Introduction

Sequential hermaphroditism (sex change) is a reproductive strategy increasingly reported in the teleost order. Three forms are known to be operative: protandry (male changes to female), protogyny (female changes to male), and bidirectional (ability to change sex multiple times). Since Liu’s (1944) initial report of sex change in the rice field eel, from female to male after the first spawning, the understanding of the reproductive anatomy and ecology of sex change has grown significantly (Yang et al. 2004; Yin et al. 2005). In contrast, knowledge of the physiology governing this process has lagged considerably; however, recent studies examining the endocrine processes of sex change have provided some insight (Chang and Liu 1998; Anderson and Forrester 2001).

Gonadal steroidogenesis is critical for the development of sexually dimorphic phenotypes in nonmammals, and the results of numerous studies have shown that activation of P-450 aromatase gene expression is important for the ovary-determining pathway in nonmammals (Akatsuka et al. 2005). P450arom is a heme-binding protein of the enzymatic complex responsible for the conversion of C19 androgens into C18 estrogens, which is encoded by the CYP19 gene. CYP19 has been cloned and characterized in fish, such as tilapia (Chang et al. 1997b), trout (Tanaka et al. 1992), catfish (Trant 1994), goldfish (Tchoudakova and Callard 1998), medaka (Tanaka et al. 1995), and zebrafish (Chiang et al. 2001). It is expressed mainly in gonads and brain. It plays critical roles in sex differentiation in gonochoristic fish (Chang et al. 1997a). CYP19 activity was elevated in natural sex differentiation of genetically female trout and female tilapia (Guiguen et al. 1999), and repression was detected in XX females masculinized by the elevation of temperature during the sex differentiation period (Kwon et al. 2000). A surge of elevated aromatase activity was observed during the sex change of protandrous black porgy (Chang et al. 1997a). Also, many experiments have shown that 17a-methyl testosterone or the aromatase inhibitor fadrozole caused sex reversal in many fish species, such as Oreochromis niloticus (Afonso et al. 2001; Kwon et al. 2000), salmon (Salmo trutta) (Piferrer et al. 1994), grouper (Epinephelus merra) (Bhandari et al. 2004, 2005), seabream (Acanthopagrus schlegeli) (Lee et al. 2002), and Japanese flounder (Paralichthys olivaceus) (Kitano et al. 2000).

The rice field eel (Monopterus albus) is a kind of protogynous hermaphrodite. The objective of this study was to clone P450arom cDNA from the rice field eel and to characterize its gene structure and putative cis-acting elements in the 5′-flanking region. Furthermore, we detected aromatase gene expression during various stages of development and in various tissues. This work will aid the elucidation of the relationship between P450arom and sex change and also provide information on the differences in regulation of P450arom in other kinds of fish.

Materials and Methods

Animals and DNA, RNA Extraction

Two adult female rice field eels with ovaries in Stage III were used to isolate P450arom cDNA. Additional animals (three per sex) were used to study the tissue expression pattern of P450arom. For developmental expression in the brain and ovary, four fry (about 30 days old) and three fry (about 50 days old) were used.

All experimental fishes were killed by decapitation after being placed under 0.01% benzocaine anesthesia. Tissues were immediately dissected for RNA extraction. Total RNA was extracted using the Tripure Isolation Reagent (Roche) according to the manufacturer’s recommendations. For developmental expression, total head and abdomen sections, containing the gonads but not the liver, were used. The quality and concentration of the RNA were assessed using a spectrophotometer and checked by running an aliquot (2 μg) on a 1% agarose/formaldehyde gel.

Genomic DNA was isolated from rice field eel muscle using a standard phenol chloroform protocol (Sambrook et al. 1989).

Oligonucleotide Primers

Oligonucleotides (Takara, Dalian, China) used as PCR primers are listed in Table 1. Degenerate primers were used to amplify both brain and ovarian samples and were designed to target sequences in the highly conserved I-helix, heme-binding region, and the conserved A region (Tanaka et al. 1995) using the CodeHop principle (Rose et al. 1998). IUB group codes were used for degenerate primers (R = A + G, N = A + C + T + G).

Table 1 Primers used in this study
Full size table

RT-PCR and RACE Cloning of Rice Field Eel P450arom cDNAs

First-strand cDNA was synthesized from 5 μg total RNA from ovary and brain using oligo (dT) primer and M-MLV reverse transcriptase (Takara) according to the manufacturer’s instructions. An aliquot (10%) of the reaction was amplified with primer pairs 1,2 and 3,2 using an Eppendorf Mastercycler. The reaction was performed in a 50 μl final volume using Taq DNA polymerase (Takara) and annealing at 58°C.

Rapid amplification of cDNA ends (RACE) was used to generate 3′ and 5′ ends of P450arom. The 3′ RACE product was amplified by PCR using primer 4 and the 3′ RACE adapter primer (3′ RACE core kit, Takara) with 20 ng cDNA as the template. Nested PCR was carried out with primer 5 and the adapter primer to increase the specificity and quantity. The 5′ RACE was carried out according to procedures described by Frohman et al. (1988). Five micrograms of total RNA was reverse-transcribed using primer 6 with M-MLV reverse transcriptase. The mRNA template was then degraded by RNase H (Takara). The reaction product was purified using a DNA extraction kit (Takara) and then appended with poly(A) using TdT (Takara) and purified again. The 5′ RACE–PCR was then carried out using the above dA-cDNA as template and primer 7 and oligo (dT) anchor primer in a procedure similar to the 3′ RACE. A nested amplification was conducted with primer 8 and the adapter primer under the same conditions as the 3′ RACE nested amplifications. Additional control reactions with only one primer were applied to identify non-specific products.

Isolation of Rice Field Eel P450arom Introns and 5′ Flanking Region

Amplification of the genomic CYP19 gene fragment was performed with the set of primers 9 and 10, designed according to the rice field eel ovarian P450arom cDNA (GenBank no. AY583785), using 200 ng of rice field eel genomic DNA as template with LA Taq (Takara). Cloning of the 5′-flanking region of rice field eel P450arom gene was performed with the Takara LA PCR in vitro cloning kit according to the manufacturer’s protocol. Briefly, aliquots of genomic DNA were digested with 80 U each of the restriction endonucleases PstI and EcoRI. Digested DNA aliquots were ligated separately to the corresponding adapter supplied with the kit. Adapter-ligated genomic DNA fragments were subjected to a primary PCR amplification with the outer adapter primer (C1 supplied with kit) and the outer gene-specific primer 11. Aliquots (1 μl) of the 50× diluted primary PCR reaction were used in the secondary PCR amplification with the nested adapter primer (C2 supplied with kit) and the nested gene-specific primer 12. Primers 11 and 12 were designed according to the rice field eel CYP19 gene (GenBank no. EF64094).

Cloning and Sequence Analysis

The fragment was purified with a Gel DNA Extraction kit (Takara) and then cloned into the pMD19-T vector (Takara) and transformed in Escherichia coli competent cells following the manufacturer’s instructions. White colonies were selected from X-Gal/IPTG ampicillin agar plates and grown in LB/ampicillin liquid medium. Plasmids were extracted by alkaline lysis. Sequencing of DNA fragments was done by Takara.

For sequence analysis, the deduced amino acid sequences of rice field eel P450arom, together with P450arom sequences reported for other vertebrate species, were aligned by Clustal W, version 1.7 (Thompson et al. 1994), using default settings (gap-opening penalty 10, gap extension penalty 0.05, gap-distance 8). The deduced amino acid sequence was also analyzed by Motif on the GenomeNet (Supercomputer Laboratory, Institute for Chemical Research, Kyoto University).

The exon–intron boundaries as well as the intron sequences were determined by comparing the genomic sequence with the cDNA sequence. The transcription start site was identified using Eponine Transcription Start Site finder (http://servlet.sanger.ac.uk:8080/eponine/). Potential binding sites for transcription factors were identified using the Transfac data matrix at the TESS website (http://www.cbil.upenn.edu/cgi-bin/tess/tess).

Fluorescent Real-time RT-PCR

In order to determine the expression pattern of P450arom, a LightCycler 1.2 (Roche, Shanghai) with SYBR green I fluorescence RT-PCR kit (Takara) was used. Primers 13 and 14 were designed according to the P450arom gene (GenBank no. AY583785). Twenty-microliter reactions were used under the conditions suggested by the SYBR real-time RT-PCR protocol: 95°C, 10 s, 20°C/s for 1 cycle; 95°C, 5 s, 20°C/s; 60°C, 5 s, 20°C/s, for 45 cycles. For the annealing temperature (T m) curve, the conditions were 95°C, 0 s, 20°C/s, 65°C, 15 s, 20°C/s, 95°C, 0 s, 0.1°C/s. In order to adjust for variations in the amount of input cDNA, the housekeeping gene β-actin was used as an internal standard. The primers (15 and 16) for amplification of β-actin were designed according to the zebrafish β-actin (GenBank no. BC067566). The amplification efficiency of P450arom and β-actin was adjusted by the standard curve of log[RNA] and C t values, which refers to the cycle number during exponential amplification at which the PCR product (measured in real time by SYBR green fluorescence) crossed a set threshold. The relative ratio of P450arom mRNA/β-actin mRNA in individual tissues could then be calculated. The ratio in female forebrain was arbitrarily set at 1.0. Student’s t-test was used to analyze the statistical difference (P < 0.05) between tissues. In order to verify the specificity of real-time RT-PCR, the T m curve was analyzed, and agarose gel electrophoresis was used to verify the product.

Results

Isolation of P450arom cDNA and Sequence Analysis

A 1,802 bp cDNA encoding P450arom (GenBank no. AY583785) was isolated by RT-PCR and RACE–PCR from the rice field eel ovary. Initially, 450 and 1,100 bp fragments were amplified from ovarian RNA and brain RNA. The DNA fragments were cloned, and three recombined clones per tissue were sequenced. Results indicated that the ovarian and brain fragments had the same sequence. The 3′ RACE also resulted in the same ∼600 bp fragment from ovary and brain. By means of 5′ RACE with ovarian RNA, a 500 bp fragment was isolated. Sequence assembly gave rise to an ovarian P450arom 1,802 bp cDNA, and the deduced amino acid sequence was based on a 1,551 bp open-reading frame (ORF) that started at an ATG codon at nt 50 and continued to a TGA stop codon (nt 1601). This ORF encoded a protein of 517 amino acid residues with a calculated molecular mass of 58.2 kDa. The 5′-untranslated region (UTR) of the P450arom gene was 49 bp, and the 3′-UTR was 202 bp in length.

Based on mutational analysis and molecular modeling, amino acids known to be essential for catalytic functions in human P450arom (I133, E302, P308, D309, T310, R435, C437) (Kao et al. 1996) were identical in the rice field eel P450arom. Consistent with the glycosylation site described at the N-terminus of human aromatase, a consensus N-glycosylation site (N-X-S/T) was identified in the amino terminal region of the rice field eel aromatase at N29, which was similar to goldfish ovarian P450arom (N30) (Gelinas et al. 1998). Further analysis indicated that the rice field eel P450arom had putative protein kinase C-dependent sites at 59, 130, 385, 410, and 503 (S/T-X-R/K).

The deduced amino acid sequence of rice field eel P450arom was compared with P450arom homologs in other representative vertebrates (including medaka ovary, goldfish brain and ovary, channel catfish brain and ovary, zebrafish brain and ovary, and human placental forms) using Clustal W (Fig. 1). The results showed that the rice field eel P450arom shared 80% sequence identity with medaka aromatase, 64 and 63% with goldfish and channel catfish ovarian aromatase, respectively, 59 and 60% with goldfish and channel catfish brain-derived aromatase, respectively, and about 50% with human aromatases. The degree of conservation was higher in the putative functional domains, including the I-helix (73% versus human), the aromatase-specific region (79% versus human), and the heme-binding region (71% versus human).

Fig. 1
figure 1

Comparison of p450arom from rice field eel (AY583785), catfish brain (AAL4612), goldfish brain (AAB39408), medaka ovary (Q92807), catfish ovary (Q92111), goldfish ovary (AAC14013), and human (NM_000103). Identical and similar amino acid residues are marked by asterisks and dots, respectively. Regions of high homology are underlined: I-helix (I), an aromatase-specific conserved region (II), and heme-binding region (III). Amino acids known to be essential for catalytic functions are in boldface type. The N-glycosylation site and putative protein kinase C phosphorylation sites are boxed

Full size image

Genomic Organization of Rice Field Eel P450arom

The exon/intron organization of the rice field eel P450arom gene was determined by the sequence isolated from DNA compared with cDNA of P450arom. The result showed that the rice field eel P450arom gene consisted of eight exons and seven introns, which were inserted at exactly the same positions as those found in human, medaka, and sea bass P450arom, but lacked an intron between the I-helix and the aromatase-specific conserved region (Fig. 2). All donor and acceptor sites of these introns were GT and AG, respectively, following the GT/AG rule as described by Breathnach and Chambon (1981). In the rice field eel, as in medaka and sea bass, the sizes of the introns were smaller than in humans (89–167, 73–213, and 89–408 bp, respectively, against 1.25–10 kb in humans) (Tanaka et al. 1995; Dalla Valle et al. 2002; Harada et al. 1990). Although the rice field eel CYP19 lacked an intron, the size of the introns, except number 7, was the longest among rice field eel, sea bass, and medaka. The total length of the introns was 918 bp in the rice field eel CYP19 gene. All introns interrupted the reading phase inside the boundary codon triplet, except for introns 6 and 7 in rice field eel and 6 and 8 in sea bass and medaka.

Fig. 2
figure 2

Genomic structure of the rice field eel CYP19 gene. Schematic diagram shows the relative positions of introns and exons. Exons are depicted as boxes, and translated exons are shaded. Introns are indicated by triangles. Numbers above the boxes indicate exon size in base pairs; numbers below the triangles indicate intron length

Full size image

Cloning and Sequence Analysis of Genomic CYP19 5′-Flanking Region in Rice Field Eel

A 1,017 bp fragment was generated from the genomic walking PCR. The sequence ligated with the CYP19 gene demonstrated the 1,065 bp upstream of the translation start site. Sequence analysis revealed that the transcription initial site was 51 bp upstream from the translation start site, which was the same as the result of the 5′ RACE. Within this region were one estrogen receptor recognition half site (ERE) (nt −62), five copies of an SRY/iSRY binding motif, and a C/EBP (CCAAT enhancer binding protein) binding site (nt −751). Unlike other ovarian CYP19 genes of fish such as goldfish and zebrafish, there was no binding sequence for steroidogenic factor-1 (SF-1) and the cAMP response element binding protein activating transcription factor family (CRE), but the gene showed sequences for the recognition of chicken ovalbumin upstream promoter-transcription factor (COUP) (nt −986) and GATA-2 (nt −186, −249), which are required for neural specificity expression. Figure 3 shows the 5′-flanking sequence of the rice field eel CYP19 gene.

Fig. 3
figure 3

Sequence of the 5′-flanking region of the rice field eel CYP19 gene. The locations of putative binding sites for specific transcription factors are labeled and underlined. Transcription site and translation start site are boxed

Full size image

Expression Pattern of P450arom mRNA

Fluorescent real-time RT-PCR was performed to study the rice field eel P450arom expression pattern in individual tissues during development. Results show that in female rice field eels (84–160 g and 33–40 cm with ovary in stage II), P450arom was expressed in the forebrain, hypothalamus, pituitary, liver, and ovary, but in males (100–200 g and 44–54 cm), it could be detected only in the pituitary and the liver (Fig. 4). The level of expression, in descending order, was female hypothalamus, female pituitary, female forebrain, ovary, female liver, male pituitary, and male liver. The expression in male fish pituitary and liver was about half that in the corresponding female parts. There were significant differences between female fish and male fish, but no significant difference was detected between female brain and ovary (Fig. 4). An analysis of the expression in fry 30 days (∼4.35 cm long) and 50 days old (∼6.80 cm long) showed that expression was higher in the brain than in the ovaries, and expression was higher in 50-days fry than in 30-days fry (Fig. 5).

Fig. 4
figure 4

Expression of P450arom in tissues of female (A) and male (B) rice field eel. Lanes F, forebrain; H, hypothalamus; P, pituitary; L, liver; O, ovary; T, testis. Expression level (C) in forebrain, hypothalamus, pituitary, liver, and gonad, expressed as mean ± SD (n = 3). Lowercase letters above the bars indicate significant differences; bars with the same letter are not significantly different (Ducan’s test, P < 0.05)

Full size image
Fig. 5
figure 5

Expression of P450arom in brain (open bar) and ovary (solid bar) during the sex formation phase of rice field eel. 1, Fry 30 days old (average length 4.35 cm); 2, Fry 50 days old (average length 6.80 cm). The results are expressed as mean ± SD (1, n = 4; 2, n = 3). Lowercase letters above the bars indicate significant differences; bars with the same letter are not significantly different (Ducan’s test, P < 0.05)

Full size image

Discussion

The CYP19 gene is present as a single copy in the haploid genome of most mammals, and tissue-specific expression of P450arom is therefore achieved by alternative splicing and/or use of a different promoter (Harada et al. 1993; Simpson et al. 1997; Golovine et al. 2003). Conversely, in teleosts, more and more evidence shows that there are two structurally and functionally different isoforms, commonly termed P450aromA and P450aromB, preferentially expressed in gonads and brain, respectively (Kwon et al. 2001; Kishida and Callard 2001; Blazquez and Piferrer 2004; Li et al. 2004). That was thought to be a consequence of an early genome duplication event in actinopterygian fish (Ohno 1970). Using RT-PCR, however, we were not able to isolate a different P450arom cDNA in the rice field eel brain. Real-time RT-PCR results show that the ovarian P450arom is also expressed in the rice field eel brain, and the expression level is higher in the brain than in the gonads of fry, which is significantly different from the other fishes that have been studied. Analysis of the 5′-flanking fragment of the rice field eel CYP19 gene shows that there are sequences for recognition of COUP and GATA-2. COUP regulation factor is found only in zebrafish brain-specific expression CYP19b, not in ovarian-specific expression CYP19a (Tong and Chung 2003). GATA-2 is found in goldfish brain-specific expression CYP19b but not in CYP19a (Tchoudakova et al. 2001). This supplies a basis for rice field eel P450arom expression in brain. Another distinguishing feature of the rice field eel CYP19 promoter region, when compared with other kinds of fish ovarian CYP19, is the absence of a recognition sequence for SF-1 and CRE. SF-1 has been reported in zebrafish (Tong and Chung 2003), Cromileptes altivelis, Gobiodon histrio, Lates caleariger (Gardner et al. 2005), goldfish (Tchoudakova et al. 2001), and medaka (Watanabe et al. 1999) ovarian CYP19 gene, as in frog (Akatsuka et al. 2005), birds (Ramachandran et al. 1999), and mammals CYP19 (Honda et al. 1994), but it is lacking in fish brain P450arom gene promoter regions. In contrast to SF-1, CRE was not found in all fish ovarian CYP19 genes. In some fishes, such as goldfish, medaka, and L. caleariger, there was no recognition sequence for CRE. Although SF-1 is the conserved regulation factor among mammals, birds, and the fish ovarian CYP19 gene, it is certain that SF-1 null mice continue to have P450arom-expressing cells in the medial preopticoamygdaloid region (Honda et al. 1994). All these facts imply that the regulation mechanism of the rice field eel CYP19 gene has its own characters. It is also possible that the rice field eel has only one aromatase gene, as in humans, because the rice field eel has a haploid genome size (0.6–0.8 pg) that is among the smallest of the vertebrates (Zhou et al. 2003), with only 24 pairs of chromosomes. Consistent with the small genome size, rice field eel CYP19 has only seven introns, lacking an intron compared with other kinds of fish and human P450arom, which all have eight introns. In response to small genome size, there is another strategy to reduce the size of introns, like medaka, which also has a small genome. The total length of the eight introns of the medaka CYP19 gene is 842 bp (Tanaka et al. 1995), and the seven introns of the rice field eel CYP19 gene are 918 bp long. The sea bass, however, has eight introns with 1,346 bp (Dalla Valle et al. 2002), and the human has eight introns with more than 30 bp (Harada et al. 1990).

Another notable character of P450arom expression in rice field eel is that the expression level in liver is higher than in other fishes that have been studied. One may speculate that since it is known that estrogen receptors are expressed highly in the liver (Socorro et al. 2000; Choi and Habibi 2003), and there is an ERE half site at the regulation region of the rice field eel, it could be inferred that the estrogen receptor can regulate liver P450arom expression, which converts androgen to estrogen. Estrogen binds to estrogen receptors to stimulate vitellogenin formation and regulate the development of germ cells.

The expression levels in male rice field eel tissues are much lower than expression levels in female fish. During sex formation, P450arom expression is higher in the brain than in the ovary, and in older fry than in younger ones. These results indicate that the P450arom expression level is related to sexual differentiation of the rice field eel.

In conclusion, the CYP19 gene in rice field eel has a different gene structure and expression pattern compared with other teleosts studied. This could explain the conservation and diversification of the CYP19 gene.