Integrated analysis of phosphoproteome and ubiquitylome in epididymal sperm of buffalo (Bubalus bubalis)

Abstract In mammals, sperm need to mature in the epididymis to gain fertilization competency. However, the molecular mechanism underlying buffalo sperm maturation remains elusive. Exploring sperm physiology at the posttranslational modification (PTM) level could help to develop our understanding of these mechanisms. Protein phosphorylation and ubiquitination are major PTMs in the regulation of many biological processes. In the present study, to our knowledge, we report the first phosphoproteome and ubiquitylome of sperm collected from the caput, corpus, and cauda segments of the epididymis using liquid chromatography–mass spectrometry combined with affinity purification. In total, 647 phosphorylation sites in 294 proteins and 1063 ubiquitination sites in 446 proteins were characterized. Some of these proteins were associated with cellular developmental processes and energy metabolic pathways. Interestingly, 84 proteins were both phosphorylated and ubiquitinated, simultaneously. Some of these proteins were involved in, for example, spermatogenesis, reproduction, and spermatid development. Taken together, these data provide a theoretical basis for further functional analysis of phosphorylation and ubiquitination in epididymal sperm of buffalo and other mammals, and serve as an important resource for exploring the physiological mechanism underlying sperm maturation.


| INTRODUCTION
To acquire progressive motility and the ability of fertilization, spermatozoa released from the testis must transit through a specialized duct called the epididymis (Marchiani et al., 2017). This tissue can be generally separated into three regions based on histological and ultrastructural differences: the caput (head), corpus (body), and cauda (tail; Sullivan et al., 2005). The caput and corpus serve as the sites for early and late sperm maturation, respectively, while the cauda region is responsible for storing functionally mature spermatozoa (Cheng et al., 2018;Dacheux & Dacheux, 2013). During epididymal transit, spermatozoa progressively lose or modify most of their surface proteins and gain new proteins in a well-organized manner. Numerous highthroughput proteomic studies have characterized proteins that are possibly involved in the appearance and maintenance of the fertility of the male gamete in the epididymis (Chauvin et al., 2012; proteasome system facilitates intracellular protein degradation and serves as the quality control system of the cell (Hochstrasser, 1995).
The phosphoproteome and ubiquitylome in epididymal sperm maturation have not been studied, to our best knowledge. Thus, comprehensively understanding phosphorylation and ubiquitination through proteome analysis in epididymal sperm is necessary.
Buffalo (Bubalus bubalis) is of considerable economic and biological interest especially in southern China; thus, a more robust understanding of the molecular mechanisms underlying epididymal sperm maturation in this ruminant species is of great value. Here, to our knowledge, we report the first phosphorylation and ubiquitination proteomic profiles of buffalo epididymal sperm using liquid chromatography-mass spectrometry (LC-MS/MS). Integrative phosphoproteome and ubiquitylome analyses suggested that epididymal sperm maturation may be partially due to the crosstalk between phosphorylation and ubiquitination. The findings further our understanding of the mechanisms underlying epididymal sperm maturation and offer a new perspective for future research into male reproduction.

| Proteomics analyses of phosphorylation and ubiquitination in epididymal sperm
In the present study, we performed global phosphorylation and ubiquitination proteome analysis of buffalo epididymal sperm using tryptic digestion, affinity enrichment, and LC-MS/MS. In total, 647 phosphorylation sites distributed on 685 peptides in 294 proteins and 1063 ubiquitination sites distributed on 1052 peptides in 446 proteins were identified with high confidence (Figure 1a). Of which, the 647 phosphorylation sites were composed of 591 phosphorylated serine (pS), 45 phosphorylated threonine (pT), and 11 phosphorylated tyrosine residues (Figure 1b). Detailed information for all identified phosphorylation and ubiquitination peptides and their matched proteins are presented in Tables S1 and S2.
We sorted the phosphorylated and ubiquitinated proteins according to the number of phosphorylation and ubiquitination sites, respectively, as shown in Figure 1c. Among the 294 phosphorylated proteins, 177 (60.2%) contained one phosphorylation site and 56 (19%) contained two phosphorylation sites. Notably, 24 (8.2%) proteins contained five or more phosphorylation sites, and eight had at least 10 sites (Table S3). Fibrous sheath interacting protein 2 was the most intensely phosphorylated protein bearing 31 different phosphorylation sites. Moreover, among the 446 ubiquitinated proteins, 251 (56%) had only one ubiquitinated site and 89 (19%) had two ubiquitinated sites. Notably, 50 (11.2%) proteins contained five or more ubiquitination sites, and 18 had at least 10 sites (Table S4).
Maestro heat like repeat family member 2B was the most intensely ubiquitinated protein bearing 29 different ubiquitination sites. The length of most phosphorylated and ubiquitinated peptides ranged from 7 to 35 amino acids, which is consistent with the property of tryptic peptides (Figure 1d). Interestingly, we compared the phosphorylation proteins with the ubiquitination data set and found that 84 phosphorylated proteins ( Figure 1e and Table 1) were simultaneously ubiquitinated.

| Functional annotation and pathway analysis of phosphorylated and ubiquitinated proteins
To investigate the possible biological roles of phosphorylated and ubiquitinated proteins, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed.
GO analysis based on the biological process category showed that spermatogenesis, reproduction, and spermatid development were the most significantly enriched in the phosphorylated proteins.
However, in the ubiquitinated proteins, the catabolic process was ranked the highest, followed by sperm-egg recognition and transport. Notably, the catabolic process, reproduction, spermatogenesis, spermatid development, fertilization, and sperm motility were enriched in both phosphorylated and ubiquitinated proteins. When enrichment analysis of the cellular components was performed, the phosphorylated proteins were significantly enriched in the sperm part, sperm flagellum, and acrosomal vesicle. These ubiquitinated proteins were enriched significantly on the cytoplasm, sperm part, and organelle. Both phosphorylated and ubiquitinated proteins were significantly enriched in the sperm part. Based on the molecular function category, protein kinase A binding, protein binding, and RNA binding were primarily enriched in the phosphorylated proteins. proteins. Additional phosphorylated proteins were mapped to KEGG pathways, including the pentose phosphate pathway, ribosome, and fructose and mannose metabolism. In addition, protein processing in the endoplasmic reticulum, ubiquitin-mediated proteolysis, and the AMP-activated protein kinase (AMPK) signaling pathway were also enriched in ubiquitinated proteins ( Figure 2).

| Integrated analysis between phosphoproteome and ubiquitylome
We compared the phosphoproteome and ubiquitylome data obtained from buffalo epididymal sperm and identified 84 proteins modified with both phosphorylation and ubiquitination ( Figure 1e and Table 1). These proteins participated in multiple biological processes, especially spermatogenesis, sexual reproduction, and the catabolic process. The most important cellular components were located in the sperm part, sperm flagellum, and motile cilium. Besides, the significantly enriched molecular functions were protein kinase A binding and small molecule binding. Furthermore, pathways including proteasome, glycolysis/gluconeogenesis, biosynthesis of amino acids, and carbon metabolism were significantly enriched in these 84 proteins ( Figure 5a).

| Confirmation of phosphorylation and ubiquitination of certain proteins
Four candidate total and phosphorylated proteins, PSMA3, PSMA3 (pS250), RAF1, and RAF1 (pS621), were confirmed by immunofluorescence analysis. As shown in Figure 6a, PSMA3 and PSMA3 (pS250) were expressed in the acrosome, neck, and tail of epididymal sperm, RAF1 was mainly expressed in the acrosome and tail, whereas RAF1 (pS621) was only located in the tail. The Western blot results also showed that the proteins PSMA3, PSMA3 (pS250), RAF1, and RAF1 (pS621) were expressed in the buffalo epididymal sperm ( Figure 6b). Furthermore, we performed immunoprecipitation (IP) and Western blot analysis to confirm the ubiquitination of two proteins (UCHL1 and HSPA2) from the ubiquitylome data ( Figure 6c). The results of the IP/Western blot analyses showed that the proteins UCHL1 and HSPA2 were presented in the ubiquitin-IP pulldown, and the protein ubiquitin was presented in the UCHL1-IP and HSPA2-IP pulldown as well, compared with the immunoglobulin G (IgG) control.
These results indicated the reliability of our phosphoproteome and ubiquitylome data set.

| DISCUSSION
In our study, we successfully identified 647 phosphorylation sites in 294 proteins and 1063 ubiquitination sites in 446 proteins in buffalo epididymal sperm by LC-MS/MS. The phosphorylated and F I G U R E 2 Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of identified phosphorylated and ubiquitinated proteins. The x-axis represents −log (p value), and the y-axis represents the name of GO categories (biology process, subcellular localization, and molecular function) and KEGG pathways F I G U R E 3 Glycolysis/gluconeogenesis pathway. Blue represents the phosphorylated proteins, red represents the ubiquitinated proteins, and yellow represent proteins both undergo phosphorylation and ubiquitination ubiquitinated proteins in epididymal sperm account for 12.5% and 19% of the total proteins (2344, data not shown), respectively. The small number of identified phosphorylated sites may be due to the single-phosphorylated enrichment method and the nature of epididymal sperm (Kuo et al., 2016). Recently, immunoaffinity reagents that are capable of capturing K-GG peptides from ubiquitin and its thousands of cellular substrates have been reported (Udeshi et al., 2013;Wagner et al., 2011). A limitation of this approach is that it does not distinguish between various types of modification, such as monoubiquitination, polyubiquitination, and the rare NEDD8 and ISG15 modifications (Bustos et al., 2012;Kim et al., 2011). However, the expression of NEDD8 in mammalian tissues was shown to be developmentally downregulated (Kamitani et al., 1997), and ISG15 expression in bovine tissues was low in the absence of interferon stimulation . Cell-culture experiments have shown that most sites identified using di-glycine-lysine-specific antibodies stem from ubiquitylated peptides. Thus, most di-Gly remnants derived from cellular peptides are derived from ubiquitinated proteins (Kim et al., 2011). Therefore, in the present study, we referred to all di-Gly modified lysines as "ubiquitination sites," even though a small portion of these sites may have been derived from a NEDD8 or an ISG15 modification.
Until recently, the phosphoproteome and ubiquitylome were most frequently reported in mammalian cells and tissues in human and murine (Danielsen et al., 2011;Huttlin et al., 2010;Kim et al., 2011;Martin-Hidalgo et al., 2020;Qi et al., 2014;Wagner et al., 2012;Zahedi et al., 2008). A comparative phosphoproteome was performed to find the phosphorylation difference between buffalo epididymal sperm (Table S1)   were common and 779 were newly identified. Further analysis demonstrated that the common ubiquitination proteins were significantly enriched in a variety of catabolic processes, while the newly identified ubiquitination proteins were associated with catabolic processes, sperm-egg recognition, sexual reproduction, fertilization, and spermatogenesis (Table S7). This indicated that phosphorylation and ubiquitination are conserved in mammalian and play both common and specific roles in different cells or tissues in different mammalian species.
We annotated phosphorylated and ubiquitinated proteins to specific cellular processes and pathways based on bioinformatics analysis. In the reproduction process, 35 proteins were phosphorylated, 37 were ubiquitinated, and 13 of these were both phosphorylated and ubiquitinated. For example, calicin (CCIN) is found in a diverse range of mammals and is a basic cytoskeletal protein of sperm cells (Paranko, 1990;Paranko et al., 2010). In our study, four phosphorylation sites (S391, S12, S58, and S97) and one ubiquitination site (K400) were found in protein CCIN. Besides, the deficiency in ROPN1 protein in mouse sperm can significantly reduce its fertility (Fiedler et al., 2013). Our previous research demonstrated that ROPN1 was specifically localized in the principal piece of buffalo spermatozoa (Huang et al., 2016). Here, we identified three phosphorylation sites and five ubiquitination sites in ROPN1. Spermatogenesis-associated protein 19 (SPATA19) is crucial for sperm mitochondrial function and male fertility (Mi et al., 2015). Eleven sites (S133, S84, S142, S91, S37, S59, S26, S116, S138, S107, and S82) were phosphorylated in this protein, while only one site (K83) was ubiquitinated. This may be due to the intense phosphorylation of protein SPATA19 during sperm to the high-affinity binding between the sperm and zona pellucid (Wen et al., 1999;Yamasaki et al., 1995). In the present study, we verified that SPA17 was ubiquitinated at K52, as well as activating phosphorylation at S54.
In addition, four proteins (CAPZA3, ROPN1, PSME4, and PRKACA) in the spermatid developmental process and two proteins (LDHC and AKAP4) participated in sperm motility were both phosphorylated and ubiquitinated. AKAP4, the most abundant protein in the fibrous sheath of the sperm tail, was mainly expressed in the postmeiotic phase of spermatogenesis. Miki et al. (2002) showed that AKAP4 deficiency will lead to the loss of effective motility in sperm.
In particular, epididymal sperm proteins carried out acrosome reactions and sperm capacitation mainly by activating phosphorylation. The glycolysis/gluconeogenesis pathway is a primary source of energy for sperm motility (Nascimento et al., 2008;Tourmente et al., 2015). GAPDH, GAPDHS, ALDOC, ALDOA, and LDHC were simultaneously phosphorylated and ubiquitinated in buffalo epididymal sperm. GAPDH activity was a parameter for determining sperm motility (Fu et al., 2017). GAPDHS, a sole GAPDH isozyme in sperm with four phosphorylated sites and six ubiquitinated sites identified, is required for sperm motility and male fertility (Miki et al., 2004).
Two additional glycolytic enzyme subunits, fructose-bisphosphate aldolase ALDOC and ALDOA, are tightly bound to the fibrous sheath of mouse spermatozoa and essential for sperm motility (Krisfalusi et al., 2006). LDHC is abundant in spermatocytes, spermatids, and sperm and is also required for male fertility (Odet et al., 2008). Three phosphorylated sites (S321, S161, and S105) and one ubiquitinated site (K232) were identified in buffalo epididymal sperm in LDHC.
Surprisingly, L-lactate dehydrogenase a chain (LDHA), an LDHC isozyme, phosphorylated at S105 and S161 rather than ubiquitinated in epididymal sperm in our study. LDHA is mainly responsible for the restoration of sperm function and fertility (Tang et al., 2013).
Moreover, our previous work showed that EEF1G was expressed in the nucleus of round spermatids in buffalo (Huang et al., 2016). Here, three ubiquitinated sites, K147, K212, and K404, were newly identified in the EEF1G in epididymal sperm.

F I G U R E 5 Continued
The proteasome participates in sperm capacitation, fertilization, and acrosome reaction (Kong et al., 2009;Morales et al., 2003;Sawada et al., 2010). A considerable portion of proteins identified in our phosphoproteome and ubiquitylome participated in the proteasome pathway, of which, eight proteins (PSMC1, PSMA4, PSMD4, PSME4, CCT3, CCT8, VCP, and OAZ3) were both phosphorylated and ubiquitinated. PSME4, a proteasome activator, is required for normal spermatogenesis and male fertility (Khor et al., 2006). A few years ago, PSME4 was identified in the mouse epididymal sperm proteome (Skerget et al., 2015). In our study, we first identified two serine phosphorylation sites (S293 and S294) and six lysine ubiquitination sites (K918, K636, K56, K36, K1501, and K574) on PSME4 that may be critical for sperm maturation. PSMD4, the 19S regulatory complex subunit with one phosphorylation site and one ubiquitination site identified, is involved in the sperm-zona pellucida penetration during fertilization (Yi et al., 2010). Meanwhile, two subunits of T-complex protein 1 (CCT3 and CCT8) are crucial for spermiogenesis (Counts et al., 2017). OAZ3 is expressed specifically in germline cells and is essential for the formation of a rigid connection between the head and tail during spermatogenesis in mice (Tokuhiro et al., 2009). Here, we identified two serine phosphorylation sites S9 and S55, one tyrosine phosphorylation site Y11, and one ubiquitination site K17 in the OAZ3 protein.
In the present study, we first confirmed that the phosphorylated protein PSMA3 (pS250) was expressed in the acrosome, neck, and tail of epididymal sperm, whereas RAF1 (pS621) was only expressed in the tail. The phosphorylation site S250 of PSMA3has not been previously reported in epididymal sperm. Serine 621 has been previously demonstrated as an important phosphorylation site of RAF1, and it plays a crucial role in catalytic activity and negatively regulates the RAF1 protein (Mischak et al., 1996). Furthermore, the ubiquitination of two proteins (UCHL1 and HSPA2) from ubiquitylome data were confirmed by IP and Western blot analysis. UCHL1 plays a key role in mitotic proliferation and differentiation of spermatogonial stem cells and fertilization (Jungkee et al., 2004;Mtango et al., 2012;Wang et al., 2006). One ubiquitination site, K4, was identified in UCHL1 in epididymal sperm. The molecular chaperone HSPA2 has been verified to play a critical role in meiosis, spermatogenesis, male fertility, and sperm-oocyte binding (Bromfield et al., 2015(Bromfield et al., , 2016Rogon et al., 2014). Here, we identified 10 ubiquitination sites (K-127, 57, 139, 160, 129, 188, 78, 109, 72, and 89) in HSPA2. The

| Trypsin digestion
Trypsin digestion was performed as previously (Wei et al., 2019) described with minor modification. Briefly, the protein solution was reduced with 5 mM dithiothreitol for 30 min at 56°C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in the dark. The protein sample was then diluted by adding 100 mM NH 4 HCO 3 to a urea concentration of less than 2 M.
Finally, trypsin was added to the diluted protein sample at 1:50 trypsin-to-protein mass ratio for the first digestion overnight and 1:100 trypsin-to-protein mass ratio for the second digestion for 4 h. The peptides thus obtained were quantified by Pierce Quantitative Fluorometric Peptide Assay (Thermo Scientific).

| Affinity enrichment of phosphorylated peptides
The phosphopeptide was enriched by TiO 2 as previously described (Wei et al., 2019). A total of 3 mg of tryptic peptides was desalted using Sep-Pak Classic C18 Columns (Waters) and phosphopeptide enrichment using a High-Select TiO 2 phosphopeptide enrichment Kit (Thermo Scientific) following the recommended protocol. Briefly, desalted lyophilized peptides were dissolved in the binding/equilibration buffer provided with the kit and centrifuged to clarify the dissolved peptides. TiO 2 spin tips were washed twice with wash buffer and equilibrated once with binding/equilibration buffer before loading peptides. Phosphopeptides were allowed to bind to the TiO 2 resin followed by sequential washing with binding buffer and wash buffer.
Finally, bound phosphopeptides were eluted using elution buffer and lyophilized quickly to avoid dephosphorylation. Tyr residue (phosphorSTY), and Gly-Gly modification for lysines as variable modifications in the phosphoproteome and ubiquitylome analysis, respectively. The false discovery rate was adjusted to less than 1%, and the minimum score for modified peptides was set at greater than 40. The site localization probability was set at greater than 0.75.

| Bioinformatic analysis
GO annotation proteome was derived from the UniProt-GOA database (http://www.ebi.ac.uk/GOA/; Huntley et al., 2014), and the lysine ubiquitination proteins were classified by GO annotation involving three categories: biological processes, molecular functions, and cellular components. The KEGG database (Kanehisa et al., 2000) was used to perform pathway analysis. The functional annotation tool of DAVID bioinformatics resources 6.8 (https://david.ncifcrf.gov; Jiao et al., 2012) was used to identify GO terms and KEGG pathways.
The protein-protein interaction network among the surveyed proteins was retrieved from the STRING database (version 10.5) with a confidence score of at least 0.7, and the interaction network was visualized in Cytoscape software (Tay et al., 2017).

| Immunofluorescence analysis
The immunofluorescence analysis was performed as previously described (Huang et al., 2016)

| IP and Western blot analysis
To confirm the ubiquitination of proteins utilizing the K-ε-GG antibody, we performed IP and Western blot analysis. The lysed cell extracts were immunoprecipitated with anti-ubiquitin antibody (ab105015; Abcam), anti-HSPA2 antibody (CSB-PA010824ESR2HU; CUSABIO BIOTECH CO.), anti-UCHL1 antibody (ab108986; Abcam), or rabbit IgG (ab205718; Abcam) using protein A+G agarose (Beyotime), after shaking at 4°C for 3 h, the supernatant was carefully removed by centrifugation at 1000g for 5 min, and the precipitate was washed five times with PBS buffer (8 mM Na 2 HPO4, 1.5 mM KH 2 PO4, 135 mM NaCl, and 2.7 mM KCl), then 20 μl 1X sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer (CWBIO) was added to the precipitate and used for electrophoresis after boiling at 100°C for 5 min. After electrophoresis, the SDS-PAGE gel was transferred onto a polyvinylidene difluoride membrane by a semidry Western blot analysis system (Bio-Rad). The membrane was blocked with 5% nonfat milk in a TBST solution for 2 h at room temperature and then incubated overnight at 4°C with primary antibody against UCHL1 (7863-1004; AbD Serotec), HSPA2 (bs-18080R; Bioss Biotechnology Inc.), and ubiquitin (bs-7944R; Bioss Biotechnology Inc.) at a dilution of 1:1000.
Membranes washed with TBST buffer three times were incubated with horseradish peroxidase-conjugated secondary antibodies (CWBIO) in TBST buffer for 1.5 h at room temperature. Bands were visualized with an ECL Detection Kit.

| CONCLUSION
In conclusion, the present study is the first to report on the phosphoproteome and ubiquitylome of epididymal sperm in adult buffalo.
All the proteins identified from our large-scale analysis were involved in numerous biological activities. Protein phosphorylation and ubiquitination could play roles as switches to control some key enzyme activities and assure the proper function of epididymal sperm development and maturation. Thus, we also provided molecular targets for the analysis of sperm maturation at the level of PTMs, but further studies are needed to elucidate the regulatory roles of phosphorylation and ubiquitination in epididymal sperm.