Cloning of two mRNA variants of brain aromatase cytochrome P450 in rainbow trout (Oncorhynchus mykiss Walbaum)

Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

Cloning of two mRNA variants of brain aromatase cytochrome P450 in rainbow trout (Oncorhynchus mykiss Walbaum) Luisa Dalla Valle∗ , Annalisa Ramina, Silvia Vianello, Paola Belvedere, Lorenzo Colombo Comparative Endocrinology Laboratory, Department of Biology, University of Padova, Via Uu Bassi 58/B, 35131 Padova, Italy Received 27 March 2002; accepted 21 June 2002

Abstract This work describes the molecular cloning of the cDNA encoding the rainbow trout (Oncorhynchus mykiss Walbaum) brain cytochrome P450arom by means of reverse transcriptase and polymerase chain reaction (RT-PCR) and 5 - and 3 -rapid amplification of cDNA ends (RACE) analyses. The results obtained demonstrate that, as in other teleost fishes, the trout genome contains, besides the gene previously identified in the ovary, a second CYP19 gene (CYP19B) expressed at high level in the brain. Moreover, two P450aromB mRNAs, forms I and II, were found to be transcribed in trout. Form I (1816 sequenced nt) contains an open reading frame (ORF) of 1464 b, a 5 -untranslated terminal region (UTR) of 124 b and at least 228 b in the 3 -UTR (incomplete, as the polyadenylation signal was not determined). Form II (1930 sequenced nt) contains an ORF of 1362 b, a 5 -UTR of 340 b and the same 3 -UTR as form I. Form II lacks the first 34 amino acids of form I, corresponding to the membrane-anchoring segment, whereas the sequence of the remaining coding region is almost the same in the two forms, resulting in proteins of 454 and 488 amino acids, respectively. Whether the two transcripts derive from the same gene by alternative splicing or are encoded by different CYP19B genes remains to be clarified. On Northern blot analyses with brain and ovary specific ORF probes and poly(A)+ -enriched RNAs from trout ovary and brain, a transcript of about 2.6 kb was identified in the ovary, as expected, whereas the full-length mRNA of brain P450arom is about 3.8 kb. The brain form is expressed in the brain and gonads, whereas expression in peripheral tissues is limited mostly to the gills. The two trout CYP19 genes are not equivalent in tissue-specific expression, indicating the possibility of distinct promoters and regulatory mechanisms. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: Brain aromatase; Cytochrome P450; Oncorhynchus mykiss

1. Introduction The catalytically active aromatase complex, responsible for the synthesis of estrogens from androgens, consists of two components: the aromatase cytochrome P450 (P450arom) protein which binds the substrate, and a ubiquitous flavoprotein, nicotinamide adenine dinucleotide phosphate (NADPH)-cytochrome P450 reductase, involved in the electrons flow required for substrate oxidation. Aromatase is expressed in a wide variety of tissues, depending on the species considered. In the human, for example, the enzyme has been found in ovary, testis, placenta, brain, liver, adipose tissue, skin, as well as in a number of tissues in the fetus [1].

∗ Corresponding author. Tel.: +39-049-8276188; fax: +39-049-8276199. E-mail address: [email protected] (L. Dalla Valle).

This cytochrome is expressed in the brain and ovary of all vertebrates studied to date [2]. Among vertebrates, teleost fish, including the rainbow trout, Oncorhynchus mykiss, are unique in having exceptionally high levels of aromatase activity in the brain, higher than in the ovary [3,4]. The production of estrogens in the brain has been correlated with many physiological and behavioural processes, including brain sexual differentiation of reproductive control centres during the development of the central nervous system [5]. Particularly, the control that locally synthesised estrogens have on neural organisation is thought to establish the different gonadotropin secretion profiles between male and female and the negative and positive feedback effects of steroid hormones on GnRH and gonadotropin secretion [6]. The neural in situ transformation of androgens to estrogens is also implied in the central activation and regulation of sexual behaviour [5]. However, estrogenic effects on the nervous system extend beyond their actions in hormonal regulation of the reproductive functions, as indicated by the detection

0960-0760/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 0 - 0 7 6 0 ( 0 2 ) 0 0 1 4 3 - 7

20

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

of aromatase protein and messenger in brain regions outside classical reproductive control centres [7]. Specifically, brain-formed estrogens affect neuronal differentiation, survival and maintenance of function [8], and improve cognitive function in both man [9] and mouse [10]. According to Callard et al. [11], the potential for high neuroestrogen biosynthesis could be related to the neuroplasticity and the neuronal regeneration capability of adult teleost fish [12]. Aromatase is encoded, in humans, by a single CYP19 gene and the control of tissue-specific expression is obtained, in this and other mammalian species studied, by the use of alternatively spliced 5 -untranslated exons associated with tissue-specific promoters. As exon I is not translated, the same protein is finally expressed in all tissues [1]. However, new evidence is currently challenging the paradigm of a single aromatase gene with tissue-specific regulation. Indeed, in pig, three different genes encode tissue-specific protein isoforms, a condition that is further complicated by the use of multiple promoters [13,14]. An aromatase pseudogene has been found to be transcribed in bovine placenta but, in this case, the transcript encodes for a nonfunctional protein [15]. In teleost fish, such as Carassius auratus [16], Danio rerio [17,18], Oreochromis niloticus [19], and Oreochromis mossambicus [20], a distinct enzyme, encoded by a novel CYP19 gene, was found to be expressed in the brain. The brain aromatase forms share higher identity between different species than with their respective ovarian counterparts. In this work, we provide evidence that the rainbow trout genome contains a second CYP19 gene, named CYP19B, besides that previously found to be expressed in the ovary [21], that two P450aromB mRNAs with different 5 -untranslated terminal region (UTR) are transcribed in trout, and that both

brain and ovary aromatase forms can be expressed in the brain, ovary and gills.

2. Materials and methods 2.1. Tissue preparation and RNA extraction Samples of brain, gills, liver and ovary were collected, during the winter months, from farm-reared 1-year-old female rainbow trouts. At the time of tissue collection, the fish were sedated in an ice bath and sacrificed by decapitation. Tissues were immediately removed by dissection, frozen in liquid nitrogen and stored at −80 ◦ C until analysed. Total RNA was extracted using the commercial product RNAzol B (Celbio, Milan, Italy) according to the manufacturer’s instructions. The RNA samples were kept at −80 ◦ C until use. The RNA was enriched in polyadenylated mRNA utilising the NucleoTrap mRNA Kit (M-Medical, Florence, Italy). 2.2. Oligonucleotides The oligonucleotides used as PCR primers are listed in Table 1. The non-specific primers (ns) were designed on highly conserved regions of P450arom cDNAs from Mozambique tilapia (brain form: AF135850; ovary form: AF135851) and goldfish (brain form: U18974; ovary form: AF020704) as well as from trout ovary [21]. Gene-specific primers (s) were then suited to our new sequence. Primers BT1 and BT2 were selected on the trout ␤-actin cDNA sequence (accession number AF157514), whereas primers 18S-1 and 18S-2 were selected on the trout 18S rRNA sequence (accession number AF308735).

Table 1 Primers used in RT-PCR and 5 - and 3 -RACE analyses for sequencing and expression analysis of rainbow trout cytochrome P450aromB Primer

Sequence

Nucleotide position

Reference sequence

ns-Rt-1 ns-Rt-2 ns-Rt-3 ns-Rt-4 s-Rt-1 s-Rt-2 s-Rt-3 s-Rt-4 s-Rt-5 s-Rt-6 s-Rt-7 AROM-1 AROM-2 AROM-3 BT-1 BT-2 18S-1 18S-2

5 -CCAGGTCCTGCAGAGCTT-3

+555 +368 +1428 +1172 +404 +1311 +1341 +43 +201 +1607 +1376 +1022 +1533 +1582 +396 +873 +396 +1014

P450aromB-I P450aromB-I P450aromB-I P450aromB-I P450aromB-I P450aromB-I P450aromB-I P450aromB-I P450aromB-II P450aromB-I P450aromB-I P450aromA P450aromA P450aromA ␤-Actin ␤-Actin 18S rRNA 18S rRNA

5 -TCTCCTCTCCGTTGATCCA-3 5 -ATCATCACCATGGCAATGTG-3 5 -GAGTCCTTAAGGTTCCATCCT-3 5 -GTCCTGAAGAGTGCCCAC-3 5 -CCAAACCCAACGAATTCAGCT-3 5 -TTGATAAAACTGTGCCCAG-3 5 -GCTAGCGGCAGACTACTCTGG-3 5 -TATTACTGTAAAACGGTGGCAA-3 5 -TCAGTAGTGGTTGTGGTTAGGT-3 5 -ACGGCTGGAAGAAACGACTGG-3 5 -CGACCAGAAGAGAAGGGGTCTACA-3 5 -CGAAAGGCTGGAAGAAACGATTAG-3 5 -ATCATCACCATGGCAATGTG-3 5 -CAGGGAGAAGATGACCCAGAT-3 5 -GATACCGCAAGACTCCATACC-3 5 -GATTCCGGAGAGGGAGCCTG-3 5 -CCTCCGACTTTCGTTCTTG-3

→ → → → → → → → → → → → → → → → → →

+537 +350 +1409 +1192 +421 +1331 +1359 +63 +222 +1586 +1356 +1356 +1510 +1563 +416 +853 +415 +996

Positions are relative to the trout P450aromB-I cDNA sequence (AJ311937), P450aromB-II (AJ311938), P450aromA [21], trout ␤-actin (AF157514), trout 18S rRNA gene (AF308735).

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

2.3. Cloning of rainbow trout brain P450aromB cDNA The rainbow trout brain P450aromB was cloned and sequenced in three steps: 5 -rapid amplification of cDNA ends (RACE) analysis, reverse transcriptase and polymerase chain reaction (RT-PCR) analysis of the coding region, and finally 3 -RACE analysis. The 5 -RACE was carried out using the 5 -RACE System (Invitrogen, Milan, Italy) following the manufacturer’s instructions. Briefly, the cDNA was synthesised by incubating 2 ␮g of total RNA of trout brain and ovary in 25 ␮l of the first-strand buffer, which was supplemented with 200 U of SuperScript II RT, 0.4 mM of dNTPs, 10 mM DTT and 0.16 ␮M of the primer ns-Rt-1, at 50 ◦ C for 1 h. Terminal transferase was used to add a homopolymeric C-tail to the 3 -end of the first-strand cDNA purified by the GlassMAX DNA isolation Spin Cartridge Purification Kit (Invitrogen). The tailed cDNA was then amplified by PCR using the primer ns-Rt-2 and the oligo dG-anchor primer (5 -GGC CAC GCG TCG ACT AGT ACG GGI IGG GII GGG IIG-3 ). The amplification procedure consisted of 2 min at 95 ◦ C followed by 40 cycles at (45 s): 95, 55, and 72 ◦ C. The extension phase of the last cycle was prolonged by 10 min. The resultant amplicons were purified from the sliced gel bands and directly sequenced. The cloning of part of the coding region was performed with a reaction of RT-PCR using a specific 5 -primer selected on the previously obtained cDNAs fragments (s-Rt-1) and a non-specific 3 -primer selected on highly conserved regions of fish P450arom cDNAs (ns-Rt-3). For this analysis we utilised 2 ␮g of total RNA of trout brain and the SuperScript One-Step RT-PCR System (Invitrogen). For the 3 -RACE, the cDNA was synthesised by incubating 2 ␮g of total RNA of trout brain in 20 ␮l of the first-strand buffer which was supplemented with 200 U of SuperScript II RT (Invitrogen), 0.5 mM of dNTPs, 10 mM DTT and 0.5 ␮M dT17 primer (5 -GAC TCG AGT CGA CAT CGA TTT TTT TTT TTT TTT TT-3 ) at 50 ◦ C for 1 h. The first-strand mixture was diluted to 200 ␮l with water and 2 ␮l were added to 50 ␮l of the PCR buffer containing 200 ␮M of dNTPs, 0.2 ␮M of the anchor primer (5 -CTG GTT CGG CCC AGA CTC GAG TCG ACA TCG-3 ), 2.5 U of Taq polymerase (Celbio), and the ns-Rt-4 primer. The cDNA obtained was further amplified by a second PCR using the s-Rt-2 primer and the PCR anchor primer. Finally a third PCR was performed with the s-Rt-3 primer and the PCR anchor primer. The amplification procedure consisted of 2 min at 95 ◦ C followed by 10 cycles at 95 ◦ C (45 s), 58 ◦ C (45 s), and 72 ◦ C (2 min) and 25 cycles at 95 ◦ C (45 s), 58 ◦ C (45 s), and 72 ◦ C (2 min) with 2 s of time increment per cycle in each extension phase. The extension phase of the last cycle was prolonged by 10 min. The fragment amplified by 3 -RACE was purified and sequenced. To define more precisely the nucleotide and amino acid sequences of trout P450aromB (forms I and II, see Section 3), cDNAs were amplified with primers flanking the open

21

reading frame (ORF) of form I (s-Rt-4 and s-Rt-6, 1565 bp) and form II (s-Rt-5 and s-Rt-6, 1521 bp). PCR products were cloned, as described in the paragraph on cRNA probe synthesis, and two independent clones of each form were sequenced. 2.4. Phylogenetic analysis Phylogenetic relationships of aromatase genes were derived by aligning deduced amino acidic sequences of trout brain- and ovary-derived P450arom forms together with P450arom sequences reported for other vertebrate species using the Clustal W program [22]. All alignment positions were included in the analysis. The tree was generated by the neighbour-joining (NJ) method of Saitou and Nei [23], as implemented in the TREECON program, version 1.3 b [24]. Insertions and deletions were not taken in account. The rainbow trout cytochrome P450c17 sequence served as an out group to root the tree. Bootstrap re-samplings [25] were also performed to test the robustness of the trees and 1000 replicates were done. Program setting is detailed in the figure legend. 2.5. cRNA probe synthesis We prepared two cRNA probes corresponding to a fragment of the coding region of P450aromA (nucleotides +1022/+1533) and of the coding region of P450aromB (nucleotides +404/+1376). The corresponding cDNAs were amplified by RT-PCR from total ovarian RNA (coding region of P450aromA, primers AROM-1 and AROM-2) and total brain RNA (coding region of P450aromB, primers s-Rt-1 and s-Rt-7) using specific primers (Table 1). The identity between each cRNA probe and the target sequence in the nonhomologous variant was always 74%. Following PCR, the amplified DNAs were resolved on a 1.2% agarose gel and the gel-purified fragments were ligated into a pGEM-T vector using a pGEM-T Vector System I according to the supplier’s recommendations (Promega, Milan, Italy). The inserts were sequenced to verify probes specificity and orientation. A recombinant plasmid for each probe was then linearized by restriction cleavage and used as a template for generating the cRNA probe. The cRNA transcripts were digoxigenin-labelled by in vitro transcription using a DIG RNA Labelling kit (Roche Diagnostics, Milan, Italy). 2.6. Northern blot analysis Poly(A)+ -enriched RNAs were obtained from female trout brain, ovary, gills, and liver, as described above. Samples in duplicate were electrophoresed on the same gel, blotted on the same filter which was cut in half for hybridisation. This was done to avoid stripping of the membrane in order to rehybridize with different aromatase probes (except for the final hybridisation with mouse ␤-actin probe).

22

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

The poly(A)+ -enriched RNAs (between 1 and 2.5 ␮g) were electrophoresed with the high range RNA ladder (MBI Fermentas, DASIT, Milan, Italy) through a 1.1% formaldehyde-denaturing gel, blotted onto a nylon membrane (Hybond-N+ , Amersham Pharmacia, Milan, Italy) and baked at 80 ◦ C for 2 h. The remaining 28S and 18S rRNAs in each poly(A)+ -enriched RNA preparation were visualised by methylene blue staining to check RNA loading and integrity. The membranes were incubated overnight at 68 ◦ C with the P450aromA and P450aromB DIG-labelled cRNA probes in 5× SSC containing 50% formamide, 0.02% sodium dodecyl sulphate (SDS), 0.1% lauroylsarcosine, 1% blocking reagent and 100 ␮g/ml of transfer RNA. Membranes were washed in 2× SSC and 0.1% SDS twice for 15 min at room temperature, and in 0.2× SSC and 0.1% SDS three times for 15 min at 68 ◦ C. They were then incubated with the Anti-DIG-AP (Roche) and the signal detected using the CPD-Star DIG Luminescent Detection Kit (Roche), according to the manufacturer’s protocol. Bands were visualised by autoradiography. Following hybridisation and detection of P450arom mRNA transcripts, the membrane blots were treated by boiling in 0.1% SDS solution to strip the probes. The membranes were then rehybridised with antisense mouse ␤-actin DIG-labelled cRNA probe. 2.7. Semi-quantitative RT-PCR assay Semi-quantitative RT-PCR assays were performed for P450aromA, P450aromB, ␤-actin messengers and 18S rRNA with three samples of poly(A)+ -enriched RNA extracted from three samples of brain, ovary and gills and one sample of liver from 1-year-old female trouts. For these analyses, we adopted the SuperScript One-Step RT-PCR System (Invitrogen), in which all components for RT-PCR are mixed in one tube and reverse transcription is automatically followed by PCR cycling without any additional steps. The concentration of MgSO4 was adjusted to 1.5 mM and primers were added to a final concentration of 0.2 ␮M. Parameters of the reactions are listed in Table 2. For each RT-PCR, a negative control was prepared by using all reagents, except RNA solution that was substituted with an equivalent volume of sterile water to check for cross-contamination. In preliminary experiments, each primer set was used to amplify equal amounts of cDNA derived from the samples Table 2 Target gene, PCR primers and conditions used in the semi-quantitative RT-PCR analyses Target gene Primers

Annealing (◦ C)

Extension time (s)

PCR cycle no.

P450aromA P450aromB ␤-Actin 18S RNA

58 58 60 58

35 60 28 35

32 25 18 32

AROM-1; AROM-3 s-RT-1; s-Rt-7 BT-1; BT-2 18S-1; 18S-2

with the highest levels of specific messenger for 10–40 cycles (50 ng of poly(A)+ -enriched RNA from ovary or brain for P450aromA and P450aromB, respectively), and, based on these analyses, a predetermined number of cycles were chosen for each primer set to maintain product accumulation in the linear range. The subsequent analyses were performed with different quantities (ranging from 50 ng to 1 ␮g) of poly(A)+ -enriched RNA. PCR products were resolved on 1% agarose gel and stained with ethidium bromide. Quantification was performed by measuring the relative intensity of the band stained by ethidium bromide, after agarose gel electrophoresis, or after Southern blot analysis (see next paragraph) using the Quantity One Quantitation Software (Bio-Rad, Milan, Italy). 2.8. Southern hybridisation analysis An aliquot (1/10) of the RT-PCR products obtained from each set of primers for aromatase analyses was electrophoresed in 1.2% agarose gel. The DNA was then transferred onto a nylon membrane (Amersham Pharmacia) and baked at 80 ◦ C for 2 h. The membrane was then incubated in a hybridisation buffer containing 5× SSC, 0.1% N-lauroylsarcosine, 0.02% SDS, 2% blocking reagent in 50% formamide for 2 h at 42 ◦ C. It was then hybridised overnight at 42 ◦ C in fresh hybridisation buffer containing digoxigenin-labelled cDNA probes synthesised by PCR (Roche). After hybridisation, the membrane was rinsed twice in 2× SSC and 0.5% SDS for 5 min at room temperature and twice in 0.1× SSC and 0.1% SDS at 68 ◦ C for 15 min. They were then incubated with the Anti-DIG-AP (Roche) and the signal detected using the CPD-Star DIG Luminescent Detection Kit (Roche), according to the manufacturer’s protocol. Bands were visualised by autoradiography. 2.9. Nucleotide sequencing Sequencing was performed on double-stranded DNA using the ABI PRISM Dye Terminator Cycle Sequencing Core Kit (Applied Biosystems, Monza, Italy). Electrophoresis of sequencing reactions was completed on the ABI PRISM model 377, version 2.1.1 automated sequencer. The homology searches were carried out using the Basic Blast program version 2.0 at http://www.ncbi.nlm.nih.gov/BLAST/, whereas the alignment was performed using the ClustalW program at http://www2.ebi.ac.uk/clustalw/. Transmembrane prediction analysis was carried out with the ExPASy Molecular Biology Server (http://www.expasy. ch/) using the Prediction of Transmembrane Regions SOSUI Software. The protein was also scanned for the occurrence of patterns stored in the PROSITE database (http://www.expasy.ch/prosite/). The molecular weights were obtained using the ProtParam tool program (http://www.expasy.ch/).

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

3. Results 3.1. Isolation of rainbow trout brain P450aromB cDNA Rainbow trout P450aromB (in a previous paper named P450arom type II, 26) was cloned using a combination of RT-PCR and 5 - and 3 -RACE analyses. Initially, we performed a 5 -RACE analysis with primers selected on highly conserved regions of P450arom cDNAs of Mozambique tilapia and goldfish brain and ovary as well as trout ovary, using total RNA extracted both from brain and ovary of 1-year-old female trouts. With ovarian RNA, we obtained a single amplicon corresponding to the 5 -UTR and part of the coding region of the previously sequenced ovarian trout P450arom (that we shall call P450aromA), except for 15 nucleotides at the 5 -end which are not present in the sequence published by Tanaka et al. [21]. With brain RNA, we obtained two amplicons of different length. Although, with different 5 -UTR, a database search with the Blast program revealed a high similarity of these PCR products (P450aromB forms I and II) to other teleost fish P450arom cDNAs. Particularly, the second form (II) lacks the first coding exon, whereas the sequence corresponding to the second coding exon is identical to that of form A. In the second step of this work, we utilised a sequencespecific primer together with a primer selected on conserved regions to amplify almost completely, by RT-PCR, the coding region of P450aromB. By means of 3 -RACE, the remaining sequence information on the coding region was obtained. As regards 3 -UTR, the sequence is incomplete, as the polyadenylation signal was not determined. To define more precisely the nucleotide and amino acid sequences of trout P450aromB forms I and II, the corresponding cDNAs were amplified with primers flanking the ORF of each cDNA: two different primers targeted to the 5 -ends of P450aromB forms I and II, and a common 3 -primer selected on the 3 -end. PCR products were cloned and two independent clones of both forms of P450aromB were fully sequenced. The two sequences contain, in the coding region, 10 differences at the nucleotidic level, leading to conservative changes of three amino acids (Fig. 1). Whether these differences are due to the presence of two CYP19B genes remains to be verified. The deduced amino acid sequence derived from the trout P450aromB-I cDNA is based on an ORF of 1464 bp, which starts from a putative initiation methionine which is 125 bp downstream from the 5 -end and continues to a stop codon TAA (1591 nt). The ATG has a nucleotide context that corresponds to the proposed consensus sequence for the initiation of translation [27]. The length of 5 -UTR of the P450aromB form I is longer than the 5 -UTR reported sequences for teleost ovarian P450arom, but comparable with those for the brain form of aromatase isolated in teleost fish, such as goldfish (109 nt) [16] or zebrafish (74 nt) [17]. The 3 -UTR is 228 bp in length and does not contain the polyadenylation signal. The sequence terminates with a stretch of A that

23

probably determined the partial result of 3 -RACE analysis. The ORF encodes a putative protein of 488 amino acid with a calculated molecular weight of 55.6 kDa. The amino acid sequence corresponds to a membrane protein with one transmembrane helix; the region that can be described as hydrophobic, the membrane-spanning region, lies between amino acids 10 and 32. The amino acid sequence corresponding to P450aromB-II derives from an ORF of 1362 bp, which starts at the first ATG in frame (341 bp) and continues to a stop codon TAA (1702 bp). The ATG has a nucleotide context that matches only partially the proposed consensus sequence for the initiation of translation [27]. In fact, it lacks the purinic base in position −3. The ORF encodes a protein of 454 amino acids with a calculated molecular weight of 52 kDa. The putative protein lacks the transmembrane helix and consequently could be described as a soluble protein. Fig. 1 shows the nucleotide and deduced amino acid sequences of the trout brain-derived P450aromB forms I and II. There is an overall 63% amino acid sequence identity between trout P450aromA and P450aromB form I, suggesting that they are only distantly related. The P450aromB-I shares 66–76% overall sequence identity with other fish P450arom brain-derived forms, (the highest identity was found with Nile and Mozambique tilapias), whereas a lower identity was found with other fish ovarian-derived aromatases (57–63%). The identity is only 53% with the aromatase sequence of the cartilaginous fish, Dasyatis sabina, while it ranges between 51 and 55% with the aromatases from other vertebrates (Table 3). The degree of conservation is higher, particularly with the brain-derived P450arom forms (75–100%), in the putative functional domains, including the I-helix region, an aromatase-specific conserved region, and the heme-binding region. The lowest identity was found with the termini of the different P450arom forms indicating that these regions are probably not relevant to the enzymatic functions (the alignment of P450arom from different vertebrate species is not shown). The identity of the sequence presented in this work is supported by a good overall homology (up to 76%) with the already known brain-derived fish sequences. 3.2. Phylogenetic analysis A phylogenetic analysis was performed to study the evolutionary relationships of the CYP19 genes using the amino acid sequence of trout cytochrome P450aromB-I and other published full-length P450arom sequences. The rainbow trout P450c17 was used as an out group to root the tree. The resulting tree (Fig. 2) shows that tetrapods, cartilaginous and teleostean fishes share a common ancestor for the CYP19 gene family. Moreover, the tree suggests that, during the evolution of the teleostean CYP19, a duplication arose within this clade giving birth to the paralogous forms of aromatase expressed in the brain and ovary. This is supported by the fact that the single orthologous CYP19 of the stingray D. sabina is strongly linked to the tetrapod forms, while

24

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

Fig. 1. Nucleotide sequence of rainbow trout P450aromB forms I and II and their deduced amino acid sequences. One-letter symbols of encoded amino acids are shown below the DNA sequence. The numbers refer to the nucleotide positions of cDNA at the end of each line. The in-frame translation start codons as well as the stop codon are given in bold type. The cDNA sequences of P450aromB-I and II have been submitted to the Genebank under the accession numbers AJ311937 and AJ311938, respectively.

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

25

Fig. 1. (Continued).

the two new paralogous teleostean genes form a very well supported independent group (bootstrap value: 980/1000). Within the teleostean CYP19, the ovarian and brain clusters are sustained by high bootstrap values (930/1000; 940/1000). 3.3. Northern blot analysis of trout brain and ovary P450arom The size and number of P450arom transcripts expressed in brain, ovary, gills and liver (as negative control) were ex-

amined by Northern blot analysis with cRNA probes specific for P450aromA and P450aromB. The Northern analyses performed with brain and ovary mRNAs are shown in Fig. 3. In each case, more RNA was loaded with the ovarian samples than with the brain samples, as shown by both ␤-actin hybridisation and methylene blue staining of 28S and 18S rRNAs. With the probe for P450aromA, a transcript of about 2.6 kb was detected with all the ovarian poly(A)+ -enriched RNAs utilised, whereas no signal of any size was detectable

26

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

Table 3 Percent of overall and partial identities of the predicted amino acid sequence of rainbow trout P450aromB-I compared to the forms isolated from other species

Nile tilapia brain Mozambique tilapia brain Goldfish brain Zebrafish brain Rainbow trout ovary Nile tilapia ovary Mozambique tilapia ovary Goldfish ovary Zebrafish ovary Japanese flounder ovary European sea bass ovary Medaka ovary Catfish ovary Stringay ovary African clawed frog ovary Chicken ovary Rat ovary Bovine ovary Equine testis Human placenta

Overall identity

I-helix region aa 282–312

Aromatase-specific conserved region aa 367–390

Heme-binding region aa 418–431

75.8 75.6 66.0 67.0 63.0 57.1 58.4 60.5 60.2 61.2 60.2 60.6 59.8 52.7 54.8 54.8 51.2 53.1 52.9 53.9

93.6 96.8 96.8 96.8 90.3 83.9 83.9 87.1 90.3 83.9 87.1 90.3 87.1 58.1 67.7 64.5 54.8 61.3 64.5 64.5

75.0 79.2 75.0 75.0 66.7 66.7 66.7 70.8 70.8 70.8 70.8 70.8 66.7 70.8 75.0 79.2 83.3 79.2 83.3 83.3

100.0 100.0 92.9 92.9 92.9 71.4 78.6 85.7 92.9 92.9 85.7 78.6 92.9 78.6 71.4 78.6 78.6 71.4 78.6 71.4

The partial identity was calculated inside three regions of high homology (as reported by Simpson et al. [53]); the amino acid interval referring to the trout sequence is reported on the first line of the table. The sequence accession numbers are reported in Fig. 2.

in the brain. The transcript length corresponds to that reported by Tanaka et al. [21] for the cytochrome P450aromA mRNA. With the probe for P450aromB, a major transcript, of about 3.8 kb, was revealed in the brain. An additional minor band of about 4 kb was also evident. With ovarian poly(A)+ -enriched RNAs, the brain probe detected two transcripts: one of about 2.6 kb and a lighter signal corresponding to the length of the brain transcript (3.8 kb). The differences in the signal intensity between samples are probably due to the different sampling time (sample 1 at the beginning of the reproductive season and sample 2 in the middle), but the number of samples is too low for any hypothesis to be framed. No signal was detected in gills and liver (results not shown). 3.4. P450arom forms A and B expression in different tissues Tissue-specific expression of P450aromB mRNA has previously been investigated, by means of a non quantitative RT-PCR protocol, in 10 different tissues (brain, gills, skeletal muscle, stomach, intestine, spleen, intermediate and posterior kidney, liver, and gonads) of male, female and XXX-triploid trouts at different ages in a study performed in parallel with the preparation of this paper [26]. The results showed that P450aromB mRNA displays a pattern of expression restricted to the brain, gonads and gills, which were found positive (although at different expression levels) in all animals studied. To understand the different expression of brain- and ovarian-derived P450arom forms in the tissues previously

found positive, poly(A)+ -enriched RNA was extracted from three samples of brain, ovary and gills and one sample of liver (as negative control) of 1-year-old female trouts and subjected to semi-quantitative RT-PCR with specific primers, followed by Southern hybridisation of the cDNAs obtained. The amplifications of ␤-actin and 18S rRNA transcripts were used as controls of gene expression and RNA loading and indicated that the quality of the RNA preparation for each tissue was similar. As expected, variations in the transcript expression between the different tissues used were visible in the ␤-actin amplification. For this reason, we used the 18S rRNA amplification, that shows minor differences, as internal standard for normalisation, as recommended by Thellin et al. [28]. P450aromA expression was clearly highest in the ovary, whereas P450aromB was highest in the brain. Nevertheless, amplifications of cytochrome P450aromA was detected on agarose gel using 50 ng of poly(A)+ -enriched RNA not only from ovary but also from all samples of brain analysed (Fig. 4). The transcript level in the brain corresponds to 14.1% ± 2.7 of specific transcript expression found in the ovary. As regards gills, a positive signal was obtained, after Southern blotting, using 250 ng of RNA in the RT-PCR analysis, resulting in 0.21% ± 0.1 of ovarian transcript expression. No positive signal was obtained with liver. Amplifications of cytochrome P450aromB were visible on agarose gel with 50 ng of brain poly(A)+ -enriched RNA and only with 1 ␮g of ovarian poly(A)+ -enriched RNA, resulting in 0.59% ± 0.3 of brain transcript (Fig. 5). With RNA from gills, the positive signal was detectable only after Southern blotting (0.14% ± 0.07 of brain transcript). A very

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

27

Fig. 2. Evolutionary relationship of the known cytochromes P450arom. The phylogenetic tree was calculated using the neighbor-joining (NJ) method with TREECON according to the following settings: (1) distance calculation: Poisson correction; (2) insertions and deletions not taken into account; (3) alignment positions: all. The distance scale is expressed in percentage. The tree is rooted with respect to rainbow trout cytochrome P450c17 (accession number CAA46675). Comparisons were made to the amino acid sequences of trout brain (AJ311937) and ovary (228574); zebrafish brain (AAK00642) and ovary (AAK00643); Nile tilapia brain (AAG18458) and ovary (P70091); Mozambique tilapia brain (AAD31030) and ovary (AAD31031); goldfish brain (P79690) and ovary (O73686); medaka ovary (Q92087); Japanese flounder ovary (BAA74777); European sea bass ovary (CAC43178); stingray ovary (AAF04617); African clawed frog ovary (BAA90529); chicken ovary (P19098); rat ovary (P22443); equine testis (O46512); bovine ovary (P46194); human placenta (P11511). Numbers indicate the values supporting the branching pattern from 1000 bootstraps. The marker of 0.1 is the length that corresponds to a 10% sequence difference. Ov: ovary; Br: brain; Te: testis; Pl: placenta.

weak signal was detected also with liver poly(A)+ -enriched RNA. As for Northern hybridisation, the differences in the amplification levels among the same types of tissues could be explained by the different sampling periods: the samples were taken in three different months during the winter.

4. Discussion Unlike human and most mammalian species, in which the gene encoding aromatase is present as a single-copy gene, there is increasing evidence that, in teleost fish, at least two CYP19 loci encode distinct P450arom isozymes that are differentially expressed in brain and ovary. In this work, the presence of a second CYP19 gene, encoding a P450aromB mRNA expressed preferentially in the

brain, was determined in the rainbow trout. Moreover, two distinct P450aromB mRNAs were identified from trout brain total RNA and named forms I and II. These two mRNAs are presumably transcribed by means of alternative splicing from this novel CYP19 gene. However, the finding of a number of nucleotide differences between these two cDNAs in the overlapping region suggests the possibility that in trout, as found in goldfish, there are more than two forms of CYP19 [29], in agreement with their known recent tetraploidization event [30,31]. To determine whether one or two CYP19B genes exist in the trout genome, a more specific study on trout genomic organisation should be carried out. Form I (1816 sequenced nt) contains an ORF of 1464 b, a 5 -UTR of 124 b and at least 228 b in the 3 -UTR (incomplete as the polyadenylation signal was not determined and the cDNA length does not correspond to the mRNA length

28

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

Fig. 3. A representative Northern blot analyses of the mRNA for P450aromA and P450aromB performed with between 1 and 2.5 ␮g of poly(A)+ -enriched RNAs from samples of ovary and brain extracted from 1-year-old rainbow trout females using trout P450aromA and P450aromB digoxigenin-labelled cRNA probes (panel A). Arrows and numbers indicate the positions and sizes of the hybridising transcripts in ovary and brain, as calculated with the RNA standards. In the panel B, the same blots were rehybridised with a mouse ␤-actin digoxigenin-labelled cRNA probe. In the panel C, methylene blue-stained 28S and 18S ribosomal RNAs of each mRNA sample are shown. Ov: ovary and Br: brain.

Fig. 4. Representative expression analysis of P450aromA mRNA in brain, ovary, gills and liver samples of female 1-year-old trouts as determined by semi-quantitative RT-PCR (upper panel) and Southern hybridisation (lower panel). Below the figure are indicated the quantities of RNA utilised in each sample. Br: brain; Ov: ovary; Gi: gills; Li: liver; MW: molecular weight; C−: negative control, (water).

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

29

Fig. 5. Representative expression analysis of P450aromB mRNA in brain, ovary, gills and liver samples of female 1-year-old trouts as determined by semi-quantitative RT-PCR (upper panel) and Southern hybridisation (lower panel). Below the figure are indicated the quantities of RNA utilised in each sample. Br: brain; Ov: ovary; Gi: gills; Li: liver; MW: molecular weight; C−: negative control, (water).

determined by Northern hybridisation). Form II (1930 sequenced nt) contains an ORF of 1362 b, a 5 -UTR of 340 b and the same 3 -UTR as form I. In agreement with other teleost brain aromatases, the N-terminal region of brain trout aromatase form I is shorter than that of teleost ovarian P450arom cDNAs and more similar in length to that of mammal, chicken, frog and stingray ovarian aromatases. Form II lacks the 34 amino acids corresponding to the first coding exon of trout brain aromatase form I. Like mammalian P450arom [32], and differently from medaka [33] and European sea bass [34] P450arom ovarian forms, a preliminary analysis of the genomic organisation of this novel CYP19B gene has found the presence of an untranslated exon I. The first coding exon encodes for the sub-domains B–D, as named by Chen and Zhou [35]. As determined by hydrophobicity analysis, the sub-domain C corresponds to the NH2 -terminal transmembrane-anchoring segment that serves to bind microsomal P450s to the endoplasmic reticulum. The active site of the protein is instead part of a large cytoplasmic domain [36]. The anchor region functions also as a signal peptide that directs the targeting into the ER, but there is very little sequence conservation in this region between the vertebrate aromatases, [35]. Lacking this segment, the brain aromatase form II, if translated, could be described as a soluble protein. Although, the deletion of the first 20 amino acids of the human P450arom determined the preservation of only 5% [37] or undetectable activity [35], the activity was almost unaffected after removing the first 10 amino acid residues [37]. Moreover, Amarneh and Simpson [38] described the expression of a recombinant derivative of P450arom with a deletion of 41 amino acid residues in insect cells, and Osawa et al. [39] the expression of an aromatase with a deletion of 38 residues in Escherichia coli; both recombinant aromatases were found to be catalytically active. Transcripts encoding brain aromatase forms lacking the sequence corresponding to exons I–III were recently found

both in rat [40] and monkey [41]. A truncated form of human aromatase corresponding to these aromatase mRNA variants has been generated by Kao et al. [42]. The enzyme was expressed in CHO cells, showing very low enzyme activity when an “in-cell” assay was used, but strong aromatase activity by means of an in vitro method with an excess of bovine liver NADPH-cytochrome P450 reductase. According to the authors, the deletion of the hydrophobic segment prevents the association of the aromatase enzyme to the ER and the interaction with the redox system, but this form is thought to be expressed in the cortex of rat brain, as intense immunoreactivity was found in this area [43]. The cytochrome P450aromB-II, the messenger of which has been described in this work, represents a truncated form of aromatase comparable to those described above. It remains to be established whether both forms of cytochrome P450aromB encode functional enzymes. The size of the trout brain aromatase transcript was estimated, by Northern hybridisation, as about 3.8 kb. This length appears to be similar to that of zebrafish (3.8 kb) [17] and goldfish brain aromatases (3 kb) [44], but longer than that of ovarian trout aromatase (2.6 kb) [21]. A second transcript greater than 4 kb was evident in the trout brain that could correspond to P450aromB form II: the transcript of this aromatase present a longer 5 -UTR than form I. The two signals detected with ovarian RNAs hybridised with the brain specific probe correspond one to the brain transcript (3.8 kb) and the second to a shorter transcript of about 2.6 kb. The specificity of the ovarian probe is demonstrated by the fact that no signal was produced with brain samples in which aromatase expression is very high. On the other hand, the specificity of the cerebral probe cannot be established in the same way, but can be deduced from the fact that it recognized the 3.8 kb band in the ovarian samples. Northern analysis shows that isoforms A and B are differentially expressed in neural and ovarian tissues, but reveals also a low degree of overlapping expression in the ovary, as both

30

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

forms were detected. No signal was detected in gills and liver: the transcripts were presumably not abundant enough to be seen in the corresponding gills samples, whereas the negative results achieved with liver confirm those obtained with RT-PCR. In a previous work [4,45], we demonstrated that the cytochrome P450aromA, whose cDNA was cloned and sequenced by Tanaka et al. [21] using ovarian RNA, is expressed in a number of peripheral tissues of male and female trouts (brain, gills, skeletal muscle, stomach, intestine, spleen, intermediate and posterior kidney and gonads), except the liver. Similar levels of amplification (obtained with a non-quantitative RT-PCR protocol) were generally obtained with brain and kidney mRNAs, whereas the signal in the other tissues was weaker. Conversely, the brain P450aromB mRNA, analysed on the same tissues and animals, displayed a more restricted pattern of expression: in fact this variant has been detected on agarose gel only in the brain, gonads and gills [26]. For this reason, we have focussed our attention on these tissues. The semi-quantitative RT-PCR analyses confirmed the expression of both forms of aromatase in brain, gills and ovary of female trout. The level of expression of cytochrome P450aromB mRNA in the brain was higher than the expression of P450aromA mRNA in the ovary (25 cycles of PCR versus 32 used to amplify the ovarian form). This result is in agreement with the high levels of aromatase activity reported in the teleostean brain [3,4]. The overlapping expression of both forms in ovaries and brain, demonstrated by RT-PCR analysis, agrees with the results obtained in zebrafish [17,18] and Nile tilapia [19]. In goldfish, both the brain and ovarian variants of P450arom are expressed in neural tissues, while in ovaries only the ovarian form was detected [16]. Contrary to the results of the other authors, in zebrafish, the expression of CYP19B was demonstrated by Chiang et al. [46] only in neural tissues. In trout, the two forms of cytochrome P450arom are co-expressed also in gills. To our knowledge, aromatase transcripts have never been demonstrated in the gills of any fish species. Only biotransformation of testosterone to 5␣-dihydrotestosterone and androstenedione has been reported in gill cells [47], while the estrogen receptor-␣ has been found, by RT-PCR, in gills of Atlantic salmon, Salmo salar [48] and trout (Ramina et al., in preparation). However, the physiological role of any estrogens formed in trout gills remains unknown. Even though the transcript abundance is low, and background or leaking gene transcription cannot be ruled out as a possible cause, the expression of cytochrome P450arom mRNA in trout gills is intriguing, especially considering the relatively large mass of branchial tissue in fish. The topology of the tree presented in this paper is in agreement with those provided by Kishida and Callard [17] and Chiang et al. [46]. The fish ovary- and brain-derived P450arom forms appear to be paralogous between themselves, while they are co-orthologous [49] to the amphibian, avian and mammalian aromatases. In accordance with the topology of the tree obtained, a genome duplication occurred

in the fish lineage after the divergence of ray- and lobe-finned fishes, the lineages leading to teleosts and tetrapods, respectively: in fact, as proposed by Gates et al. [50], the products of a duplication after divergence would be more closely related to each other than they are to their ortholog in the nonduplicated lineage. This hypothesis is supported by recent studies on linkage maps for zebrafish and hox clusters of zebrafish and Fugu rubripes, suggesting that the corresponding chromosome was doubled by an additional whole genome duplication after the divergence of the teleostean and tetrapod lineages [51]. This process provides duplicated genes to evolve new adaptive functions and/or, as in this case, differential regulation of tissue expression, resulting in the preservation of both members of the pair. Both teleost CYP19 genes maintain enzymatic activity [16,46,52], but present distinct expression profiles suggesting differences in their physiological functions, differences that, in mammals, could be obtained with the use of alternative promoters and untranslated first exons. In conclusion, we have demonstrated the presence of a second CYP19 gene in the trout genome and confirmed that teleost fish contain at least two separate and distinct CYP19 loci with different expression domains.

Acknowledgements The authors wish to thank Dr. E. Negrisolo for help in phylogenetic analysis and comments. Research was supported by Grant no. 5C 117 from the Ministry of Agriculture and Forestry Policies of Italy, in the purview of the Fifth Triennial Plan for Fisheries and Aquaculture in Marine and Brackish Waters. References [1] E.R. Simpson, M.D. Michael, V.R. Agarwal, M.M. Hinshelwood, S.E. Bulun, Y. Zhao, Expression of the CYP19 (aromatase) gene: an unusual case of alternative promoter usage, FASEB J. 11 (1) (1997) 29–36. [2] G.V. Callard, A. Tchoudakova, Evolutionary and functional significance of two CYP19 genes differentially expressed in brain and ovary of goldfish, J. Steroid Biochem. Mol. Biol. 61 (3–6) (1997) 387–392. [3] M. Pasmanik, G.V. Callard, Aromatase and 5␣-reductase in teleost brain, spinal cord, and pituitary gland, Gen. Comp. Endocrinol. 60 (2) (1985) 244–251. [4] P. Belvedere, L. Dalla Valle, A. Lucchetti, A. Ramina, S. Vianello, L. Colombo, Extraglandular expression of genes encoding steroidogenic cytochromes in rainbow trout (Oncorhynchus mykiss Walbaum), Ann. N. Y. Acad. Sci. 839 (1) (1998) 559–589. [5] E.D. Lephart, A review of brain aromatase cytochrome P450, Brain Res. Brain Res. Rev. 22 (1) (1996) 1–26. [6] N.J. MacLusky, F. Naftolin, Sexual differentiation of the central nervous system, Science 211 (1981) 1294–1302. [7] B.S. McEwen, S.E. Alves, Estrogen actions in the central nervous system, Endocr. Rev. 20 (3) (1999) 279–307. [8] C.D. Toran-Allerand, Mechanisms of estrogen action during neural development: mediation by interactions with the neurotrophins and

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25] [26]

their receptors? J. Steroid Biochem. Mol. Biol. 56 (1–6) (1996) 169– 178. S.T. Mahajan, A.B. Pinto, D.B. Williams, The additional benefits of hormone replacement therapy, Prim. Care Update Ob. Gyns. 8 (6) (2001) 260–263. H.N. Fugger, T.C. Foster, J. Gustafsson, E.F. Rissman, Novel effects of estradiol and estrogen receptor-␣ and ␤ on cognitive function, Brain Res. 883 (2) (2000) 258–264. G.V. Callard, A. Tchoudakova, M. Kishida, E. Wood, Differential tissue distribution, developmental programming, estrogen regulation and promoter characteristics of CYP19 genes in teleost fish, J. Steroid Biochem. Mol. Biol. 79 (1–5) (2001) 305–314. S.C. Clint, G.K.H. Zupanc, Neuronal regeneration in the cerebellum of adult teleost fish, Apteronotus leptorhynchus: guidance of migrating young cells by radial glia, Brain Res. Dev. Brain Res. 130 (1) (2001) 15–23. I. Choi, D.L. Troyer, D.L. Cornwell, K.R. Kirby-Dobbels, W.R. Collante, F.A. Simmen, Closely related genes encode developmental and tissue isoforms of porcine cytochrome P450 aromatase, DNA Cell Biol. 16 (6) (1997) 769–777. L.G. Graddy, A.A. Kowalski, F.A. Simmen, S.L. Davis, W.W. Baumgartner, R.C. Simmen, Multiple isoforms of porcine aromatase are encoded by three distinct genes, J. Steroid Biochem. Mol. Biol. 73 (1/2) (2000) 49–57. R. Fürbaß, J. Vanselow, An aromatase pseudogene is transcribed in the bovine placenta, Gene 154 (2) (1995) 287–291. A. Tchoudakova, G.V. Callard, Identification of multiple CYP19 genes encoding different cytochrome P450 aromatase isozymes in brain and ovary, Endocrinology 139 (4) (1998) 2179–2189. M. Kishida, G.V. Callard, Distinct cytochrome P450 aromatase isoforms in zebrafish (Danio rerio) brain and ovary are differentially programmed and estrogen regulated during early development, Endocrinology 142 (2) (2001) 740–750. J.M. Trant, S. Gavasso, J. Ackers, B.C. Chung, A.R. Place, Developmental expression of cytochrome P450 aromatase genes (CYP19a and CYP19b) in zebrafish fry (Danio rerio), J. Exp. Zool. 290 (5) (2001) 475–483. J.Y. Kwon, B.J. McAndrew, D.J. Penman, Cloning of brain aromatase gene and expression of brain and ovarian aromatase genes during sexual differentiation in genetic male and female Nile tilapia Oreochromis niloticus, Mol. Reprod. Dev. 59 (4) (2001) 359–370. M. Cruz, A.V.M. Canario, cDNA cloning of brain and ovary aromatase in tilapia, Oreochromis mossambicus, in: B. Norberg, O.S. Kjesbu, G.L. Taranger, E. Andersson, S.O. Stefansson (Eds.), Proceedings of the Fish Symposium 1999 on Reproductive Physiology of Fish, Bergen, Norway, p. 193. M. Tanaka, T.M. Telecky, S. Fukada, S. Adachi, S. Chen, Y. Nagahama, Cloning and sequence analysis of the cDNA encoding P-450 aromatase (P450arom) from a rainbow trout (Oncorhynchus mykiss) ovary relationship between the amount of P450arom mRNA and the production of oestradiol-17␤ in the ovary, J. Mol. Endocrinol. 8 (1) (1992) 53–61. J.D. Thompson, D.G. Higgings, T.J. Gibson, W. Clustal, Improving the sensitivity of progressive multiple sequence alignment through sequence weighing, Nucleic Acid Res. 22 (22) (1994) 4673–4680. N. Saitou, M. Nei, The neighbor-joining method: a new method for reconstructing phylogenetic trees, Mol. Biol. Evol. 4 (4) (1987) 406–425. Y. Van de Peer, R. De Wachter, TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment, Comput. Appl. Biosci. 10 (5) (1994) 569–570. J. Felsenstein, Confidence limits on phylogenies: an approach using the bootstrap, Evolution 39 (1985) 783–791. L.D. Valle, A. Ramina, P. Belvedere, L. Colombo, Tissue distribution of cytochrome P450arom type II in peripheral tissues and gonads of rainbow trout, in: H.J.T. Goos, R.K. Rastogi, H. Vaudry, P.

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36] [37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

31

Pierantoni (Eds.), Perspective in Comparative Endocrinology: Unity and Diversity, Monduzzi Editore, Bologna, 2001, pp. 1061–1067. M. Kozak, Point mutations define a sequence flanking the AUG initiator codon that modulates translation by eukaryotic ribosomes, Cell 44 (2) (1986) 283–292. O. Thellin, W. Zorzi, B. Lakaye, B. De Borman, B. Coumans, G. Hennen, T. Grisar, A. Igout, E. Heinen, Housekeeping genes as internal standards: use and limits, J. Biotechnol. 75 (2/3) (1999) 291–295. A. Tchoudakova, M. Kishida, E. Wood, G.V. Callard, Promoter characteristics of two CYP19 genes differentially expressed in the brain and ovary of teleost fish, J. Steroid Biochem. Mol. Biol. 78 (5) (2001) 427–439. C. Risinger, D. Larhammar, Multiple loci for synapse protein SNAP-25 in the tetraploid goldfish, Proc. Natl. Acad. Sci. U.S.A. 90 (22) (1993) 10598–10602. F.W. Allendorf, G.H. Thorgaard, Tetraploidy and the evolution of salmoid fishes, in: B.J. Turner (Ed.), Evolution Genetics of Fishes, Plenum Press, New York, 1994, pp. 1–53. G.D. Means, M.S. Mahendroo, C.J. Corbin, J.M. Mathis, F.E. Powell, C.R. Mendelson, E.R. Simpson, Structural analysis of the gene encoding human aromatase cytochrome P-450, the enzyme responsible for estrogen biosynthesis, J. Biol. Chem. 264 (32) (1989) 19385–19391. M. Tanaka, S. Fukada, M. Matsuyama, Y. Nagahama, Structure and promoter analysis of the cytochrome P-450 aromatase gene of the teleost fish, medaka (Oryzias latipes), J. Biochem. 117 (4) (1995) 719–725. L.D. Valle, L. Lunardi, L. Colombo, P. Belvedere, European sea bass (Dicentrarchus labrax L.) cytochrome P450arom: cDNA cloning, J. Steroid Biochem. Mol. Biol. 80 (1) (2002) 25–34. S. Chen, D. Zhou, Functional domains of aromatase cytochrome P450 inferred from comparative analyses of amino acid sequences and substantiated by site-directed mutagenesis experiments, J. Biol. Chem. 267 (31) (1992) 22587–22594. S.D. Black, Membrane topology of the mammalian P450 cytochromes, FASEB J. 6 (2) (1992) 680–685. B. Amarneh, C.J. Corbin, J.A. Peterson, E.R. Simpson, S. Graham-Lorence, Functional domains of human aromatase cytochrome P450 characterised by linear alignment and site-directed mutagenesis, Mol. Endocrinol. 7 (12) (1993) 1617–1624. B. Amarneh, E.R. Simpson, Expression of a recombinant derivative of human aromatase P450 in insect cells utilising the baculovirus vector system, Mol. Cell. Endocrinol. 109 (2) (1995) R1–R5. Y. Osawa, T. Higashiyama, Y. Toma, C. Yarborough, Diverse function of aromatase and the N-terminal sequence deleted form, J. Steroid Biochem. Mol. Biol. 61 (3–6) (1997) 117–126. J. Kato, N. Yamada-Mouri, S. Hirata, Structure of aromatase mRNA in the rat brain, J. Steroid Biochem. Mol. Biol. 61 (3–6) (1997) 381–385. C.E. Roselli, J.A. Resko, Cytochrome P450 aromatase (CYP19) in the non-human primate brain: distribution, regulation, and functional significance, J. Steroid Biochem. Mol. Biol. 79 (1–5) (2001) 247– 253. Y.C. Kao, T. Higashiyama, C. Yarborough, Y. Osawa, S. Chen, Functional characterisation of 102-amino acid-deleted form of human aromatase (delta102-aromatase), Steroids 64 (6) (1999) 422–429. M.K. Sanghera, E.R. Simpson, M.J. McPhaul, G. Kozlowski, A.J. Conley, E.D. Lephart, Immunocytochemical distribution of aromatase cytochrome P450 in the rat brain using peptide-generated polyclonal antibodies, Endocrinology 129 (6) (1991) 2834–2844. D. Gelinas, G.A. Pitoc, G.V. Callard, Isolation of a goldfish brain cytochrome P450 aromatase cDNA: mRNA expression during the seasonal cycle and after steroid treatment, Mol. Cell. Endocrinol. 138 (1/2) (1998) 81–93. A. Ramina, L.D. Valle, M. Bacchin, P. Belvedere, L. Colombo, Identification of a distinct form of cytochrome P450arom expressed

32

[46]

[47]

[48]

[49]

L. Dalla Valle et al. / Journal of Steroid Biochemistry & Molecular Biology 82 (2002) 19–32 in rainbow trout brain, in: Proceedings of the 20th Conference of European Comparative Endocrinologists, Faro, Portugal, 5–10 September 2000, p. 63. E.F. Chiang, Y.L. Yan, Y. Guiguen, J. Postlethwait, B.C. Chung, Two CYP19 (P450 aromatase) genes on duplicated zebrafish chromosomes are expressed in ovary or brain, Mol. Biol. Evol. 18 (4) (2001) 542– 550. I. Leguen, C. Carlsson, E. Perdu-Durand, P. Prunet, P. Pärt, J.P. Cravedi, Xenobiotic and steroid biotransformation activities in rainbow trout gill epithelial cells in culture, Aquat. Toxicol. 48 (2/3) (2000) 165–176. S.A. Rogers, L. Llewellyn, T. Wigham, G.E. Sweeney, Cloning of the Atlantic salmon (Salmo salar) estrogen receptor-␣ gene, Comp. Biochem. Physiol. B Biochem. Mol. Biol. 125 (3) (2000) 379–385. J.S. Taylor, Y. Van de Peer, A. Meyer, Genome duplication, divergent resolution and speciation, Trends Genet. 17 (6) (2001) 299–301.

[50] M.A. Gates, L. Kim, E.S. Egan, T. Cardozo, H.I. Sirotkin, S.T. Dougan, D. Lashkari, R. Abagyan, A.F. Schier, W.S. Talbot, A genetic linkage map for zebrafish: comparative analysis and localization of genes and expressed sequences, Genome Res. 9 (4) (1999) 334–347. [51] S. Aparicio, Vertebrate evolution, recent perspectives from fish, Trends Genet. 16 (2) (2000) 54–56. [52] J. Zhao, P. Mak, A. Tchoudakova, G.V. Callard, S. Chen, Different catalytic properties and inhibitor responses of the goldfish brain and ovary aromatase isozymes, Gen. Comp. Endocrinol. 123 (2) (2001) 180–191. [53] E.R. Simpson, M.S. Mahendroo, G.D. Means, M.W. Kilgore, M.M. Hinshelwood, S. Graham-Lorence, B. Amarneh, Y. Ito, C.R. Fisher, M.D. Michael, C.R. Mendelson, S.E. Bulun, Aromatase cytochrome P450, the enzyme responsible for estrogen biosynthesis, Endocr. Rev. 15 (3) (1994) 342–355.