Duality of serotonin-N-acetyltransferase in the gilthead seabream (Sparus aurata): molecular cloning and characterization of recombinant enzymes

General and Comparative Endocrinology 138 (2004) 139–147

www.elsevier.com/locate/ygcen

Duality of serotonin-N-acetyltransferase in the gilthead seabream (Sparus aurata): molecular cloning and characterization

of recombinant enzymes�

Bina Zilberman-Peled,a Itai Benhar,b Steven L. Coon,c Benny Ron,d and Yoav Gothilfa,¤

a Department of Zoology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Tel Aviv, Israelb Department of Molecular Microbiology and Biotechnology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Israel

c Section on Neuroendocrinology, Laboratory of Developmental Neurobiology, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA

d National Center for Mariculture, Israel Oceanographic and Limnological Research Institute, Eilat, Israel

Received 2 February 2004; revised 14 April 2004; accepted 18 May 2004

Abstract

Serotonin-N-acetyltransferase (arylalkylamine-N-acetyltransferase, AANAT) is the key enzyme in the biosynthesis of melatoninin the pineal gland and retinal photoreceptors. Rhythmic AANAT activity drives rhythmic melatonin production in these tissues.The presence of two AANATs, AANAT1 and AANAT2, has been previously demonstrated in three fresh water teleosts. This dual-ity, the result of early gene duplication, is unique to teleost species. In this study, the cDNAs encoding for AANAT1 and AANAT2were cloned from a marine Wsh, the gilthead seabream (sb, Sparus aurata). Northern blot hybridization analysis indicates that sbA-ANAT1 and sbAANAT2 are exclusively expressed in the retina and pineal gland, respectively. Bacterially expressed recombinantsbAANATs exhibit diVerential enzyme kinetics. Recombinant retinal sbAANAT1 has relatively high substrate aYnity and lowactivity rate; it is inhibited by high substrate and product concentrations. In contrast, recombinant pineal sbAANAT2 exhibits lowsubstrate aYnity and high activity rate and is not inhibited by substrates or products. The two recombinant enzymes also exhibitdiVerential substrate preference. Retinal sbAANAT1 acetylates a range of arylalkylamines while pineal sbAANAT2 preferentiallyacetylates indoleethylamines, especially serotonin. The diVerent spatial expression patterns, enzyme kinetics, and substrate prefer-ences of the two sbAANATs support the hypothesis that, as a consequence of gene duplication, teleosts have acquired two AANATswith diVerent functions. Pineal AANAT2 specializes in the production of large amounts of melatonin that is released into the circu-lation and exerts an endocrine role. Retinal AANAT1, on the other hand, is involved in producing low levels of melatonin that exe-cute a paracrine function. In addition, retinal AANAT1 may carry out an as yet unknown function that involves acetylation ofarylalkylamines other than serotonin. 2004 Elsevier Inc. All rights reserved.

Keywords: Melatonin; Serotonin; Indoleethylamine; Arylalkylamine-N-acetyltransferase; Pineal gland; Retina; Seabream

1. Introduction

The hormone melatonin is an important componentof the circadian system in vertebrates. Melatonin pro-duction takes place during the night at two major sites:

� The nucleotide sequences reported in this paper have been sub-mitted to GenBank with Accession Nos. AY533402–AY533403.

¤ Corresponding author. Fax: +972-3-6409403.E-mail address: [email protected] (Y. Gothilf).

0016-6480/$ - see front matter 2004 Elsevier Inc. All rights reserved.doi:10.1016/j.ygcen.2004.05.007

the pineal gland and retinal photoreceptors cells. Thepineal gland is the source of circulating melatoninwhich provides a night-time signal and plays anendocrine role in the regulation of a variety of dailyand annual physiological rhythms (Arendt, 1995).Retinal melatonin, on the other hand, plays aparacrine role related to photoreceptor adaptation todarkness, and generally does not contribute to thecirculating levels of melatonin (Cahill and Besharse,1995).

140 B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147

Melatonin is derived from serotonin by two enzy-matic reactions: N-acetylation of serotonin followed byO-methylation of N-acetylserotonin. The Wrst step, trans-fer of the acetyl group from acetyl-CoA to serotonin, iscatalyzed by serotonin-N-acetyltransferase (arylalkyl-amine-N-acetyltransferase, AANAT). The temporalchanges in the rate of melatonin production are con-trolled by the activity of AANAT. Increased melatoninproduction at night reXects increased AANAT activityand termination of melatonin production during the dayreXects proteasomal degradation of the enzyme (Gastelet al., 1998). In some species, AANAT gene expression isrhythmic (Klein et al., 1997). Rhythmic AANAT geneexpression, AANAT activity, and hence melatonin pro-duction, are driven by an internal circadian clock and byexternal light signals (Foulkes et al., 1997; Ganguly et al.,2002; Klein et al., 1997; Li et al., 1998; Tosini and Fuku-hara, 2003).

In mammals and chicken, only a single AANAT genehas been identiWed and is expressed in both the pinealgland and retina. However, two AANAT genes—AANAT1 and AANAT2—have been discovered inthree fresh water teleosts, Northern pike, Esox lucius(Coon et al., 1999); trout, Oncorhynchus mykiss (Begayet al., 1998; Mizusawa et al., 1998); and zebraWsh, Daniorerio (Falcon et al., 2003; Gothilf et al., 1999). These twoenzymes exhibit diVerent spatial expression pattern,enzyme kinetic, and substrate preference. In pike andtrout, AANAT1 is expressed only in the retina andAANAT2 is expressed only in the pineal gland (Coon etal., 1999; Mizusawa et al., 2000). Retinal AANAT1,which has a higher similarity to the mammalian andchicken AANATs than AANAT2, acetylates arylalkyl-amines, while AANAT2 preferentially acetylates indole-ethylamines. In addition, retinal AANAT1 exhibitshigher aYnity for its substrates than does AANAT2(Benyassi et al., 2000; Coon et al., 1999). Based on theseobservations, it has been proposed that the selectiveadvantage of having two AANATs may be related to theparacrine role of retinal melatonin versus the endocrinerole of pineal melatonin (Coon et al., 1999; Falcon et al.,2003).

Nevertheless, the signiWcance of the AANAT dualityin teleosts is still unknown. Towards resolving this ques-tion and establishing ground work for further investiga-tion we have cloned two AANATs from a commerciallyimportant aquaculture species—the gilthead seabream(sb, Sparus aurata). Here we show that in this marineWsh, AANAT1 (sbAANAT1) is expressed only in theretina while AANAT2 (sbAANAT2) is expressed only inthe pineal gland. Using recombinant enzymes, we dem-onstrate that the two sbAANATs exhibit diVerentkinetic parameters and substrate preference. Our resultsstrengthen the hypothesis that the AANAT duality hasevolved to serve two melatonin generating systems withdiVerent kinetic requirements. An extension of this

hypothesis, presented herein, is that the retinal sbA-ANAT1 performs other functions in addition to its rolein melatonin synthesis.

2. Materials and methods

2.1. Isolation of full length sbAANAT1 and sbAANAT2cDNAs

Adult seabream pineal glands and retinas were col-lected at night and total RNA was isolated using TRIzolreagent (Gibco-BRL, Grand Island, NY). MessengerRNA was isolated using Oligo(dT)-magnetic beads(Dynal, Oslo, Norway) and was used as a template tosynthesize Wrst strand cDNAs on the beads according tomanufacturer’s instructions. PCR was performed on thepineal and retina Wrst strand cDNAs using a set ofdegenerate primers DF1 (cgmcacacrytnccngcyagygag)and DR1 (ggagccyytgccctgctgbcg). AmpliWcation reac-tions were performed in 50 �l containing 2 �M DF1 and1 �M DR1, 200 �M dNTPs, 2 �l Wrst strand cDNA, and1 U Taq DNA polymerase (Boehringer–Mannheim,Indianapolis, IN). Reaction conditions were 95 °C for2 min followed by 40 cycles of denaturation at 94 °C for30 s, annealing at 55 °C for 1 min, and extension at 72 °Cfor 20 s. PCR products (327 bp) were sub-cloned intopGEM-T Easy (Promega, Madison, WI) and the cloneswere sequenced.

Seabream AANAT1-speciWc primers, 5sbnat1a (gtaccgtgaggcttgtggagcgtcaggg) and 3sbnat1a (gtggctttcatcatcggttcactttggg), were designed according to the327 bp fragment isolated from retina and were used in5�- and 3�-RACE PCR ampliWcation according to man-ufacturer’s instructions (SMART RACE cDNA ampliW-cation kit, Clontech, Palo Alto, CA) yielding 550 and1100 bp products, respectively. AmpliWcation withnested primers, 5sbnat1b (atccgtggtaagcctctcctggt) and3sbnat1b (gaccaggagaggcttaccacgga), was performed onthe products and the nested PCR products were sub-cloned into pGEM-T Easy and sequenced. Overall,1.4 kb of the sbAANAT1 cDNA was isolated.

The 5� end of sbAANAT2 was ampliWed from seab-ream pineal poly(A+)RNA using a 5�-RACE kit (Gibco-BRL) according to manufacturer’s instructions. Firststrand cDNA was synthesized using SB2R1 primer(tcagggcacagggtgaggaagtgac) and 5�-RACE was per-formed using SB2R2 primer (actctccagacacagagacaaatgc) and a universal primer (Gibco-BRL). The prod-uct was sub-cloned into pGEM-T Easy (Promega) andsequenced.

The 3� end of sbAANAT2 was ampliWed from Wrststrand pineal cDNA using oligo(dT) and nested primersSBF1 (acaaagaggagctggaggag) and SBF2 (agggacacacaggtcagc), which correspond to sequences upstream ofthe putative ATG translation start site (according to

B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147 141

information obtained by 5�-RACE). The 3�-RACE PCRproduct, 1.4 kb in length, was cloned into pGEM-T Easyand sequenced. This clone, sbAANAT2c, contains210 bp of 5� UTR, the entire coding region and 3� UTRof sbAANAT2 cDNA.

2.2. Northern blot hybridization

Seabream pineal gland and retina total RNA wereextracted as described above and fractionated througha 1.5% agarose gel containing formaldehyde, trans-ferred onto a Hybond-N nylon membrane (Amersham,Buckinghamshire, UK) by capillary blotting, and UV-cross-linked to the membrane. The 327 bp fragment ofsbAANAT1 and sbAANAT2 cDNAs were labeled byrandom priming with [�-32P]dCTP (Amersham) andused as probes to detect the corresponding mRNAs byhybridization at stringent conditions: hybridization at68 °C with Quikhyb (Stratagene, La Jolla, CA) and Wnalwashes at 60 °C with 0.1£ SSC. The membranes wereexposed to X-ray Wlm for 24 h.

2.3. Preparation of GST-sbAANAT1 and GST-sbAANAT2 fusion proteins

The coding region of sbAANAT1 was PCR ampliWedusing a proofreading enzyme mixture, DyNAzyme EXT(Finnzymes, Espoo, Finland) and a set of speciWc primerscontaining restriction enzyme sites, sbNAT1exf (gaaagatgaattcatgtcggtggtgagcgc) and sbNAT1exr (gtccctcgcggccgctcaacagccgctgttcc). Similarly, the coding region of sbA-ANAT2 was PCR ampliWed using sbNAT2exf (gtattgaggatccatgacacagcaggtcag) and sbNAT2exr (gtagggagcggccgcctagcagccgctgttcc). The AANAT1 PCR product wasdigested with EcoRI and NotI, and ligated into EcoRI/NotI-cut pGEX-4T-1 vector (Amersham), giving rise topsb1ex. The AANAT2 PCR product was digested withBamHI and NotI, and ligated into BamHI/NotI-cutpGEX-4T-1 vector, giving rise to psb2ex. These con-structs, psb1ex and psb2ex, generate sbAANAT1 andsbAANAT2 fused to glutathione S-transferase (GST).The clones were transformed into BL21 Escherichiacoli and bacteria were grown at 37 °C until the OD600was 0.5. Recombinant protein production was inducedwith 0.2 mM IPTG, overnight at 16 °C (psb1ex-trans-formed) or for 5 h at 24 °C (psb2ex-transformed). Thebacteria were pelleted, resuspended in 4 ml of buVer A(2£ PBS, 10 mM DTT, 10% glycerol, and 0.1% TritonX-100, pH 6.8) per 50 ml growth medium, and soni-cated. The sonicate was centrifuged (14,000g, 4 °C) for10 min and the supernatant was mixed with glutathioneSepharose beads (Amersham). The beads were washedtwice with buVer A and twice with buVer B (100 mMsodium citrate, 50 mM Tris–HCl, 5 mM DTT, and 10%glycerol, pH 8). The GST-sbAANAT fusion proteinswere eluted from the beads with 1 ml buVer B containing

10 mM glutathione. Protein was stored in aliquots at¡80 °C.

2.4. Colorimetric assay of AANAT activity

The colorimetric assay for AANAT activity is basedon the detection of CoASH generated during acetyltransfer. In this assay, CoASH reacts with the thiolreagent 5,5�-dithio-bis(2-nitrobenzoic acid) (DTNB)and forms a colored product with maximum absor-bance at 405 nm (De angelis et al., 1998; Riddles et al.,1983). Standard reactions were performed in 100 �l bycombining 25 �l of diluted enzyme (5 �g/ml) and 75 �lphosphate/NaCl assay buVer (Table 1) containing BSA(0.05 mg/ml), 2 mM EDTA, acetyl-CoA, and aminesubstrates at indicated concentrations. Incubation timewas 40 min at 36 °C for sbAANAT1 and 27 °C for sbA-ANAT2. Optimal conditions (see Table 1) were usedunless otherwise stated. Enzymes were inactivated bythe addition of 150 �l stop solution (0.1 M phosphatebuVer, pH 6.8, containing 1 mM DTNB, 10 mMEDTA, and 3 M guanidine hydrochloride) and the col-ored product was measured at 405 nm after 5 min ofincubation at room temperature. Background level wasdetermined by adding the enzymes immediately afterthe addition of stop solution. Assays were performedtwo or three times in duplicates. Km and Vmax valueswere calculated after linearization of the activityvalues according to the Lineweaver–Burk plot (1/V vs1/S).

3. Results

3.1. Tissue distribution of AANAT1 and AANAT2

Using Northern blot analysis AANAT1 mRNA wasdetected as a single 2.2 kb band only in the retina butwas not detected in the pineal gland. AANAT2 mRNAwas exclusively detected as a single 1.4 kb band in thepineal gland (Fig. 1). This diVerential spatial expressionpattern is in accordance with the tissue distribution ofthe two AANATs in pike and trout (Coon et al., 1999;Mizusawa et al., 2000).

Table 1Optimal parameters for sbAANAT activity colorimetric assays

For details, see Section 2.

Validation parameter sbAANAT1 sbAANAT2

Molarity (M) 0.1 phosphate/0.5 NaCl

0.3 phosphate

pH 6.5 6.5Temperature (°C) 36 27Incubation time (min) 40 40Enzyme concentration

(�g/ml)1.2 1.2

142 B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147

3.2. Sequence analysis

The observed size of the retinal sbAANAT mRNA(2.2 kb) is larger than its cloned sequence (1.4 kb). Thismay reXect a large 3�-polyA tail or an incomplete cDNA,i.e., 5� or 3� regions are missing. Nevertheless, the entirecoding region has been cloned and its analysis predicts a202 amino acid protein (Fig. 2A) that has 79–88% iden-tity to Wsh AANAT1 proteins, 61–68% identity to mam-malian AANATs, 76% identity to chicken AANAT, and66–68% identity to Wsh AANAT2 proteins. Phylogeneticanalysis places the cloned retinal sbAANAT with theWsh AANAT1 group which is related to chicken, Xeno-pus, and mammalian AANATs (Fig. 2B).

The pineal sbAANAT cDNA encodes a 205 aminoacid protein (Fig. 2A) that has a high identity (85–88%)with Wsh AANAT2s and lower identity to mammalian(54–58%) and chicken (63%) AANATs and WshAANAT1s (60–67%). Phylogenetic analysis placespineal sbAANAT together with the Wsh AANAT2s inone group that is distinct, but related to, all other AAN-ATs (Fig. 2B).

Accordingly, sbAANAT1 and sbAANAT2 have arelativity low (66%) amino acid similarity to each other(Fig. 2A). Both proteins contain amino acid residuesthat are conserved in all known AANATs and are

Fig. 1. Northern blot analysis of seabream AANAT1 and AANAT2mRNA expression. Total RNA was extracted from seabream retina(30 �g) and pineal glands (5 �g), blotted, and hybridized using radio-labeled sbAANAT1 or sbAANAT2 probes. The tissue, retina orpineal, is given above each lines and the probes are given at the bottomof the gel.

involved in the catalytic activity and the regulation ofthe enzyme (Fig. 2A). These residues include two proteinkinase A (PKA) phosphorylation sites, sbAANAT1Arg23–Thr26 and Arg197–Ser200 and sbAANAT2Arg26–Thr29 and Arg200–Ser203, which may mediate theobserved cAMP-regulation of AANAT proteasomaldegradation (Falcon et al., 2001; Ganguly et al., 2002),and two histidines, sbAANAT1 His115 and His117 andsbAANAT2 His118 and His120, which form a catalyticsite (Hickman et al., 1999). In addition, sbAANAT1 con-tains eight highly conserved amino acid residues—Phe51,Pro59, Leu104, Met154, Ile176, Val178, Leu181, and Phe183—that probably form a hydrophobic pocket which func-tions as the serotonin recognition and binding site(Hickman et al., 1999). Interestingly, two of these residuesare diVerent in sbAANAT2 and other Wsh AANAT2s:Leu104 was replaced by Met and Met154 was replaced byIle (Fig. 2A). These changes may explain the diVerencesin the substrate preference and aYnity of the twoenzymes. Another notable diVerence is the lack of a Lysresidue near the N-terminal end of sbAANAT1 (Lys12 insbAANAT2). This Lys is conserved in all AANATsexcept for Xenopus AANAT (Accession No. AY316296and AY316297) and sbAANAT1 (current study) and isconsidered a putative ubiquitination site for the prote-asomal degradation of AANAT (Ganguly et al., 2002;Gastel et al., 1998; Klein et al., 1999). The lack of this res-idue in the retinal sbAANAT1 may have an implicationfor its regulation, presumably, a diVerent response tolight.

3.3. Production of recombinant enzyme and optimizationof assay conditions

Recombinant sbAANATs were expressed as GST-fusion proteins in E. coli and puriWed using glutathioneaYnity chromatography. The puriWed products were ofthe predicted molecular mass (approximately 53 kDa) asdetermined by SDS–polyacrylamide gel electrophoresis(data not shown). Both recombinant proteins exhibitedenzymatic activity: i.e., transfer of the acetyl group fromacetyl-CoA to serotonin using the AANAT colorimetricassay. Based on the similar kinetics of native and recom-binant GST-AANATs from sheep and pike (Coon et al.,1999; De angelis et al., 1998) it is likely that kinetic char-acterizations of recombinant GST-sbAANAT fusionproteins also reXect the native enzymes.

Validations of assay parameters indicate linear activ-ity of both sbAANAT1 and sbAANAT2 throughout thetested range of enzyme concentrations (0.2–3 �g/ml; Wnalconcentration) and linear activity as a function of incu-bation time (up to 120 min for sbAANAT1 and up to90 min for sbAANAT2). The optimum pH for bothenzymes was 6.5 with sbAANAT2 being more tolerantto pH variations than sbAANAT1 (data not shown).Testing a range of NaCl and phosphate concentrations

B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147 143

Fig. 2. Sequence analysis of seabream AANAT1 and AANAT2. (A) Comparison of deduced amino acid sequences of sbAANAT1 (GenBank Acces-sion No. AY33402) and sbAANAT2 (GenBank Accession No. AY33403). Residues linked by vertical lines are identical; vertical dots indicate simi-larity. Putative PKA target sites are underlined. Conserved histidines which form a catalytic site are labeled with dots. The N-terminal lysine which isassumed to serve as an ubiquitination site is boxed. Conserved residues which are hypothesized to form the hydrophobic substrate binding pocketare labeled with asterisks. Residues within the substrate binding pocket that are diVerent in sbAANAT2 are labeled ‘$.’ (B) Dendrogram showing thephylogenetic relationships of seabream AANATs to other AANATs. Peptide sequences were used to generate the dendrogram; all residues up to andincluding the N-terminal PKA site, and residues including and following the C-terminal PKA site (see A) were removed prior to alignment.Sequences were aligned using Clustal W and the dendrogram was generated using Phylip software. The tree was rooted with yeast (Saccharomycescerevisiae) as the outgroup. GenBank accession numbers for the peptides used to generate the dendrogram: human (Homo sapiens; NP_001079); rat(Rattus norvigicus; NP_036950); chicken (Gallus gallus; AAB40942); Xenopus-1a1 (Xenopus laevis; AY316296); Xenopus-1a2 (X. laevis; AY316297);zebraWsh-1 (Danio rerio; AAQ54582); zebraWsh-2 (D. rerio; NP_571486); puVerWsh-2 (Sphoeroides nephelus; AAL73048); pike-1 (Esox lucius;AAD21316); pike-2 (E. lucius; AAD21317); trout-1 (Oncorhynchus mykiss; BAA34809); trout-2 (O. mykiss; AAD25333); seabream-1 (Sparus aurata;AY33402); seabream-2 (S. aurata; AY33403); and yeast (S. cerevisiae; NP_010356). The number following each Wsh label identiWes the form ofAANAT to which that sequence belongs.

144 B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147

revealed an optimum activity of sbAANAT1 at 0.1 Mphosphate/0.5 M NaCl and of sbAANAT2 at 0.3 Mphosphate (Table 1).

Fig. 3. AANAT activity as a function of incubation temperature.Enzyme activities were measured at a range of temperatures (0–50 °C).Reactions were performed in the presence of 1 mM acetyl-CoA andtryptamine at a concentration of 0.5 mM (sbAANAT1 assays) or5 mM (sbAANAT2 assays). All other variables were kept at their opti-mum (Table 1). Experiments were conducted three times; the graphsrepresent one experiment.

Fig. 4. AANAT activity as a function of acetyl-CoA concentration.Enzyme activities were measured at saturating concentration of trypt-amine (0.5 mM in sbAANAT1 and 10 mM in sbAANAT2 assays) andincreasing acetyl-CoA concentrations (0.1–3 mM). All other variableswere kept at their optimum (Table 1). Background level was deter-mined for each diVerent acetyl-CoA concentration to account for thepresence of free thiol contaminant in the acetyl-CoA solution. Experi-ments were conducted twice. The graph represent one experiment.Inset: Lineweaver–Burk plot, r2 D 0.99 for both enzymes.

As has been documented for the pike AANATs,the eVect of temperature on the activity of the twoenzymes is diVerent. Maximal activity for sbAANAT1was at 36 °C and for sbAANAT2 at 27 °C. SeabreamAANAT1 activity was temperature-sensitive while sbA-ANAT2 was active in a wide range of temperatures andexhibited about 50% of its maximal activity at 0 °C(Fig. 3).

Based on the above experiments, further kinetic char-acterizations were performed with a Wnal enzyme con-centration of 1.2 �g/ml, 40 min incubation time, optimaltemperature for each enzyme, and optimal assay buVersfor each enzyme (Table 1).

3.4. Activity as a function of acetyl-CoA concentration

Enzyme activities were measured at a range of acetyl-CoA concentrations and a saturated concentration oftryptamine. The results of both sbAANAT1 and sbA-ANAT2 Wt the classical Michaelis–Menten equation andkinetic parameters were determined after transformationof the data to Lineweaver–Burk plot. Seabream AANAT1has a Km of 0.62mM and Vmax of 0.49nmol/h/ng; sbA-ANAT2 has a Km of 1.11mM and Vmax of 0.93nmol/h/ng(Fig. 4 and Table 2).

3.5. Substrate preference

Acetylation of diVerent amine substrates by the twoenzymes was measured in the presence of indoleethylam-ines (serotonin and tryptamine) and phenylethylamines(tyramine and phenylethylamine) at a range of concen-trations, and a saturated concentration of acetyl-CoA.This has revealed striking diVerence between the twoenzymes. Seabream AANAT1 acetylated all the testedsubstrates with similar velocities (Fig. 5 and Table 2).Seabream AANAT2, on the other hand, preferentiallyacetylated the indoleethylamines, serotonin, and trypt-amine, at high velocities and exhibited insigniWcantactivity in the presence of the phenylethylamines sub-strates, phenylethylamine and tyramine (Fig. 5 andTable 2). Moreover, sbAANAT2 acetylated serotonin10-fold faster than sbAANAT1: sbAANAT1 Vmax is

Table 2Comparison of Km and Vmax values of sbAANAT1 and sbAANAT2

sbAANAT Km and Vmax values were calculated after linearization of the activity values (depicted in Figs. 4 and 5) according to the Lineweaver–Burk plot (1/V vs 1/S). Values are the mean of two separate experiments. Km and Vmax values of sbAANAT2 for the phenylethylamine substratescould not be calculated because the low enzyme activity did not follow the Michaelis–Menten equation.

Substrate Km (mM) Vmax (nmol/h/ng protein)

sbAANAT1 sbAANAT2 sbAANAT1 sbAANAT2

Acetyl-CoA 0.62 1.11 0.49 0.93Serotonin 0.05 2.05 0.25 2.55Tryptamine 0.10 1.40 0.24 0.81Phenylethylamine 0.34 — 0.32 —Tyramine 1.25 — 0.30 —

B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147 145

0.25 nmol/h/ng; sbAANAT2 Vmax is 2.55 nmol/h/ng(Fig. 5 and Table 2). These results indicate high prefer-ence of AANAT2 for indoleethylamines, especially forserotonin.

Although sbAANAT1 activity is relativity slow, itsaYnity for its substrates, especially for serotonin is high(Table 2). The Km values of sbAANAT1 for tryptamineand serotonin are 0.1 and 0.05 mM, respectively, 14 and41 times lower than that of sbAANAT2 (1.4 and2.05 mM for tryptamine and serotonin, respectively).These diVerences in aYnity indicate the ability of sbA-ANAT1 to act at low substrate concentrations. Simi-larly, the Km values of sbAANAT1 for phenylethylamine(0.34 mM) and tyramine (1.25 mM) reXect relatively highaYnity also to the phenylethylamines substrates.

Another apparent kinetic diVerence is the inhibitionof sbAANAT1 by indoleethylamine substrates (seroto-

nin and tryptamine) at concentrations above 1 mM; suchinhibition was not observed with sbAANAT2. In orderto determine whether this is substrate or product inhibi-tion, enzyme assays were performed in the presence ofvarying concentrations of the product N-acetylserotonin(0.01–1.0 mM). While N-acetylserotonin had very slighteVect on sbAANAT2, it clearly inhibited sbAANAT1activity (Fig. 6A). This inhibition occurred at concentra-tions (IC50 D 0.68 mM) which resemble those that wereformed when high substrate concentrations were used.This indicates that the apparent reduction of sbA-ANAT1 activity (Fig. 5) reXects product inhibition. Sim-ilarly, melatonin clearly inhibited sbAANAT1 activity(IC50 D 0.26 mM) but had almost no eVect on sbA-ANAT2 activity (Fig. 6B). This may reXect an autoregu-latory mechanism of melatonin synthesis (Zheng andCole, 2002).

Fig. 5. DiVerential preference for amine substrates by sbAANATs. Activity was measured in optimal conditions for each enzyme (Table 1) in thepresence of saturating levels of acetyl-CoA (1 mM for sbAANAT1 and 2 mM for sbAANAT2 assays) and increasing substrate concentration (0.05–10 mM) of serotonin (closed boxes), tryptamine (closed triangles), phenylethylamine (open boxes), and tyramine (open triangles). Experiments wereperformed twice. Each graph represents one experiment.

Fig. 6. Product inhibition of sbAANAT activities. (A) EVect of increasing levels of N-acetylserotonin (0–1.0 mM) on sbAANAT activities. Activitieswere measured at optimal conditions (Table 1) in the presence of low (Km or lower) serotonin concentrations (0.05 mM for sbAANAT1; 0.5 mM forsbAANAT2 assays) and 1 mM acetyl-CoA. (B) EVect of increasing levels of melatonin (0–1.0 mM) on sbAANAT activities. Activities were measuredas described above in the presence of 0.25 mM tryptamine for sbAANAT1 and 0.5 mM tryptamine for sbAANAT2 assays. Experiments were per-formed twice; each graph represents one experiment.

146 B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147

4. Discussion

This study demonstrates the existence of two seab-ream AANATs with distinct spatial expression patternsand diVerent enzyme kinetics, substrate preferences, andtemperature optima. Previously, two functional AAN-ATs were demonstrated in three species, pike, trout, andzebraWsh, all of which represent evolutionarily ancientteleostean orders. Seabream is a member of an evolu-tionarily modern order—perciformae. Thus, the AANATduality is now extended to neoteleosti suggesting that itmay be a feature of all teleosts. This duality appears tobe the result of a genome duplication that occurred inthe Wsh lineage after its emergence from the vertebrateline (Amores et al., 1998; Wittbrodt et al., 1998) and pro-vided a unique opportunity for each AANAT to evolveseparately and attain its speciWc temporal and spatialexpression pattern, kinetic characteristics, regulation,and, possibly, specialized functions.

Like in the other Wsh species, sbAANAT1 is expressedonly in the retina while sbAANAT2 is preferentiallyexpressed in the pineal gland. Accordingly, the pattern ofexpression may reXect unique tissue-dependent require-ments for melatonin production in these tissues. Melato-nin is produced mainly in the pineal gland and, in lowerlevels, in the retina. The rate of melatonin production isthe reXection of AANAT activity and availability of thesubstrate. Accordingly, levels of serotonin are high in thepineal gland and low in the retina. The aYnity of eachenzyme for serotonin enables the enzymes to act and toproduce melatonin according to tissue requirements andin diVerent concentrations of the substrate: The low Kmof the retinal AANAT1 allows production of melatoninwhere concentration of serotonin is low while the higherKm of the pineal AANAT2 is more consistent with thehigh levels of serotonin present in this tissue (O’Brienand Klein, 1986). The velocities of the two enzymes alsoreXect the role of the end product, melatonin, in each tis-sue: in accord with the role of the pineal gland in pro-ducing circulating levels of melatonin, AANAT2 iscapable of acetylating serotonin about 10 times fasterthan AANAT1. Thus, the kinetics exhibited by sbA-ANAT1 and sbAANAT2 are in agreement with the par-acrine and endocrine functions of the retinal and pinealmelatonin, respectively.

As part of the assay optimization, a diVerent tempera-ture–activity relationship was described for each sbA-ANAT. Similar diVerences in optimal temperature existalso in pike and trout, suggesting that these diVerencesmay be related to a conserved physiological feature.Nonetheless, considering the life history of the giltheadseabream, this diVerence may be of physiological conse-quences in this species. This Wsh inhabits coastal lagoonsmost of the year. Towards the winter spawning season itmigrates out to the open sea where it spawns daily for 2–3 months. In the spring, the end of the spawning season,

adults migrate back to the coastal lagoons. Thus, thismigrating species is subjected to large temperature Xuc-tuations throughout the year, from 10 to 14 °C duringthe winter in the open sea, to over 30 °C during the sum-mer in coastal lagoon (Ben-Tuvia, 1979). Consequently,pineal sbAANAT2, which is active throughout the rangeof the environmental temperatures, transduces photope-riodic information in the form of circulating melatoninrhythms throughout the year. Retinal sbAANAT1, onthe other hand, which is temperature-sensitive, would beplaying a role in visual adaptation to darkness only athigh temperatures during the summer, when Wsh are sub-jected to strong day-light radiation in the shallowcoastal lagoon. Further analysis of the yearly variationsin circulating and retinal levels of melatonin and inpineal and retinal AANAT activity will be required inorder to corroborate this speculation.

A notable diVerence between the two enzymes is theirsubstrate preference. Pineal sbAANAT2, like other WshAANAT2s, preferentially acetylates indoleethylamines,especially serotonin, while retinal sbAANAT1, likemammalian and chicken AANATs, is capable of acety-lating a wide range of arylalkylamines. These diVerencesmay be attributed to structural diVerences in the sub-strate binding pocket.

Based on these kinetic diVerences, we propose that inaddition to melatonin production, retinal AANATscarry out a yet unknown function that involves acetyla-tion of arylalkylamines other than serotonin. This idea issupported by evolutionary considerations. First, the sub-strate preference displayed by each of the two enzymes isconserved among evolutionary distant Wsh species,which occupy diVerent habitats and display diVerent lifehistories, suggesting a physiologically important func-tional diVerence between the two enzymes. Second, theability of retinal AANATs to acetylate various aryl-alkylamines has been conserved throughout vertebratesand was described for a yeast ortholog of vertebrateAANATs (Ganguly et al., 2001). This putative ancestralenzyme acetylates a wide range of arylalkylamine sub-strates and has been suggested to be important for aminedetoxiWcation (Ganguly et al., 2001). An N-acetyltrans-ferase (involved in melatonin synthesis but is not anortholog of vertebrate AANAT) with a broad range ofsubstrates has evolved in parallel in insects (Amherdet al., 2000; Hintermann et al., 1996). Such evolutionaryconservation may suggest that broad arylalkylamineacetylation constitutes an important component for reti-nal function. In Wsh, genome duplication allowed oneenzyme, retinal AANAT1, to keep this broad ancestralactivity, and the other enzyme, pineal AANAT2, tobecome highly specialized in serotonin acetylation andmelatonin production. Evidently, this resulted in the twoenzymes serving two functions—pineal melatonin pro-duction and retinal production of melatonin and a yetunknown acetylated amine. In higher vertebrates, on the

B. Zilberman-Peled et al. / General and Comparative Endocrinology 138 (2004) 139–147 147

other hand, the sole AANAT had to keep its ancestralcharacteristics of broad substrate capabilities in order toserve these two functions.

Acknowledgments

The authors thank Dr. David C. Klein, NationalInstitutes of Health, for useful suggestions throughoutthe research. This research was supported by Grant No.232/00-17.2 from the Israel Science Foundation, Jerusa-lem, Israel.

References

Amherd, R., Hintermann, E., Walz, D., AVolter, M., Meyer, U.A., 2000.PuriWcation, cloning, and characterization of a second arylalkyl-amine N-acetyltransferase from Drosophila melanogaster. DNACell Biol. 19, 697–705.

Amores, A., Force, A., Yan, Y.L., Joly, L., Amemiya, C., Fritz, A., Ho,R.K., Langeland, J., Prince, V., Wang, Y.L., WesterWeld, M., Ekker,M., Postlethwait, J.H., 1998. ZebraWsh hox clusters and vertebrategenome evolution. Science 282, 1711–1714.

Arendt, J., 1995. Melatonin and the Mammalian Pineal Gland. Chap-man & Hall, London, UK.

Begay, V., Falcon, J., Cahill, G.M., Klein, D.C., Coon, S.L., 1998. Tran-scripts encoding two melatonin synthesis enzymes in the teleostpineal organ: circadian regulation in pike and zebraWsh, but not introut. Endocrinology 139, 905–912.

Ben-Tuvia, A., 1979. Studies on the population and Wsheries of Sparusaurata in the Bardawil Lagoon, eastern Mediterranean. Invest.Pesq. 43, 43–67.

Benyassi, A., Schwartz, C., Coon, S.L., Klein, D.C., Falcon, J., 2000.Melatonin synthesis: arylalkylamine N-acetyltransferases in troutretina and pineal organ are diVerent. Neuroreport 27, 255–258.

Cahill, G.M., Besharse, J.C., 1995. Circadian rhythmicity in vertebrateretinas: regulation by photoreceptor oscillator. Prog. Retin. EyeRes. 14, 267–291.

Coon, S.L., Begay, V., Deurloo, D., Falcon, J., Klein, D.C., 1999. Twoarylalkylamine N-acetyltransferase genes mediate melatonin syn-thesis in Wsh. J. Biol. Chem. 274, 9076–9082.

De angelis, D.J., Gastel, J., Klein, D.C., Cole, P.A., 1998. Kinetic analy-sis of the catalytic mechanism of serotonin N-acetyltransferase. J.Biol. Chem. 273, 3045–3050.

Falcon, J., Galarneau, K.M., Weller, J.L., Ron, B., Chen, G., Coon,S.L., Klein, D.C., 2001. Regulation of arylalkylamine N-acetyl-transferase-2 (AANAT2, EC 2.3.1.87) in the Wsh pineal organ: evi-dence for a role of proteasomal proteolysis. Endocrinology 142,1804–1813.

Falcon, J., Gothilf, Y., Coon, S.L., Boeuf, G., Klein, D.C., 2003.Genetic, temporal and developmental diVerences between melato-nin rhythm generating systems in the teleost Wsh pineal organ andretina. J. Neuroendocrinol. 15, 378–382.

Foulkes, N.S., Whitmore, D., Sassone-Corsi, P., 1997. Rhythmic tran-scription: the molecular basis of circadian melatonin synthesis.Biol. Cell 89, 487–494.

Ganguly, S., Coon, S.L., Klein, D.C., 2002. Control of melatonin syn-thesis in the mammalian pineal gland: the critical role of serotoninacetylation. Cell Tissue Res. 309, 127–137.

Ganguly, S., Mummaneni, P., Steinbach, P.J., Klein, D.C., Coon, S.L.,2001. Characterization of the Saccharomyces cerevisiae homolog ofthe melatonin rhythm enzyme arylalkylamine N-acetyltransferase.J. Biol. Chem. 276, 47239–47247.

Gastel, J.A., Roseboom, P.H., Rinaldi, P.A., Weller, J.L., Klein, D.C.,1998. Melatonin production: proteasomal proteolysis in serotoninN-acetyltransferase regulation. Science 279, 1358–1360.

Gothilf, Y., Coon, S.L., Toyama, R., Chitnis, A., Namboodiri, M.A.A.,Klein, D.C., 1999. ZebraWsh serotonin N-acetyltransferase-2:marker for development of pineal photoreceptors and circadianclock function. Endocrinology 140, 4895–4903.

Hickman, A.B., Namboodiri, M.A., Klein, D.C., Dyda, F., 1999. Thestructural basis of ordered substrate binding by serotonin N-acetyl-transferase: enzyme complex at 1.8 Å resolution with a bisubstrateanalog. Cell 97, 361–369.

Hintermann, E., Grieder, N.C., Amherd, R., Brodbeck, D., Meyer,U.A., 1996. Cloning of an arylalkylamine N-acetyltransferase(aaNAT1) from Drosophila melanogaster expressed in the nervoussystem and the gut. Proc. Natl. Acad. Sci. USA 93, 12315–12320.

Klein, D.C., Baler, R., Roseboom, P.H., Weller, J.L., Bernard, M.,Gastel, J.A., Zatz, M., Iuvone, P.M., Begay, V., Falcon, J., Cahill, G.,Cassone, V.M., Coon, S.L., 1999. The molecular basis of the pinealmelatonin rhythm: regulation of the serotonin N-acetylation. In:Lydic, R., Baghdoyan, H. (Eds.), Handbook of Behavioral StateControl: Cellular and Molecular Mechanisms. CRC Press, BocaRaton, pp. 45–59.

Klein, D.C., Coon, S.L., Roseboom, P.H., Weller, J.L., Bernard, M., Gastel,J.A., Zatz, M., Iuvone, P.M., Rodriguez, I.R., Begay, V., Falcon, J.,Cahill, G.M., Cassone, V.M., Baler, R., 1997. The melatonin rhythm-generating enzyme: molecular regulation of serotonin N-acetyltrans-ferase in the pineal gland. Recent Prog. Horm. Res. 52, 307–357.

Li, X., Borjigin, J., Snyder, S.H., 1998. Molecular rhythms in the pinealgland. Curr. Opin. Neurobiol. 8, 648–651.

Mizusawa, K., Iigo, M., Masuda, T., Aida, K., 2000. Photic regulationof arylalkylamine N-acetyltransferase 1 mRNA in trout retina.Neuroreport 11, 3473–3477.

Mizusawa, K., Iigo, M., Suetake, H., Yoshiura, Y., Gen, K., Kikuchi,K., Okano, T., Fukada, Y., Aida, K., 1998. Molecular cloning andcharacterization of a cDNA encoding the retinal arylalkylamine N-acetyltransferase of the rainbow trout, Oncorhynchus mykiss. Zool.Sci. 15, 345–351.

O’Brien, P.J., Klein, D.C., 1986. Pineal and Retinal Relationships. Aca-demic Press, Orlando, FL.

Riddles, P.W., Blakeley, R.L., Zerner, B., 1983. Reassessment of Ell-man’s reagent. Methods Enzymol. 91, 49–60.

Tosini, G., Fukuhara, C., 2003. Photic and circadian regulation of reti-nal melatonin in mammals. J. Neuroendocrinol. 15, 364–369.

Wittbrodt, J., Meyer, A., Schartl, M., 1998. More genes in Wsh? BioEs-says 20, 511–515.

Zheng, W., Cole, P.A., 2002. Serotonin N-acetyltransferase: mechanismand inhibition. Curr. Med. Chem. 9, 1187–1199.