Introduction

Fishes in a wide variety of aquatic habitats around the world show various types of sex determination patterns (Desjardins and Fernald 2009). Sex determination in fish is a flexible and complex process, with some species exhibiting the genetic sex determination (GSD) pattern: the sex of the organism is determined by its genotype. Fishes with GSD in which the sex determination and differentiation is influenced by environmental factors are referred to as GSD + EE (environmental effects) species (Ospina-Alvarez and Piferrer 2008; Penman and Piferrer 2008). Regardless of the variation among species, it is hypothesized that genes involved in sex determination and differentiation are conserved throughout evolution. The steroidogenic enzyme cytochrome P450 aromatase (P450arom, cyp19a gene) and its transcription factors have been shown related to mammal and fish gonadal differentiation. Among the transcription factors, forkhead box L2 (foxl2) plays a central role in fish ovarian differentiation, and doublesex- and mab-3-related transcription factor 1 (dmrt1), mullerian inhibitory substance (mis), and Sry-related HMG box-9 (sox9) could play important roles in testicular differentiation in most fish species (Vernetti et al. 2013).

The nr0b1 gene, also called dax1 (dosage-sensitive sex reversal, adrenal hypoplasia congenital critical region, on chromosome X, gene 1), and the nr5a subfamily, also known as Fushitarazu factor-1 (ftz-f1), are also important transcriptional regulators of the expression and activity of cyp19a during sex differentiation in many vertebrates (Gurates et al. 2003; Nakamoto et al. 2007). The nr0b1 gene is an unusual member of the orphan nuclear receptor superfamily. Mammalian nr0b1 is repressed by sry during male sex determination (Swain et al. 1998) and acts as a negative regulator of the genesis of steroids through inhibiting nr5a1 transcription activity (Park and Jameson 2005). Fish nr0b1 has been identified and characterized in several species including Oryzias latipes, Oreochromis niloticus, Dicentrarchus labrax, and Oncorhynchus mykiss (Wang et al. 2002; Baron et al. 2005; Martins et al. 2007; Nakamoto et al. 2007), but its role during gonadal differentiation is unclear. The nr5a subfamily genes are regulators in steroidogenic organs and play important roles in the transcriptional activation of steroidogenic enzymes (Hammer and Ingraham 1999). Homologue genes of nr5a have also been identified in teleosts including Oryzias latipes, Cynoglossus semilaevis, Epinephelus coioides, Oreochromis niloticus, and Sebastes schlegelii (Watanabe et al. 1999; Yoshiura et al. 2003; Zhang et al. 2004; Deng et al. 2008; Shafi et al. 2013). Investigations have shown that teleost nr5a genes may be involved in gonadal differentiation through regulating the transcription of cyp19a.

Olive flounder Paralichthys olivaceus is an important maricultured fish in Japan, Korea, and China. The female flounder generally exhibits more rapid growth than the male (Yamamoto 1999), so there has been considerable attention given to all-female production and the mechanisms of flounder sex determination and gonadal differentiation. It is possible to reverse the phenotypic female to male in larvae reared at high water temperatures during the thermosensitive period of gonadal differentiation (Yamamoto 1999; Sun et al. 2013). Hence, the flounder is a good model for the study of gonadal differentiation in fish. The cyp19a gene and its transcriptional factors, including foxl2, mis, dmrt1, and sox9, have been reported to be involved in flounder gonadal differentiation (Jo et al. 2007; Kitano et al. 1999; Wen et al. 2011; Yamaguchi et al. 2007; Yoshinaga et al. 2004). Our recent transcriptome study of flounder gonads showed that flounder nr0b1 is a female-biased gene (Fan et al. 2014), but its full-length complementary DNA (cDNA) sequence and its expression patterns during the flounder gonadal differentiation period and at different gonadal development stages remain unclear. Moreover, nr5a genes have not been reported in flounder.

In the present study, flounder nr0b1 and nr5a2 cDNA sequences were isolated and characterized, and their expression levels in adult tissues and gonads at different development stages were analyzed. The nr0b1 and nr5a2 expression patterns in gonads of gynogenetic and high-temperature-treated gynogenetic flounder during the gonadal differentiation period, and the relationship between these two genes and cyp19a during this period, were also studied. The goals of the research were to determine whether nr0b1 and nr5a2 are involved in flounder gonadal differentiation and to provide basic data for further investigation of the molecular mechanism of sex phenotype formation in the flounder and other fishes.

Materials and methods

Animals

Adult wild-type male and female flounder (37–45 cm in total length, TL) and specimens at development stages I to V (14–45 cm TL) were purchased from Nanshan market (Qingdao, China) or Shenghang fish farm (Weihai, China). The gonadal development stages I–V were determined by histological evaluation according to Sun et al. (2009) and Blazer (2002). The ovarian development stages were based on ovarian follicle growth and maturation, while testicular stages were based on the relative proportions of spermatocytes, spermatids, and spermatozoa.

Gynogenesis and thermal treatment

Artificially induced gynogenesis was conducted to obtain the diploid XX flounder (You et al. 2001). When the juveniles reached 1.2–1.5 cm TL, they were sorted into a control (18 ± 0.5 °C) and a high-temperature treatment group (28 ± 0.5 °C). Each group, comprising 200 randomly selected individuals, was placed in a 90-L plastic tank with aerated seawater. Water temperature in the treatment group was raised from 18 to 28 °C at a rate of 2 °C per day. Photoperiod was maintained at 14L:10D. The juveniles were fed with a commercial diet two to four times daily until they reached 15 cm TL to ensure the verification of sex by histology. The experiment was duplicated.

Total RNA isolation and cDNA synthesis

Tissue samples, including testis, ovary, kidney, heart, liver, brain, spleen, kidney, head kidney, stomach, intestine, muscle, eye, and gill, were collected from five male and five female wild-type adult fish. Three testis and three ovary tissue samples from wild-type flounder were also collected at each development stage after histological evaluation. Samples were immediately immersed in liquid nitrogen and stored at −80 °C. Total RNA of adult tissues and gonads was extracted using TRIzol Reagent (Invitrogen, USA) and treated with DNase I (MBI Fermentas, Canada). Quantity and purity of the RNA was checked using the ratio A260:A280 nm (NanoDrop ND-1000 Spectrophotometer, USA). The first strand cDNA was synthesized from 1 μg of total RNA using SuperScript II Reverse Transcriptase (Invitrogen, USA). Gonads of gynogenetic juveniles at 1.5, 2, 3, 3.8, 4, 6, 7, 8, and 10 cm TL (n = 6 in each TL group) were sampled under a stereomicroscope (Leica MZ125, Germany). To investigate differences in gene expression patterns of gynogenetic female (XX) vs. wild-type male vs. XX neo-male, gonads from five juveniles (12 cm TL) of each type were sampled. Half of each gonad was used for histological evaluation to verify the sex, and the other half was used for quantitative real-time PCR (qRT-PCR). The samples for qRT-PCR were immediately immersed in liquid nitrogen and stored at −80 °C. Total RNA of gonads was isolated using E.Z.N.A. MicroElute Total RNA Kit (OMEGA, USA). First strand cDNA was synthesized using PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Japan).

Molecular cloning and analysis of nr0b1 and nr5a2

The full-length cDNA sequences of flounder nr0b1 and nr5a2 were obtained using the rapid amplification of cDNA ends (RACE) method with the SMART RACE cDNA Amplification Kit (Clontech, Japan) in accordance with the manufacturer’s instructions. Primers are listed in Table 1 (P1-P6 for nr0b1 and P7-P12 for nr5a2). Alignments of the amino acid sequences of several species retrieved from GenBank were conducted with the multiple alignment software ClustalW (http://www.ebi.ac.uk/clustalw). The phylogenetic tree was reconstructed using the neighbor-joining method implemented in MEGA v. 5.0, and the branch supports were assessed with 1000 bootstrap replications.

Table 1 PCR primers used in the present study
Full size table

Expression pattern analysis of nr0b1, nr5a2, and cyp19a

The transcript levels of nr0b1 and nr5a2 in organ tissues and gonads at different development stages were analyzed by qRT-PCR. In order to better understand the relationship between cyp19a and its transcriptional regulators nr0b1 and nr5a2, the temporal expression of cyp19a over the course of the flounder gonadal differentiation period in the 18 °C group was analyzed. Flounder-specific qRT-PCR expression primers were used for nr5a2 (Table 1, P13 and P14), nr0b1 (Table 1, P17 and P18), and cyp19a (Table 1, P19 and P20, Wen et al. 2014). The β-actin was selected as the reference gene (Table 1, P15 and P16) based on a previous study (Zheng and Sun 2011).

The qRT-PCR was conducted using Platinum SYBR Green qPCR SuperMix-UDG (Invitrogen, USA) with an Eppendorf real-time detection system following manufacturers’ instructions. Amplification consisted of an initial denaturation step at 95 °C for 5 min; 35 cycles at 95 °C for 30 s, 58 °C for 30 s, and 60 °C for 60 s; followed by a final extension at 72 °C for 10 min. A single-cycle melting curve analysis of amplified product was performed following each qRT-PCR to confirm that a single PCR product was amplified and detected. Samples were run in triplicate, and relative gene expression levels were calculated by the 2−ΔΔCt method. Significance among groups was tested by ANOVA or the independent samples t test using SPSS software package (P < 0.05).

Results

Characterization of flounder nr0b1 and nr5a2

The full-length nr0b1 cDNA sequence of 1446 bp was obtained from flounder gonads. The cDNA contained a 5′ untranslated region (UTR) of 219 bp, an open reading frame (ORF) of 897 bp, and a 3′ UTR of 330 bp (HQ380020, Fig. S1A). The deduced protein possessed 298 amino acid residues with a calculated molecular mass of 33.08 kDa. A typical polyadenylation signal (AATAAA) was found at 14 bp upstream of the poly (A) tail. The analysis of conserved domains revealed the presence of a DNA-binding domain. Multiple sequence alignment and phylogenetic analysis (Fig. 1a) suggested that the predicted flounder Nr0b1 protein exhibits high sequence identity to other fish Nr0b1 proteins. The alignment also revealed that Nr0b1 is more conserved in the C-terminal region than in the N-terminal.

Fig. 1
figure 1

Neighbor-joining trees of Paralichthys olivaceus Nr0b1 (a) and Nr5a2 (b) with their homologues in other vertebrates based on information from GenBank. The phylogenetic tree constructed using the NJ method and bootstrap analysis with 1000 replicates was used to assess the strength of nodes in the tree. GenBank accession numbers of the sequences used in the analysis are indicated in the figures

Full size image

Flounder full-length nr5a2 cDNA encoding 523 amino acids consists of 2425 bp with a 5′ UTR of 573 bp, an ORF of 1572 bp, and a 3′ UTR of 280 bp (JX999939, Fig. S1B). The deduced Nr5a2 protein contained all domains found in other Nr5a/Ftz-f1 proteins, including the highly conserved DNA-binding and ligand-binding regions I, II, and III and Ftz-f1box, as well as the activation function-2 hexamer. Multiple sequence alignment revealed that the predicted flounder Nr5a2 protein showed the highest similarity to Acanthopagrus schlegelii Ff1a (98 %). Phylogenetic analysis (Fig. 1b) resulted in four groups of homologous Nr5a/Ftz-f1: Ff1a, Ff1b, Ff1c, and Ff1d. The flounder Nr5a2 clustered with other teleost Ff1a, indicating that the flounder nr5a2 gene should be classified in the ff1a group.

Quantitative expression of nr0b1 and nr5a2 in wild-type flounder tissues

The expression levels of flounder nr0b1 and nr5a2 in ovary were higher than those in testis, demonstrating a sexually dimorphic expression pattern (Fig. 2). With the exception of female gill, all organs analyzed showed nr0b1 gene expression, with gonad, spleen, and brain exhibiting the highest levels. Ten of the twelve examined organs, but not head kidney and muscle, showed significantly different expression levels (P < 0.05; using independent samples t test) in the sexes. Higher expression in the female was detected in gonad, heart, liver, and intestine, whereas the male organs exhibited higher levels in brain, stomach, kidney, spleen, gill, and eye (Fig. 2a). The nr5a2 gene was expressed in all examined tissues except head kidney. The highest levels were detected in liver, intestine, and stomach. Only gonad, brain, liver, intestine, stomach, and eye showed significantly different expression levels (P < 0.05; independent samples t test) in the sexes. All tissues except brain showed female-enhanced expression levels (Fig. 2b).

Fig. 2
figure 2

Expression patterns of nr0b1 (a) and nr5a2 (b) in wild-type adult female and male flounder organ tissues. qRT-PCR of nr0b1 and nr5a2 relative to β-actin in tissues from five females or five males. Values are means ± S.D. An asterisk indicates significant difference (P < 0.05, independent samples t test) between females and males

Full size image

The expression of nr0b1 and nr5a2 during flounder gonadal differentiation

The sex of XX flounders reared at 18 and 28 °C was determined histologically when the juveniles reached 15 cm TL. The percent of males (15–20 cm TL) in the 18 and 28 °C groups was 0 % (n = 28) and 97.9 % (n = 49), respectively.

The expression patterns of nr0b1 and nr5a2 indicated that both genes were expressed in primitive gonads and the ensuing gonadal differentiation period (Fig. 3). The expression level of nr0b1 increased rapidly in primitive gonad prior to differentiation (2 cm ± 0.1 cm TL) in both temperature groups, with the expression peak observed at 2.0 ± 0.1 cm TL. During the gonadal differentiation period, the nr0b1 messenger RNA (mRNA) level decreased sharply from 2 to 4 cm TL, then remained at a consistently low level. The lowest expression level of nr0b1 in the 18 °C group and the 28 °C group was observed at 6 and 4 cm TL, respectively (Fig. 3a). Prior to gonadal differentiation, flounder nr5a2 was initially highly expressed and subsequently decreased markedly, with the expression peak observed at 1.5 cm ± 0.1 cm TL. During gonadal differentiation, expression patterns showed a variation trend similar to nr0b1, with the expression levels in the 18 °C group consistently slightly higher than seen in the 28 °C group. At TL 8 cm, a significantly higher level of expression was observed in the 18 °C group compared to that in the 28 °C group. Similar to nr0b1, the lowest expression of nr5a2 in the 18 and 28 °C groups appeared at 6 and 4 cm TL, respectively (Fig. 3b).

Fig. 3
figure 3

qRT-PCR of nr0b1, nr5a2, and cyp19a relative to β-actin in gynogenetic flounder gonads during the gonadal differentiation period. a nr0b1 gene expression during gonadal differentiation. b nr5a1 gene expression during gonadal differentiation. c Comparison between cyp19a mRNA level (histogram) and the ratio of nr5a2/nr0b1 (line) in the 18 °C group. Larvae grown at 18 and 28 °C resulted in 100 and 2.1 % females, respectively. All values are means ± S.D. (n = 6). An asterisk indicates significant difference (P < 0.05, independent samples t test) between the 18 and 28 °C groups

Full size image

Because nr0b1 represses nr5a target genes by regulating nr5a transactivation, one might expect that the nr5a2/nr0b1 ratio would determine target gene transcription. Hence, the ratio of nr5a2/nr0b1 and expression level of its target gene, cyp19a, in flounder gonad were analyzed during the gonadal differentiation period. A positive relationship was observed between the expression of the cyp19a and the ratio of nr5a2/nr0b1 during ovarian differentiation in the 18 °C group (Fig. 3c).

We compared gene expression patterns in gynogenetic females (XX) vs. wild-type males (assumed to be XY) vs. XX neo-males. Neither nr0b1 nor nr5a2 showed significantly different expression in wild-type males from that in XX neo-males (Fig. 4).

Fig. 4
figure 4

Expression levels of nr0b1 (a) and nr5a2 (b) in gynogenetic female (XX), wild-type male, and XX neo-male gonads. qRT-PCR of nr0b1 and nr5a2 relative to β-actin in flounder gonads. All values are means ± S.D. (n = 3). Groups with different letters show significant differences (P < 0.05, ANOVA)

Full size image

The expression of nr0b1 and nr5a2 in wild-type flounder gonads at five development stages

In general, both nr0b1 and nr5a2 genes were more highly expressed in ovary than in testis at all five development stages (Fig. 5). At stages II, IV, and V, nr0b1 expression levels in ovary were significantly higher than observed in testis (P < 0.05), with no significant differences at stages I and III (Fig. 5a). Expression of nr5a2 in ovary was significantly higher than in testis at stages I, II, III, and IV (P < 0.05). There were no significant differences between ovaries and testes at stage V (Fig. 5b).

Fig. 5
figure 5

Expression levels of nr0b1 (a) and nr5a2 (b) in flounder gonads at five developmental stages. qRT-PCR of nr0b1 and nr5a2 relative to β-actin in wild-type flounder gonads. Values are means ± S.D. (n = 3). An asterisk indicates significant differences (P < 0.05, independent samples t test) between the ovary and testis

Full size image

Discussion

The flounder nr0b1 has a partially conserved Kozak sequence in the 5′ UTR and a putative polyadenylation signal in the 3′ UTR. The deduced flounder Nr0b1 protein has two distinct domains. Its C-terminus is homologous to the ligand-binding domain of the nuclear hormone receptor family. The unique N-terminus lacks the characteristic zinc-finger motif and contains only one of the three LXXLL motifs present in mammalian Nr0b1, which is involved in protein interactions as a nuclear receptor binding domain (Guo et al. 1996). Flounder Nr5a2 has highly conserved regions of I, II, III, FTZ-F1 box, and AF-2 hexamer, as seen in other fishes such as Oryzias latipes, E. coioides, and Oreochromis niloticus. The phylogenetic analysis showed flounder Nr5a2 to cluster with other teleost Ff1a. As several reported ftz-f1 in other fish species, flounder nr5a2 cannot be categorized into either the ad4bp/sf-1 group or lrh-1/ftf group (Kuo et al. 2005).

In order to investigate if there was a gene expression difference between wild-type males (eggs fertilized with spermatozoa of males with no treatment and juveniles cultured at ambient temperature, assumed to be XY type) and XX neo-males, we studied the nr0b1 and nr5a1 expression patterns in gynogenetic females (XX) vs. wild-type males vs. XX neo-males. Neither nr0b1 nor nr5a2 showed significantly different expression in wild-type males vs. XX neo-males. This indicated that the XX neo-male was a good model for studying the mechanism of flounder sex phenotype formation.

Flounder nr0b1 was widely expressed in most adult organs, as were its orthologs in Oreochromis niloticus (Wang et al. 2002) and D. labrax (Martins et al. 2007), whereas flounder nr5a2 transcripts were primarily observed in ovary, testis, brain, liver, intestine, and stomach of adult fish. The two genes were co-expressed in flounder gonad, brain, liver, intestine, and stomach. A study of Clarias gariepinus showed that the expression pattern of ftz-f1 (nr5a) in brain during gonad ontogeny and the ovarian reproductive cycle correlates with cyp19b expression in brain and cyp19a expression in ovary; hence ftz-f1 might implicate potential roles in ovarian differentiation and development via the brain-pituitary-gonad axis (Sridevi et al. 2011). High levels of flounder nr0b1 and nr5a2 are co-expressed in brain and gonad, suggesting a potential role in flounder gonad development via endocrine feedback. Further study is needed to confirm this. The expression levels of both nr0b1 and nr5a2 were higher in ovary than in testis. Our previous transcriptome study of flounder gonads identified several genes with expression profiles differing in ovary and testis and showed that nr0b1 was overexpressed in ovary (Fan et al. 2014). The results of the present study were consistent with these findings. However, no valid information about nr5a2 expression was detected in the transcriptome data, as the assembled contig of this gene was of insufficient length (unpublished data).

The gonadal differentiation period in flounder is closely related to its total length. According to Sun et al. (2009), when juveniles reached 38.0 ± 1.7 mm TL, the ovarian cavity was observed in the developing ovary. When TL reached 86.5 ± 5.9 mm, oocytes were detected. Differentiation of testis was initially observed in juvenile flounders at 63.5 ± 3.4 mm TL, when spermatogenic cells proliferated rapidly and the sperm duct was formed. The seminal lobule appeared when juveniles reached approximately 76.0 mm TL. In the present study, the high expression of nr0b1 and nr5a2 in undifferentiated flounder gonad suggested that these genes may have an effect on primitive gonad development. When juveniles reached 3–4 cm TL, at onset of ovarian differentiation, the expression levels of nr0b1 and nr5a2 in the 18 °C group were higher than in the 28 °C group. When the larvae reached 6–8 cm TL, at onset of testicular differentiation, the nr0b1 and nr5a2 expression was higher in the 28 °C group. A study of A. schlegelii showed that the testicular development required cooperative function of nr0b1 and nr5a4 and that they exhibit antagonistic interaction in oocyte development (Wu et al. 2008). A similar mechanism may exist in flounder nr0b1 and nr5a2.

Although nr0b1 and nr5a are important for fish reproduction, they show species-dependent expression patterns in gonad. Oncorhynchus mykiss nr0b1 is characterized as an early expressed gene involved in testicular differentiation (Baron et al. 2005). In contrast, Oreochromis niloticus nr0b1 may be involved in the regulation of ovarian differentiation (Wang et al. 2002). No dimorphic pattern of nr0b1 was detected at the mRNA level during either the thermosensitive period or gonadal differentiation period in D. labrax (Wang et al. 2002). Expression levels of nr5a are higher in testis than in ovary in S. schlegelii, C. gariepinus, and A. schlegelii (Liu et al. 2004; Shafi et al. 2013; Sridevi et al. 2011), whereas its expression in E. coioides and Oreochromis niloticus is higher in ovary (Yoshiura et al. 2003; Zhang et al. 2004), as it is in flounder. It was reported that the most important functions of mammalian nr0b1 during gonad development are repressing nr5a1-mediated transcription of steroidogenic gene cyp19a and promoting the development of testis by regulating mis transcription (Shen et al. 1994). However, the roles of nr0b1 and nr5a in fish gonad development have not been elucidated in detail. An important mechanism in which nr0b1 exerts its negative transcription effect is associated with its interaction with nr5a; hence, the nr0b1/nr5a ratio could be a factor in the fine-tuning of nr5a target gene transcription. Several reports have shown that nr5a could be a transcription regulator of cyp19a expression and activity in fish such as Oncorhynchus mykiss and Oreochromis niloticus (Watanabe et al. 1999; Yoshiura et al. 2003). It is well known that cyp19a is responsible for conversion of androgens to estrogens and that the estradiol-17β (E2) produced by it is required for ovary development. In this study, we found a positive relationship of the ratio of nr5a2/nr0b1 expression with the expression level of cyp19a during the flounder gonadal differentiation period. The ratio of nr5a2/nr0b1 and the expression of cyp19a both increased at 3 cm TL in the 18 °C group immediately before the rapid increase of E2 at the onset of ovarian differentiation (Sun et al. 2013). Following the onset of ovarian differentiation, the ratio of nr5a2/nr0b1 and the cyp19a expression decreased to a lower level, and the E2 concentration dropped correspondingly. The expression ratios of nr5a2/nr0b1 and cyp19a expression level in the 28 °C group changed slightly during testicular differentiation (data not shown), similar to the tendency of E2 concentration reported by Sun et al. (2013). Thus, nr0b1 and nr5a2 may affect E2 levels indirectly and be involved in flounder gonadal differentiation by regulating the cyp19a expression. Future study is needed to investigate how nr0b1 and nr5a2 regulate gonadal differentiation in flounder.