Two Figla homologues have disparate functions during sex differentiation in half-smooth tongue sole (Cynoglossus semilaevis)

Figla is a germ-cell-specific transcription factor associated with ovary development and differentiation. In vertebrates, one transcriptional form of Figla is commonly found. However, besides the common form of this gene (named Figla_tv1), a new transcriptional form (named Figla_tv2) was identified in half-smooth tongue sole (Cynoglossus semilaevis). The full-length cDNA of Figla_tv1 was 1057 bp long with a 591-bp open reading frame encoding a predicted 196 amino acid protein, while Figla_tv2 encoded a 125 amino acid protein. Figla_tv1 and Figla_tv2 expression in various tissues was detected by qRT-PCR. Figla_tv1 was expressed mainly in ovary, skin and liver, while Figla_tv2 was expressed in all examined tissues. In the gonads, Figla_tv1 was expressed in ovary, while Figla_tv2 was predominately expressed in testis of pseudomales. Further, in situ hybridization located Figla_tv1 only in oocytes and Figla_tv2 mainly in germ cells of pseudomale testis. After knocking down Figla_tv2 in a pseudomale testis cell line, the expression of two steroid hormone-encoding genes, StAR and P450scc, was significantly up-regulated (P < 0.05). Our findings suggest that Figla_tv1 has a conserved function in folliculogenesis, as in other vertebrates, and that Figla_tv2 may have a role in the spermatogenesis of pseudomales by regulating the synthesis and metabolism of steroid hormones.


Figla is a germ-cell-specific transcription factor associated with ovary development and differentiation. In vertebrates, one transcriptional form of Figla is commonly found. However, besides the common form of this gene (named Figla_tv1), a new transcriptional form (named Figla_tv2) was identified in half-
Basic helix-loop-helix (bHLH) proteins are members of a large superfamily that regulates a number of developmental and metabolic processes, including sex determination, cell differentiation, nervous system development, oncogenesis, and cholesterol metabolism 1,2 . Studies in many species, including Drosophila, Saccharomyces cerevisiae, Homo sapiens, and Arabidopsis, have provided a great deal of information about the structure and function of this family of proteins [2][3][4][5][6][7][8] .
The factor in the germline alpha (Figla) gene encodes a germ cell-specific bHLH transcription factor that mediates several developmental processes, including fertilization and early embryogenesis 9 . Many studies have focused on the important roles of Figla in gonad development and differentiation. In Mus musculus, Figla expression was first detected in the ovary at embryo day 13 (E13) 10 . Its expression increased dramatically at the end of embryo development and peaked at two days postpartum, when oocytes have become enclosed in primordial follicles 10 , suggesting the probable involvement of Figla in ovary follicle development. Furthermore, Figla was shown to regulate the expression of three zona pellucida genes (Zp1, Zp2, and Zp3) that encode proteins necessary for zona pellucida formation in growing oocytes 11 , indicating a crucial role for Figla in folliculogenesis. Moreover, targeted mutations of Figla in female M. musculus resulted in abnormal ovarian gonadogenesis, including failure to form primordial follicles, massive depletion of oocytes, and subsequent female sterility 10 . However, in mutated males, gonad development appeared to be normal, and these mice were fertile 10 . Together, these data indicate that Figla is indispensable only for ovary folliculogenesis and is not essential for testis development. In a subsequent study, it was suggested that Figla may balance sexually dimorphic gene expression in the postnatal ovary because Figla knockout resulted in the enhanced expression of many testis-specific genes in the oocytes of newborn M. musculus, while in male germ cells, the ectopic expression of Figla down-regulated a subset of these testis-specific genes 12 .
In teleosts, Figla has commonly been regarded as a marker gene of ovary development or early oocyte differentiation, but studies examining the regulation and roles of Figla in gonad development are limited [13][14][15] . Half-smooth tongue sole (Cynoglossus semilaevis) is an economically important marine flatfish in China. Because of their sex-dimorphic growth (females grow 2 to 4 times faster than males 16 ), increasing the proportion of females in cultivated populations will be economically beneficial. To achieve this goal, understanding the sex-determination and differentiation mechanisms of C. semilaevis is particularly important. The primary sex of C. semilaevis is determined by the sex chromosomes: females (ZW) harbor a large W sex chromosome, while males possess two ZZ sex chromosomes 17 . Approximately 14% of ZW genetic females were shown to be sex-reversed to phenotypic males, the so-called pseudomales 17 . Interestingly, these pseudomales are fertile and can mate with the normal females to produce the viable offspring. A number of sex-related genes, including Dmrt1, wnt4a, Sox9, Foxl2, tesk1, piwil2, and Gadd45g, have been isolated from C. semilaevis; however, the molecular mechanisms associated with sex determination and differentiation are still poorly understood [18][19][20][21][22][23]  ) in our RNA-seq data from adult gonads of C. semilaevis and found that the methylation levels of these genes were closely related to gonad development 24 . However, the gonad expression patterns and functions of the two Figla transcripts during development are still unclear. In the present study, we cloned the full-length cDNAs of the two Figla isoforms in C. semilaevis by rapid amplification of cDNA ends (RACE) and used quantitative real-time polymerase chain reaction (qRT-PCR) and in situ hybridization (ISH) to detect the spatial and gonad expression of the two genes. Furthermore, the expression of genes that may be regulated by Figla_tv2 was analyzed after RNA interference (RNAi) knockdown of Figla_tv2. In the present study, we aimed to identify the sequences of the Figla_tv1 and Figla_tv2 transcripts, examine their sex-dimorphic expression profiles, and illustrate their functional diversity during gonad development. To analyse the similarities/differences between the two Figla nucleotide sequences, a multiple sequence alignment was conducted by Clustal X. The 714-bp nucleotide sequences at the 3′ -ends were nearly identical in the two transcripts, while the remaining sequences did not show much similarity ( Figure S2). In order to illustrate the alternative splicing of the two transcripts, the cDNA sequences were successfully mapped to the genomic sequence. As shown in Fig. 1A, the genomic sequence of Figla was comprised of six exons (1-6) that were separated by five introns (A-E). The two transcripts were produced by alternatively splicing the different exons. Figla_tv1 was generated by splicing out the second exon, while Figla_tv2 was generated by splicing out the first exon, and the other exons were spliced together (Fig. 1B).

Sequence characteristics of two Figla homologues in C. semilaevis. In order to characterise the
To evaluate the similarities/differences of vertebrate Figla proteins, a multiple sequence alignment including Figlas from C. semilaevis and other vertebrates was conducted. All the vertebrate Figla proteins, including Figla_tv1 from C. semilaevis, had a typical conserved bHLH region composed of 57 aa (Fig. 2). However, Figla_tv2 from C. semilaevis lacked the basic region and six N-terminal amino acids from the HLH region (Fig. 2)

Tissue expression of Figla_tv1 and Figla_tv2.
To determine the expression patterns of the two Figla isoforms, we measured mRNA levels in 11 tissues of C. semilaevis. Figla_tv1 mRNA was predominately expressed in the skin, liver and brain of male and the gonad of female C. semilaevis and was negligibly expressed in the other tissues examined (Fig. 3). On the other hand, Figla_tv2 mRNA was expressed widely in the tissues of both sexes, including skin, liver, muscle and gill (Fig. 3), although there were some differences between the tissues from males and females. expression was detected in ovary, testis and pseudomale testis at different stages by qRT-PCR. As shown in Fig. 4, Figla_tv1 mRNA was first detected at 120 days after hatching (dah) in the ovary and then continued to be expressed until it reached a maximum at 1 year after hatching (yah) before declining sharply at 2 yah. Conversely, Figla_tv1 mRNA was almost undetectable in the gonads of both male and pseudomale C. semilaevis during all the developmental stages tested (Fig. 4). Figla_tv2 mRNA was exclusively expressed in the pseudomale testis ( Fig. 4), where it was initially detected at 120 dah and then rapidly increased to reach a maximum at 180 dah before declining sharply at 1 yah. Figla_tv2 mRNA persisted in the adult testis at 2 yah (Fig. 4). we measured the expression of Sox9a, wt1a and tesk1, three spermatogenesis-related genes in C. semilaevis 20,21,25 . However, no differential expression (P > 0.05) between the treated group and the control was observed for the three genes as detected by qRT-PCR (data not shown). Besides, the mRNA levels of two steroid hormone-encoding genes, steroidogenic acute regulatory protein (StAR) and cytochrome P450 side-chain cleavage (P450scc), were detected after RNAi silencing of

Discussion
The Figla gene has been cloned in many mammalian and fish species, but until now only one transcript has been reported in each species. In the present study, in addition to the common form (Figla_tv1) of the gene, we isolated and characterized another transcript (Figla_tv2) from C. semilaevis. The highly conserved bHLH region is thought to be required for Figla binding to the E-box motif of the target protein 10,11 . Because Figla_tv2 lacked the conserved bHLH region (Fig. 2), we speculate that Figla_tv2 may be incapable of interacting with DNA directly and may function by associating with other molecules.
In vertebrates such as Bos taurus, M. musculus, O. latipes, O. niloticus and H. sapiens, Figla was expressed mainly in the gonads 9,11,13,15,26 . In this study, we also detected C. semilaevis Figla_tv1 expression mainly in the gonads of one-year-old female fish (Fig. 3), indicating the conserved role of Figla_tv1 in gonad tissue during development of the fish. We also noted that Figla_tv1 was expressed in other tissues, such as the skin and liver of male C. semilaevis (Fig. 3). However, the function of Figla in these tissues is not yet clearly understood. The robust expression of Figla in the skin and liver of C. semilaevis may indicate a role for Figla in immune functions, as previously reported in several teleosts [27][28][29] . It is very likely that the previous studies of Figla tissue distribution reported overlapping expression of both alternative splicing forms. For example, G. rarus Figla showed a similar composition to C. semilaevis Figla_tv1, but its expression pattern was the same as Figla_tv2 (Fig. 3) 14 . To evaluate whether there are multiple Figla isoforms in other teleosts too, we searched the transcriptome data in GenBank and Ensemble. We found that in most teleosts there was only one Figla isoform (Figla_tv1) except that Takifugu rubripes expressed three Figla isoforms. Therefore, we speculate the multiple Figla isoforms probably exist in some fish species or even are generally present in the teleosts, while they are undetectable due to the sequencing depth or coverage. More studies are needed to test these possibilities.
In C. semilaevis, the histological and cellular differentiation of the ovary is not synchronous; histological differentiation was reported to begin 56-62 dah, whereas cellular differentiation occurred at 120 dah with the appearance of the ovarian cavity 18 . In this study, Figla_tv1 mRNA was first detected at 120 dah in the ovary and persisted into adulthood (Fig. 4). This finding is reasonably consistent with the cellular differentiation period of the ovary, suggesting that  In order to evaluate whether Figla_tv2 was involved in sex-reversal, we measured the levels of Figla_tv2 transcripts in the gonads of males, females and pseudomales. Figla_tv2 was expressed exclusively in the testis of pseudomales, with weak or no expression in the testis of males or in the ovary of females (Fig. 4), suggesting it is probably a pseudomale-biased gene that may play an important role in the sex-reversal process.
The ISH results showed that Figla_tv2 mRNA was mainly located in the germ cells of pseudomale testis (Fig. 5i,l), suggesting an important role in regulating spermatogenesis. However, the underlying regulatory mechanism is still unknown. Proteins that belong to the bHLH superfamily generally function as transcriptional enhancers or inhibitors of target genes 1,2 . Despite the lack of a complete bHLH region, we speculate that Figla_tv2 may regulate the development of pseudomale testis by repressing or activating its target genes. Steroid hormones are commonly present in the gonads of vertebrates, where they are known to play critical roles in gonad development [30][31][32] . StAR and P450scc are key regulatory proteins in the synthesis and metabolism of steroid hormones in teleosts [33][34][35] . StAR mediates the transport of cholesterol across the mitochondrial membrane, a rate-limiting step in steroid hormone synthesis [32][33][34][35] , and P450scc catalyzes the conversion of cholesterol to pregnenolone, a pivotal step in the initiation of steroidogenesis [33][34][35] . After the in vitro knockdown of Figla_tv2 in cultured pseudomale testis cells, qRT-PCR showed that StAR and P450scc mRNA levels were significantly up-regulated (Fig. 6B), suggesting C. semilaevis Figla_tv2 may be involved in spermatogenesis by regulating steroid hormones synthesis and metabolism in the gonads of pseudomales.

Conclusions
Two functional homologues, Figla_tv1 and Figla_tv2, were successfully isolated from C. semilaevis. The two transcripts had different tissue expression patterns in males and females, implying their probably functional diversity. Figla_tv1 exhibited predominant "vertebrate-like" ovarian expression, suggesting a conserved role in folliculogenesis. Figla_tv2 was expressed mainly in the testis of pseudomales and is likely to play a role in the spermatogenesis of pseudomales by influencing the synthesis and/or metabolism of steroid hormones. Fish and sample collection. All half-smooth tongue sole (C. semilaevis) used in these experiments were purchased from Huanghai Aquaculture Ltd. (Haiyang, China). One-year-old fish (three individuals of each sex) were randomly sampled. Spleen, skin, heart, intestine, brain, head kidney, liver, muscle, gill, ovary, and testis tissues (11 in all) were collected and immediately transferred to liquid nitrogen and were then stored at − 80 °C until RNA extraction. In addition, parts of the gonad were fixed in 4% paraformaldehyde (pH 7.4) at 4 °C for 24 h, then dehydrated through a methanol series (25, 50, 75, and 100%) and finally stored in 100% methanol at − 20 °C for ISH analysis. Moreover, three independent gonadal samples at 35, 65, 80, 120, 150, and 180 dah and at 1 and 2 yah were sampled within one family. To determine the genetic sex of the fish, parts of the caudal fins were collected and fixed in 100% ethanol for DNA extraction.
cDNA synthesis and identification of the genetic sex. Total RNA was extracted from the frozen tissues with TRIzol reagents (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Then, the RNA was transcribed to cDNA using a PrimeScript ™ RT reagent kit (TaKaRa Bio Inc., Otsu, Japan) with gDNA Eraser to avoid contamination by genomic DNA.
Genomic DNA was extracted from the fins using the phenol-chloroform method as described previously 16 . Female-specific primer pairs (CseF382F and CseF382R) (Table S1) were used to identify the genetic sex of the fish 23 . After PCR amplification with the conditions: 94 °C for 5 min, followed by 35 cycles (94 °C for 30 s, 54 °C for 30 s, and 72 °C for 30 s), and 72 °C for 10 min, the products were run on a 1.2% agarose gel, and samples with a single 291 bp band were identified as genomic females, while samples with no bands were identified as genomic males.
Phenotypic sex was determined by histological analysis following procedures described previously 36 . Pseudomales are individuals for which the genomic sex is identified as female, but the phenotypic sex is male.
Isolation of Figla full-length cDNA from C. semilaevis tissues. RACE-ready first-strand cDNA was synthesized from total RNA using a SMART ™ RACE cDNA amplification kit (Invitrogen) according to the manufacturer's instructions. The specific primers for the outer and nest amplifications were based on the mRNA sequence that determined from whole-genome sequencing 18 and the RNA-seq data 24 of C. semilaevis. The sequences of all primers used for RACE amplifications are listed in Table S1. The universal primer UPM and the outer primers were used for the first 5′ -and 3′ -RACE amplifications. The products were then diluted 100 times with ddH 2 O and used as templates for the nested PCR reactions with the nest primers and the universal primer NUP. The outer and nest amplifications were both performed using the same touchdown PCR procedures: denaturation at 94 °C for 5 min, followed by 15 cycles (94 °C for 30 s, first five cycles 68 °C for 30 s, and subsequently the temperature was reduced by − 2 °C per five cycles, and 72 °C for 2 min), 20 cycles (94 °C for 30 s, 60 °C for 30 s, and 72 °C for 2 min), and then incubation at 72 °C for 10 min. The nested PCR products were separated on a 1.2% agarose gel and purified with a DNA purification kit (TaKaRa Bio Inc.). The purified 5′ -and 3′ -RACE products were subcloned into a PMD-18T vector (TaKaRa Bio Inc.) and sequenced. qRT-PCR analysis. The expression levels of the two Figla genes in the tissues and gonads at differential developmental stages were determined by qRT-PCR as described previously 23 . Two pairs of primers (Table S1) were designed based on the specific gene regions of Figla_tv1 and Figla_tv2. The specificity of the primers was verified by a single distinct peak obtained in a melting curve analysis. The β-actin was identified previously as a reliable internal reference gene in various tissue samples of C. semilaevis 39 and was used as the first reference gene. In addition, Rpl13α was validated as a suitable reference gene for quantifying gene expression in teleost species such as Danio rerio 40 . Our recent studies have found it was expressed with a relatively stable level in various tissues of C. semilaevis (data not shown). Therefore, it was employed as the second reference gene. The geometric means of β-actin and Rpl13α expression levels were used to normalize the data. Quantification was conducted using an ABI 7500 detection system (Applied Biosystems, Foster City, CA, USA) with SYBR Green Master Mix (TaKaRa Bio Inc.). The qRT-PCR procedure was as follows: 94 °C for 30 s, followed by 40 cycles of 94 °C for 5 s, and 60 °C for 34 s. Tissues and gonads were sampled from three individuals, and triplicate assays for each sample were conducted. The ABI 7500 system SDS software version 1.4 (Applied Biosystems) was used to analyze the qRT-PCR data, and baseline and cycle threshold values were set automatically. The relative mRNA expression of target genes was calculated by the 2 −△△Ct method, as described previously 41 .
All data are presented as the mean ± SEM of three samples with three parallel repetitions. Differences between means were tested by one-way analysis of variance (ANOVA) followed by Duncan's post-hoc test. (SPSS software version 18.0; SPSS Inc., Chicago, IL), and the significance level was set at P < 0.05. All assays in the qRT-PCR complied with the MIQE guidelines 42 .  (Table S1), respectively. The PCR products were sub-cloned into the pEASY-T5-Zero vector and verified by sequencing. The recombinant plasmids with the correct insertion were extracted with a plasmid extraction kit (Tiangen Bio Inc., Beijing, China) and linearized by BamHI or EcoRV. The linearized plasmids were used as templates and the probes were synthesized with a DIG RNA Labeling Kit (Roche, Mannheim, Germany). For ISH analysis, the stored gonads were dehydrated in an ascending gradient of ethanol and embedded in paraffin wax, and 6-μ m sections were cut. ISH was performed as described previously 43 using three replicate samples. The sections were examined and photographed with a Nikon E80i microscope (Nikon Co., Tokyo, Japan).  China). The cells used for the RNAi treatment were from the CSPMG cell line that was established previously 44 and stored in liquid nitrogen. After thawing by incubating in a 42 °C water bath for 5 min, the CSPMG cells were harvested by centrifugation at 180 × g for 10 min. Then, the cells were incubated at 24 °C in 25-cm 2 cell culture flasks with 20% FBS-DMEM/F12 medium. After an approximately 3 days of culturing, the cells formed a confluent monolayer. The cells were transferred to six-well plates and cultivated at 24 °C for 12 h to ensure complete attachment to the plates. The siRNA was transfected into the CSPMG cells using Lipofectamine 2000 reagent (Invitrogen) following the manufacturer's instructions. An NC siRNA labeled by Cy3 was used to assess the transfection efficiency, and the transfection efficiency was evaluated by calculating the ratio of cells expressing red fluorescence signals to all the cells used for transfection. The average transfection efficiency of the cells used for assay was calculated to be about 90%. The treated groups were transfected with siRNA (si-cse-Figla 001 and si-cse-Figla 002) at a final concentration of 30 nM. The controls were transfected with NC siRNA at the same concentration. The cells from the treated and control groups were cultivated at 24 °C for 48 h and were then harvested by centrifugation at 180 × g for 10 min. After washing three times with phosphate-buffered saline, the cells were stored in the same TRIzol reagents that were used for the RNA extraction. Three replicates were conducted for each group. Total RNA extraction and the re-transcriptions were performed according to the methods described above. The expression of Figla_tv2, StAR, and P450scc was measured by qRT-PCR using the primers listed in Table S1. The qRT-PCR data were analyzed as described above.