Polyubiquitin gene Ubb is required for upregulation of Piwi protein level during mouse testis development

Testis development, including early embryonic gonad formation and late postnatal spermatogenesis, is essential for the reproduction of higher metazoans to generate fertile gametes, called sperm. We have previously reported that the polyubiquitin gene Ubb is required for fertility in both male and female mice. In particular, the Ubb-null male mice showed an azoospermia phenotype due to arrest of spermatogenesis at the pachytene stage. Here, we analyzed the whole testis proteome at postnatal day 20 to define the molecular mediators of the male-infertility phenotype caused by Ubb knockout. From the identified proteome, 564 proteins were significantly and differentially expressed in Ubb-knockout testes and, among these, 36 downregulated proteins were involved at different stages of spermatogenesis. We also found that levels of piRNA metabolic process-related proteins, including Piwil2 and Tdrd1, were downregulated in Ubb-null testes through functional gene ontology analysis. Further, protein–protein interaction mapping revealed that 24 testis development-related proteins, including Hsp90aa1, Eef1a1, and Pabpc1, were directly influenced by the depletion of ubiquitin. In addition, the reduced mRNA levels of these proteins were observed in Ubb-knockout testes, which closely resembled the global downregulation of piRNA-metabolic gene expression at the transcriptional and post-transcriptional levels. Together with proteomic and transcriptional analyses, our data suggest that Ubb expression is essential for the maintenance of testicular RNA-binding regulators and piRNA-metabolic proteins to complete spermatogenesis in mice.


INTRODUCTION
Ubiquitin (Ub) is a highly conserved eukaryotic protein encoded by mono-and polyubiquitin genes in mammals [1][2][3]. In mice, two monoubiquitin genes (Uba52 and Uba80) encode a single Ub fused with each ribosomal protein, and two polyubiquitin genes (Ubb and Ubc) comprise tandem repeats of 4 or 9 Ub-coding units, respectively [4]. Cellular Ub, which is synthesized de novo, is freely available for well-studied biochemical reactions defined as ubiquitination-catalyzed by E1, E2, and E3 enzymes [5]. So far, many studies have shown that hundreds of different intracellular proteins are monoubiquitinated or polyubiquitinated [6]. These complicated Ub signals are accurately recognized by receptor proteins that have Ub-binding domains, performing a wide range of biological functions, such as DNA repair, cell cycle regulation, protein degradation, and cell signal transduction [6]. Therefore, in order to maintain the intracellular Ub levels, monoubiquitin genes are always highly expressed, and polyubiquitin genes are also expressed at high levels, which may be regulated depending on the cellular condition [7][8][9]. In contrast, depletion of the intracellular Ub levels by knockout of Uba52 or Ubc leads to failure in embryonic development in the early stages [10,11].
From the early embryonic stage to postnatal adulthood, mouse testes continue to develop through several important steps [12].
In the early embryonic stage, migrating primordial germ cells (PGCs) arise from a specific region of the mesoderm and subsequently participate in sex determination [13]. In addition, testicular somatic cells differentiate into Sertoli cells and form embryonic gonads with PGCs [14]. After birth, the testicular somatic and germ cells proliferate and mature, and spermatogonia subsequently undergo spermatogenesis, including meiosis. Furthermore, epigenetic regulation, cell signaling, and endocrine systems are tightly controlled to achieve testicular development and successfully produce mature sperm [15].
Recent studies in the field of infertility have focused only on a single spermatogenic factor [16][17][18]. Few studies have profiled reproduction-related factors that are affected by Ubb at the protein level in mammalian testes. Lan-Tao et al. recently showed that the mutation in the D-box region of piwi-like protein (Piwil) reduced the binding of the Rnf8 during late sperm generation, resulting in degradation of Piwil and infertility in mice [17]. Li et al. confirmed that the interaction between Dazl and Pabpc1 using immunoprecipitation mass spectrometry [16]. The translation of many sperm-forming factors that are key to the progression of sperm formation is regulated by Pabpc1. In addition, the association between infertility and Dazl was confirmed at the single-protein level of cells arrested in the zygotene/pachytene-transition stage owing to the lack of Dazl. Huang et al. confirmed that 44 ubiquitinated proteins are associated with spermatogenesis using liquid chromatography-mass spectrometry (LC-MS) analysis of ubiquitinated peptides obtained through the affinity-enrichment method in the testes of adult buffalo [18].
We have previously reported that Ubb-knockout mice are viable but show two striking phenotypes: adult-onset hypothalamic neurodegeneration and infertility in both male and female mice [19,20]. The infertility in Ubb-knockout mice is caused by the failure of gametes to form owing to developmental defects in the gonads following birth. In particular, the testes and ovaries of Ubbknockout mice are smaller than those of wild-type (WT) mice, and the size differences increase as they become sexually mature [20]. Furthermore, meiotic cell cycle arrest in germ cells leads to azoospermia in Ubb-null male mice, whereas other somatic cells in the testes are normal. The expression level of Ubb in testicular germ cells is relatively higher than that in testicular somatic cells, leading to a more significant decrease in Ub levels [20]. Although there have been thorough investigations into the mechanism of developmental defects in Ubb-knockout testes, including transcriptome analysis, it is still unclear how reduced Ub levels lead to meiotic cell cycle arrest [21].
Here, we focused on WT and Ubb −/− testes at postnatal day 20 (P20), which began to generate sperm, and at embryonic day 14.5 (E14.5) to investigate the molecular perturbation caused by cellular Ub deficiency. We used state-of-the-art mass spectrometry-based quantitative proteomics between WT and Ubb −/− testes to identify key proteins related to meiotic cell cycle arrest in Ub-deficient spermatogonia. Our data provide a rich resource for profiling the mouse testis proteome and mapping the role of Ubb in various stages of spermatogenesis.

RESULTS
Profiling of global proteins expressed in the testis of the Ubbknockout mouse model Before proteomic analysis to identify the changes underlying Ubb disruption, we isolated testis tissues from four WT and three Ubb −/− male mice at postnatal day 20. All seven samples were lysed, digested, labeled with TMT, and then pooled and analyzed using LC-MS/MS (Fig. 1a). A total of 8105 testicular proteins were identified, of which 6511 proteins could be quantified for relative comparison. Next, a two-sample t-test was performed and the proteins with a p value of < 0.05 and a ±1.2-fold change in their expression were considered significant and selected as DEPs. In total, 564 proteins were differentially expressed in Ubb −/− testes compared with WT controls, including 277 upregulated and 287 downregulated proteins that satisfied the p value threshold (Fig.  1b, Supplementary table 2). To gain insights into the proteinranked list influenced by Ubb disruption, the proteins that showed a difference in expression levels were ranked based on the value of log 2 fold change (log 2 FC). Many of the proteins related to spermatogenesis (such as Piwil1, Tdrd6, Piwil2, Tdrd1, and Pabpc1) showed decreased expression levels in Ubb −/− and were mostly ranked toward the edge (Fig. 1c). Comparing our data with that of Wei et al., we found that 5863 proteins matched with our data [22], and 403 proteins in the spermatogenesis database were identified [23] (Fig. 1d). Among the 403 spermatogenesis-related proteins, 108 proteins involved in the meiosis stage and 104 proteins involved in the post meiotic stage were identified using LC-MS/MS analysis (Fig. 1e, Supplementary table 3). Intriguingly, 36 proteins that were downregulated because of Ubb disruption were involved in various stages of spermatogenesis, consistent with our previous report on meiotic arrest due to Ub deficiency [16,18,21,23,24]. In addition, 11 proteins, including Eif4g3, Hsp90aa1, Tdrd9, and Sycp2, were involved with spermatocytes and showed abnormally lower levels of expression as compared to that of the WT control. The levels of eight proteins, including Ddx4, Tdrd6, Tdrd7, and Rnf17, involved in the spermatid stage of sperm maturation, were also downregulated (Supplementary  table 3). Taken together, our proteomic profiling revealed that Ubb disruption perturbs a wide range of spermatogenesis-related biological functions.
Functional enrichment gene ontology analysis of DEPs (WT vs. Ubb-knockout) Next, we performed enrichment analysis to characterize the testicular proteins associated with Ubb disruption. GO analysis showed enrichment of DEPs in biological processes, cellular components, and various pathways in the Ubb-null state. For GO analysis, an opensource software, g:profiler, was used, and functional profiling of Ubbrelated proteins was performed by statistical enrichment analysis sing the g:scs algorithm. Proteins with reduced expression owing to deletion of Ubb were primarily enriched in cilium-related biological processes, cellular components, and pathways, and the p-value obtained using g:SCS was very low. The enriched biological processes, such as sexual reproduction, cilium organization, and piRNA metabolic process, were actively downregulated in Ubb −/− testes (Fig. 2a), whereas the upregulated proteins were mainly enriched in processes such as cellular component organization, developmental process, and cytoskeleton organization (Supplementary table 4). We also observed cellular components enriched in proteins with decreased abundance, including meiotic spindle, cilium, ribonucleoprotein complex, chromatoid body, and p-granule (Fig. 2b). Furthermore, the upregulated-proteins were associated with cell junctions, cell-cell contact zones, membrane regions, and extracellular matrix (Supplementary table 4). Notably, we observed decreased protein abundance-related to glycolysis/glucogenesis and cilium assembly. In addition, protein-enriched in Hedgehog "off"state reactome encompassed the downregulated proteins (Fig. 2c). In contrast, upregulated proteins were enriched in extracellular matrix interaction, focal adhesion, and prostaglandin synthesis and regulation pathways. In addition, the proteins with increased abundance in our knockout model were related to the apoptotic execution phase and signaling by platelet-derived growth factor (PDGF) (Supplementary table 4).
Interaction networks between spermatogenesis and Ubbrelated testis proteins Gene ontology analysis showed that, Ubb-knockout affected the expression of proteins required for piRNA metabolism and sperm formation. To investigate the link between sperm formation and Ub, we constructed a protein-interaction network with the Ubbencoded protein (Ubb) using the STRING database and Cytoscape software. Of the testis proteins identified by LC-MS/MS analysis, 296 interacted with Ubb (Supplementary table 5), of which 24 proteins (including Psma8, Hsp90aa1, Rps6, Eef1a1, Uchl5, and Eef2) are closely related to Ubb deletion (Fig. 3a). Interestingly, 24 DEPs that interact with Ubb were mainly enriched with molecular functions of RNA binding, organelle organization, cilium, reproduction, and ribonucleoprotein granules. In particular, it was found that Hsp90aa1, Eef1a1, Eef2, and Eif3f may be important for the maturation of meiotic cells and spermatocytes, through information from enriched biological processes, spermatogenesis online databases, and previous studies [18,[25][26][27] (Fig. 3b). From the protein-protein interaction network and proteome analysis, the following functional associations were observed: in our results, Hsp90aa1 protein interacting with Ubb, which upon deletion induces apoptosis of germ cells in adult mice, inhibits growth at the pachytene stage, and inhibits gamete formation [28], showed a significantly downregulated expression level in the Ubb-null model, and levels of proteins related to piRNA pathway components interacting with Hsp90aa1 were also affected by Ubb elimination (see Fig. 3b). Our results also implicated a wider role for Eef2, Eef1a1, and Eif3f, which interact with Ubb. Fig. 1 Overall experimental process and global proteome profiling results. a An overview of the method followed to process the tissue samples from the testis of four WT and three Ubb −/− (KO) mice for proteome analysis. b The volcano plot representation for the identified proteome and differentially expressed proteins. Red and blue dots represent significantly up-and downregulated proteins, respectively, with statistical criteria (student's t-test, p-value < 0.05; red is log2fc > 1, blue is downregulated). c Protein rank plot for downregulated proteins, based on their log2-transformed fold change. d Venn-diagram for proteins in Spermatogenesis database (red circle), proteins-identified in the current study by LC-MS/MS analysis (yellow circle), and proteins-identified in Wei et al., (blue circle). e Spermatogenesis Online databasebased characterization of testicular cell type of profiled proteins.
Among the proteins that interact with these proteins, the expression levels of Hspa2 and Eif4g3 proteins were significantly reduced in the infertile group because of Ubb-knockout as compared to that in WT mice [29,30].
Altered expression of DEPs at the RNA level, including Piwil1 and Piwil2 Among the proteins affected by Ubb disruption, Hsp90aa1, Eef1a1, Eef2, and Eif3f are those that are largely related to spermatogenesis. All four proteins are associated with spermatogenesis markers, as reported in many studies [16,18,24]. In particular, levels of TDRD family proteins and Piwil1/2, which induce pachytene arrest when deficient in spermatocytes, were generally downregulated, and all these proteins interacted with the four key proteins mentioned above. Moreover, several DEPs in our data showed DNA-binding characteristics, including germ-cell-specific transcription factors, and misregulation of their gene expression may be critical for the failure of testicular development and growth arrest at the pachytene stage during meiosis in Ubb −/− mice. In order to test this, we examined the altered expression of several up-and downregulated DEPs at the RNA level, based on the results of proteomic analysis. Upregulated proteins included extracellular matrix constituents and proteins related to PDGF signaling, which were highly expressed in testicular somatic cells, and downregulated proteins were related to piRNA metabolism and ciliogenesis, which are essential for meiotic progression of germ cells and spermatogenesis.
We found that the RNA expression levels of Piwil1 and Piwil2, which are involved in piRNA metabolism, and Dync2h1 and Dynlrb2, involved in ciliogenesis, were significantly reduced in Ubb −/− testes (Fig. 4a). However, there was no significant difference in the expression levels of the extracellular matrix constituents and the proteins involved in PDGF signaling, including Col6a1 and Thbs3, between WT and Ubb −/− testes (Fig. 4b). Their expression increased only at the protein level, which was probably due to the difference in the ratio of cellular composition in tissues by the relative increase in testicular somatic cell numbers and the decrease in germ-cell numbers in Ubb −/− testes and the direct misregulation of the protein level owing to Ub deficiency. In fact, it was confirmed that the expression of Tubb2a, which was one of the upregulated proteins in Ubb −/− testes, although it is involved in ciliogenesis, was also slightly, but not significantly, increased at the RNA level (Fig. 4c). Overall, our qRT-PCR results suggest that the expression of Piwil1 and Piwil2, which are highly expressed in the germ cells of developing testes, significantly decreased not only at the protein level but also at the RNA level in Ubb −/− testes. Among the mouse Piwi protein family, Piwil1 is expressed only in differentiating sperm during the late stage of spermatogenesis, while Piwil2 is expressed throughout spermatogenesis [31]. Accordingly, the decreased expression of Piwil2 may precede that of Piwil1 during spermatogenesis in Ubb −/− testes. In contrast to Piwil1/2, PDGF signaling proteins and extracellular matrix constituents, which are relatively highly expressed in testicular somatic cells (such as Sertoli cells and Leydig cells), were only increased at the protein level in Ubb −/− testes.
Increased expression of Ubb is essential for upregulation of Piwil2 level after birth We found that levels of not only proteins essential for ciliogenesis (such as Dync2h1 and Dynlrb2) but also of late-expressed Piwi Fig. 2 Ubb influence on the expression of spermatogenesis proteins. a, b Dot plot of gene ontology analysis results from downregulated DEPs enriched in biological processes and cellular components. Color indicates the g:scs correction -log(adjusted pvalue), the circle size indicates the number of genes in that term, and the ratio of genes is in the category over total downregulated proteins. Categories related to spermatogenesis are shown in red. c Dot plot for Kegg pathway, reactome, and Wiki pathway-based categories enriched with downregulated proteins.
protein, especially Piwil1, decreased in Ubb −/− testes. Because Ubb-null spermatogonia cannot progress into meiosis owing to pachytene arrest [20], downregulation of levels of these proteins at P20, when they are supposed to increase at the late stage of spermatogenesis, may result from the different populations of testicular germ cells between WT and Ubb −/− testes. Furthermore, it is unclear whether Ubb expression is required for the upregulation of critical protein levels, including Piwil2, to pass the appropriate meiosis stages in testicular germ cells.
To investigate this, we isolated the gonadal ridges from WT and Ubb-null male embryos at 14.5 dpc and analyzed the RNA expression levels for Piwil2 and Piwil4. In this period, Piwil1 was not expressed, Piwil2 was expressed at levels lower than that at P20, and Piwil4 was highly expressed in the developing embryonic gonad [31]. We found that the expression levels of Piwil2 and Piwil4 were not significantly different between WT and Ubb −/− gonadal ridges (Fig. 5a). Interestingly, the expression level of not only Piwil4, which is highly expressed from early embryonic development through birth, but also that of Piwil2, was not altered, despite Ubb disruption (Fig. 5a). Furthermore, we observed that Ubb and Piwil2 expression was significantly upregulated in P20 testes as compared to that in embryonic gonadal ridges (Fig. 5b, c). This observation and our proteomic data raised the possibility that the infertility phenotype of Ubb −/− male mice was caused by the misregulation or lack of dramatic upregulation of Piwil2 expression during testis development between E14.5 and P20 (Fig. 5c). In fact, Piwil2-knockout mice also showed the pachytene-arrest phenotype during spermatogenesis [32]. Interestingly, Piwil4 expression was downregulated during testis development, regardless of genotypes (Fig. 5d). In the interaction map, the thickness of the edge is proportional to the combined score. All interactions contribute to the combined score; higher scores correlate to more biologically meaningful evidence. The node color scale indicates the fold-change values, and nodes for proteins without significant differential expression are indicated with gray. b A network of Hsp90aa1, Eef1a1, and Eef2 proteins affected by the presence of Ubb and proteins related to fertility. The orange edges are the interaction between Ubb and the first shell, while green edges refer to the interacting partners of Hsp90aa1, Eef1, and Eif4g3. The proteins involved in RNA binding, piRNA metabolic process, cilia, and reproductive process are indicated with black circles and squares. This network has been created with STRING in cytoscape.
Based on these results and proteomic data, we suggest that upregulation of Ubb expression during postnatal testicular development is required for meiotic progression and increased germ cell numbers through the upregulation of levels of proteins in the Piwi family and piRNA metabolism.

DISCUSSION
Polyubiquitin genes in mice are highly expressed in almost all cell types and throughout their life. When cells encounter specific environmental stimuli, such as oxidative stress, these genes are further upregulated to supplement readily available free Ub. Even under complex and elaborate processes such as embryonic development, the expression of polyubiquitin genes is essential for the differentiation of stem cells and tissue development. We previously reported that the straight phenotypes of the polyubiquitin gene Ubb-knockout mice are hypothalamic neurodegeneration and infertility in both male and female mice. Of the two polyubiquitin genes in mice, the expression levels of Ubb are relatively higher than that of Ubc in the germ cells of the testes; thus, Ub deficiency causes fatal defects in the pachytene stage during meiotic progression.
Here, we investigated the proteome of WT and Ubb-knockout testes at P20 to analyze the DEPs and their functions during the developmental processes. Contrary to our speculation that levels of a significant number of proteins were supposed to be upregulated in Ub-deficient testes, as the general function of Ub in cells is the degradation and quality control of proteins, only 562 proteins were significantly differentially expressed. We also found downregulated levels of a significant number of proteins (287), which were involved in piRNA metabolism and spermatogenesis. When a protein-protein interaction map was created, it showed that Ubb interacted with 267 proteins, of which, levels of 22 proteins were downregulated. Among the proteins with downregulated levels, Hsp90aa1, Eef2, Eef1a1, and Eif3f are critical and play different roles in spermatogenesis.
As reported previously, deletion of Hsp90aa1 induces apoptosis of germ cells in adult mice, arrests growth at the pachytene stage, and inhibits gamete formation [28,33]. Also, Hsp90aa1 is associated with the biogenesis of Piwil through regulation of phosphorylation and piRNA loading into Piwil [25,28]. In our results, the Hsp90aa1 protein that interacted with Ubb exhibited a significantly downregulated expression level in the Ubb-null model, and levels of proteins related to piRNA pathway components interacting with HSP90aa1 were also affected by Ubb elimination (see Fig. 3b). As shown in Fig. 3b, Hsp90aa1 closely interacted with Piwil1/2, which is involved in piRNA production, Ddx4, which is related to sperm motility, and piRNA pathway components, such as Tdrd family proteins, which are required for normal sperm formation, with a high combined score. In addition to Piwil1/2, Tdrd1, Tdrd9, and Hspa2 interacted with the Hsp90aa1 protein. These proteins are known to be essential for the maturation of germ cells into late spermatocytes and spermatids and, in the Ubb −/− model, their levels were remarkably reduced [29,34,35] (Supplementary table 3). Conversely, levels of the Mov10l, which is highly expressed in normal pachytene cells Fig. 4 Altered gene expression of testicular DEPs. Based on proteomic analysis, several mRNA levels of Ubb +/+ (WT) and Ubb −/− testis were determined using qRT-PCR. a Among downregulated proteins, expression levels of their genes involved in the piRNA metabolic process (Piwil1, Piwil2, and Piwil4) and cilium assembly (Dync2h1 and Dynlrb2) were compared between Ubb +/+ and Ubb −/− testis (n = 3 each). b Among upexpressed proteins, expression levels of their genes involved in PDGF signaling (Col6a1, Col6a2, and Thbs3) were compared between Ubb +/+ and Ubb −/− testis (n = 3 each). c The expression level of Tubb2a, which was involved in cilium assembly, increased in Ubb −/− testis as compared to that in WT testis (n = 3 each). All mRNAs used were isolated from Ubb +/+ or Ubb −/− testes, which were collected from mice at postnatal day 20 (P20). All expression levels of genes were normalized to Gapdh level, and data are expressed as the mean ± SEM of indicated number of samples. *P < 0.05; **P < 0.01 vs. Ubb +/+ . NS not significant. and stabilizes the structure of Piwil protein [36,37], and Fkbp6, which promotes recycling of Piwi protein, were not significantly different in the Ubb deletion model (see Supplementary table 2).
Our results also implicated a wider role for Eef2, Eef1a1, and Eif3f, which interact with Ubb. Eef2 and Eef1a1 interact with Ubb and show a significant difference in expression and are associated with Upf1 in one of the germ cell-specific perinuclear structures called chromatin bodies (CB) [38][39][40]. In addition, Rpl10l is essential for sperm formation, and its level was downregulated in the Ubb −/− model. Rpl10l is a testis-specific retrogene and is an essential protein when transitioning from prophase to metaphase (meiosis I) in spermatocytes [41]. The level of Rpl10l, which is associated with all of Eef1a1, Eef2, and Eif3f, perpetuates transition from prophase to metaphase in spermatocytes, and also interacts with 60 s and 40 s ribosomal proteins downregulated in the Ubb −/− model. The Rpl3/4/6 protein has also been shown to interact with the Pabpc1 protein, which regulates the translational activation of sperm-generating mRNA in the pachytene stage [42] (see Fig. 3a). Moreover, Eef1a1 interacted with Tdrd1, which regulates the recognition of piRNA mediators or cofactors for protein-protein interactions in the piRNA pathway by binding to Piwil2 of the Tudor family [34,43,44]. Eef1a1, Eef2, and Hsp90aa1 are commonly associated with Hspa2, which is also an essential protein in spermatogenesis and the development of spermatids [45]. Hspa2 interacts with Cdk1 and plays an essential role in the formation of the Cdk1/cyclin B1 complex in sperm cells; therefore, deficiency of Hspa2 indicated prophase arrest of pachytene spermatocytes during meiosis I [29].
Eif3f is one of the constituent proteins of the Piwi/Eif3f/ Elavl1 supercomplex that regulates mRNA translation of Tudor domain family proteins in CB [42,46,47] and shows interactions with Ubb, but is independent of Ubb deletion [27]. Meanwhile, Elavl1, which interacts with Eif3f, also did not change its expression levels owing to the deletion of Ubb, but the expression of Eif4g3 and Pabpc1 proteins that form a supercomplex with Elavl1 was influenced by Ubb. Eif4g3 functions in the translation of Hspa2 and is required for male germ cells [30,48]. Deletion of Eif4g3 affects Hspa2 translation and inhibits sperm formation during the meiosis stage, leading to infertility [30]. Pabpc1 showed the highest expression level in the pachytene and roundspermatid (RS) stages, and its deficiency can cause infertility [42]. Thus, our results showed a significant decrease in Pabpc1 protein level as compared to that in WT mice causing infertility because of Ubb-knockout, suggesting that this decrease in expression level may induce infertility along with changes in other protein levels. In addition, Pabpc1 interacts with the poly A tail of Dazl mRNA and acts as a translation regulator of proteins necessary for maturation in spermigonesis and spermatogonia, as well as in the meiosis stage [16].
Regarding pathways, expression of PDGF and extracellular matrix-related proteins was only different at the protein level, which could be linked to different numbers of cells, as Ubbknockout caused a reduction in testis size. However, expression of piRNA metabolism and ciliogenesis-related genes was downregulated at the transcriptome level. This highlights the alteration of levels of proteins that regulate the transcription of proteins involved in piRNA metabolism and ciliogenesis. Piwi proteins are primarily involved in genome maintenance and regulation of the expression of various target genes [47]. Piwil2 expression was found to be downregulated at both the RNA and protein levels. This points to the plausible role of Ubb-mediated proteins in the transcriptional regulation of Piwil2. In summary, this study provided the proteome of Ubb −/− testes and furnished a mechanism for Ubb knockout-mediated infertility. It also highlights the unappreciated role of Piwi-related proteins in spermatogenesis and testicular growth. were measured using qRT-PCR and normalized to Gapdh level. b Ubb expression levels of E14.5 and P20 Ubb +/+ testis (n = 3 each) were determined using qRT-PCR and normalized to Gapdh level. c, d For the relative comparison of mRNA levels between E14.5 gonadal ridges and P20 testis, the measured levels of Piwil2 and Piwil4 expression in Ubb +/+ and Ubb −/− testis (n = 3 each) were presented as fold change relative to that in WT at E14.5. Data are expressed as the mean ± SEM of indicated number of samples. **P < 0.01; ***P < 0.001 vs. Ubb +/+ at E14.5. NS, not significant.

MATERIALS AND METHODS Mouse
Mice were maintained in plastic cages with ad libitum access to food and water. Ubb +/+ and Ubb −/− mice were obtained from the interbreeding of Ubb +/− mice. All experiments were performed in accordance with the relevant guidelines and regulations approved by the University of Seoul Institutional Animal Care and Use Committee (UOS IACUC). All experimental protocols were approved by the UOS IACUC (UOS-170517-1, UOS IACUC-2020-03-A).
Isolation testis and embryonic male gonad from Ub +/+ and Ubb −/− mice Testes were isolated from Ubb +/+ (WT) and Ubb −/− mice at postnatal day 20 (P20) and placed in cold PBS to remove other adjacent tissues. Embryonic male gonads were isolated from WT and Ubb −/− mice 14.5 days post coitum (dpc) with a dissection microscope and placed in cold PBS to remove other embryonic tissues. After isolation, P20 testes and E14.5 male gonads were immediately frozen in liquid nitrogen and stored at −80°C for further experiments.
TMT labeling of peptides and basic pH reverse-phase fractionation Peptides were labeled with seven different mass tags among the Tandem Mass Tag (TMT) reagents (90111; Thermo Fisher Scientific, USA) as follows: the four peptide samples from WT testis tissue were labeled with 126 N, 127[N/C], and 128 N TMT reagents, and the three peptide samples from Ubb −/− testis tissue were labeled with 128 C, 129 C, and 130 N TMT reagents. The prepared TMT reagent was transferred to the peptide sample, and the mixture was incubated for 1 h at RT following brief vortexing. The reaction was quenched with 5% hydroxylamine and incubated for 30 min at RT. All TMT-labeled peptides in the batch were pooled and concentrated by vacuum centrifugation. Labeled peptides were loaded on an analytical column (XBridge Peptide BEH C18 Column;  [50]. The separated peptides were collected every 1 min into 96 tubes and nonconsecutively pooled into 12 fractions by combining eight parts. The resulting 12 fractions were desalted using a Pierce C18 spin column (89870; Thermo Fisher Scientific, USA) (Fig. 1a).

LC-MS/MS analysis for global proteome
For LC-MS analysis, a nanoACQUITY UPLC system (Waters™, USA) was coupled to a Q Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Fisher Scientific, Germany) and equipped with a trap column (Acclaim™ PepMap™ 100 C18 LC Column, C18, 75 μm × 2 cm, 5 μm, 164564; Thermo Fisher Scientific, USA) for cleanup followed by an analytical column (EASY-Spray™ LC Columns, C18, 75 μm × 50 cm, 2 μm, ES803A; Thermo Fisher Scientific, USA). The peptides were separated using a mobile phase comprising solvent A (0.1% formic acid and 2% acetonitrile in water) and solvent B (0.1% formic acid in 90% acetonitrile). For global proteome-digested peptides, the labeled peptides were trapped using 100% solvent A for 4 min at a flow rate of 3 μL/min, and the optimized linear-gradient elution program was set as follows (T min:% of solvent B): 0:3, 4:3, 7:12, 143:40, 144:80, 154:80, 155:5, and 180:5. The flow rate was 300 nL/min throughout the run time. Full MS scans were acquired for the mass range of 350-1,800 m/z at a resolution of 70,000 in MS1 level and an automatic gain control (AGC) target of 1e6, and MS/MS analysis was performed by data-dependent acquisition in positive mode. The electric potential of electrospray ionization was set to 2.2 kV, and the temperature of the capillary was set to 250°C. The top 12 precursor peaks were selected from the MS1 scan and separated for fragmentation. For MS2 acquisition with high-energy collisional dissociation (HCD), the resolution was set to 35,000 with a fixed first m/z of 100 m/z, an AGC target of 1e6, an isolation window of 2.0 m/z, an isolation offset of 0.5 m/z, and a normalized collision energy (NCE) of 32%. The charge states of unassigned, 1, or >6 were discarded, and dynamic exclusion for 30 s was enabled.

Protein database searching and quantification of global proteomic data
The MS/MS spectra were searched against a composite database of Uniprot Mus musculus reference (May 2020; 21,989 entries) using the Sequest HT search engine. The search was restricted to fully tryptic peptides, allowing up to two missed cleavage sites. In modification, carbamidomethylation, alkylation of disulfide bonds in cysteine (+57.021 Da), TMT modification of lysine, and N-termination (+229.163 Da) were noted as static modifications. Oxidation of methionine (+15.995 Da) was used as a variable modification. The false-discovery rate (FDR) for peptide level was set to 0.01 for removing false-positive data and to quantify each reporter ion, using the "peptide and protein quantifier" method in Proteome Discoverer 2.4 (Thermo Fisher Scientific, USA).

RNA isolation and quantitative reverse transcription-polymerase chain reaction
Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was performed as previously described, with slight modifications [11]. Briefly, total RNA was isolated from cells using Tri-reagent (Molecular Research Center, USA) following the manufacturer's protocol, and 10 ng of total RNA was used as a template for reverse transcription using SuperScript reverse transcriptase (Enzynomics, Republic of Korea) and oligo-dT (Cosmogenetech, Republic of Korea). For qRT-PCR, we used SYBR-Green qPCR Mastermix (Enzynomics, Republic of Korea) and an iCycler system (Bio-Rad, USA). The mRNA expression levels of target genes, including Ubb, Piwil1, Piwil2, Piwil4, Dync2h1, Dynlrb2, and Thbs3, were normalized to Gapdh levels. The sequences of the primers used for qRT-PCR are listed in Supplementary table S1.

Statistical analysis and gene enrichment analysis
For the enrichment pathway analysis of all identified proteins, we classified proteins with more than ±1.2-fold expression changes in a two-tailed Student's t-test as differentially expressed proteins (DEPs) with a p-value < 0.05. Annotation enrichment processes were performed for the significantly upregulated or downregulated DEPs. Gene Ontology (GO) searches were performed to explore the biological processes and cellular components affected by Ubb-knockout using g:Profiler [51]. The GO biological process enhanced by DEPs was found to be FDR < 0.05. To reconstruct the network model for DEPs, protein-protein interaction information was collected from the STRING v11.0 public database [52]. The network model used Cytoscape to depict the enhanced process and interaction data.