Molecular characterization of monoamine oxidase in zebrafish (Danio rerio)

Comparative Biochemistry and Physiology, Part B 140 (2005) 153 – 161 www.elsevier.com/locate/cbpb

Molecular characterization of monoamine oxidase in zebrafish (Danio rerio) Andrea Setini*, Federica Pierucci, Ornella Senatori, Antonietta Nicotra Department of Animal and Human Biology, University of Rome I, Viale dell’Universita`, 32, Rome 00185, Italy Received 24 July 2004; received in revised form 30 September 2004; accepted 2 October 2004

Abstract Monoamine oxidase (MAO) is responsible for the degradation of a number of neurotransmitters and other biogenic amines. In terrestrial vertebrates, two forms of the enzyme, named MAO A and B, were found in which mammals are coded by two similar but distinct genes. In teleosts, the biochemical data obtained so far indicate that enzyme activity is due to a single form, whose sequence, obtained for trout, displays 70% identity with mammal MAO A and B. In this paper, we carried out an investigation of zebrafish MAO (Z-MAO) to shed further light on the nature of the MAO form present in aquatic vertebrates. Sequencing studies have revealed an open reading frame 522-amino-acids long with MW 58.7 kDa, displaying 84% identity with trout MAO and about 70% identity with mammal MAO A and MAO B. Analysis of the sequence and of the predicted secondary structure shows that also in Z-MAO principal domains characterizing the MAOs are present. The domain linking the FAD is very well conserved, while the transmembrane domain sequence linking the enzyme to the external mitochondrial membrane does not appear to be conserved even with respect to trout MAO. Comparison with the amino acids which, according to the human MAO B and rat MAO A models, line the substrate-binding site shows that in Z-MAO, several residues (V172, N173, F200, L327) differ from MAO B but are similar or identical to the corresponding ones present in rat MAO A, as well as in trout MAO. A three-dimensional model is reported of the substrate-binding site of Z-MAO obtained by comparative modeling. Our observations support the hypothesis that the MAO form present in aquatic vertebrates is a MAO A-like form. Experiments performed to test the effect of selective MAO A (clorgyline) and MAO B (deprenyl) inhibitors on the enzyme’s activity in liver and brain confirm the presence of a single form of MAO in zebrafish. D 2004 Elsevier Inc. All rights reserved. Keywords: Monoamine oxidase (MAO); Zebrafish; Sequence comparison; Three-dimensional modeling; Activity; Clorgyline; Deprenyl

1. Introduction Monoamine oxidase (MAO, EC, 1.4.3.4.) is a flavoprotein of the outer mitochondrial membrane that is involved in the degradation of a variety of biogenic and xenobiotic amines. MAO plays an important physiological role in the nervous system by regulating the levels of classical neurotransmitters like serotonin, dopamine, and noradrenalin and inactivating trace amines able to act as false neurotransmitters. In peripheral organs, the enzyme deaminates monoamines with endocrine and paracrine effects or, * Corresponding author. Tel.: +39 6 49914772; fax: +39 6 4958259. E-mail address: [email protected] (A. Setini). 1096-4959/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.cbpc.2004.10.002

as in liver and intestine, it functions as a detoxifying enzyme, preventing the accumulation of vasoactive dietary amines up to toxic levels (Youdim et al., 1988; Berry et al., 1994). Based on substrate and inhibitor selectivity, two functional forms of the enzyme, namely, MAO A and MAO B, were first identified in mammals (Fowler and Ross, 1984; Kinemuchi et al., 1984; Youdim and Finberg, 1991). A different sensitivity to a specific inhibitor is associated with a different affinity for given substrates and thus a different physiological function. MAO A preferentially deaminates serotonin and noradrenalin, and its activity is inhibited by nanomolar concentrations of clorgyline. MAO B is highly sensitive to inhibition by deprenyl, with its preferred

154

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

substrates being h-phenylethylamine (PEA) and benzylamine. Both forms usually catalyze dopamine, tryptamine, and tyramine deamination (Sullivan et al., 1986; Youdim et al., 1988). By cloning MAO cDNA, it was subsequently demonstrated that MAO A and MAO B are two distinct proteins with a high degree of sequence identity, encoded by different genes having identical exon–intron organization (Shih et al., 1999). Chimeric enzymes and site-directed mutagenesis studies have attempted to elucidate the mechanisms of the different substrates and inhibitor recognition by MAO A and MAO B (Tsugeno and Ito, 1997; Cesura et al., 1998; Shih et al., 1998; Veselovsky et al., 1998; Geha et al., 2000). An outstanding contribution in this direction was provided by studies on the secondary and tertiary structure of the two enzymes (Wouters, 1998; Binda et al., 2002; Binda et al., 2003; Ma et al., 2004; Veselovsky et al., 2004). In contrast to the large number of data reported for MAO in mammals, there have been relatively few studies to date on MAO in non mammalian vertebrates. The occurrence of MAO A and MAO B activities was demonstrated in other terrestrial vertebrates like reptiles and birds (Hall and Uruen˜a, 1983a; Pintar et al., 1983). Different situations were found in the case of amphibians. In Urodela, which are better adapted to an aquatic life style, MAO type A predominates (Hall and Uruen˜a, 1983b). In terrestrial anurans, both MAO A and MAO B were detected (Kobayashi et al., 1981; Hall and Uruen˜a 1983a; Nicotra and Senatori, 1988). Interestingly, however, in toad and bullfrog tadpoles, during premetamorphic stages, MAO A activity predominates, while MAO B increases rapidly only during metamorphosis. These findings suggest that MAO B could be interpreted as an adaptive physiological response to a terrestrial life style. In this context, papers devoted to the characterization of monoamine oxidase in teleosts are of considerable interest. Substrate affinity and inhibitor sensitivity indicated the occurrence of only one enzyme form in this taxon, which in some species has been considered as an MAO A-like form, and in others as a novel type of MAO distinct from both MAO A and MAO B (Hall and Uruen˜a, 1983a; Yoshino et al., 1984; Nicotra and Senatori, 1989; Kumazawa et al., 1998; Salles et al., 2001). In rainbow trout, the presence of one MAO form was confirmed by cloning a liver MAO cDNA (Chen et al., 1994). The deduced amino acid sequence showed 70% and 71% identity with those of H- and R-MAO A and human Hand rat R-MAO B, respectively. To provide more information on the structure–function evolution of this important enzyme in vertebrates, we investigated the nature of MAO in the freshwater cyprinid zebrafish (Danio rerio), a suitable model system for different biological areas including regulatory physiology (Briggs, 2002). To this end, the zebrafish MAO (Z-MAO) nucleotidic sequence was obtained by reverse transcriptasepolymerase chain reaction (RT-PCR) using total RNA from liver, and the deduced amino acid sequence was compared

with MAO sequences from trout and mammals. Data on the secondary structure predictions and on computer modeling of the active site, based on the recently described threedimensional structure of human MAO B (Binda et al., 2002) and rat MAO A (Ma et al., 2004), are also reported. In a second series of experiments, we assayed the effects of selective inhibitors for MAO A and MAO B on enzyme activity from zebrafish liver and brain homogenates.

2. Material and methods 2.1. Animals and tissue preparation Adult zebrafish (D. rerio) were purchased from a commercial supplier and maintained under standard conditions at 28 8C. Samples about 3.5 cm long were anaesthetized in 0.04% tricaine methane sulfonate (MS222), and after dissection, liver and brain were quickly weighed and stored at 70 8C. 2.2. Sequencing strategy Sequencing was carried out on RT-PCR fragments. Oligonucleotide primers (Table 1) were designed on the basis of the conserved sequences of trout MAO (Chen et al., 1994), and on two sequences both encoding a putative zebrafish amine oxidase (GenBank sequences, EST Data Base, accession numbersCD760288 and CD601387). The latter were used to design primers in the relatively nonconserved regions flanking the 5V and 3V of the open reading frame. The amplified fragments obtained were then sequenced directly. Total RNA was extracted from zebrafish liver by TriPure Isolation Reagent according to the manufacturer’s protocol (Roche Diagnostics, Germany). RT-PCR was performed in a single-step protocol using the Ready to go RT-PCR Beads kit (Amersham Biosciences, Piscataway, NJ). To 50 AL of reaction mix containing RT Moloney Murine Leukemia

Table 1 Sense and antisense oligonucleotide primers used for RT-PCR reactions Sequence 5V GCCACTGAACAGACACGGAC 3V 5VAGGGCTGAGTGCAGCTAAGC 3V 5V GAAGTCCTACCCATTCAAAGG 3V 5V GAGGTTCCAGCCAGATCAGTG 3V 5V ACAGTTTGGCAGGGTCTTG 3V 5V CCTTTGAATGGGTATGACTTTC 3V 5V TTACGAACAGAGTGGCAAAG 3V 5V CCTCCAGAATACTCCTCCTCAC 3V 5V AGAAGGTGGTGACAAACGGGAG 3V 5V CTGATTAAGTGATTGGATGC 3V a b c d

Position is referred to CD760288. Position is referred to Chen et al. (1994). From our cDNA Z-MAO preliminary fragments. Position is referred to CD601387.

Position 62–81a 150–169b 387–407b 739–760b c

375–396a 602–583a 1291–1271b 1549–1528b 250–231d

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

Virus, Taq polymerase, dNTPs, ribonuclease inhibitors, BSA RNase/DNase-free, 1 Ag of total RNA and 100 pmol of each specific primer were added. RT was performed at 42 8C for 40 min. Thirty-five PCR cycles were performed as follows: denaturation (95 8C1 min), annealing (56 8C1 min), extension (72 8C1 min) followed by one cycle of extension at 72 8C for 10 min. The quality of PCR products was visualized by agarose gel (1.2%) with 0.005% (W/V) ethidium bromide, and the products with the expected length were purified by GFX PCR DNA purification kit (Amersham Biosciences). Both strands of the RT-PCR products were sequenced directly by the Dye Terminator Perkin Elmer automated method. For cDNA synthesis, we used a MAO specific primer also employed as antisense primer for PCR amplification. Primers (Table 1) designed by GeneRunner software (Hasting software) were custom synthesized by MWG Biotech. 2.3. Determination of MAO activity Tissues were homogenized in 0.1 M phosphate buffer pH 7.4 (PB), and after a short centrifugation at 3000g, enzyme activity against serotonin (5-hydroxytryptamine,5HT) and h-phenylethylamine (PEA) was measured by sensitive nonisotopic spectrophotometric assay (Ko¨chli and von Wartburg, 1978). This method allows the H2O2 produced by MAO activity to be estimated by recording the peroxidase-catalyzed H2O2-dependent oxidation of 2V,7Vdichlorofluorescein at 502 nm. In total volume of 1 mL of PB 1 mg of homogenate was preincubated with horseradish peroxidase 80 Ag/mL (HRP type II, Sigma) and 2V,7V-dichlorodihydrofluorescein diacetate 2.510 4 M (DCFH-DA from Sigma, dissolved before use in 0.01 N NaOH ) for 5 min before the substrate (100 AM) was added. The inhibitory effect of clorgyline and deprenyl (RBI) was analyzed by incubating the homogenates for 30 min with the inhibitor over a concentration range of 10 9M–10 4 M before HRP and DCFH-DA addition. Incubations were carried out at 37 8C in normal air under constant stirring. The reaction mixture also contained sodium azide (3 mM) as an inhibitor of endogenous catalases and semicarbazide (1 mM) to inhibit semicarbazide-sensitive amine oxidases. Ten minutes after the addition of the substrate, the increase in absorbance due to DCF generation was measured at 502 nm. According to Ko¨chli and von Wartburg (1978), and as evaluated by our previous measurements, the H2O2 produced was measured using a molar extinction coefficient for DCF equal to 91,000 M 1 cm 1, and a reaction stoichiometry of 5.3 mol of DCF formed from each mole of H2O2. All assays were performed in duplicate, and blank values were obtained using heat-inactivated homogenate. In preliminary experiments, it was ascertained that under the experimental conditions adopted, enzyme activity was linearly proportional to incubation time (5–20 min) and milligram of homogenate (0.5–2 mg/mL).

155

3. Results 3.1. Sequence analysis of zebrafish MAO (Z-MAO) and comparison with other MAOs Most combinations between sense and antisense primers shown in Table 1 produced single fragments of the expected size in RT-PCR reactions performed using liver total RNA. The reliability of the sequence was based on the sequencing of both strands for each fragment obtained by forward and reverse primers and on the overlapping of different PCR products so as to perform sequencing at least four times for the entire length of the open reading frame. The sequencing strategy adopted allowed us to rule out that more than one form of MAO may be expressed. In fact, in the overlapping sequences among the fragments we obtained, there were no bhot spotsQ indicating uncertainties in the sequence due to the overlapping of different nucleotides present in different forms of MAO. A 1616-bp nucleotide sequence was obtained (Fig. 1) containing an ATG start codon at position 10 and a TAA stop codon at position 1578, the open reading frame encompassing 1569 bp (GenBank, accession number: AY185211). When the Z-MAO nucleotide sequence was matched (GeneRunner software) with the nucleotide sequences of trout MAO (T-MAO), human MAO A (HMAO A), rat MAO A (R-MAO A), human MAO B (HMAO B), and rat MAO B (R-MAO B), the identities were 78%, 66%, 66%, 68%, and 68%, respectively. Fig. 1 also shows the Z-MAO deduced amino acid sequence. The open reading frame encoded a 522-amino acid protein with a molecular mass of about 58.7 kDa. The deduced amino acid sequences of Z-MAO and of MAOs from other species are aligned in Fig. 2. As expected, the amino acid identity in most cases was slightly higher than the nucleotide identity. In fact, Z-MAO displayed 84%, 68%, 70%, 68%, 69% identity to T-MAO, H-MAO A, RMAO A, H-MAO B, and R-MAO B, respectively. Identities between Z-MAO versus MAO A and MAO B were similar to those shown by T-MAO with respect to the same MAO forms (about 70% and 71%, respectively). Two highly conserved regions, described in all the known MAO sequences, corresponding to a dinucleotide-binding region and to a region containing the FAD-binding pentapeptide, appeared to be highly conserved also in Z-MAO, where they were located at the positions 7–44 and 381–452, respectively. Another region that appeared to be conserved, in all the MAOs that we considered, was the 168–221 one. The reason for this amino acidic sequence similarity and function of the latter region remains to be accounted for. Fig. 2 shows the essential amino acids that line the substrate-binding site of H-MAO B and R-MAO A in accordance with Binda et al. (2002) and Ma et al. (2004). From the matching of MAOs, we observed that only residues Y60, F168, Y188, I198, Q206, F343, Y398, and Y435 of H-MAO B and R-MAO B coincided with those of

156

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

Fig. 1. Nucleotide and deduced amino acid sequences of zebrafish MAO cDNA. Start codon and stop codon are boxed. The numbers of positions are indicated for each line.

fish-MAO (Z-MAO, T-MAO), H-MAO A, and R-MAO A. Leucine 171 instead was replaced by valine in fishMAO and by isoleucine in MAO A. These differences are probably not significant, because the three amino acids have similar properties. Non polar cysteine at position 172 and the non aromatic residue isoleucine at position 199 in H-MAO B were replaced in both fish-MAO and MAO A by a polar, uncharged asparagine and by the aromatic amino acid phenylalanine, respectively. Position 326 of H-MAO B is represented by tyrosine, while leucine and isoleucine are present in the corresponding position in fish-MAO and in MAO A, respectively. Threonine in position 314 of MAO B is replaced by a cysteine 323 in MAO A and serine 315 in Z-MAO. Therefore, of the 13 amino acids lining the substratebinding site, 8 were shared by all the MAOs examined, 1 retained similar biochemical features (L171), 3 (C172, I199, Y326) were represented by similar or identical amino acids in fish-MAO and MAO A, and 1 (T314) is different in Z-MAO and MAO A. From the analysis of the deduced amino acid sequence, we propose that the region included between 493 and 510

residues is a transmembrane region (underlined in Fig. 2). The transmembrane helices were predicted by a system of neural networks that worked by integrating a multiple alignment of membrane proteins. The final prediction has an expected per-residue accuracy of about 95% (Rost et al., 1995, Rost, 1996). The transmembrane a-helix, 18 amino acid long, may represent a putative membrane-binding domain anchoring the enzyme to the outer mitochondrial membrane. The sequence of this domain does not seem to be conserved between zebrafish and trout (residues 493– 513; Swiss Protein P49253: Bairoch 1998, unpublished observations), neither of which is conserved with respect to mammal MAOs. In mammals, the C-terminus transmembrane a-helix sequence is conserved among the MAO A or MAO B from different species, but is not conserved between the two forms. 3.2. Secondary structure prediction of Z-MAO and threedimensional modeling of the substrate-binding site For a better characterization and fuller understanding of the role of structural motifs present in the protein, a

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

157

Fig. 2. Amino acid sequence alignment and comparison of zebrafish MAO (Z-MAO, AY185211) with trout MAO (T-MAO, P49253), human MAO A (HMAOA, M68840) and B (H-MAOB, M69177), rat MAO A (R-MAOA, D00688) and B (R-MAOB, M23601). Accession numbers are from GenBank Data Base, except T-MAO which is from Swiss Prot. Boxed amino acids correspond to two highly conserved regions involved in FAD binding. The FAD-binding pentapeptide is bold-faced. The amino acids lining the substrate-binding site in H-MAO B are marked with an asterisk. The C-terminal putative transmembrane domain of Z-MAO is underlined.

secondary structure prediction was obtained using an algorithm by PROF server (Ouali and King, 2000). The expected average accuracy of the subset PROFsec prediction is N82% for all the residues. The comparison of different types of MAO revealed a high similarity among all the MAOs considered, and Z-MAO did not display any

selective similarity with either MAO A or MAO B (result not shown). An interesting region was located between residues 98 and 111 in Z-MAO. This region was compatible with a loop described for the corresponding region 99–112 in H-MAO B (see below). In addition, the prediction of the secondary structure confirmed that the N-terminus segment

Fig. 3. Stereoview of the model of Z-MAO substrate-binding site obtained through the computerized integration of the Z-MAO sequence with the known threedimensional models of H-MAO B and R-MAO A. The model illustrates the three-dimensional spatial distribution of the essential amino acid residues lining the substrate-binding site. Residues other than from H-MAO B are underlined.

158

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

more interesting to consider in particular the architecture of the substrate-binding site (Fig. 3). The model illustrated the three-dimensional distribution of the essential amino acid residues lining the substrate-binding site, mentioned when comparing the primary structure of MAOs of different species. The groove where the substrate is probably housed is located between flavin ring and V172, N173, L327, F200, Q207 residues. 3.3. Effects of selective inhibitors on MAO activity Fig. 4. MAO-specific activities (nmol/mg tissue/min) measured in liver and brain homogenates. MAO activity was measured using 5-HT and PEA as substrates (100 AM). Values are the meansFS.E.M. of three duplicate determinations.

was folded in a h-a-h substructure as the dinucleotidebinding domain region (Rossmann fold) conserved in all MAOs. Proteins can be superimposed to deduce structural alignments and compare their tertiary structure and in particular the active sites. We thus exploited the comparative modeling method to build a three-dimensional model of Z-MAO. We used as template the known structures of H-MAO B and R-MAO A (1GOS and 1o5wA, respectively, Swiss Protein Data Bank) and as software the Deep View Swiss-PdbViewer. The overall tertiary structure of Z-MAO has not been reported since, given the high sequence identity, it appeared very similar to the other MAOs used for modeling. It was instead

Under the adopted experimental conditions in liver and brain, 5-HT appeared more actively deaminated than PEA (Fig. 4). Enzyme activity against 5-HT was slightly higher in liver than in brain, and the opposite was true for activity against PEA, but in both cases the differences were not significant. The inhibition dose–response curves obtained using clorgyline (a selective inhibitor of MAO A activity) and deprenyl (a selective inhibitor of MAO B activity) are reported in Fig. 5. The inhibition curves obtained were simple sigmoidal ones, and both inhibitors were able to completely block MAO activity when present in relatively high concentrations (10 5–10 4 M). In liver, no evident selectivity was observed with respect to the two inhibitors regardless of the substrate used. These findings were confirmed by IC50 values (drug concentration necessary to give 50% enzyme inhibition) reported in Table 2. In brain, MAO activity against 5-HT was equally sensitive to the inhibitors, while MAO activity against PEA appeared

Fig. 5. Inhibition of liver and brain MAO activity by different concentrations of clorgyline (solid line) and deprenyl (dashed line). 5-HT and PEA were used as substrates (100 AM). The percentages of MAO activity inhibition of the control samples are plotted as a function of the negative log of molar concentrations of inhibitor. Values are the meansFS.E.M. of three duplicate determinations.

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

159

Table 2 IC50 values calculated by nonlinear regression analysis of data presented in Fig. 5 Liver

Brain

[Clorgyline] (M) 5-HT PEA

8.10F0.5910 4.96F2.5610

7 7

[Deprenyl] (M) 2.80F1.7110 6.22F1.5810

slightly more sensitive to clorgyline, the IC50 values being 3.45F0.4910 7 and 1.54F0.2210 6, respectively.

4. Discussion All MAOs sequenced so far, including Z-MAO, display three domains: a flavin-binding domain, a substrate-binding domain, and a membrane-binding domain anchoring the enzyme to the outer mitochondrial membrane. The amino acid sequence we obtained for Z-MAO displays a high degree (84%) of identity with T-MAO and like the latter around 70% identity with human and rat MAO A and MAO B. MAO A and MAO B have been sequenced in various mammal species (Shih et al., 1999; Abell and Kwan, 2001). In Fig. 2, we have limited the comparison to human and rat MAOs as they are the ones most frequently used in genetic manipulation studies. Furthermore, for human MAO B and rat MAO A, there exists a three-dimensional model based on the crystallographic structure (Binda et al., 2002; Ma et al., 2004). Thanks to these more recent studies, today we can not only compare the overall sequence identity, but also check whether the amino acids with key roles in substrate and inhibitor preference are conserved. The overall folding structure of H-MAO B and R-MAO A appears to be very similar, although differences are reported in the substratebinding site organization. According to the H-MAO B model (Binda et al., 2002), the substrate-binding domain appears to be divided into two chambers: a smaller bentrance cavityQ that functions as a passage way for diffusion of the substrate into the bsubstrate cavityQ. The access to the entrance cavity involves the movement of loop 99–112. The diffusion of the substrate into the active site is then enabled by the transient movement of a gate consisting of the four residues separating the two cavities (Y326, I199, L171, F168) and that form part of the residues lining the substrate-binding site as reported in the Results section. The structure of R-MAO A substrate-binding site (Ma et al., 2004) does not reveal any definite entrance pathway for the substrate, and it has been suggested that the substrate passage into the active site might require a conformational change of the molecule. The study of the deduced amino acid sequence and secondary structure prediction confirmed that also in ZMAO, the flavin-binding domain is highly conserved in both its two parts located at the carboxyl-terminus and at the amino-terminus regions. Besides, the study of the predicted secondary structure of Z-MAO has shown that Z-MAO has

[Clorgyline] (M) 6 7

4.76F0.5110 3.45F0.4910

7 7

[Deprenyl] (M) 8.03F0.4510 1.54F0.2210

7 6

a structure compatible with a loop (residues 98–111) corresponding to the loop 99–112 of H-MAO B. A comparison of the amino acids lining the substratebinding site of H-MAO B and R-MAO A shows that four of these (L171/I180, C172/N181, I199/F208, Y326/I335) in fish-MAO are different from H-MAO B but similar or identical to those of R-MAO A and H-MAO A. It must be emphasized that in H-MAO B, these amino acids, except for C172, form part of the gate separating the two chambers. As has already been pointed out, the residue L171 is represented in corresponding position in fish-MAO and MAO A by amino acids with similar properties. However, for residues I199 and Y326 in H-MAO B or F208 and I 335 in R-MAO A, several considerations must be made. In particular, residue I199/F208 is deemed to be a key residue as it is responsible for important differences between MAO A and MAO B. The fact that unlike H-MAO B in the corresponding position of MAO A, as well as also in fishMAO, there is a less hydrophobic phenylalanine, could account for the different biochemical behaviour of the various forms of MAO. In fact, the specific inhibitor for MAO B,1,4-diphenyl-2-butene is excluded form the MAO A pocket by the F208 residue (Ma et al., 2004). This is also supported by the fact that in rat, reciprocally switching I199 in MAO B with F208 in MAO A was sufficient to switch their substrate and inhibitor preferences (Tsugeno and Ito, 1997). However, the same is not true for H-MAOs (Geha et al., 2000), where substrate and inhibitor preference are not determined by the above mentioned residue. Moreover, also bovine MAO B displays a phenylalanine in position 199 (Abell and Kwan, 2001). This kind of substitution within the MAO B group could, among other things, account for the different efficacy of some inhibitors observed in MAO B from different mammal species (Krueger et al., 1995). It should nevertheless be borne in mind that in the R-MAOs, other differences compared with the H-MAOs are also present. For example, there are additional cysteine residues capable of generating disulphide bridges and proline residues which induce a turn. Therefore, in addition to the residues lining the substrate-binding site, also other residues could be important in the molecule’s conformation. A significant change could also be represented by the occurrence of a small non polar leucine at position 327 in ZMAO instead of the more bulky and polar tyrosine 326 in MAO B. Leucine-like characteristics are displayed by the isoleucine present in the corresponding position (I335) in MAO A. Geha et al. (2002) have demonstrated that switching I335 and Y326 in H-MAO A and B, respectively,

160

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161

allows the substrate and inhibitor selectivity to be switched in MAO A and B. Further confirmation of the importance of this residue also comes from the evidence of Ma et al. (2004), who suggests that Y326 of MAO B but not I335 of MAO A possibly interferes with the 5 HT bond with MAO. A particular influence could be exerted on the affinity for substrate and the sensitivity to inhibitors by residue C172 of MAO B, as it is located in a region of the substrate cavity that is capable of interacting with the benzene ring present in many substrates. Both in fish-MAO and MAO A, this residue is represented by an asparagine which, as it is polar, could favour the affinity for substrates bearing OH groups linked to the benzene ring, such as serotonin, noradrenalin, and other MAO A-preferred substrates. The occurrence of a putative membrane-binding domain in Z-MAO confirms that also fish-MAO is associated with the mitochondrial fraction (Kinemuchi et al., 1983). As reported above, the sequence of this domain is not conserved with respect to the one occurring in other MAOs, including trout MAO. In H-MAO B, it has been suggested that membrane attachment requires as many as four additional hydrophobic side chains (F481, L482, L486, P487) in the elongated polypeptide stretch 481–488 preceding the Cterminal helix. Unlike the C-terminal transmembrane segment, these residues are conserved in all the MAOs, including Z-MAO, except for L 482 which is replaced by tryptophan in fish-MAO and in H-MAO A. The function of the conserved region corresponding in zebrafish to residues 168–221 is not known, as we have reported. Based on the position of this region in a threedimensional model, it may be postulated that some of the residues contained in it are involved in the interactions with the mitochondrial membrane. This region could also prove indispensable for the dimerization of the MAO molecules. Biochemical studies to characterize enzyme activity have been described in various different bony fish species. Using selective inhibitors for MAO A and MAO B activity like clorgyline and deprenyl as diagnostic tools, in this taxon the presence of only one form of MAO has been demonstrated. In some cases, for instance, as in trout (Chen et al., 1994), catfish (Kumazawa et al., 1998), and pacu (Salles et al., 2001), MAO activity displayed greater sensitivity to MAO A inhibitors, although lower than that normally observed for mammal MAO A. However, in other cases, as in carp (Kinemuchi et al., 1983; Nicotra and Senatori, 1989) and in pike (Senatori et al., 1990), MAO displayed the same sensitivity to the two inhibitors. In carp, a cyprinid like zebrafish, this was further confirmed also using other selective and potent MAO A and MAO B inhibitors, such as FLA 788(+), FLA 336(+), and MD 780236 (Yoshino et al., 1984). Similar studies on zebrafish reported in this paper indicate that also in this species, both in liver and brain, MAO activity is due to a single form and argue against the possibility that it can be ascribed to MAO A or MAO B. The presence of a single form of enzyme is indicated by the pattern in the inhibition curves. The finding that in brain,

enzyme activity against PEA is slightly more sensitive to clorgyline (Fig. 4) could be the result of a different membrane environment. The mitochondrial membrane is supposed to play a role in enzyme activity (Binda et al., 2004; Ma et al., 2004). The biochemical behaviour of the MAO in zebrafish is not exactly the same even as that reported for trout (Chen et al., 1994). It must be assumed that this is differently affected by membrane environment or by the configuration taken on by the enzyme because of a few residues external to the active site and different in the various MAOs. In conclusion, currently available evidence indicates that, in fish, there is a single MAO form more similar to MAO A than to MAO B. Z-MAO seems to display both a substratebinding site that is almost identical to MAO A (Fig. 3) and a secondary structure in the portion 98–111 similar to MAO B. Taking into account the fact that the specificity seems to be given by residues 200 and 327, and that the conformational changes typical of MAO A are possible in Z-MAO, the latter should be considered more similar to MAO A than to MAO B. These findings seem to support the hypothesis that the common ancestor of the two forms existing in the terrestrial vertebrates could be a MAO A-like form typical of aquatic vertebrates. Further comparative studies on the molecular characteristics of MAO in a larger number of vertebrate species are thus deemed necessary to support this view and to better elucidate MAO changes occurred during evolution.

Acknowledgment This research was funded by Italian MURST grants.

References Abell, C.W., Kwan, S.W., 2001. Molecular characterization of monoamine oxidases A and B. Prog. Nucleic Acid Res. Mol. Biol. 65, 129 – 156. Berry, M.D., Juorio, A.V., Paterson, I.A., 1994. The functional role of monoamine oxidase A and B in the mammalian central nervous system. Prog. Neurobiol. 42, 375 – 391. Binda, C., Newton-Vinson, P., Hubalek, F., Edmondson, D.E., Mattevi, A., 2002. Structure of human monoamine oxidase B a drug target for the treatment of neurological disorders. Nat. Struct. Biol. 9, 22 – 26. Binda, C., Li, M., Hubalek, F., Restelli, N., Edmondson, D.E., Mattevi, A., 2003. Insights into the mode of inhibition of human mitochondrial monoamine oxidase B from high-resolution crystal structures. Proc. Natl. Acad. Sci. U. S. A. 100, 9750 – 9755. Binda, C., Hubalek, F., Li, M., Edmondson, D.E., Mattevi, A., 2004. Crystal structure of human monoamine oxidase B, a drug target enzyme monotopically inserted into the mitochondrial outer membrane. FEBS Lett. 564, 225 – 228. Briggs, J.P., 2002. The zebrafish: a new model organism for integrative physiology. Am. J. Physiol. 282, R3 – R9. Cesura, A.M., Gottowik, J., Lang, G., Malherbe, P., Da Prada, M., 1998. Structure–function relationships of mitochondrial monoamine oxidase A and B: chimeric enzymes and site-directed mutagenesis studies. J. Neural Transm., Suppl. 52, 189 – 200.

A. Setini et al. / Comparative Biochemistry and Physiology, Part B 140 (2005) 153–161 Chen, K., Wu, H.F., Grimsby, J., Shih, J.C., 1994. Cloning of a novel monoamine oxidase cDNA from trout liver. Mol. Pharmacol. 46, 1226 – 1233. Fowler, C.J., Ross, S.B., 1984. Selective inhibition of monoamine oxidase A and B: biochemical, pharmacological and clinical properties. Med. Res. Rev. 4, 323 – 358. Geha, R.M., Chen, K., Shih, J.C., 2000. Phe208 and Ile199 in human monoamine oxidase A and B do not determine substrate and inhibitor specificities as in rat. J. Neurochem. 75, 1304 – 1309. Geha, R.M., Chen, K., Wouters, J., Ooms, F., Shih, J.C., 2002. Analysis of conserved active site residues in monamine oxidase A and B and their three-dimensional molecular modeling. J. Biol. Chem. 177, 17209 – 17216. Hall, T.R., Uruen˜a, G., 1983a. Pharmacology and physiology of monoamine oxidase activity in vertebrates. A comparative study. Comp. Biochem. Physiol., B 76, 393 – 397. Hall, T.R., Uruen˜a, G., 1983b. Distribution of monoamine oxidase activity in tissues of the urodeles Ambystoma tigrinum (tiger salamander) and Necturus maculosus (mud puppy). Comp. Biochem. Physiol., C 74, 35 – 39. Kinemuchi, H., Sudo, M., Yoshino, N., Kawaguchi, T., Sunami, Y., Kamijo, K., 1983. A new type of mitochondrial monoamine oxidase distinct from type-A and type-B. Life Sci. 32, 517 – 524. Kinemuchi, H., Fowler, C.J., Tipton, K.T., 1984. Substrate specificities of the two forms of monoamine oxidase. In: Tipton, K.F., Dostert, P., Strolin Benedetti, M. (Eds.), Monoamine oxidase and disease. Academic Press, New York, pp. 53 – 62. Kobayashi, S., Takahara, K., Kamijo, K., 1981. Monoamine oxidase in frog liver and brain. Comp. Biochem. Physiol., C 69, 179 – 183. Kfchli, H., von Wartburg, J.P., 1978. A sensitive photometric assay for monoamine oxidase. Anal. Biochem. 84, 127 – 135. Krueger, M.J., Manouz, F., Ramsay, R.R., Milcent, R., Singer, T.P., 1995. Dramatic species differences in the susceptibility of monoamine oxidase B to a group of powerful inhibitors. Biochem. Biophys. Res. Commun. 206, 556 – 562. Kumazawa, T., Seno, H., Ishii, A., 1998. Monoamine oxidase activities in catfish (Parasilurus asotus) tissues. J. Enzyme Inhib. 13, 377 – 384. Ma, J., Yoshimura, M., Yamashita, E., Nakagawa, A., Ito, A., Tsukihara, T., 2004. Structure of rat monoamine oxidase A and its specific recognitions for substrates and inhibitors. J. Mol. Biol. 338, 103 – 114. Nicotra, A., Senatori, O., 1988. Changes in monoamine oxidase activity by mitochondria isolated from late embryos of Bufo bufo. Comp. Biochem. Physiol., C 89, 5 – 9. Nicotra, A., Senatori, O., 1989. Some characteristics of mitochondrial monoamine oxidase activity in carp (Cyprinus carpio) and in trout eggs (Salmo gairdneri). Comp. Biochem. Physiol., C 92, 401 – 404.

161

Ouali, M., King, R.D., 2000. Cascaded multiple classifiers for secondary structure prediction. Protein Sci. 9, 1162 – 1176. Pintar, J.E., Maxwell, G.D., Breakefield, X.O., 1983. Characterization of monoamine oxidase activity during early stages of quail embryogenesis. Int. J. Dev. Neurosci. 1, 49 – 57. Rost, B., 1996. Predicting one-dimensional protein structure by profile based neural networks. Methods Enzymol. 266, 525 – 539. Rost, B., Casadio, R., Fariselli, P., Sander, C., 1995. Transmembrane helices predicted at 95% accuracy. Protein Sci. 4, 521 – 533. Salles, C., Souza, R.H.S., Salles, J.B., Cunha Bastos, V.L.F., Dias, R.A., Cunha Bastos, J., 2001. Biochemical characteristics of liver and brain monoamine oxidase from pacu. J. Fish Biol. 58, 1301 – 1310. Senatori, O., Nicotra, A., Scopelliti, R., 1990. Monoamine oxidase activity in embryos of pike. Comp. Biochem. Physiol., C 96, 87 – 90. Shih, J.C., Chen, K., Geha, R.M., 1998. Determination of regions important for monoamine oxidase (MAO) A and B substrate and inhibitor selectivities. J. Neural Transm., Suppl. 52, 1 – 8. Shih, J.C., Chen, K., Ridd, M.J., 1999. Monoamine oxidase: from genes to behavior. Annu. Rev. Neurosci. 22, 197 – 217. Sullivan, J.P., McDonnel, L., Hardiman, O.M., Farrel, M.A., Phillips, J.P., Tipton, K.F., 1986. The oxidation of tryptamine by the two forms of monoamine oxidase in human tissues. Biochem. Pharmacol. 35, 3255 – 3260. Tsugeno, Y., Ito, A., 1997. A key amino acid responsible for substrate selectivity of monoamine oxidase A and B. J. Biol. Chem. 272, 14033 – 14036. Veselovsky, A.V., Ivanov, A.S., Medvedev, A.E., 1998. Is one amino acid responsible for substrate specificity of monoamine oxidase A and B? Biochemistry (Moscow) 63, 1695 – 1701. Veselovsky, A.V., Ivanov, A.S., Medvedev, A.E., 2004. Computer modeling and visualization of active site of monoamine oxidases. Neurotoxicology 25, 37 – 46. Wouters, J., 1998. Structural aspects of monoamine oxidase and its reversible inhibition. Curr. Med. Chem. 5, 137 – 162. Yoshino, M., Obata, T., Sho, S., Kinemuchi, H., 1984. Enzymic and molecular characteristics of a new form of monoamine oxidase, distinct from form A and form B. Jpn. J. Pharmacol. 35, 105 – 115. Youdim, M.B.H., Finberg, J.P.M., 1991. New directions in monoamine oxidase A and B selective inhibitors and substrates. Biochem. Pharmacol. 41, 155 – 162. Youdim, M.B.H., Finberg, J.P.M., Tipton, K.F., 1988. Monoamine oxidase. In: Trendelenburg, U., Weiner, N. (Eds.), Handbook of Experimental Pharmacology 90. Springer-Verlag, Berlin, pp. 119 – 192.