rams and bucks, uncovering many differentially abundant proteins related to sperm capacitation and metabolism [17].

Compared to the extensive studies conducted on vertebrate and invertebrate model animals, there are relatively few studies focusing on the proteomic characteristics of mature spermatozoa and eggs in cnidarians. The reef-building coral Galaxea fascicularis (clade: Complexa; family: Euphyllidae) is widespread in the Indo-Pacific region and holds significant potential for reef ecosystem stability [18]. The populations of G. fascicularis consist of female and hermaphroditic colonies, and the former produce red eggs and the latter spawn spermatozoa and lipid-filled white eggs. There are no obvious morphological differences between the two colony types [19]. White eggs contain a large amount of lipid that enhances gamete buoyancy, allowing them to carry mature spermatozoa in buoyant bundles to the sea surface [19]. Yolk proteins related to vitellogenesis (such as GfEP-1, 2, and 3) have been discovered in red eggs but not in white eggs [20,21], and another novel yolk protein, GfEP-4, has been identified in white eggs, where its concentration is much lower than in red eggs [22]. There are two distinct results on whether white eggs are fertile [23]: on the Great Barrier Reef (GBR), white eggs produced by hermaphroditic colonies cannot be fertilized [19], while white eggs from hermaphroditic coral populations of Taiwan can be fertilized [24]. Nevertheless, red eggs from female colonies always have high fertilization rates after binding with mature spermatozoa, though the intrinsic mechanism of sexual reproduction remains unclear [19,25].

To better understand the molecular mechanism of sexual reproduction, G. fascicularis with large polyps was selected as the model species to analyze the proteome profiles of fertile red eggs and spermatozoa based on the DIA-MS approach, and we revealed the reproduction-related key proteins. This study aimed to uncover gamete maturation and fertilization-related molecular information, providing greater insight into the sexual reproduction mechanism in hermatypic corals. These findings would promote the development of the artificial breeding of reef-building corals for the purpose of the preservation of coral germplasm resources and the protection of coral reefs.

2. Materials and Methods

2.1. Experimental Coral and Gamete Collection

Adult scleractinian corals of the species G. fascicularis, with a mean diameter greater than 10 cm, were collected in May 2023 from a depth of 3–4 m at the coral nursery in Wuzhizhou (109°45′ E, 18°180′ N) in Sanya, China. Six adult coral colonies (three females and three hermaphrodites) were carefully transported to the coral culture laboratory at the School of Marine Biology and Fisheries and cultivated in tanks with recirculating seawater. Intensive daily observations and examinations of gamete release conditions were conducted after sunset. Upon spawning, gamete samples were collected.

For the female colonies, mature red eggs were collected immediately after release, briefly rinsed with filtered seawater, and cryopreserved. Hermaphroditic colonies were transferred to separate beakers containing filtered seawater before spawning. The released sperm–egg mixtures were suspended in the beakers and rinsed through a 25 μm screen to separate white eggs and debris from the spermatozoa. This washing process was repeated 5–6 times until no white eggs were visible in the spermatozoa samples under the microscope [15]. Red eggs and spermatozoa pools were collected from three female and three hermaphroditic individuals, separately, and three biological replicates of each gamete type were collected. The collected red eggs and spermatozoa samples were kept at −80 °C before protein extraction experiments.

2.2. Protein Extraction and Enzymatic Hydrolysis

Taking an appropriate amount of each sample (100 mg), the proteins from each sample were extracted using the following procedure. First, an appropriate amount of extraction buffer (with a protease inhibitor cocktail) was added to each sample and treated with ultrasonic treatment (50 HZ/s, 120 s). Subsequently, the supernatants were centrifugated at 25,000× g for 20 min at 4 °C. The supernatants were collected, and then treated with 10 mM DTT at 56 °C for 1 h. Next, 55 mM IAM was added, and the mixture was kept in the dark at room temperature for 45 min. Subsequently, after the addition of four times the volume of pre-cooled acetone, precipitation was carried out at −20 °C for 2 h, and this step was repeated three times. The mixture was centrifuged for 15 min at a speed of 25,000× g at 4 °C, and then the supernatant was discarded. The collected precipitate was allowed to air dry; then we added an appropriate amount of lysis buffer, and the solution was treated with ultrasonic treatment (50 HZ/s, 120 s) again to facilitate protein dissolution. Finally, the samples were centrifuged for 15 min at a speed of 25,000× g at 4 °C for the purpose of collecting supernatants.

The protein concentration of each sample was obtained by using the Bradford assay with BSA as a standard, and the quality of the extracted protein was examined by SDS-PAGE. An amount of 100 μg protein of each sample was treated with modified sequence-grade trypsin (Promega, Madison, WI, USA) at 37 °C (4 h) for digestion at a substrate-to-enzyme ratio of 40:1 (w/w). After that, the protein peptides were desalted using a Strata X C18 column (Phenomenex, Torrance, CA, USA) and vacuum-dried.

2.3. High-pH Reverse-Phase Separation

Equal amounts of protein peptides from each sample were extracted and mixed. The mixed peptide samples were then dissolved in 5% acetonitrile (buffer A, pH 9.8) and injected into a Shimadzu LC-20AD HPLC pump system, which was connected to a Gemini C18 column (5 µm, 4.6 × 250 mm) for peptide separation using a linear gradient method. Using 95% ACN (mobile phase B, pH 9.8), the peptide mixture was eluted according to the following steps: 0 to10 min, 5% buffer B; 10 to 50 min, 5% to 35% buffer B; 35% to 95% buffer B for 1 min; maintenance in 95% ACN for 3 min; and re-equilibration for 10 min with the initial concentration of buffer B. The flow rate of gradient elution was set at 1 mL/min.

We monitored the elution peak at a wavelength of 214 nm, and fractions were collected with an interval time of 1 min. The peptides were combined according to the elution peak of the chromatogram to obtain 10 fractions, which were then freeze-dried.

2.4. DDA and DIA Analyses by Nano-LC-MS/MS

Dried peptides were reconstituted with 0.1% formic acid (mobile phase C), and after centrifugation for 10 min at a speed of 20,000× g, the supernatant was collected for subsequent injection. Separation was performed using a Bruker nanoElute system fitted with an analytical self-packed C18 column (1.8 μm, 75 μm × 25 cm). The samples were first enriched and desalted, followed by separation with a gradient from 2 to 80% buffer D (0.1% formic acid in ACN solution) at a flow rate of 300 nL/min. The end of the LC separator was connected to the mass spectrometer and the following parameters were used.

For the construction of the spectral library, the peptide mixtures separated by LC were ionized using a nanoESI source. These ionized peptides were analyzed by a timsTOF Pro tandem mass spectrometer (MS/MS) operating in DDA mode, with the following parameters: (1) Ion source voltage = 1.6 kV; MS1 mass spectrometer scanning range = 100–1700 (m/z); ion mobility range = 0.6–1.60 V.S/cm²; cumulative scan time = 100 ms. (2) MS2 mass spectrometer scanning range = 100–1700 (m/z); cumulative scan time = 100 ms. The precursors for a satisfactory MS2 scan were as follows: charge range of 0 to 5⁺, top 10 precursors with peak intensity exceeding 10,000, and the peak intensity above 2500. The ion fragmentation mode was CID, and fragment ions were detected in TOF. The dynamic exclusion time was set to 30 s.

The peptides separated by LC were ionized by a nanoESI source and then analyzed by a timsTOF Pro MS/MS in DIA mode for quantitation. The main parameter settings were as follows: the ion source voltage was 1.6 kV; the ion mobility range was 0.6–1.60 V.S/cm²; and the MS1 scanning range was set to 302–1077 m/z. A peak intensity above 2500 could be detected; the 302–1077 m/z scanning range was divided into four steps, and each of them was composed of eight windows for continuous window fragmentation and signal acquisition. The fragmentation mode was CID, the fragmentation energy was 10 eV, and the mass width of each window was 25. The cycle time of the DIA scan was 3.3 s.

2.5. Data Analysis

The data obtained from DDA-MS were processed using MaxQuant software (version 1.5.3.30) incorporating the Andromeda search engine [26]. The MS/MS spectra results were searched against a protein database predicted by the mature ovary and ovotestis transcriptomes CDS (unpublished) of G. fascicularis. The method used to predict the protein sequences was as follows: first, we performed CDS forecasting, and the candidate coding regions in unigene were detected by the Transdecoder (v3.0.1) software with default setting; then, we aligned these regions to SwissProt using BLAST. Subsequently, using Hmmscan, a search for homologous sequences of the Pfam protein and prediction of the coding regions were conducted. Finally, we converted the CDS sequences to protein sequences. In addition, common laboratory contaminant sequences from the cRAPome and symbiondinium proteome were also included in the search database. Protein identifications were accepted if the protein was composed of at least two identified peptides. The false discovery rate (FDR) threshold was set at 1% for both peptide spectrum matches and protein levels. The output file of MaxQuant was used as a standard spectral library for DIA-MS analysis. The raw data of DIA-MS were analyzed by Spectronaut (v.12.4, Biognosis, Schlieren, Switzerland), which utilized iRT peptides to calibrate retention time [27]. The target-decoy-based strategy was applied to control the FDR at the desired level (1%), yielding significant quantitative results. The correction and normalization of the intra-system error of each sample were achieved by the MSstats software package [28]. Significance test between the red eggs and spermatozoa groups was performed based on the model, evaluating significantly differentially abundant proteins (DAPs) between the two types of gametes using the filtration criteria of fold change (FC) ≥ 2 and p value < 0.05.

2.6. Bioinformatics Analysis

By searching against the Nr database (NCBI nonredundant protein sequences), KOG/COG, Swiss-Prot, and the KEGG database, the functions of the identified proteins were annotated. To better understand the biological characteristics of DAPs in eggs and spermatozoa, GO and KEGG enrichment analyses were performed by using OmicShare tools at www.omicshare.com/tools (accessed on 5 October 2024), and Bioinformatics at http://www.bioinformatics.com.cn/ (accessed on 7 October 2024). Hierarchical clustering was performed using the R package heatmap. Venn diagrams and volcano plots were drawn using the OmicShare tools at www.omicshare.com/tools, and GO chord diagrams were plotted in https://www.bioinformatics.com.cn.

The prediction of interacting partners in protein–protein interactions (PPI network) was performed by the STRING server (http://string.embl.de) (accessed on 10 October 2024). A medium confidence level (≥0.4) was applied for the minimum interaction score condition of the PPI network. The degree of each protein was calculated by Cytoscape (version 3.9.1) with default settings. The hub proteins were screened by betweenness centrality according to the cytoHubba plug-in, and the proteins with the top 10% of nodes were considered hub genes [29].

For the protein EUPHY, multi-sequence alignment was performed with the online tool Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/, accessed on 10 July 2024). Protein structure was determined using AlphaFold prediction [30,31]. A conserved domain search was performed with SMART. The molecular weight of the protein EUPHY and theoretical isoelectric point (pI) were acquired by the online tool at https://web.expasy.org/compute_pi/ (accessed on 10 July 2024). Orthologues of EUPHY were identified by Orthoscope. BLASTP at NCBI was used to calculate the sequence identities among aligned sequences, and MEGA 7.0 software was applied to build a phylogenetic tree with the neighbor-joining (NJ) algorithm using 1000 bootstrap replicates.

3. Results

3.1. Spawning Events, and Morphology and Size of Mature Fertile Gametes

Under laboratory conditions, the spawning of mature adult G. fascicularis began around 21:00 on 14 April 2023 (nine nights after the full moon) and continued for 2–3 days for each colony. Approximately 10–15 min before spawning, the red egg masses in female colonies and white egg–sperm bundles in hermaphroditic colonies became visible around the oral disk. Once spawning had started, the polyps’ mouths contracted, releasing the gametes simultaneously. When gametes were released into the water column, the red egg masses separated from one another, and the white egg–sperm bundles dissociated completely within a short time. In addition, split spawning events were observed, where polyps in the same colony spawned on different nights (Figure 1).

According to result of the scanning electron microscopy (SEM) observations, the mature red eggs were round or oval, with a longer diameter of approximately 301.36 ± 39.80 μm and a shorter diameter of approximately 265.63 ± 8.60 μm (mean ± SD, n = 4) (Figure 2B). The surface of the red eggs was not smooth and was covered with microvilli, each ending in a bulb-like bulge (Figure 2C). Mature spermatozoa consisted of a bullet-shaped head measuring approximately 1.91 ± 0.13 μm and a flagellum measuring approximately 34.06 ± 8.65 μm in length (mean ± SD, n = 4). The spermatozoa featured a cytoplasmic collar at the base surrounding the anterior portion of the flagellum (Figure 2E,F). The SEM samples of spermatozoa were acquired directly from the white egg–sperm mixture to keep their structure intact. Interestingly, the surface of white eggs we observed in egg–sperm mixture samples occasionally exhibited a distinctly different microvillus structure with more spacing compared to the red eggs.

3.2. Proteomics Profiles and Functional Annotation of Mature Red Eggs and Spermatozoa in G. fascicularis

To understand the reproductive molecular mechanisms of reef-building corals, the protein identification and quantification of red eggs and spermatozoa in G. fascicularis were accomplished through DIA-based proteomics. The SDS-PAGE results of extracted proteins showed distinct protein bands between red eggs and spermatozoa samples, indicating significant differences at the protein level (Figure S1).

In this study, a total of 90,299 peptides, corresponding to 8336 proteins, were identified and quantified by DIA-based proteomics. All these identified proteins had two or more unique peptides (Figure 3A), and most proteins had molecular weight (MW) values ranging from 20 to 60 kDa (Figure 3B). Additionally, out of the 8336 proteins, 7741 were present in red eggs, and 7279 were detected in spermatozoa. Among these proteins, 6683 (80.17% of the total) were shared between the two types of fertile gametes. Venn diagram analysis indicated that 596 proteins were spermatozoa-specific, and 1056 were egg-specific (Figure 3C). These sex-specific proteins will be a valuable resource for the future development of sex-specific markers for germ cells. All proteins identified in the red eggs and spermatozoa are listed in Tables S1 and S2, respectively.

To better understand the functional characteristics of these identified proteins, putative functions were annotated based on the GO database. Due to the vast similarity in protein components, the function annotation results between spermatozoa and eggs were similar. Among the 7741 proteins identified in red eggs, 4963 proteins had GO annotations and could be divided into 19, 19, and 12 terms for the biological process (BP), cellular component (CC), and molecular function (MF) categories, respectively (Figure S2A). Similarly, for mature spermatozoa, 4453 out of 7279 proteins had GO annotations, and could be divided into 18, 19, and 12 terms for the BP, CC, and MF categories, respectively (Figure S2B).

3.3. Functional Analysis of DAPs in Red Eggs and Spermatozoa

To identify candidate proteins involved in gamete maturation and fertilization, we firstly investigated proteins that were significantly differentially abundant in red eggs and spermatozoa. Our results revealed a total of 4413 proteins that were significantly differentially abundant in red eggs and spermatozoa (Table S3). Of these DAPs, there were 3121 up-regulated in red eggs, while 1292 proteins were up-regulated in spermatozoa (Figure 4A). The hierarchical cluster analysis demonstrated consistent abundance levels of proteins within three biological repeats of each gamete and clear distinctions between the two types of gametes (Figure 4B).

Then, we further performed GO and KEGG enrichment analysis for these DAPs separately. The enrichment analysis for significant DAPs in red eggs and spermatozoa revealed enriched GO terms (Figure 5A,B). For significant DAPs in red eggs, considerably enriched GO terms were included in the MF category, such as ATP binding, cadherin binding, and calmodulin binding; in the CC category, the cytoplasm and nucleus were the most enriched GO terms; in the BP category, the enriched GO terms were the protein metabolic process, embryo development ending in birth or egg hatching, and chordate embryo development (Figure 4A and Table S4). For significant DAPs in spermatozoa, membrane-bound organelles, the microtubule associated complex, and the dynein complex were considerably enriched in cellular components; transporter activity and proton transmembrane transporter activity were enriched in molecular function; and cellular processes, cilium movement, cilium organization, and cilium assembly were enriched in biological processes, respectively (Figure 4B and Table S5).

To further elucidate the associated biological pathways of these identified significant DAPs, we mapped them to reference canonical pathways in the KEGG database. For significant DAPs in red eggs, the highly enriched pathways demonstrated were DNA replication, progesterone-mediated oocyte maturation, lectins, and focal adhesion (Figure 5C and Table S6). Meanwhile, for significant DAPs in spermatozoa, pathways like oxidative phosphorylation, the citrate cycle (TCA cycle), insulin secretion, and cell motility were significantly enriched, suggesting that different pathways were involved during the maturation period of the two types of gametes (Figure 5D and Table S7).

3.4. Construction of a PPI

Based on the annotation result of DAPs, a relevant PPI network was constructed for core protein detection. In this study, the construction of the PPI network was based on the 238 proteins from 20 KEGG pathways associated with the cell cycle and endocrine and energy metabolism, yielding 2738 edges and 238 nodes. The PPI network identified hub proteins which mainly related to energy metabolism, including ATP synthase subunit alpha (ATP5A1), ATP synthase subunit gamma (ATP5C1), NADH dehydrogenase [ubiquinone] iron–sulfur protein 8 (NDUFS8), NDUFS2 and NDUFS3, succinate dehydrogenase A (SDHA), SDHB, and succinyl-CoA synthetase subunit α (SUCLG1). In addition, key proteins which might be involved in the cell cycle and reproduction were also identified, such as cyclin-dependent kinase 1 (CDK1), CDK2, cell division cycle protein 6 (CDC6), mitotic checkpoint serine/threonine-protein kinase BUB1-like (BUB1), BUB, proliferating cell nuclear antigen (PCNA), and wee1-like protein kinase 1-A (WEEL) (Figure 6).

3.5. Key Proteins Involved in Gamete Maturation and Fertilization

By searching specific GO keywords (such as reproduction GO:0000003, germ cell development GO:0007281, cell cycle GO:0007049, spermatogenesis GO:0007283, flagellated sperm motility GO:0030317, spindle assembly GO:0051225, and cilium assembly GO:0060271), as well as putative proteins/genes described in the bibliography that are involved in sexual reproduction in corals and other invertebrates, several candidate proteins potentially related to gamete maturation and reproduction were identified in the present study. In red eggs, at least 30 proteins were primarily involved in oocyte maturation and reproduction, including proteins like red fluorescent protein (RFP), vitellogenin proteins (VG), egg protein (EP), EUPHY, ATP-dependent RNA helicase bel (DDX3X), and low-density lipoprotein receptor-related protein 4 (LRP4), which have high abundance in red eggs (

Table 1. Candidate proteins highly abundant in red eggs, possibly relevant for oocyte maturation and reproduction; log2foldchange value (eggs/spermatozoa) with star (*) represent proteins only identified in red eggs.

Protein_ID	Swissprot Accession	Gene Symbol	Log2 FC	Protein Name
TRINITY_DN2957_c0_g1_i4-Sp	Q9BXT4	TDRD1	3.64	Tudor domain-containing protein 1
TRINITY_DN210_c0_g1_i1-Oc	O60671	RAD1	1.59	Cell cycle checkpoint protein RAD1
TRINITY_DN2472_c0_g1_i1-Oc	Q8K396	MND1	3.17	Meiotic nuclear division protein 1 homolog
TRINITY_DN1390_c1_g1_i1-Oc	Q9DGA5	CDK1	2.25	Cyclin-dependent kinase 1
TRINITY_DN15566_c0_g1_i1-Sp	P50613	CDK7	2.69	Cyclin-dependent kinase 7
TRINITY_DN3082_c0_g2_i1-Oc	P43450	CDK2	1.37	Cyclin-dependent kinase 2
TRINITY_DN1611_c0_g1_i1-Oc	Q00534	CDK6	2.28	Cyclin-dependent kinase 6
TRINITY_DN12844_c0_g2_i1-Sp	Q62623	CDC20	2.72	Cell division cycle protein 20 homolog
TRINITY_DN1422_c0_g1_i1-Oc	O43683	BUB1	2.82	Mitotic checkpoint serine/threonine-protein kinase BUB1-like
TRINITY_DN13050_c0_g1_i1-Oc	O43684	BUB3	3.13	Mitotic checkpoint protein BUB3
TRINITY_DN12_c0_g1_i1-Oc	P39963	CYCB	2.75	G2/mitotic-specific cyclin-B
TRINITY_DN10269_c0_g1_i2-Sp	Q95P04	RFP	2.33	Red fluorescent protein
TRINITY_DN1316_c0_g1_i1-Oc	Q9JMB7	PIWI1	3.02	Piwi-like protein 1
TRINITY_DN3806_c0_g1_i1-Sp	Q9VHP0	DDX3	3.48	ATP-dependent RNA helicase bel
TRINITY_DN808_c0_g1_i1-Sp	P61258	PCNA	2.48	Proliferating cell nuclear antigen
TRINITY_DN892_c0_g1_i2-Sp	Q90243	VG1	6.06	Vitellogenin (Fragment)
TRINITY_DN9308_c0_g2_i1-Oc	Q94637	VG3	9.58	Vitellogenin
TRINITY_DN821_c0_g1_i5-Oc	A2RUV0	NOTC1	7.70	Neurogenic locus notch homolog protein 1 (EUPHY)
TRINITY_DN17701_c0_g1_i1-Oc	B8VIU6	EP	3.64	Egg protein
TRINITY_DN17981_c0_g1_i1-Oc	P98157	LRP1	*	Low-density lipoprotein receptor-related protein 1
TRINITY_DN26695_c0_g1_i1-Sp	Q07954	PLRP1	1.92	Prolow-density lipoprotein receptor-related protein 1
TRINITY_DN26695_c0_g2_i1-Sp	Q8VI56	LRP4	3.87	Low-density lipoprotein receptor-related protein 4
TRINITY_DN18356_c0_g1_i1-Oc	G8HTB6	ZP	4.72	ZP domain-containing protein
TRINITY_DN3598_c0_g1_i12-Sp	P38529	HSF1	1.52	Heat shock factor protein 1
TRINITY_DN685_c0_g1_i1-Sp	O02705	HS90A	2.36	Heat shock protein HSP 90-alpha
TRINITY_DN4812_c0_g1_i1-Sp	Q3MSQ8	DDX4	3.23	Probable ATP-dependent RNA helicase DDX4
TRINITY_DN1577_c0_g1_i1-Oc	Q93105	INSR	*	Insulin-like receptor
TRINITY_DN2507_c0_g1_i1-Oc	Q8AYK6	WEE1	3.22	Wee1-like protein kinase 1-A
TRINITY_DN884_c0_g1_i7-Sp	P05556	ITB1	1.10	Integrin beta-1
TRINITY_DN2904_c0_g1_i1-Sp	A7S338	LIS1	1.85	lissencephaly-1 homolog

). Egg-specific proteins, such as Insulin-like receptor (LRP1) and low-density lipoprotein receptor-related protein 1 (INSR), were also identified. In addition, 18 candidate proteins were randomly selected, and the chord diagram shows the relationship between these candidate proteins and key GO terms (Figure 7A).

Furthermore, more than 27 proteins related to spermatogenesis, sperm motility, and fertilization were specifically or highly abundant in spermatozoa. These include spermatogenesis-associated protein family members (SPATAs), sperm-associated antigen family members (SPAGs), sperm flagellar proteins 1 and 2 (SPEF1 and SPEF2), EF-hand domain-containing protein 1-like (EFHC1), EFHC2, and other potential candidate proteins. Testis-specific serine/threonine-protein kinase (TSSK1, TSSK2, and TSSK4) proteins were spermatozoa-specific proteins (

Table 2. Candidate proteins highly abundant in spermatozoa, possibly relevant for spermatogenesis, sperm motility, and fertilization; log2foldchange value (eggs/sperms) with star (*) represent proteins only identified in spermatozoa.

Protein_ID	Swissprot Accession	Gene Symbol	Log2 FC	Protein Name
TRINITY_DN16925_c0_g1_i2-Sp	Q3SZW1	TSSK1	*	Testis-specific serine/threonine-protein kinase 1
TRINITY_DN8133_c1_g1_i1-Sp	O54863	TSSK2	*	Testis-specific serine/threonine-protein kinase 2
TRINITY_DN62707_c0_g1_i1-Sp	Q9D411	TSSK4	*	Testis-specific serine/threonine-protein kinase 4
TRINITY_DN2043_c0_g2_i1-Sp	Q9JLI7	SPAG6	−6.23	Sperm-associated antigen 6
TRINITY_DN2182_c0_g1_i1-Oc	Q3V0Q6	SPAG8	−8.38	Sperm-associated antigen 8
TRINITY_DN10173_c0_g1_i1-Oc	Q8K450	SPAG16	−3.61	Sperm-associated antigen 16
TRINITY_DN2541_c0_g1_i1-Oc	Q6Q759	SPAG17	−4.34	Sperm-associated antigen 17
TRINITY_DN1595_c0_g1_i3-Sp	Q6DMN8	SPATA4	−4.42	Spermatogenesis-associated protein 4
TRINITY_DN643_c0_g1_i2-Oc	Q9NWH7	SPATA6	−5.29	Spermatogenesis-associated protein 6
TRINITY_DN6811_c0_g2_i2-Sp	Q9D552	SPATA17	*	Spermatogenesis-associated protein 17
TRINITY_DN4634_c0_g4_i3-Sp	Q8TBZ9	TEX47	−6.37	Testis-expressed protein 47
TRINITY_DN129_c0_g1_i2-Oc	Q4R642	TCTE1	−3.91	T-complex-associated testis-expressed protein 1
TRINITY_DN295_c0_g1_i7-Sp	Q08CI4	CCYL1	−1.34	Cyclin-Y-like protein 1
TRINITY_DN11196_c0_g1_i1-Oc	Q9R095	SPEF2	−4.13	Sperm flagellar protein 2
TRINITY_DN16400_c0_g1_i2-Sp	Q0IH24	SPEF1	−4.27	Sperm flagellar protein 1
TRINITY_DN37130_c0_g1_i1-Sp	Q3UGF1	WDR19	−3.05	WD repeat-containing protein 19
TRINITY_DN1108_c0_g1_i1-Oc	Q969V4	TEKT1	−7.19	Tektin-1
TRINITY_DN1471_c0_g1_i1-Sp	Q9UIF3	TEKT2	*	Tektin-2
TRINITY_DN13549_c0_g1_i1-Oc	Q63164	DYH1	−5.13	Dynein heavy chain 1, axonemal
TRINITY_DN14153_c0_g3_i1-Sp	Q9P225	DYH2	−5.38	Dynein heavy chain 2, axonemal
TRINITY_DN1745_c0_g1_i1-Oc	Q9CQ46	EFCAB2	−4.17	EF-hand calcium-binding domain-containing protein 2
TRINITY_DN2971_c0_g1_i1-Sp	Q2KIU7	RSPH9	−4.45	Radial spoke head protein 9 homolog
TRINITY_DN140_c0_g5_i1-Sp	P92177	YWHAE	−1.45	14-3-3 protein epsilon
TRINITY_DN3137_c0_g1_i17-Sp	P62332	ARF6	−1.31	ADP-ribosylation factor 6
TRINITY_DN3184_c0_g1_i1-Sp	O35594	ITF81	−3.76	Intraflagellar transport protein 81 homolog
TRINITY_DN11002_c0_g2_i1-Sp	EFHC1	Q5JVL4	−6.40	EF-hand domain-containing protein 1-like
TRINITY_DN33226_c0_g1_i1-Sp	EFHC2	Q32TF8	−5.32	EF-hand domain-containing family member C2

). The relationships between 18 candidate proteins and key GO terms are depicted by a chord plot (Figure 7B).

3.6. Analysis of Protein EUPHY in G. fascicularis

Compared to the egg yolk proteins VG and EP, the protein EUPHY (Protein_ID: TRINITY_DN821_c0_g1_i5-Oc), which was annotated as “neurogenic locus notch homolog protein 1” in the Swissprot database, also had a high abundance in red eggs; thus, its biology characteristics were further investigated. For EUPHY, 65 unique peptides were identified, and the percent coverage was 75.52%. The protein EUPHY consisted of 321 amino acid residues, with a predicted molecular weight of 48.25 kDa and a deduced isoelectric point of 6.57. The multi-sequence alignment result showed that EUPHY in G. fascicularis shared a high sequence identity (89.38%) with Euphy (AKR17060.1) of Fimbriaphyllia ancora, and a moderate sequence similarity (46.60%) with NOTC1 (XP_020612749.1) of Orbicella faveolate (Figure S3). The 3D structure showed a high degree of structural similarity between EUPHY of G. fascicularis and Euphy of F. ancora, while they were distinct from NOTC1 of O. faveolate (Figure 8A). In addition, the protein domain composition was almost identical between the two euphy proteins mentioned above, and a little different from NOTC1 of O. faveolate based on the domain prediction result (Figure 8B). The sequence of EUPHY did not exhibit sequence similarity to the yolk proteins EP and VG in G. fascicularis. The phylogenetic analysis further revealed that EUPHY of G. fascicularis had a close relationship with Euphy of F. ancora (Figure 8C). According to a previous report, the protein Euphy is the third ovarian somatic-derived yolk protein in F. ancora, and it is transported to and accumulates in eggs during the process of oogenesis [32]. Thus, our results imply that the EUPHY protein we detected might be a novel yolk protein in G. fascicularis.

4. Discussion

The large-scale cultivation of corals offers an alternative method to wild harvest for ornamental trade and has great potential for restoring reefs and preserving biodiversity [33]. Currently, coral reefs face severe threats from multiple factors, including ocean warming [34,35], ocean acidification [36], and microplastics [37]. These stressors can impair coral reefs by decreasing live coral coverage directly. Meanwhile, more and more studies show that the sexual reproduction of corals is compromised, although some species may survive under certain environments. These stressors can reduce the quality and quantity of gametes, subsequently affecting embryo development, larval settlement, and overall reef resilience. Thus, understanding the reproductive biology of reef-building corals is essential for the successful sexual propagation of corals and forms a crucial basis for scientifically effective coral reef restoration. Consequently, characterizing the protein profiles of fertile gametes and identifying key proteins involved in reproduction are essential for evaluating gamete quality and reproductive success in corals.

A comparative transcriptome analysis between mature spermatozoa and eggs in Montipora capitata was performed and detected only 247 differentially expressed genes, which suggests that the mature eggs and spermatozoa of corals are not highly differentiated with respect to functional capability [38]. The DIA-based proteome offers the advantages of quantifying thousands of proteins with low variation and high reproducibility [39]. In this study, we identified a total of 7741 and 7279 proteins in mature red egg and spermatozoa samples, respectively. Of these, 6683 proteins were common to both gamete types, while 596 and 1056 proteins were specific to spermatozoa and eggs, respectively. Due to the fact that the majority of proteins in both spermatozoa and eggs are identical, their functional annotation results are very similar, which is consistent with the result in M. capitata [38]. Additionally, 4413 DAPs out of the 6683 shared proteins were detected when comparing the protein abundance in spermatozoa and red eggs. Therefore, the utilization of a high-precision DIA-MS method for the first time, to characterize the molecular identities of proteins in fertile female and male gametes of hermatypic corals, enabled us to uncover gamete maturation- and reproduction-related key proteins, paving the way for coral sexual propagation efforts and offering valuable insights for coral conservation research.

Sexual reproduction is believed to be regulated by the endocrine system in vertebrates, including mammals, fish, and birds [40,41,42]. Understanding the gamete maturation and reproduction regulation mechanisms is very important for sexual propagation. Because the regulatory mechanisms of gamete maturation and reproduction are quite clear in fishes, the gonadotropin-releasing hormone (GnRH), Human Chorionic Gonadotropin (HCG), and Common Carp Pituitary Gland (CPG) are effectively used in the artificial breeding of fishes [43]. In corals, the endocrine system is a significant endogenous factor regulating reproduction and spawning [44,45,46,47]. The annual profiles of sex steroids, aromatase, and immunoreactive gonadotropin-releasing hormone (irGnRH) have been characterized in E. ancora previously, and it has been inferred that irGnRH and glucuronided estradiol may play crucial roles in the regulation of reproduction and mass spawning in corals [48,49]. In the broadcast spawner coral Acropora tenuis, estradiol-17β (E₂) is considered to be involved in the process of oogenesis, but is not responsible for the regulation of vitellogenin synthesis [50]. It has been confirmed that the homogenates of M. capitata could convert estradiol to estrone and testosterone to androstenedione and androstenedione, which is evidence that 17β-hydroxysteroid dehydrogenase and 5α-reductase are present in M. capitata [46]. Although some studies have reported that the levels of sex steroids increase in the tissues of several corals during the onset of spawning [45,49,50,51], and hypothesize that sex steroids are associated with gamete maturation, the endogenous regulatory mechanisms remain unknown. In this study, significant DAPs in red eggs were significantly enriched in progesterone-mediated oocyte maturation pathways, revealing one of the regulatory pathways of oocyte maturation and providing valuable information for further study. In addition, the spermatozoa maturation process was not only regulated by the endocrine system (e.g., insulin secretion), but was also accompanied by the process of energy metabolism, as revealed by the KEGG enrichment result.

The oocyte maturation process comprises the stages of oogonia, previtellogenesis, and vitellogenesis [52]. Vitellogenesis occurs as an important process for the accumulation of yolk proteins, fats, and other nutrients. The accumulated nutrients, especially yolk proteins, are essential for embryo development and larval survival [53,54]. VG, a precursor protein of the yolk, is processed into the most abundant component of the yolk and serves as a major energy source supporting embryogenesis [55]. To date, VG and the egg protein (EP), the two primary types of major yolk protein precursors, have been identified in F. ancora and G. fascicularis. Studies have shown that these proteins are synthesized by somatic cells in the mesenteries, accumulated in eggs, and consumed during embryonic development [20,21,56]. In G. fascicularis, the two types of yolk protein precursors are encoded by different genes: GfVg encodes a putative vitellogenin (VG)-like precursor and is stored as GfEP-1 to -3 in mature eggs, and GfEP-1 to -3 share high sequence similarity with VG of other oviparous vertebrates and invertebrates, whereas GfEP-4 encodes a distinct yolk protein (EP) with a different sequence from VG [20,22]. Subsequently, a third ovarian somatic-derived yolk protein precursor, Euphy, is identified in the eggs of F. ancora. This Euphy protein contains three alternating repeats of fibronectin domain 2 (FN2), epidermal growth factor (EGF)-like domains, and an additional calcium-binding EGF-like domain (EGF-CA) [32]. In the present study, we identified that VG and EP proteins were abundantly stored in red eggs, which was consistent with previous reports. Furthermore, we discovered another protein, EUPHY, which was also highly abundant in red eggs. In addition, EUPHY consisted of three alternating repeats of FN2- and EGF-like domains, along with an EGF-CA domain, and had a high degree of structural similarity and a close phylogenetic relationship with Euphy of F. ancora, indicating the presence of a third yolk protein in G. fascicularis. It is reported that the euphy-like protein might be specific to some scleractinian taxa in cnidarian, and evolves in a unique way [32]. This is the first time the third yolk proteins in G. fascicularis have been reported, which not only provides valuable molecular information for oocyte maturation, but also has potential for future biological applications.

In organisms like sea bass, rain trout, and insects, vitellogenin receptor (VGR) is exclusively localized in oocytes and has a high affinity for VG; additionally, the expression level of VGR transcript is up-regulated during the initial phase of the uptake of VG, indicating that VGR is important for yolk protein absorption and oocyte maturation [57,58,59]. Meanwhile, the process of VG uptake in many corals has yet to be clearly elucidated. In A. tenuis, low-density lipoprotein receptor (LDLR) is found in ovarian and mesentery tissues and is highly expressed during the vitellogenic stages, suggesting its role in transporting vitellogenin from mesenterial somatic cells to oocytes [50]. In the present study, LDLR was specifically present in red eggs, and PLRP1 (pro-low-density lipoprotein receptor-related protein 1) and LRP4 (low density lipoprotein receptor) were significantly differentially abundant in red eggs. Our results provide a basis for further exploration of the VG uptake process in G. fascicularis. Furthermore, other candidate proteins that might be closely connected with oocyte maturation and reproduction were also identified. Fluorescent proteins (FPs) are important color determinants in reef-building corals [60]. Previous studies have shown that FPs are present in the oocytes of Pocillopora verrucosa and M. capitata, and they have confirmed the expression of a novel RFP in E. ancora, which plays a significant role in coral oocyte maturation and protects oocytes from oxidative stress [61,62,63]. In this study, we observed a highly abundant level of RFP in red eggs, suggesting its potential involvement in oocyte maturation in G. fascicularis.

Spermiogenesis is a highly complex and tightly regulated process involving acrosomal biogenesis, chromatin condensation, flagellar assembly, and the disposal of excess cytoplasm [64]. According to previous reports, the TSSK family plays a critical role in spermatogenesis in vertebrates, with most members being sperm-specific proteins [65,66,67]. For instance, deficiencies in TSSK1 and TSSK2 have been shown to lead to male sterility in mice, while the depletion of TSSK4 results in sub-fertility due to severely reduced sperm motility [68,69,70]. In giant clams Tridacna squamosa, three TSSKs (TSSK1, TSSK4, and TSSK5) have been detected, and they exhibit high expression levels in male gonads, indicating the important role of TSSKs in giant clam spermatogenesis [71]. In Bay Scallops Argopecten irradians, five TSSKs (TSSK1/2, TSSK3, TSSK4, TSSK5) have been identified in mature testes, and they are exclusively localized in the spermatids and spermatozoa, suggesting that TSSKs may play a crucial role in spermiogenesis [72]. In the present study, TSSK proteins were abundant and exclusively presented in mature spermatozoa. This result was similar to those found in A. irradians [72], indicating significant roles of TSSKs in coral spermatogenesis.

In broadcast-spawning corals, spermatozoa and eggs are released into the water column, and fertilization only occurs when the spermatozoa reach the eggs through flagellar propulsion; thus, in spermatozoa, flagellar motility is necessary for fertilization [73]. In the coral Acropora spp., it has been demonstrated that the presence of chemical signals secreted by eggs could activate sperm flagellar motility, and this motility initiation by eggs is species-specific [74]. The elevation of intracellular pH and the regulation of intracellular Ca²⁺ concentration are associated with the sperm motility regulation cascade [74]. In ascidian Ciona intestinalis, the Ca²⁺-binding proteins EFHC1 and EFHC2 were detected in sperm and localized in axonemes, with the authors speculating that they are involved in the Ca²⁺-dependent regulation of sperm motility [75]. In the present study, 13 DAPs were significantly enriched in GO terms of “proton transmembrane transporter activity” (Figure 4B). In addition, we identified proteins (EFHC1, EFHC2) with EF-hand domains that could potentially be Ca²⁺-binding proteins involved in the Ca²⁺-dependent regulation of sperm motility. The molecular functions of these proteins need to be further explored to find key proteins that affect the regulation of sperm motility.

5. Conclusions

So far, the intrinsic characteristics of reproduction is elusive in reef-building corals, highlights the necessity of conducting reproductive biology research on them. In this study, we comprehensively elucidated the proteome profiles of fertile red eggs and spermatozoa, revealing the intricate protein components in the fertile gametes of G. fascicularis. Our results showed that a large number of proteins (6683 proteins) were common to both gametes, suggesting that red eggs and sperm are not highly differentiated with respect to functional capability. In addition, a lot of sex-biased or sex-specific proteins were detected and shed light on the possible molecular mechanisms of gamete maturation and fertilization. Moreover, enrichment analyses of these sex-biased proteins indicated that different regulation pathways were involved in the maturation processes of sperm and red eggs. Our study and its generated datasets provide a foundation for future studies regarding gamete maturation and fertilization from molecular perspectives. Furthermore, our proteomic data will be a useful reference for the future development of sex-specific markers for germ cells for use in coral aquaculture and ecological studies.