primary structures for a mammalian cellular and serum copper

THE JOURNAL of BIOLOGICAL CHEUISTEY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 13, Issue ofApril 1, pp. 9926-9932, 1994 Printed in U S A .

Primary Structures for a Mammalian Cellular and Serum Copper Amine Oxidase*

(Received for publication, August 27, 1993, and in revised form, December 13, 1993)

David Mu$, Katalin F. Medzihradszkys, Gregory W. AdamsJ, Petra Mayerfj, Wade M. Hiness, Alma L. Burlingames, Alan J. Smith%, Danying CaiS, and Judith P. KlinmanSII From the wepartment of Chemistry, University of California, Berkeley, California 94720, the $Department of Pharmaceuticul C h e ~ i s t ~ , Uni~ersity of Californ~a, Sun Francisco, California 94143, and the W c k m a n Center, Stanford University, Medical Center, Stanford, California 94305

The 6-hydroxydopa quinone-containing active site peptide from bovine serum amine oxidase has been found to be highly homologous to a segment of a cloned human kidney a~i~oride-binding protein (Barbry, P., Champe, M., Chassande, O,, Munemitsu, S., Champigny, G., Lingueglia, E., Maes, P,, Frelin, C., Tartar, A., Ullrich, A, and Lazdunski, M. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 7347-7361). Additionally, a second 38-residue tryptic peptide (peptide X I ) isolated from bovine serum amine oxidase shows 82% identity with a portion near the carboxyl terminus of the human kidney ~iloride-binding protein. When an extended active site peptide was isolated from porcine kidney diamine oxidase (Janes, S. M., Palcic, M. M., Scaman, C. H., Smith, A. J., Brown, D. E., Dooley, D. M., Mure, M., and Klinman, J. P. (1992) Bio- chemistry 31,12147-12154), it was found to be fully contained in the human kidney amiloride-binding protein. Examination of amiloride binding to bovine s e m amine oxidase and porcine kidney diamine oxidase reveals dissociation constants of 196 and 9.1 p ~ , respectively. Taken together, these findings indicate that the cDNA isolated for human kidney amiloride-binding protein encodes a human kidney diamine oxidase. Two oli- gonucleotides, based on the tryptic peptide X I and active-site peptide of bovine serum amine oxidase, were used to amplify a portion of cDNA from a commercial bovine liver cDNA library through the use of the polymerase chain reaction. A full-length clone (2.7 kilobase pairs) for bovine serum amine oxidase was subsequently obtained through screening of the same cDNA library with the amplified 0.7-kilobase pair cDNA. These studies provide the first primary sequences for a mammalian cellular and serum copper amine oxidase. Computer alignment of amine oxidase cDNA-derived protein sequences reveals three conserved histidine residues, which are likely to be ligands to copper.

Bovine serum amine oxidase and porcine kidney diamine oxidase belong to the class of enzymes designated copper amine oxidases (EC 1.4.3.6). These enzymes catalyze the oxidation of amines to aldehyde and ammonium ion, followed by a two- electron reduction of dioxygen to hydrogen peroxide. Although copper amine oxidases are distributed in a wide range of or-

Grant GM39296 (to J. P. K.) and the National Center for Research * Financial support was provided by the National Institutes of Health

Resources, National Institutes of Health, Grant RR 01614 and Tobacco- related Disease Research Program, number 2RT0089 (to A. L. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisernent” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

I/ To whom correspondence should be addressed.

ganisms and tissues (for a recent review, see Ref. 11, the physiological functions and the precise sites of biosynthesis of the mammalian enzymes remain unresolved. A novel organic redox cofactor, 6-hydroxydopa (TOPA)’ quinone, was first discovered in bovine serum amine oxidase (2) and subsequently shown to be the organic cofactor in other copper amine oxidases including porcine kidney diamine oxidase (3). The identification of TOPA quinone cofactor has led to detailed enzyme mechanistic studies (e.g. Ref. 4 and references therein), as well as the study of the chemistry of TOPA quinone model compounds (e.g. Ref. 5).

Although the biochemical properties of copper amine oxidases are unfolding rapidly, the molecular biological aspects of these proteins have received relatively little attention. Among the large number of copper amine oxidases, only a few non- mammalian enzymes have been cloned; these enzymes include yeast amine oxidase from ~ u n s e n u ~ u p o Z y ~ o r p ~ u (61, lentil seedlings amine oxidase (7), and bacterial methylamine oxidase (8). Although mammalian lysyl oxidases have also been cloned and sequenced (9-111, lysyl oxidase is small (32 kDa) and shares little homology with other enzymes of the copper amine oxidase class (80-95 kDdsubunit).

In order to describe the consensus sequence surrounding the TOPA quinone cofactor, active site peptides from a range of copper amine oxidases have been isolated and sequenced by high performance tandem mass spectrometry or Edman degradation (3). Routine comparison of the resulting sequences to the Protein Identification Resource database (12) revealed a high degree of similarity between amine oxidase active site-derived peptides and a region of human kidney amiloride-binding protein cloned recently (13). Human kidney amiloride-bin~ng protein is a glycoprotein appearing as a single band at 97 kDa on SDS-polyacrylamide gel electrophoresis, but migrating under nonreducing conditions at approximately 180 kDa. In this paper, we show that the designated amiloride-binding protein from human kidney (13) is in actuality a human kidney diamine oxidase. This fortuitous finding provided a frame of ref- erence for the subsequent cloning of a serum-derived mammalian copper amine oxidase. As described herein, a full-length cDNA for bovine serum amine oxidase has been isolated from a bovine liver cDNA library, resolving the long standing question regarding at least one of the tissue sources of circulating copper amine oxidase. Sequencing of the bovine serum amine oxidase EDNA allows, for the first time, a comparison of primary structures for a cellular and serum copper amine oxidase. Sequence alignment suggests separate protein domains with a carboxyl-

(also known as 6-hydroxydopa) MOPAC, mixed oligonucleotide-primed The abbreviations used are: TOPA, 2,4,5-trihydroxyphenylalanine

amplification of cDNA, cDNA, complementary DNA, kb, kilobase pairfs); ATPyS, adenosine 5‘-O-(thiotnphosphate).

9926

9927 Structures of Mammalian Copper Amine Oxidases TABLE I

Comparison of peptide sequences for copper amine oxidases to human kidney amiloride-binding protein Sequences for human kidney amiloride-binding protein are from Barbry et al. (13).

Active site peptides Human kidney amiloride-binding protein: 453TTSTVYNYDYIWDFIFYPNGVMEAK47' Porcine kidney diamine oxidase:",* TTSTVYNYDYIWDFIFY(Y?) (N?) Bovine serum amine oxidase (peptide VII):*,' SVSTMLNYDYVWDMVFYPNGAIEVK Porcine serum amine oxidase:",b SVSTMLNYDYVWDMIFHP

Human kidney amiloride-binding protein: 209GYFLHPTGLELLVDHL'' Bovine serum amine oxidase (peptide IV): GPYLHPVGLELLVDH Human kidney amiloride-binding protein: 662DLVAWVTVGFLHIPHSEDIPNTATPGNSVGFLLRPFNF6g3 Bovine serum amine oxidase (peptide XI): DLVAWVTAGFLHIPHAEDIPNTVTVGNGVGFFLRPYNF

Nonactive site peptides

Janes et al. (3). * The Dreviouslv unidentifiable residue at Dosition 12 is now shown to be tryptophan (47).

Mu e; aZ. (22)"and Janes et al. (3).

terminal domain encoding the cofactor consensus sequence and putative copper binding sites, whereas the more divergent amino-terminal domain is proposed to accommodate the differing physiological functions and substrate specificities of these proteins.

MATERIALS AND METHODS Enzyme Isolation and Amino Acid Sequencing-Bovine serum amine

oxidase (14) and porcine kidney diamine oxidase (15) were purified as referenced. The former concentrations were estimated by the Bradford method (16), using bovine serum albumin as a protein standard and a molecular mass of 170 kDa; the latter concentrations were determined spectrophotometrically a t 280 nm (17) and using a molecular mass of 170 kDa. NH,-terminal sequence data for bovine serum amine oxidase were obtained by Edman degradation on an Applied Biosystem 470A automated microsequencer. Sequencing of bovine serum amine oxidase- derived peptides was performed either by Edman sequencing or by tandem mass spectrometry using a Kratos Concept IIHH (Manchester, United Kingdom) four sector instrument of EBEB geometry with mul- tichannel array detection (18).

Competitiue Inhibition by Amiloride-Amiloride hydrochloride was from Sigma, and putrescine hydrochloride was from Fluka. Amiloride concentrations were determined spectrophotometrically at 361 nm (Et:,,, = 630, Merck Index, 10th edition, New York, 1983). Bovine serum amine oxidase activity was assayed spectrophotometrically according to the method of Neumann et al. (19) on either a Cary 118 or a Hewlett- Packard vectra 286112 spectrophotometer. Porcine kidney diamine oxidase activity was determined by oxygen uptake on a Yellow Springs Instrument polargraphic oxygen electrode model 53.

Isolation and Characterization of Bovine Serum Amine Oxidase cDNA CZones-A polymerase chain reaction-based mixed oligonucleotide-primed amplification of cDNA (MOPAC) (20,21) was used to amplify a partial bovine serum amine oxidase cDNA from an adult female bovine liver cDNA library in A g t l O vectors (Clontech, Palo Alto, CA). The sense primer (5'-TA(C/T)GA(C/T)TA(CPT)GT(C/G)TGGGA(C/T)AT- GGT-3') was synthesized to the amino acid residue 8-15 (TOPA-Asp- Tyr-Val-Trp-Asp-Met-Val) of the bovine serum amine oxidase tryptic active site peptide (22). The antisense primer (5'-AT(G/A)TC(T/C)TCG- GC(G/A)TG(GPT)GG(T/G/A)AT(G/A)TG-3') was synthesized to the residue 12-19 (His-Ile-Pro-His-Ala-Glu-Asp-Ile) of a bovine serum amine oxidase tryptic peptide XI (reported herein) near the carboxyl terminus. A typical MOPAC experiment consisted of the following: 50% glycerol, 20 pl; 100 m~ Tris, pH 8.3, 500 m~ KCl, 10 pl; sense primer, 120-150 pmol (5 pl); antisense primer, 120-150 pmol (5 pl); 2 pl of bovine liver cDNA library (lo6 plaque-forming units, frozen and thawed twice) mixed with 52 pl water; 4 m~ of dNTP, 5 p1; 1-3 units of Taq DNA polymerase (Perkin Elmer), 1 pl. The thermal cycles of amplification were carried out in a Perkin-Elmer DNA thermal cycler using this program: step i, 94 "C for 5 min; step ii, 94 "C for 30 s, 44-51 "C (vari- able annealing temperatures) for 30 s, 72 "C for 1 min; step iii, 72 "C for 1 min. Step ii was repeated 30 times. The MOPAC-generated cDNA was sequenced by the thermal cycle sequencing reactions (Stratagene). Sub- sequent screening of the bovine liver cDNA library with the MOPAC- derived probe employed the random-primed labeling method of Fein- berg and Vogelstein (23) and was performed according to Chaplin et al. (24). For clones giving initial positive hybridization signals, gel analyses of MOPAC were carried out to eliminate false positives. The insert cDNA (2.7 kb) of an authentic recombinant was cloned into pGEM3Z

(Promega) in two orientations at an EcoRI restriction site. Exonuclease 111-directed deletion of pGEM3Z plasmid DNA was performed according to the procedure of Sambrook et al. (25), and each deleted subclone was transformed into JMlO9-competent cells. The interval between adja- cent deleted subclones ranged from 180 to 230 base pairs. Double- stranded DNA of each deleted subclone was prepared from a plasmid miniprep kit (Qiagen). All regions of both strands of the cDNA were sequenced by the 2',3'-dideoxy chain-termination method (26) with Se- quenase 2.0 from U. S. Biochemical Corp. and [ C Y - ~ ~ S I ~ A T P ~ S (specific activity 1000 Ci/mmol) from Amersham. Secondary structure rich regions of the gene were unwound either by substituting dITP for dGTP in the reaction mixture or by adding 30-40% formamide to sequencing gels.

Hydropathy plot analyses of the cDNA-derived bovine serum amine oxidase sequence was conducted according to Kyte and Doolittle (27) using a window size of 11 residues for calculation. Sequence alignment was carried out with a Sun Microsystem IntelliGenetics Sequence Anal- ysis Program (IntelleGenetics, Inc., Mountain View, CA) or Genetics Computer Group Sequence Analysis Software (Genetics Computer Group, Madison, WI) . Protein sequences were aligned according to identity or similarity match.

RESULTS AND DISCUSSION

Human Kidney Diamine Oxidase-Initial screening of the existing Protein Identification Resource data base revealed considerable sequence homology between a cofactor-containing extended tryptic peptide isolated from bovine serum amine oxidase (3) and a human kidney amiloride-binding protein (13). Although a human kidney amine oxidase has not been characterized, an amine oxidase from porcine kidney has been purified and partially sequenced. A comparison was, therefore, made between an extended active site tryptic peptide from porcine kidney diamine oxidase and human kidney amiloride- binding protein. As indicated in Table I, alignment of sequences for these two proteins indicates virtual identity, with the exception of the 2 amino acids at the end of the porcine kidney diamine oxidase sequence attributed to low yields for PTH derivatives during the final rounds of Edman sequencing. For the purposes of comparison of the cDNA-derived human sequence and the protein-derived porcine sequence, the topa quinone precursor amino acid, tyrosine (22), is shown at position 8 in the porcine kidney sequence. We note the presence of tryptophan a t position 12 of the porcine kidney peptide, in contrast to earlier studies that had indicated an unidentifiable amino acid at this position (47).

Although the identity between human kidney amiloride- binding protein and bovine serum amine oxidase is less than that for porcine kidney diamine oxidase, there is still substan- tial overlap between these sequences. As indicated in Table I, a 60 and 61% identity exists between the active site peptides for bovine and porcine serum amine oxidases and human kidney amiloride-binding protein. This is not unexpected in light of previous studies showing a 56% identity between amine oxidases from the same species but different tissues (e.g. a 56%

Structures of Mammalian Copper Amine Oxidases

800

A

.5 .3 .1 1 3 5 7 9

1/1putrescine], m~ -' FIG. 1. A, reciprocal plots of rates of benzaldehyde produced with

bovine serum amine oxidase in the absence I + I and presence ( ' . : I of 115 VM amiloride. Assay conditions were 0.1 M potassium phosphate, pH 7.2.

tion with porcine kidney diamine oxidase in the absence 1' . I and pres- 1.0-12 mM benzylamine. 13. reciprocal plots o f rates o f oxygen consump-

ence ( + ) of 31 p~ amiloride. Assays conditions were 0.1 potassium phosphate, pH 7.2, 0.22-4.5 mv putrescine.

identity exists between porcine serum amine oxidase and porcine kidney diamine oxidase (3) ) . There appears to be a greater conservation of structure between amine oxidases from the same tissue but different organisms (e.g. an 89% identity was observed between porcine and bovine serum amine oxidases (3)). In the course of efforts to sequence the entire protein for bovine serum amine oxidase, additional nonactive site peptides have been obtained that reveal high homology to human kidney amiloride-binding protein. Thes are designated as peptides IV and XI in Table I. Of particular note is peptide XI, which shows 82% identity with a portion near the carboxyl-terminus of amiloride-binding protein.

As a result of the above findings, the ability of amiloride to inhibit bovine serum and porcine kidney enzymes was investi- gated (Fig. 1). We find that amiloride inhibits these enzymes in a competitive manner, with Ki values of 196 and 9.1 p>i, respectively. These can he compared with values of 0.5 and 4.2 p~

1 2 3 4 5 6

" "

determined by equilibrium binding rxprrimrnts using nmiloride-binding protein purified from porcine kidnry mrmbrnnrs (28). We especially note the similarity hrtwrrn K, = 9.1 p>t, obtained by activity inhihition mrnsuremrnts \vith pnrc.int. kidney diamine oxidasr and a K,, = 4.2 uv. drtrrminrd hy rquilih- rium hinding dialysis binding experimrnts with nmiloridr- binding protein (2x1. Thr fnilurr to src thr highrr affinity amiloride site (0.5 p~ 1 in porcinr kidnry diamine oxidnsr suggests that this si te is uncorrelatrd with rnzymr activity.

Although the isolated nativr nmiloridr-hinding protrin has been characterized as n Na' chnnnrl, ruknryotic crlls trans- fected with the cloned gene showed nmiloride-hinding in t h r cell membrane fraction, hut no amiloride-sensitivr Na' channrl activity (13). Furthermore, the hydropathy plot of thc cloned amiloride-binding protcin failed to reveal any mrmhranr-spnn- ning segment. I t was noted by Rarbry c t nl. ( 18 I that human kidney amiloride-binding protrin Iackrd homology to prrvi- ously cloned receptors or ion channel protrins. Thr cDNA of a second amiloride-sensitive sodium channel has reccntly hern isolated from the epithelial cells of rat distal colon (29. 4x1 . The deduced protein sequence shows two putativr mrmhranr spnn- ning segments, lacking any statistically simificant homology to the human kidney amiloride-binding protrin. In this cnsr. protein expression in X~nop11.s oocyte rcconstitutrs thr activity of the highly selective amiloridr-hlockahle sodium channrl (29. 48). These findings. together with I i I the high homology hr- tween peptides from bovine serum nminr oxidase and amiloride-binding protein (Table I ) , ( i i ) the idrnt i ty hr twrrn thr porcine kidney diamine oxidase active site prptidr and amiloridr- binding protein (Table I ), and r i i i I thr tight hinding of nmiloridr to porcine kidney diamine oxidase (Fit. I R l , Irad to t h r con- clusion tha t the cloned putative human kidnry amiloridr-hinding protein (13) is in actuality a human kidney dinminr oxidase. As a consequencr, thr complrtr primnv structure of n cellular mammalian copper amine oxidnsr is now nvnilnhlr.

BoLtine Serum Amine Oxi~n.~r-Considrrahlr difficulty was initially encountered in efforts to obtain n clonr for hovinr serum amine oxidase. To begin with. the tissuc source of this enzyme was uncertain, and scrrening of commrrcial cDSA l i -

Structures of Mammalian Copper Amine Oxidases 9929

braries from bovine liver and small intestine with a degenerate tide I). This sequence appears at residue 17 (Fig. 3) and is oligonucleotide based on the sequence of an active site peptide located behind a 16-residue sequence which shows all the typi- yielded several false positives. The finding that human amilo- cal features of a signal peptide (32) except for the lack of ride-binding protein is an amine oxidase, with high homology charged residues. A similar case has been reported in the lit- to peptides VI1 and XI of bovine serum amine oxidase, provided erature for a 24-residue sequence preceding the amino termi-

active site peptide V 1 I : S V S T M L N ~ Y P N G A I E V K - .eueprimr IEVKLHATGYISSAFLF ...

near C-terminus peptide XI: DLVAWYTAGFL-NTVTVGNGVG ...... ... NNETIAGKDLVAWYTAG '3-

an alternate strategy. Based on the reasonable assumption that the relative position of peptides VI1 and XI of bovine serum amine oxidase would be the same as in human amiloride-binding protein, a polymerase chain reaction-based method was used to amplify the region of cDNA of bovine serum amine oxidase flanked by these two sequences (20, 21). For the pur- pose of reducing primer degeneracy, primer syntheses were based on the segments of peptides VI1 and XI containing amino acids with the fewest numbers of codon (see "Materials and Methods").

Proceeding on the assumption that liver was a more likely source than small intestine for a plasma protein, the polymerase chain reaction-based cloning experiment was performed using a cDNA library from bovine liver. Initially, an annealing temperature of 44 "C was employed, leading to the expected 0.7-kb product DNA(Fig. 2, lane 1 ) . From the gel in Fig. 2 it can be seen that unknown DNA, a t higher molecular weight, was amplified as well. To check the authenticity of the 0.7-kb DNA, experiments were repeated under more stringent conditions of elevated annealing temperatures (46-51 "C). As shown in Fig. 1 (lanes 2 6 ) , the band intensities at higher molecular weight decreased to zero, whereas the band of interest remained domi- nant through the experiments, indicating that the 0.7-kb domi- nant band was likely to be the DNA of interest. This was confirmed by sequencing the ends of the 0.7-kb amplified cDNA and comparing the translated amino acid sequence with the original peptide sequence (see above).

The available 0.7-kb partial clone for bovine serum amine oxidase was subsequently used to screen lo6 recombinants of the same liver cDNA library, leading to 25 positive clones. Thir- teen of the 25 clones failed to give an 0.7-kb amplified DNA in polymerase chain reaction experiments. One of the 12 positive clones was found to be full length, as judged by the insert size (2.7 kb) and the nucleotide sequence at the 5' end. The nucleotide sequence of this clone is shown in Fig. 3. The first ATG was assigned as the initiation codon, since this is in a context described by Kozak (30); the sequence GCGATGT ~ has already been described as a start codon for a cloned human liver aldehyde dehydrogenase (31). The authenticity of the bovine serum amine oxidase cDNA was confirmed by the exact matches between peptide sequences of native enzyme and the deduced protein sequence. As shown in Fig. 3, 11 peptide sequences (I-XI),' including the amino-terminal sequence (peptide I) and active site peptide VII, are identical to different regions in the deduced sequence. This demonstrates for the first time that liver is a source of mammalian serum amine oxidases. The amino- terminal sequence of native bovine serum amine oxidase identified by Edman degradation is REEGGVGSEEGVGKQ (pep-

' G. W. Adams, P. Mayer, K. F. Medzirhadszky, A. H. Hines, and A. L. Burlingame, manuscript in preparation.

nus of mature human serum protease inhibitor (al-antitryp- sin) (33).

The 2664-base pair cDNA of bovine serum amine oxidase contains a open reading frame of 2289 nucleotides which can be translated to a protein sequence of 762 amino acids. The cal- culated molecular weight (82,836 excluding the signal peptide and 84,750 including the signal sequence) is close to the observed value of 85,00O/subunit (34). This is consistent with an estimate by Yamada et al. (35) of 4% glycosylation in mature protein. Tyrosine 470 corresponds to the post-translationally modified redox cofactor, 6-hydroxydopa quinone (21, consistent with our previous finding that TOPA quinone cofactor is generated from tyrosine in yeast (22). The entire sequence contains three possible N-linked glycosylation sites (136NVT, 231NIT, and 665NET). The hydropathy plot of the deduced bovine serum amine oxidase sequence (Fig. 4) shows that there are more hydrophilic than hydrophobic regions. This result is very similar to the results obtained by Barbry et al. (13) with human kidney amiloride-binding protein (now designated human kidney diamine oxidase).

Structural Homologies among Eukaryotic Copper Amine Oxidases-TOPA quinone of amine oxidases, contained in active site peptides with a consensus sequence, is believed to derive from a tyrosine precursor. (22, 3). This feature is confirmed in the present study with bovine serum amine oxidase. As illustrated in Fig. 5, the consensus sequence NYD(E) for TOPA quinone occurs in the carboxyl-terminal portion of each protein whose cDNA sequence is available. As pointed out pre- viously (3), aspartate flanks TOPA at the +1 position in all eukaryotes studied; the only exception is the presence of glu- tamate in the amine oxidase from yeast.

Numerous studies point toward the presence of 3 histidines as ligands to copper in the amine oxidases (3g38). In an earlier study, comparison of cDNA-derived protein sequences for a yeast and a plant amine oxidase led to the identity of 3 conserved histidines at positions 8, 246, and 357 (7). With the present availability of sequences for two mammalian proteins, homology comparisons become more rigorous. These sequences were aligned pairwise and part of the alignments is shown in Fig. 5. One His-X-His motif 40-50 residues toward the carboxyl terminus from the cofactor consensus sequence is conserved in all alignments. This most likely contains 2 of the histidine ligands to copper, based on the observation that histidines in His-X-His motifs have been identified as ligands to type I1 or I11 copper in the crystal structures of several copper proteins (39). It is interesting to note that the His-X-His motif in yeast amine oxidase is contained in a stretch of sequence HNHQH. HQH is the conserved motif when aligned with lentil seedling amine oxidase, but it is HNH when aligned with bovine serum amine oxidase or the human kidney diamine oxidase. We are therefore not certain which motif constitutes ligands to copper in the

Structures of Mammalian Copper Amine Oxidases GGA GAA CGG GTT GCA TGT TGC ATT CTG ATT CAA GAG CCT GAA GAG TCA AGA 51

M F I F I F L S L W T L L V M G ~ ~ GTT TTA GCG ATG TTC ATC TTC ATT TTT CTG TCC TTG TGG ACT CTT CTG GTG ATG GGC 1 0 8

R E E G G V G S E E G V G K O AGG GAG GAA GGT GGT GTT GGG AGT GAG GAG GGA GTT GGG AAG CAA TGT CAT CCC AGC 1 6 5

7 C H P S 3 5

I

CTG CCT CCC CGC TGC CCC TCC AGA TCC CCT AGT GAC CAG CCC TGG ACA CAC CCT GAC 2 2 2 L P P R C P S R S P S D Q P W T H P D 5 4

Q S Q L F A D L S R E E L T T V M S F 7 3 CAG AGC CAG CTG TTT GCA GAC CTG AGC CGA GAA GAG CTG ACA ACT GTG ATG AGC T T C 2 7 9

CTG ACT CAG CAG CTG GGG CCA GAC CTG GTG GAT GCA GCC CAG GCC CGA CCC TCA GAC 3 3 6 L T Q Q L G P D L V D A A Q A R P S D 9 2

N C V F S V E L Q L P P K B A A L A H l l l AAC TGT GTC TTC TCG GTA GAA CTT CAG CTG CCC CCC AAG GCT GCA GCC CTG GCC CAC 3 9 3

I1 L D B G S P P P A R E A L A I V F F G 1 3 0 CTG GAC AGG GGG AGC CCC CCA CCT GCC CGG GAG GCA CTG GCC ATC GTC TTC TTT GGC 4 5 0

V G Q P Q P N V T E L V V G P L P Q P S l 4 9 GGA CAA CCC CAG CCC AAT GTG ACT GAG CTG GTA GTG GGG CCG CTG CCC CAG CCC TCC 5 0 7

Y M R D V T V E R H G G P L P Y Y R R l 6 8 TAC ATG CGG GAT GTG ACC GTG GAG CGT CAT GGC GGC CCC CTG CCC TAT TAC CGA CGC 5 6 4

I11 P V L L R E Y L D I D Q M I F N R E L 1 8 7 CCC GTG CTT CTC CGA GAG TAC CTG GAC ATA GAC CAG ATG ATC TTC AAC AGA GAG CTG 6 2 1

P Q A A G V L H H C C S Y K Q G G Q K 2 0 6 CCC CAG GCT GCT GGT GTC CTG CAC CAC TGC TGC TCC TAC AAA CAA GGA GGA CAG AAA 6 7 8

L L T M N S A P R G V Q S G D R S T W 2 2 5 CTG TTG ACC ATG AAC TCA GCT CCC CGT GGA GTG CAG TCA GGT GAT AGG TCC ACT TGG 7 3 5

V F G I Y Y N I T K G G P Y L H P V G L TTT GGC ATC TAC TAT AAC ATC ACA AAG GGT GGG CCT TAC CTG CAC CCC GTG GGG TTG 7 9 2

2 4 4

IV E L L V D H K A L D P A D W T V Q K V 2 6 3 GAG CTT CTG GTA GAC CAT AAG GCT CTG GAC CCT GCC GAT TGG ACC GTC CAG AAG GTG 8 4 9

F F Q G R Y Y E N L A Q L E E Q F E A Z E Z TTC TTT CAA GGC CGC TAC TAT GAA AAT CTG GCC CAG CTG GAG GAG CAG TTT GAG GCT 9 0 6

G Q V N V V V I P D D G T G G F W S L 3 0 1 GGC CAG GTG AAT GTG GTG GTG ATC CCA GAC GAT GGC ACA GGT GGG TTC TGG TCC CTG 9 6 3

K S Q V P P G P T P P L Q F H P Q G P 3 2 0 AAG TCC CAG GTG CCT CCG GGT CCA ACT CCC CCT CTG CAG TTC CAT CCT CAG GGC CCC 1 0 2 0

R F S V O G N R V A S S L W T F S F G 3 3 9 CGC TTC AGT GTC CAG GGC AAT CGA GTG GCC TCC TCA TTG TGG ACT TTC TCC TTT GGC 1 0 7 7

V L G A F S G P R V F D V R F Q G E R L 3 5 8 CTC GGA GCT TTC AGT GGT CCT AGG GTC TTT GAC GTT CGA TTC CAG GGA GAA CGG CTG 1 1 3 4

A Y E I S L Q E A G A V Y G G N T P A 3 7 7 GCT TAT GAG ATC AGC CTG CAA GAG GCT GGT GCT GTC TAC GGT GGG AAT ACT CCA GCA 1 1 9 1

A M L T R Y M D S G F G M G Y F A T P 3 9 6 GCA ATG CTC ACT CGC TAT ATG GAT AGT GGC TTT GGC ATG GGT TAC TTC GCC ACA CCC 1 2 4 8

L I R G V D C P Y L A T Y M D W H F V 4 1 5 CTG ATT CGT GGC GTA GAC TGC CCT TAC CTG GCC ACC TAC ATG GAC TGG CAC T T C G T T 1 3 0 5

V E S Q T P K T L H D A F C V F E Q N 4 3 4 GTG GAG TCC CAA ACC CCC AAG ACA CTA CAT GAT GCC TTT TGT GTG TTT GAG CAG AAC 1 3 6 2

K G L P L R R AAG GGC CTG CCC CTG AGG CGA CAC CAC TCA GAT TTT CTT TCC CAC TAT TTT GGG GGC 1 4 1 9

H H S D F L S H Y F G G 4 5 3

VI V A Q T V L V F R S V S T M L N Y D Y GTT GCA CAG ACA GTG CTG GTC TTC AGG TCT GTG TCC ACG ATG CTC AAC TAT GAC TAT 1 4 7 6

4 7 2

VI1 V W D M V F Y P N G A I E V K GTG TGG GAT ATG GTC TTC TAC CCC AAT GGG GCC ATA GAA GTC AAA TTG CAT GCC ACG 1533

L H A T 4 9 1

G Y I S S A F L F G A A R R Y G N Q V S l D GGC TAC ATC AGC TCA GCG TTC CTC TTT GGT GCT GCC CGA AGA TAC GGA AAC CAG GTT 1 5 9 0

FIG. 3. Nucleotide sequence of bovine @e- amine oxidase and the derived protein sequence. Nucleotides are numbered in the 5' to 3' direction. Numbering of the amino acid residues be- gins with the first methionine. Amino acid residues underlined correspond to the following: peptide I, from NH2-terminal Ed- man sequencing in this study; peptide VII, cofactor-containing peptide (22, 3); peptides 11-VI and VIII-XI, tryptic bovine serum amine oxidase peptides sequenced by mass spectrometric analyses in this study. The asparagine residues la- beled with V on top represent possible N-linked glycosylation sites. The arrow at Gly-16 indicates the putative signal sequence cleavage site. The common polyadenylation signal sequence -AATAAA- is not present, with the closest match -A?TAAA- 25 nucleotides upstream from the polyadenylation site.

G E H T L G P V H T H S A H Y K V D L 5 2 9 GGG GAA CAC ACG CTG GGC CCC GTC CAC ACC CAC AGT GCC CAC TAC AAG GTG GAT CTG 1 6 4 7

D V G G L E N W V W A E D M A F V P T 5 4 8 GAC GTG GGA GGA CTG GAA AAC TGG GTC TGG GCT GAG GAT ATG GCT TTT GTC CCC ACG 1 7 0 4

A I P W S P E H Q I Q R L O V T R K Q S 6 7 GCG ATA CCC TGG AGC CCT GAG CAC CAG ATA CAG AGG CTG CAG GTG ACC CGG AAG CAG 1 7 6 1

VI11

Structures of Mammalian Copper Amine Oxidases 9931 L E l E E Q A A F P L G G A S P R Y L 5 8 6 CTG GAG ACC GAG GAG CAG GCT GCC TTC CCC CTG GGA GGG GCT TCC CCT CGC TAC CTG 1818

y L A S K Q S N K W G H P R G Y R I Q 6 0 5 TAC CTG GCC AGC AAG CAG AGC AAC AAG TGG GGG CAC CCT CGC GGC TAT CGC ATC CAG 1 8 7 5

IX

FIG. 3-continued

T V S F A G G P M F Q N S P M E R A F 6 2 4 A C C G T C A G C T T T G C T G G G G G G C C G A T G C C C C A G A A C A G C C C C A T G G A G A G A G C C T T C 1 9 3 2

S W G R Y O L A I T O R AGC TGG GGG AGG TAC CAG CTG GCC ATA ACC CAG AGG AAG GAG ACA GAG CCC AGC AGC 1 9 8 9

K E T E P S S 6 4 3

X S S V F N Q N D P W T P T V D F S D F 6 6 2 TCC AGC GTC TTC AAT CAG AAT GAC CCC TGG ACT CCC ACC GTG GAC TTT TCT GAC TTC 2 0 4 6

I N N E T I A G K D L V A W V T A G F 6 8 1 ATC AAC AAT GAG ACC ATT GCT GGA AAG GAC TTG GTG FCC TGG GTG ACA GCC GGT TTC 2 1 0 3

V

XI L H I P B A E 3 I P N T V T V G N G V 7 0 0 CTG CAC ATC CCA CAT GCA GAG GAT ATT CCC AAC ACG GTG ACG GTG GGG AAC GGT GTG 2160 G F F L R P Y , - N F F D Q E P S M D S A S 1 9 GGC T'TC T T C T T G AGA CCC TAC h;>C T T C T T T GAC C A G G A G C C C T C C A T G G A T T C T G C T 2 2 1 7

G A C T C C A T C T A C T T C C G G G A R G G C C A G G A T G C T G G G T C C T G T G A G A T C A A C C C C C T G 2 2 7 4 D S I Y F f i E G Q D h G S C E I N F L 7 3 8

A C L P Q A A T C A P D L P V F S H G 7 5 7 GCT TGC CTG CCC CAG GCT GCT ACC T G T G C C C C T G A C C T C C C T G T C T T C T C C C A C G G G 2 1 3 1

GGC TAC CCT GAG TAC TAG CTG GTC CTT GGG GCG AGG AAT GTG TCT AGG ACC CAA AGA CAA GGG 2 3 9 4 XGT GTG GAG AGC GGA GGG tiCT GGG CAC TGA GTT TCT CCC AGC TCC CAC CCC AGG TTC CTC CCT CTC 2 4 6 0 CCA TCT TCT GCC CTT GCC TCC TCA ACC CTG CTG GAA GTA TCC ATA GCC TGT GCT GCC TGA CCC ATG 2 5 2 6 SGC TTC TCA ACT CTG TTG ACT CCT GIiC CTG CTG CAT TCC TAT CGC AGA GAG AAA AGG RAT GGC CAA 2592

AAA CCC CTG GGA GTC TAG AAT GAT GAT TAA ACC TTT CZG GAA GAC TTA ACT TCG GCA AAA AAA AAA ARA AAA 26b8

2 6 6 4

G Y F E Y . 7 6 2

phob 3.0 I

phi I

0 153 305 458 6 10 763

Residue Number FIG. 4. Hydropathy plot of the deduced bovine serum amine oxidase protein sequence. A highly hydrophobic potential signal peptide

can be seen at the amino terminus. Numbering starts with the first methionine residue, including the putative signal sequence. Values indicating hydrophobicity and hydrophilicity are above and below zero, respectively. The arrow indicates the location of the possible copper-binding ligands His-519 and His-521.

BSAO 436GLPLRRH-HSDFLS--HYFGGVAQTVLVFRSVSTMLNYDYVWDHVFYPNGAIEVKLHATG--YISSAFLFGRRRR-YGNQVGEHTL-GPVHTIIS-AHYKVD5Z8 HKDAO 422GVPLRRHFNSNFKGGFNFYAGLKGQVLVLRTTSTVYHYDYIWDFIFYPNGVMEAKMHATG--YVHATF-YTPEG--CARHSPAHPP-DWQHTHSLVHYRVD517 YAO 3 7 1 G L L F K H S D F R D N F A T S L V T R A - T K L W S Q I F T A A N Y E Y Q H L F S L R I D 4 6 5 LNSAO 3 7 0 N I M W R H T E T G I P N E S I E E S R T E V D ~ I R T W T V G N l l D R F Y I Y Y L D 4 6 9

FIG. 5. Sequence homology near the active site of eukaryotic copper amine oxidases. The cofactor-containing consensus sequence and conserved histidines are in boldface, and a gap is marked with dash. All numbering includes the signal peptide. BSAO, bovine serum amine oxidase; HKDAO, human kidney diamine oxidase; YAO, yeast amine oxidase; LNSAO, lentil seedlings amine oxidase.

yeast enzyme. In addition to the histidines found in a His-X-His motif, another histidine 20-30 residues toward the amino terminus from the cofactor consensus sequence is also conserved in all sequences. The 3 histidines described above are the only invariant ones found in all four amine oxidase sequences, which leads us to propose that they are ligands to copper in amine oxidases. An additional histidine near the protein amino terminus was reportedly conserved when the cDNA-derived

protein sequences of amine oxidases from yeast and lentil seedlings were aligned with that of the amine oxidase from Ar- throbactor PI (40,8). However, this does not hold true when the mammalian enzymes are included in the alignments (data not shown).

We note that the conserved histidines are embedded in hydrophilic segments and hence do not violate the empirical observation that metal ions are ligated by a shell of hydrophilic

9932 Structures of Mammalian Copper Amine Oxidases

groups (41). Also, the positions of conserved histidines relative to the cofactor are well preserved despite variations in length of the polypeptide chains (587 residues for lentil seedlings enzyme, 692 for the yeast enzyme, 713 for human kidney diamine oxidase, and 762 for bovine serum amine oxidase). Therefore we believe that the overall structure around copper and the cofactor is likely to be very similar for amine oxidases.

From differential scanning calorimetry studies, previous in- vestigators (42) have speculated that each subunit of bovine serum amine oxidase folds into two domains, comprising about 60 and 40% of the polypeptide, respectively. This is consistent with the cDNA-deduced protein sequence (Fig. 3) in which the active-site cofactor 6-hydroxydopa quinone (%-470) is flanked by 60% of polypeptide from the amino terminus and 40% from the carboxyl terminus. It suggests that the morphology for copper amine oxidases is analogous to numerous other enzymes characterized by substrate binding sites at the interface of two domains.

Functional roles for copper amine oxidases are easier to de- fine for prokaryotes than eukaryotes, with the former yielding enzyme induction in response to amines in the growth medium as carbon or nitrogen source. The literature is, in fact, replete with confusing terminology regarding the mammalian copper amine oxidases. Based on the apparent properties of the isolated enzyme, copper amine oxidases have been assigned names such as benzylamine oxidase, semicarbazide-sensitive amine oxidase, plasma amine oxidase, diamine oxidase, his- taminase, etc. (cf. Ref. 1). The data reported in this paper show, at a minimum, two classes of mammalian copper amine oxidases. A possible role for the cellular amine oxidase is the regulation of histamine and polyamine levels, whereas the serum proteins have been speculated to control the level of circulating biogenic amines such as dopamine and phenethyl- amine (1). Major questions to be addressed in the future include the tissue distribution of mRNA's for copper amine oxidases and the chromosome locations of their precursor DNA's. Although bovine serum amine oxidase has been cloned from a liver library, previous investigations have suggested that serum proteins may arise from the small intestine (43) and vascular smooth muscle (44, 45). The available cDNA's for a serum (bovine serum amine oxidase) and cell associated (human kidney diamine oxidase) amine oxidase will greatly facili- tate screening for mRNA levels in a range of cell types. Regard- ing chromosome localization, Barbry et al. (46) have already shown that human kidney diamine oxidase is located on chromosome 7, in a region flanking the cystic fibrosis locus.

Acknowledgments-We thank Fred C. Walls for expertise in carrying out tandem mass spectrometry and Sharon Walker for obtaining some preliminary mass spectrometric sequence information. We also thank Corey Schwartz and Zhonghua Yu for obtaining Edman data.

REFERENCES 1. McIntire, W. S., and Hartman, C. (1993) in Principles and Applications of

Quinoproteins (Davidson, V. L., ed) pp. 97-171, Marcel Dekker, he., New York

2. Janes, S. M., Mu, D., Wemmer, D., Smith, A. J., Kaur, S., Malthy, D., Burl- ingame, A. L., and Klinman, J. P. (1990) Science 248,981-987

3. Janes, S. M., Palcic, M. M., Scaman, C. H., Smith, A. J., Brown, D. E., Dooley, D. M., Mure, M., and Klinman, J. P. (1992) Biochemistry 31, 12147-12154

4. Hartmann, C., Brzovic, P., and Klinman, J. P. (1993) Biochemistry 32,2234- 2241

5. Mure, M., and Klinman, J. P. (1993) J. Am. Chem. Soc. 115, 7117-7127

6. Bruinenherg, P. G., Evers, M., Waterham, H. R., Kuipers, J., Amberg, A. C.,

7. Rossi, A,, Petruzzelli, R., and Finazzi-Agro, A. (1992) FEBS Lett. 301,253-257 8. Zhang, X., Fuller, J. H., and McIntire, W. S. (1993)J. Bacteriol. 176,56174627 9. Trackman, P. C., Pratt, A. M., Wolanski, A,, Tang, S. S., Offner, G. D., Troxler,

R. F., and Kagan, H. M. (1990) Biochemistry 29,4863-4870 10. Trackman, P. C., Pratt, A. M., Wolanski, T., Tang, S. S., Offner, G. D., Troxler,

R. F., and Kagan, H. M. (1991) Biochemistry 30,8282 11. Hamalainen, E. R., Jones, T. A., Sheer, D., Taskinen, K, Pihlajaniemi, T., and

Kivirikki, K. I. (1991) Genomics 11,50%516 12. George, D. G., Barker, W. C., and Hunt, L. T. (1986) Nucleic Acids Res. 14,

11-15 13. Barhry, P., Champe, M., Chassande, O., Munemitsu, S., Champigny, G.,

Lingueglia, E., Maes, P., Frelin, C., Tartar, A., Ullrich, A,, and Lazdunski, M. (1990) Proc. Natl. Acad. Sei. U. S. A. 87,7347-7351

and A B , G. (1989) Biochim. Biophys. Acta 1008, 157-167

14. Janes, S. M., and Klinman, J. P. (1991) Biochemistry 30,45994605 15. Tamura, H., Horike, K., Fukuda, H., and Watanabe, T. (1989) J. Biochem.

16. Bradford, M. M. (1976) Anal. Biochem. 72,248-254 17. Zeidan, H., Watanabe, K., Piette, L. H., and Yasunobu, K. T. (1980) J. Biol.

18. Wen, D. X., Livingston, B. D., Medzihradszky, K. F., Kelm, S., Burlingame, A. Chem. 255, 7621-7626

19. Neumann, R., Hevey, R., and Abeles, R. H. (1975) J. Biol. Chem. 250,6362- L., and Paulson, J. C., (1992) J. Biol. Chem. 267,21011-21019

20. Lee, C. C., and Caskey, C. T. (1990) in PCR Protoco2s:A Guide to Methods and 6367

Applications (Innis, M. A., Gelfand, D. H., Sninsky, J. J., and White, T. J., eds) pp. 4G53, Academic Press, Inc., San Diego, CA

21. Lee, C. C., Wu, X., Gihbs, R. A,, Cook, R. G., Muzny, D. M., and Caskey, C. T. (1988) Science 239,1288-1291

22. Mu, D., Janes, S. M., Smith, A. J., Brown, D. E., Dooley, D. M., and Klinman, J. P. (1992) J. Biol. Chem. 267, 7979-7982

23. Feinberg, A,, and Vogelstein, B. (1983) Anal. Biochem. 132, G13 24. Chaplin, D. D., and Brownstein, B. H. (1990) in Current Protocols in Molecular

Biology (Ausuhel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A,, Struhl, K., eds) Vol. 1, pp. 6.0.3-6.11.16, John Wiley &

25. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1988) in Molecular Cloning: A Sons, Inc

Laboratory Manual pp. 13.1-13.104 Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

26. Sanger, F., Nicklen, S., and Coulson,A. R. (1977)Proc. Natl. Acad. Sci. U. S. A. 74, 5463-5467

28. Barhry, P., Chassande, O., Duval, D., Rousseau, B., Frelin, C., and Lazdunski, 27. Kyte, J., and Doolittle, R. F. (1982) J. Mol. Biol. 157, 105-132

29. Canessa, C. M., Horisberger, J.-D., and Rossier, B. (1993) Nature 361,467470 30. Kozak, M. (1987) Nucleic Acids Res. 15,8125-8148 31. Braun, T., Bober, E., Singh, S., Agarwal, D. P., and Goedde, H. W. (1987)

32. von Heijne, G. (1985) J. Mol. Biol. 184, 99-105 33. Leicht, M., Long, G. L., Chandra, T., Kurachi, K., Kidd, V. J., Mace, M., Davie,

34. Yasunobu, K. T., Ishitaki, H., and Minamiura, N. (1976) Mol. Cell. Biochem.

35. Yamada, H., Gee, P., Ehata, M., and Yasunohu, K. (1964) Biochim. Biophys.

36. Scott, R. A,, and Dooley, D. M. (1985) J. Am. Chem. Soc. 107,4348-4350 37. Barker, G. J., Knowles, P. F., Pandeya, K. B., and Rayner, J. B. (1986)Biochem.

(Tokyo) 105,299-306

M. (1989) Biochemistry 28,3744-3749

Nucleic Acids Res. 16, 3179

E. W., and Woo, S. L. C. (1982) Nature 297,655-659

13,3-29

Acta 81, 165-171

J. 237,609-612 38. McCracken, J., Peisach, J., and Dooley, D. M. (1987) J. Am. Chem. Soc. 109,

4064-4072 39. Adman, E. T. (1991) Adu. Protein Chem. 42, 144-197 40. Dooley, D. M., Brown, D. E., Clague, A. W., Kemsley, J. N., McCahon, C.,

A. J. (1993) in Bioinorganic Chemistry ofCopper (Karlin, K. D., and Tyklhr, McGuirl, M. A,, Turowski, P. N., McIntire, W. S., Farrar, J. A,, and Thomson,

Z., eds) pp. 459-470, Chapman and Hall, New York 41. Yamashita, M. M., Wesson, L., Eisenman, G., and Eisenberg, D. (1990) Proc.

42. Giartosio, A., Agostinelli, E., and Mondovi, B. (1988) Biochem. Biophys. Res. Commun. 154,66-72

43. DAgostino, L., Pignata, S., Daniele, B., Ventriglia, R., Ferrari, G., Ferraro, C., Spaguolo, S., Lucchelli, P., and Mazzacca, G. (1989) Biochim. Biophys. Acta

Natl. Acad. Sci. U. S. A. 87, 5648-5652

993,228-232 44. Lewinsohn, R. (1981) J. Pharm. Pharmacol. 33,569472 45. Hysmith, R. M., and Boor, P. J. (1987) Biochem. Cell Biol. 66,821-829 46. Barbry, P., Simon-Bouy, B., Matthi, M.-G., Le Guern, E., Jaume-Roig, B., Chas-

47. Mu, D. (1993) Ph.D. Thesis Structural Studies of the 6-Hydroxydopa Quinone- sande, 0.. Ullrich, A., and Lazdunaki, M. (1990) Hum. Genet. 85,587-589

containing Copper Amine Oxidases. Ph.D. thesis, University of California, Berkeley, CA

48. Lingueglia, E., Voilley, N., Waldmann, R., Lazdunski, M., and Barhry, P. (1993) Febs Lett. 318,95-96

primary structures for a mammalian cellular and serum copper

Documents