Potential Involvement of ewsr1-w Gene in Ovarian Development of Chinese Tongue Sole, Cynoglossus semilaevis

Simple Summary Sexual dimorphism is a phenomenon commonly existing in animals. Chinese tongue sole Cynoglossus semilaevis is an economical marine fish with obvious female-biased size dimorphism. So, it is important to explore the molecular mechanism beyond gonadal development for sex control in aquaculture industry. RNA-binding protein Ewing Sarcoma protein-like (ewsr1) gene is important for mouse gonadal development and reproduction, however there are limited studies on this gene in teleost. In this study, two ewsr1 genes were cloned and characterized from C. semilaevis. The ewsr1-w gene, located in W chromosomes, showed female-biased expression during C. semilaevis gonadal development. In addition, knock-down effect and transcriptional regulation of Cs-ewsr1-w further suggested its essential role in ovarian development. This study broadened our understanding on ewsr1 function in teleost, and provided genetic resources for the further development of sex control breeding techniques in C. semilaevis aquaculture. Abstract Ewsr1 encodes a protein that acts as a multifunctional molecule in a variety of cellular processes. The full-length of Cs-ewsr1-w and Cs-ewsr1-z were cloned in Chinese tongue sole (Cynoglossus semilaevis). The open reading frame (ORF) of Cs-ewsr1-w was 1,767 bp that encoded 589 amino acids, while Cs-ewsr1-z was 1,794 bp that encoded 598 amino acids. Real-time PCR assays showed that Cs-ewsr1-w exhibited significant female-biased expression and could be hardly detected in male. It has the most abundant expression in ovaries among eight healthy tissues. Its expression in ovary increased gradually from 90 d to 3 y with C. semilaevis ovarian development and reached the peak at 3 y. After Cs-ewsr1-w knockdown with siRNA interference, several genes related to gonadal development including foxl2, sox9b and pou5f1 were down-regulated in ovarian cell line, suggesting the possible participation of Cs-ewsr1-w in C. semilaevis ovarian development. The dual-luciferase reporter assay revealed that the -733/-154 bp Cs-ewsr1-w promoter fragment exhibited strong transcription activity human embryonic kidney (HEK) 293T cell line. The mutation of a MAF BZIP Transcription Factor K (Mafk) binding site located in this fragment suggested that transcription factor Mafk might play an important role in Cs-ewsr1-w basal transcription. Our results will provide clues on the gene expression level, transcriptional regulation and knock-down effect of ewsr1 gene during ovarian development in teleost.


Characterization of Cs-ewsr1-w and Cs-ewsr1-z
The characters including open reading frame (ORF), amino acid sequence, molecular weight, protein domains, and phosphorylation sites were predicted and analyzed by DNAstar (V7.  [34], Philadelphia, PA, USA) was used to construct phylogenetic tree by neighbour-joining algorithm (NJ). The NCBI accession numbers of amino acid sequences used in this study were listed in Table 2. Table 2. Accession numbers of Ewsr1 proteins used in this study.

Species
Accession No.

Gene Expression Patterns of Cs-ewsr1-w and Cs-ewsr1-z in Different Tissues and Stages
The expressions of Cs-ewsr1-w and Cs-ewsr1-z in different development stages and tissues were analyzed with gene specific primers (Table 1) via qPCR assays on a 7500 Fast Real Time PCR platform (Applied Biosystems, Foster City, CA, USA). β-actin was set as the internal control. The 20 µL reactions contained 10 µL SYBR Premix Ex TaqTM (TaKaRa, Tokyo, Japan), 2 µL cDNA, 0.4 µL of each sense and anti-sense primers, and 0.4 µL ROX Dye II. The qPCR program was set as default settings followed by the dissociation curve, that is, 95 • C for 30 s, 40 cycles of 95 • C for 5 s, 60 • C for 30 s. The relative mRNA expression of Cs-ewsr1-w and Cs-ewsr1-z were processed by using 2 −∆∆Ct method [35]. The data were analyzed by one-way ANOVA followed by Duncan's multiple comparison in SPSS 25.0 (IBM Corp., Armonk, NY, USA), and the differences were considered significant when p < 0.05.

Promoter Activities Analysis of Cs-ewsr1-w
Based on the TFs prediction, six promoter plasmids with luciferase report were constructed by serial-deletion to detect the promoting activity of regulatory elements of Cs-ewsr1-w. The primers were listed in Table 1. The fragments were inserted into pGL3-basic vector (Promega, Madison, WI, USA) for the recombinant plasmid construction of pGL3-Cs-ewsr1-w-F1~F6 by using TSV-S1 Trelief ® SoSoo Cloning Kit (Tsingke, Beijing, China).
Human embryonal kidney (HEK) 293T cells were maintained in DME/F-12 containing 10% fetal bovine serum (FBS, Gibco, New York, NY, USA) and 1% bFGF (Invitrogen, Carlsbad, CA, USA) in 5% CO 2 at 37 • C. The pGL3-Cs-ewsr1-w~F1~F6 were transfected into HEK293T cells by using Lipo8000TM Transfection Reagent (Beyotime, Shanghai, China). Meanwhile, pGL3-basic and PGL3-control plasmids were used as the negative control and the positive control, respectively. The pRL-TK plasmid was transfected at the mean time as the internal reference. Dual Luciferase Reporter Gene Assay Kit (Beyotime, Shanghai, China) was employed to measure the promoter activities. Each experiment was performed in triplicates following the standard protocol provided by the manufacturer. Data obtained from Varioskan Flash spectral scanning multimode reader (Thermo Fisher Scientific, Vantaa, Finland) were analyzed by LSD (Least-significant difference) in SPSS 25.0, and the significance was regarded at p < 0.05.

The Knockdown Effect of Cs-ewsr1-w siRNA in C. semilaevis Ovarian Cells
The specific siRNA of Cs-ewsr1-w gene, the negative control siRNA, and siR transfect control (5cy3) were synthesized in Sangon Biotech (Sangon, Shanghai, China). C. semilaevis ovarian cells were cultured in L-15 medium supplemented with 1% bFGF and 15% FBS at 24 • C. Cs-ewsr1-w siRNA was transfected into the cells by using riboFECT TM CP Transfection Kit (Ribobio, Beijing, China) following the protocol described in the previous study [13]. Three replicates were set for both Cs-ewsr1-w-siRNA and negative control (NC) groups. At 48 h post transfection, the cellular status and the florescence of 5cy3-transfected cells would be checked. When it reached 90-95% of cell confluency and the percentage of transfection reached~80%, it would be a good timing to harvest ovarian cells for the following experiments. After cell collection and RNA extraction, reverse transcription and qPCR assays were performed following the methods mentioned above. The relative expression levels of sex-related genes, such as Forkhead Box L2 (foxl2), SRY-box transcription factor 9b (sox9b), POU Class 5 Homeobox 1 (pou5f1) were measured with the primers listed in Table 1.

Gene Cloning and Characterization of Cs-ewsr1s
Cs-ewsr1-w (GenBank accession no. 103397238) located in W chromosomes, with the full length of 2297 bp containing the ORF region of 1767 bp encoding 588 amino acids ( Figure 1A). Functional domain prediction showed that Cs-Ewsr1-W contained RNA recognition motif in 313-393 residues and a Ran binding protein zinc finger domain in 454-480 residues. Cs-ewsr1-z (GenBank accession no. 103398620), located in Z chromosome, was 2230 bp in full length. It contains 1794 bp ORF region encoding 598 amino acids Cs-ewsr1-w (GenBank accession no. 103397238) located in W chromosomes, with the full length of 2,297 bp containing the ORF region of 1,767 bp encoding 588 amino acids ( Figure 1A). Functional domain prediction showed that Cs-Ewsr1-W contained RNA recognition motif in 313-393 residues and a Ran binding protein zinc finger domain in 454-480 residues. Cs-ewsr1-z (GenBank accession no. 103398620), located in Z chromosome, was 2,230 bp in full length. It contains 1,794 bp ORF region encoding 598 amino acids ( Figure 1B). The same functional domains were located at 313-399 and 460-486 in Cs-Ewsr1-Z, respectively. The phylogenetic tree was constructed by using Ewsr1 proteins from 17 different species. The results showed that two Cs-Ewsr1 proteins were clustered together and then embedded with the teleost clade with Paralichthys olivaceus, Scophthalmus maximus and Poecilia formosa. The mammalian and the other species were clustered together ( Figure 2). RNA recognition motif and Ran binding protein zinc finger domain were conserved in all the aligned species, including teleost, amphibians, and mammalians.
species. The results showed that two Cs-Ewsr1 proteins were clustered together and th embedded with the teleost clade with Paralichthys olivaceus, Scophthalmus maximus a Poecilia formosa. The mammalian and the other species were clustered together (Figure RNA recognition motif and Ran binding protein zinc finger domain were conserved in the aligned species, including teleost, amphibians, and mammalians. " represents RN recognition motif, " " respresents Ran binding protein zinc finger domain, "▲" represents Ew proteins in C. semilaevis

The Expression Patterns of Cs-ewsr1s in Different Tissues and Developmental Stages
qPCR assays revealed that the sex-biased expression patterns of two Cs-ewsr1s in C nese tongue sole. Cs-ewsr1-w was expressed in all tissues of female tongue sole with t highest expression in gonad, but was hardly detected in any tissue of male tongue so ( Figure 3A). In comparison, Cs-ewsr1-z was prevalently expressed in all tissues of fema and male tongue sole ( Figure 3B). It exhibited the highest expression in male gonad, f lowed by female gonad, brains of male and female, livers of male and female, and fema heart and intestine ( Figure 3). the Ran binding protein zinc finger domain is represented by a dot-dash underline.
The phylogenetic tree was constructed by using Ewsr1 proteins from 17 different species. The results showed that two Cs-Ewsr1 proteins were clustered together and then embedded with the teleost clade with Paralichthys olivaceus, Scophthalmus maximus and Poecilia formosa. The mammalian and the other species were clustered together ( Figure 2). RNA recognition motif and Ran binding protein zinc finger domain were conserved in all the aligned species, including teleost, amphibians, and mammalians. " represents RNA recognition motif, " " respresents Ran binding protein zinc finger domain, "▲" represents Ewsr1 proteins in C. semilaevis

The Expression Patterns of Cs-ewsr1s in Different Tissues and Developmental Stages
qPCR assays revealed that the sex-biased expression patterns of two Cs-ewsr1s in Chinese tongue sole. Cs-ewsr1-w was expressed in all tissues of female tongue sole with the highest expression in gonad, but was hardly detected in any tissue of male tongue sole ( Figure 3A). In comparison, Cs-ewsr1-z was prevalently expressed in all tissues of female and male tongue sole ( Figure 3B). It exhibited the highest expression in male gonad, followed by female gonad, brains of male and female, livers of male and female, and female heart and intestine (Figure 3).
" represents RNA recognition motif, " stop codon at the end of the ORF. The RNA recognition motif is represented by an underscore and the Ran binding protein zinc finger domain is represented by a dot-dash underline.
The phylogenetic tree was constructed by using Ewsr1 proteins from 17 different species. The results showed that two Cs-Ewsr1 proteins were clustered together and then embedded with the teleost clade with Paralichthys olivaceus, Scophthalmus maximus and Poecilia formosa. The mammalian and the other species were clustered together ( Figure 2). RNA recognition motif and Ran binding protein zinc finger domain were conserved in all the aligned species, including teleost, amphibians, and mammalians. " represents RNA recognition motif, " " respresents Ran binding protein zinc finger domain, "▲" represents Ewsr1 proteins in C. semilaevis

The Expression Patterns of Cs-ewsr1s in Different Tissues and Developmental Stages
qPCR assays revealed that the sex-biased expression patterns of two Cs-ewsr1s in Chinese tongue sole. Cs-ewsr1-w was expressed in all tissues of female tongue sole with the highest expression in gonad, but was hardly detected in any tissue of male tongue sole ( Figure 3A). In comparison, Cs-ewsr1-z was prevalently expressed in all tissues of female and male tongue sole ( Figure 3B). It exhibited the highest expression in male gonad, followed by female gonad, brains of male and female, livers of male and female, and female heart and intestine (Figure 3).

The Expression Patterns of Cs-ewsr1s in Different Tissues and Developmental Stages
qPCR assays revealed that the sex-biased expression patterns of two Cs-ewsr1s in Chinese tongue sole. Cs-ewsr1-w was expressed in all tissues of female tongue sole with the highest expression in gonad, but was hardly detected in any tissue of male tongue sole ( Figure 3A). In comparison, Cs-ewsr1-z was prevalently expressed in all tissues of female and male tongue sole ( Figure 3B). It exhibited the highest expression in male gonad, followed by female gonad, brains of male and female, livers of male and female, and female heart and intestine ( Figure 3).  Cs-ewsr1-w gradually increased with ovary development, and reached the peak at 3 y ( Figure 4A). However, its expression was hardly detected during testis development because of no expression in testis. Cs-ewsr1-z gene was expressed in all tested develop-  Cs-ewsr1-w gradually increased with ovary development, and reached the peak at 3 y ( Figure 4A). However, its expression was hardly detected during testis development because of no expression in testis. Cs-ewsr1-z gene was expressed in all tested developmental stages of ovaries and testes. Its expression was relatively low in 6 m female and male, and was significantly higher in testes of 1.5 y male ( Figure 4B).

Promoter Activity of Cs-ewsr1-w Detection and Analysis
The Cs-ewsr1-w promoter sequence of 2,690 bp (-2,590/+99) was cloned by geno DNA with specific primers Cs-ewsr1-w-P-F/R (Table 1). Cs-ewsr1-w promoter region h only one CpG island, which was located from -1,215 to -1,101 bp.
A series of promoter fragments with different length deletion were generated to plore the promoter activity of Cs-ewsr1-w gene. The promoter activities of all Cs-ewsr fragments were significantly higher than that of pGL3-basic (p < 0.05, Figure 5), amo which the activity of Cs-ewsr1-w-P-F2/R fragment was the highest. The relative activ significantly decreased by 2.7-fold from Cs-ewsr1-w-P-F5/R fragment to Cs-ewsr1-w-P-F fragment (p < 0.05, Figure 5), indicated region -733 to -154 positively affected the promo activity of Cs-ewsr1-w gene. Similarly, the positive effect was detected from other two gion as well, which were -2,190 to -1,692 bp and -154 to +99 bp ( Figure 5).

Promoter Activity of Cs-ewsr1-w Detection and Analysis
The Cs-ewsr1-w promoter sequence of 2690 bp (−2590/+99) was cloned by genomic DNA with specific primers Cs-ewsr1-w-P-F/R (Table 1). Cs-ewsr1-w promoter region had only one CpG island, which was located from −1215 to −1101 bp.
A series of promoter fragments with different length deletion were generated to explore the promoter activity of Cs-ewsr1-w gene. The promoter activities of all Cs-ewsr1-w fragments were significantly higher than that of pGL3-basic (p < 0.05, Figure 5), among which the activity of Cs-ewsr1-w-P-F2/R fragment was the highest. The relative activity significantly decreased by 2.7-fold from Cs-ewsr1-w-P-F5/R fragment to Cs-ewsr1-w-P-F6/R fragment (p < 0.05, Figure 5), indicated region −733 to −154 positively affected the promoter activity of Cs-ewsr1-w gene. Similarly, the positive effect was detected from other two region as well, which were −2190 to −1692 bp and −154 to +99 bp ( Figure 5). fragments were significantly higher than that of pGL3-basic (p < 0.05, Figure 5), among which the activity of Cs-ewsr1-w-P-F2/R fragment was the highest. The relative activity significantly decreased by 2.7-fold from Cs-ewsr1-w-P-F5/R fragment to Cs-ewsr1-w-P-F6/R fragment (p < 0.05, Figure 5), indicated region -733 to -154 positively affected the promoter activity of Cs-ewsr1-w gene. Similarly, the positive effect was detected from other two region as well, which were -2,190 to -1,692 bp and -154 to +99 bp ( Figure 5). The results indicated mutation of Mafk binding site led to the significant decrease by 46% in the activity of Cs-ewsr1-w-P-F5/R fragment (p < 0.05, Figure 7), which showed no significant difference compared with the activity of Cs-ewsr1-w-P-F6/R fragment. Mutations on other TF binding sites showed no significant effect (Figure 7). Prediction of TF binding sites in the -733/-154 bp interval of Cs-ewsr1-w promoter revealed numerous TFs binding sites, including STAT4, HOXA3, c-Myc, GAGA factor, MAC1, POU1F1a, Mafk, and PRA ( Figure 6). The interval containing TF Mafk, c-MYC, MAC1 and POU1F1a were mutated and transfected into 293T cells for detection after 48 h. The results indicated mutation of Mafk binding site led to the significant decrease by 46% in the activity of Cs-ewsr1-w-P-F5/R fragment (p < 0.05, Figure 7), which showed no significant difference compared with the activity of Cs-ewsr1-w-P-F6/R fragment. Mutations on other TF binding sites showed no significant effect (Figure 7).   . Nucleotide sequence of Cs-ewsr1-w promoter region (−911/+50) and the predicted transcription factor binding sites. The highly active region (-733/-154) was highlighted in yellow. Star codon "ATG" was labelled in red. Underlined boldface letters indicated the predicted transcription factors. Figure 7. Fluorescence activity of the mutated transcription factor binding sites in Cs-ewsr1-w promoter compared with that of Cs-ewsr1-w promoter region with different deletion. Different letters represent significant differences among species (p < 0.05).

Expression Patterns of Sex-Related Genes in Cs-ewsr1-w Knockdown Ovarian Cells
Cs-ewsr1-w expression was significantly reduced by 80% (p < 0.05) after in vitro Cs-ewsr1-w siRNA interference (RNAi), while no significant variation of Cs-ewsr1-z gene was detected. The down-regulation of sex-related genes was detected after Cs-ewsr1-w RNAi (p < 0.05), including foxl2, sox9b and pou5f1. Among that, foxl2 expression significantly dropped by18.8-fold compared with the control group. The expression levels of sox9b and pou5f1 were significantly reduced by half (p < 0.05, Figure 8).

Expression Patterns of Sex-Related Genes in Cs-ewsr1-w Knockdown Ovarian Cells
Cs-ewsr1-w expression was significantly reduced by 80% (p < 0.05) after in vitro Cs-ewsr1-w siRNA interference (RNAi), while no significant variation of Cs-ewsr1-z gene was detected. The down-regulation of sex-related genes was detected after Cs-ewsr1-w RNAi (p < 0.05), including foxl2, sox9b and pou5f1. Among that, foxl2 expression significantly dropped by18.8-fold compared with the control group. The expression levels of sox9b and pou5f1 were significantly reduced by half (p < 0.05, Figure 8).

Discussion
Ewsr protein, a key player in cancer, is involved in RNA metabolism and DNA repair [38]. However, studies on teleost EWS protein are rare. Based on the comparative transcriptome analysis on early developmental stages of gonad in C. semilaevis, ewsr1-w was specifically expressed in ovary and continuously up-regulated with ovarian differentiation [15]. The amino acid sequences of Cs-Ewsr1-w and Cs-Ewsr1-z proteins have high similarity of 91.41%, both of which contained the conserved RNA recognition motif and Ran binding protein-zinc finger domain, suggesting that Cs-Ewsr1s might have similar function with Ewsr1s in other vertebrates. RNA recognition motif is associated with the interaction of protein with RNA. Meanwhile, this protein has a variable number of RGG (arginine-glycine-glycine) repeats that are regarded as a RNA-binding region as well [39].

Discussion
Ewsr protein, a key player in cancer, is involved in RNA metabolism and DNA repair [38]. However, studies on teleost EWS protein are rare. Based on the comparative transcriptome analysis on early developmental stages of gonad in C. semilaevis, ewsr1-w was specifically expressed in ovary and continuously up-regulated with ovarian differentiation [15]. The amino acid sequences of Cs-Ewsr1-w and Cs-Ewsr1-z proteins have high similarity of 91.41%, both of which contained the conserved RNA recognition motif and Ran binding protein-zinc finger domain, suggesting that Cs-Ewsr1s might have similar function with Ewsr1s in other vertebrates. RNA recognition motif is associated with the interaction of protein with RNA. Meanwhile, this protein has a variable number of RGG (arginine-glycine-glycine) repeats that are regarded as a RNA-binding region as well [39]. Based on the phylogenetic analysis, the sequences we obtained from C. semilaevis fell in a well-supported clade, suggesting that we obtained the ewsr1 gene orthologs.
Based on our qPCR results, Cs-ewsr1-w gene was uniquely expressed in females with the highest transcriptional level in ovary. Its expression increased gradually with ovarian development from 90 d to 3 y. These results indicated the possible involvement of Cs-ewsr1-w gene in ovarian development. After Cs-ewsr1-w gene expression was interfered in the ovarian cells of C. semilaevis, several gonadal development-related genes were down-regulated, including foxl2, sox9b and pou5f1. Foxl2 gene, belonging to winged helix transcription factor, is one of the crucial players in ovarian development [40]. Many studies have shown that this gene functions in sex differentiation and gonadal development in teleost [41][42][43][44]. C. semilaevis foxl2 was significantly expressed in 12-month old ovary of phase II fish ovary development stages, suggesting its possible involvement in oocyte development [45]. Sox9b, another important gene related to gonadal development, were significantly expressed in ovaries of fugu (Takifugu rubripes) and zebrafish (Danio rerio) [46,47]. It facilitated sex differentiation and gonadal development in medaka and Japanese flounder [48][49][50]. Besides, prominent expression of sox9b was detected in gonads of early-stage C. semilaevis, indicating its potential involvement in gonadal differentiation [51]. Pou5f1 (also known as oct4) is a key TF regulating embryonic stem cell pluripotency, primordial germ cell formation, early embryonic and gonadal germ cell development [27]. Pou5f1 analogue in teleost plays a post-embryonic role in adult gonad and gametes development [52][53][54]. Based on the suppressive effect of Cs-ewsr1-w knockdown on several sex-related genes, we proposed that it might be a positive regulator in ovarian development of C. semilaevis. In the future, analysis including in vivo trials would be conducted for further investigation on its mechanism.
Promoters contain sequence-specific binding sites for many TFs, and the TFs could recognize and bind target sequences to guide underlying transcription and regulate transcriptional activity [55][56][57]. In Cs-ewsr1-w promoter −733 to −154 bp was the core region that had a great effect on transcription regulation. After site-direct mutagenesis on the TF-binding sites, the activity of TF Mafk-binding site decreased significantly, suggesting potential involvement of Mafk in Cs-ewsr1-w transcription. Mafk, a member of small MAFs family, was essential for mice embryonic development [58]. It regulates genes involved in several cellular processes, including ubiquitination/proteasome [59]. Numerous ubiquitinconjugating enzyme genes showed female-biased gene expressions in early developmental stages of C. semilaevis, revealing the indispensable involvement of ubiquitination pathway in female differentiation [15]. Based on the above analysis, we deduced that TF Mafk might be a positive regulator in Cs-ewsr1-w transcription during C. semilaevis ovarian development. The regulation mechanism between them are worthy for further functional studies. In the future, further studies will be performed to get a clearer picture on the regulation of Cs-ewsr1-w by TF Mafk during female differentiation and ovarian development.

Conclusions
In this study, two ewsr1 genes were cloned and characterized from Chinese tongue sole (Cs-ewsr1-w and Cs-ewsr1-z). The female-biased gonad expression of Cs-ewsr1-w was observed from 90 d to 3 y, suggesting its potential roles in ovarian development. Its knockdown significantly down-regulated the expressions of foxl2, sox9b and pou5f1. The activity analysis, and the prediction and verification of transcription factors for Cs-ewsr1w promoter shed some lights on the transcription regulation of this gene. Our findings suggested the potential roles of Cs-ewsr1-w in C. semilaevis ovarian development, providing fundamental information for further exploration on its biological functions in teleost.