THEJOURNAL OF BIOLOGICAL CHEMISTRY

Vol. 266, No. 14, Issue of May 15,pp. 87848789,1991 Printed in U.S.A .

0 1991 by The American Society for Biochemistry and Molecular Biology, Inc

Nucleoside Diphosphate Kinase from Human Erythrocytes STRUCTURAL CHARACTERIZATION OF THE TWOPOLYPEPTIDE HETEROGENEITY OF THE HEXAMERIC ENZYME*

CHAINS RESPONSIBLE FOR

(Received for publication, November 5, 1990)

Anne-Marie Gillesl, Elena Presecans, Alin Vonicas, and Ioan Lascusli From the $Unite deBiochimie des Regulations Cellulaires,Znstitut Pasteur, 75724 Paris Ceden 15, France and the §Department of Biochemistry, Medicaland Pharmaceutical Institute, 3400 Cluj-Napoca, Romania

EXPERIMENTALPROCEDURES

Chemicals-Coupling enzymes and nucleotides were purchased Nucleoside-diphosphate kinase (NDP kinase, EC 2.7.4.6)’ from Boehringer Mannheim. 8-BrIDP was synthesized by oxidative catalyzes phosphorylation of nucleoside 5”diphosphates to deamination of 8-BrADP (13). Cyanogen bromide (CNBr) andtrypsin corresponding triphosphates, following a ping-pong mecha- (TPCK-treated) were from Sigma. Acetonitrile for HPLC was obtained from British Drug House (Great Britain). Materials for chronism, matography were from Pharmacia LKB Biotechnology Inc. Cibacron N1TP E Ft NIDP + E-P (1) red 3B-P (Ciba-Geigy,Switzerland) was coupled to Sepharose CL-GB as described previously (14). E-P NZDP NzTP + E (2) Purification of NDP Kinases-Partially purified NDP kinase from The enzyme fromdifferent sources was purified to apparent human erythrocytes (40-70% purity as judged by SDS-PAGE) was in a single step by chromatography on Cibacron red 3B-Phomogeneity. Very often, however, NDP kinase preparations prepared Sepharose CL-GB by scaling up the procedure described by Presecan were heterogeneous with respect to theirisoelectric point (1). et al. (2). Gel permeation chromatography in the presence of 6 M urea Moreover, human (2, 3), rat (4), mouse (5), brewers’ yeast,’ allowed obtention of pure enzyme with a rather good separation of and microtubule-associated (6) NDP kinase displayed two NDP kinase from contaminating adenylate kinase, despite the small protein bands when analyzed by SDS-PAGE. Such heteroge- difference in molecular mass of the two polypeptides. Fractions containing NDPkinase were pooled,diluted 10-fold with 50 mM Tris* This work was supported in part by Grant URA 1129 from the HC1 (pH 7.6) and 1mM magnesium acetate, and thendialyzed against Centre National de la Recherche Scientifique and by the Fondation the same buffer. The enzyme was precipitated by dialysis against a pour la Recherche Medicale. The costs of publication of this article saturated solution of ammonium sulfate and stored at 4 “C. Before were defrayed in part by the payment of page charges. This article use, ammonium sulfate precipitated NDP kinase was sedimented by must therefore be hereby marked “advertisement” in accordance with centrifugation at 10,000 X g for 5 min in 1.5-ml Eppendorf tubes, dissolved in 50 mM Tris-HC1 (pH 7.6) and 1 mM magnesium acetate, 18 U.S.C. Section 1734 solelyto indicate this fact. 7l To whom correspondence should be addressed. Present address: and then desalted on 1 X 10-cm Sephadex G-25 column equilibrated Unite de Biochimie Cellulaire, Institut Pasteur, 28, rue du Docteur with the same buffer. Carboxymethylation of NDP Kinase-Pure A6 and B6 isoenzymes Roux, 75724 Paris Cedex 15, France. The abbreviations used are: NDP kinase, nucleotide-diphosphate were reduced with dithiothreitol and S-carboxymethylated with iokinase; CNBr, cyanogen bromide; A, acidic; B, basic; A6 and B6 refer doacetic acid as described by Waxdal et al. (15). Chemical and Enzymatic Digestion of NDP Kinase and Purification to the homogeneous renatured monoisozymic NDP kinases; SDSPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; of Peptides by HPLC-Carboxymethylated A6 and B6 isoenzymes TPCK, L-1-tosylamido-2-phenylethylchloromethyl ketone; 8- were digested at 37 ”C for 16 h with 1%(w/w) TPCK-trypsin. CarBrADP, 8-bromoadenosine 5’-diphosphate; 8-BrIDP, 8-bromoinosine boxymethylated B6 form was dissolvedin 70% formic acid and cleaved with a 5-fold excess (w/w) of CNBr for 24 h at room temperature. 5”diphosphate; HPLC, high performance liquid chromatography. Peptides were purified by reverse-phase HPLC using a Perkin-Elmer I. Lascu, E. Presecan, and A. Vonica, unpublished results.

+

+

8784

Downloaded from www.jbc.org at CNRS on October 9, 2006

Human erythrocyte nucleoside-diphosphate kinase neity was considered as resulting from proteolysis (7), phos(NDP kinase) is a hexameric enzyme consisting of two phorylation (4), or even to represent distinct enzymes with kinds of polypeptide chains, A and B. By random as- similar physico-chemical properties (8).Monoisozymic NDP sociation (A6,A5B.. .AB5,B6) these polypeptides kinases prepared from beef and sheep brain (7,9), may likely form isoenzymes differing in their isoelectric point. result from selective loss of other isoenzymes. Recently, genes Chains A and B of NDP kinase were purified by ion- encoding NDP kinase from Myxococcus xanthus (lo), Dicexchange chromatographyunder denaturing condi- tyosteliurn discoideum (ll),and rat (12) were cloned and their tions. Upon mixing and renaturation, the isozymic pat- nucleotide sequences determined. tern of NDP kinase obtained by conventional methods Since NDP kinases from eukaryotes are hexamers, we was restored. Antibodies raised against purified chains supposed that various isoenzymes might result from random showed significant cross-reactivity, both in immunoblot experiments and activity inhibition studies. Se- association of two kinds of polypeptides, as was shown much quence determination showed that both chains con- earlier for the tetrameric enzyme, lactate dehydrogenase. In sisted of 152 amino acid residues corresponding to M , this paper, we show that human erythrocyte NDP kinase or 17,143 (chain A) and17,294 (chain B), respectively. consists of two distinct polypeptide chains. A and B, which There was high homology between the two sequences can be obtained in nearly homogeneous form by ion-exchange (88% identity). The phosphorylationsite on the enzyme chromatography under denaturing conditions. Mixture of is located at His-118. Chain A was identical with hu- these polypeptides and renaturation restore the isoenzymic man Nm23 protein, which has been reported as a po- pattern of NDP kinase purified by conventional methods. We tential suppressor protein in tumor metastasis and have also established the complete amino acid sequence of chain B was identical with Nm23-HZ protein. both polypeptides.

Erythrocyte NDP Kinase

8785

SI 2

3 4 5 6

i

t

FRICTIOI IUWEER

Op\, 8r

C

I

2

/‘

/‘

10

20 30 40 FRACTION NUMBER

50

FIG. 2. Analysis of the renatured NDP kinases on DEAESepharose CL-GB. Aliquots of the renatured NDP kinases (see the “ExperimentalProcedures”) were desalted and applied to DEAESepharose CL-GB columns (0.9 X 8 cm) equilibratedwith 50 mM Tris-HCI (pH 8.6), 1 mM EDTA and eluted witha linear gradient of 0-0.3 M NaCl in the samebuffer, a t a rate of 8 ml/h, and fractions of 1 ml were collected. The dashed line represents the slope of the gradient. A large mixing device was used to simultaneously feed all columns. A, NDP kinase renatured frompolypeptide chain A; B, NDP kinase renatured from the polypeptide chain B; C, NDP kinase renatured from a mixture of polypeptides A and B. apparatus series 410 equipped with a LC 135 diode array detector. The flow rates were 1 ml/min, and measurements were made a t 230 and 280 nm. Supernatant of the tryptic digest was injected onto a

Elution

time cmln)

FIG. 4. Separationoftryptic peptides by reverse-phase HPLCon a nucleosil C-18 column. Solublepeptidesresulting from 1 mg of digested protein were eluted with a linear gradientof 060% acetonitrile (- - - -) and detected by their absorbance a t 230 nm (-). The flow rate was 1 ml/min. Peptides arenumbered according to their position in the sequence, starting from the N terminus. A, digest of polypeptide chain A; B, digest of polypeptide chain B.

nucleosil c-18column (5 pm, 4.6 X 250 mm) equilibrated in 25 mM ammonium acetate buffer (pH 6). Tryptic core and CNBr peptides were dissolved in trifluofoacetic acid 0.1% and injected onto a nucleosil C-4 column (300 A, 4.6 x 250 mm) equilibrated in trifluoroaceticacid 0.1%. In both cases,peptides were eluted withalinear gradient of ammonium acetate or trifluoroacetic acid to 60% acetonitrile. Fractions suspected of containing more than onepeptide were rechromatographed on the same column in 5 mM potassium phosphate (pH 6) usinga lineargradient of 0-60% acetonitrile. The peptides were dried by lyophilization. Amino Acid and Sequence Analysis-The protein or dried peptides were hydrolyzed in 100 pl of HCI 6 N for 20 h a t 110 “C, and the analyses were performed on a Biotronik LC 5001 amino acid analyzer or a Beckman system 6300 high performance analyzer. Manual sequencing of the isolated peptides was conducted by the 4-(dimethyl-

Downloaded from www.jbc.org at CNRS on October 9, 2006

FIG. 1. Separation of polypeptide chains A and B of NDP FIG. 3. SDS-PAGE (lunes I and 2, 8 pg of protein) and kinase. NDP kinase (about 20 mg) was equilibrated by gel filtration immunoblot analysis (lanes 3-6, 0.4 pg of protein) ofpurified in 50 mM Tris-HCI (pH 8.6), 50 mM glycine, 50 mM P-mercaptoeth- chains A (lanes I , 3, and 5 ) and B (lanes2 , 4 , and 6) of human anol, 1 mM EDTA, 6 M urea and applied to a DEAE-Sepharose CL- erythrocyte NDP kinase. In lanes 3 and 4 proteins were reacted 6B column (1 X 10 cm) equilibrated with the same buffer a t a rate of with immune serum raised against peptide A, whereas in lanes 5 and 20 ml/h. Thecolumn was washed withthe equilibration buffer at the 6 proteins were reacted with immune serumraised against polypeptide same rate until all the polypeptide chain B was eluted, and 1.5-ml B. Lane S contains the molecular mass markers: Q, phosphorylase a fractions were collected. The polypeptide chain A was eluted by the (94,000); b, bovine serum albumin (67,000); c, ovalbumin (43,000); d, addition of 0.1 M NaCl to the above buffer. 0,absorbance a t 280 nm; carbonic anhydrase (30,000); e, soybean trypsin inhibitor (20,300); f, D, NDP kinase activity was measured after dilution of aliquots of the lysozyme (14,400). fractions 10-fold in 50 mM Tris-HCI (pH 7.4), 5 mM MgCI?, and 2-h incubation a t room temperature. The high base-line absorbance is due to theoxidation product(s) of (3-mercaptoethanol.

8786

Erythrocyte NDP Kinase 1 10 20 30 40 M A N C E R T F I A I K P D G V Q R G L V G E I I K R F E Q K G F R L V G L K F M Q A S E D L L K

Nm 2 3 NDPK A

T1

I

T7

TI

NDPK B

TZ

T3

T4 ,

II

1

TZ

L-r

60

U T5

1 - U - A

I,

50

W

T4

T3

L T6

l

J

T7

M l L L R

80

70

TB

L

E H-

90

Nm 2 3

E H Y V D L K D R P F F A G L V K Y M H S G P V V A M V W E G L N V V K T G R V ~ L G E T N P A D TB -1

T9

NDPK A

T9

T10

NDPK B

Nm 2 3 NDPK A NDPK B

TI2

T11

T10 1 ' 1

I1

TI 1

TI2

L~-~---I~"-PLL-N

I

U

TI3 I

100 110 120 130 S K P G T I R G D F C I Q V G R N I I H G S D S V E S A E K E I G L W F H P E E L ~ D Y T ~ ~ A Q TI 2

I

T13 I

T13 I

I

T14

T14

TI 5 II

140

T15

K J I H

T16

TI 7

T18

I u I L S - K

150 Nm 2 3

I

T18

NDPK B

D-V-

FIG. 5. Primary structure of polypeptide chains A (NDPK A ) and B (NDPK E ) from human erythrocytes. The sequence is based on analysisof tryptic peptides and alignment with the amino acid sequence of human Nm23 protein,as deduced from the nucleotide sequence of the correspondinggene (26). amino)azobenzene-4'-isothiocyanate/phenylisothiocyanate doublecoupling technique (16). Immunochemical Techniques-Rabbits received 1 mg of purified polypeptides A or B in complete Freund's adjuvant by injection in the foot pads. Sixweeks later, the rabbitswere boosted with the same antigen in incomplete Freund's adjuvant. At varying time intervals beginning 12 days after thebooster, rabbits were bled. Antisera were stored at 4 "C in the presenceof 0.1% sodium azide. Analytical Procedures-NDP kinase activity was measured with 1.0 mM ATP as donor nucleotide and 0.3 mM 8-BrIDP as acceptor nucleotide as previously reported (17, 18), except thatthefinal reaction volume was 0.52 ml and thatbovine serum albumin (0.5 mg/ ml) was included in the assay mixture. One unit of enzyme activity corresponds to 1 Fmol of ADP formed per min a t 25 "C. Proteins were determined by the procedure of Bradford (19) or of Gornall et al. (20). SDS-PAGEwas performed according to Laemmli (21)using 12.5% gels. After electrophoresis, the proteins were transferred to nitrocellulose sheets asdescribed by Towbin etal. (22). Nitrocellulose was incubated 30 min a t room temperature with 1:1500 or 1:3000 dilution of antisera (forB and A, respectively). The immunochemical detection was performed with alkaline-phosphatase-labeled anti-rabbit immunoglobulins.

of NDP kinase reactivation following treatment with urea were notexamined in greatdetail, we observed that this process was rapid and occurred in good yields even at relatively high (5-20 PM) concentrations of protein. This might be due to the fact that the small monomeric units of NDP kinase probably consist of a single structural domain,whereas other phosphotransferaseswhose three-dimensional structure is known consist of two or three structural domains(23,24). When isolatedpolypeptides were renatured separately, then resubmitted to ion-exchange chromatography under nondenaturing conditions (Fig. 2, A and B ) ,they migrated as single distinct species. The molecular mass of each protein, determined by gel permeation chromatography under nondenaturing conditions,was around 105 kDa, confirming the hexameric of A6 and B6 formsof NDP kinasewere structure. The peaks coincident, which indicatesthat differences in molecular masses observed in SDS-PAGE are rather artefactual. When polypeptides A and B were renatured together,several species with retention times on DEAE-Sepharose intermediary between A6 and B6 (Fig. 2C) and very similar to that of the RESULTS AND DISCUSSION Purification under Denaturing Conditions of Two Polypep- native NDP kinase were generated (data not shown). This experiment favors the hypothesis that NDP kinase isoentideChains from Human Erythrocyte N D P Kinase-NDP zymes result from random association of two different kinds kinase purified, as described under"ExperimentalProcedures" (specific activity 800 units/mg of protein), was heter- of polypeptides.Isoenzymes generated by renaturation of ogeneous by ion-exchange chromatography on DEAE-Seph- polypeptides A and B in mixture and partially separated by SDSarose under native conditions. SDS-PAGE analysis revealed ion-exchange chromatography were re-analyzedby the existence of two distinct polypeptides with apparent mo- PAGE. The expected pattern emerged, i.e. the basic isoenlecular masses of 20.5 and 19 kDa (2). To establish whether zymes were richer in B polypeptide, whereas the acidic isothese polypeptides are different entities or whether they ariseenzymes were richer in A polypeptide (not shown). Catalytic and Immunochemical Properties of A6 and B6 by covalent modification such proteolysis or phosphorylation of a single species, they were isolated by ion-exchange chro- Isoenzymes-The kinetic properties ofA6 and B6 forms of matography on DEAE-SepharoseCL-GB in the presence of 6 erythrocyte NDP kinase were quite similar. The specific activities determined with 1 mM ATP and 0.3 mM 8-BrIDP M urea (Fig. 1). Fractions containing NDP kinase in urea (which we called A for acidic and B for basic) were diluted were of 550 units/mg of protein for A6 and 740 units/mg of with 50 mM Tris-HC1 (pH7.6) plus 1mM magnesium acetate protein for B6. At a fixed concentration of ATP (1 mM), the and dialyzed against the same buffer, allowing recovery of apparent K,,, for 8-BrIDP was 33 PM for A6 and 60 pM for B6. more than two-thirdsof initial activity. Although the kinetics Excess nucleoside diphosphate inhibited both enzyme activi-

Downloaded from www.jbc.org at CNRS on October 9, 2006

NDPK A

N W I Y E 2'15

Erythrocyte NDP Kinase

8787

Downloaded from www.jbc.org at CNRS on October 9, 2006

Peptides Resulted from TrypticDigest-Amino acid composition of chain A and B revealed many similarities (not shown). However, the content inbasic residues was higher in the case of form B. Moreover, the tryptic map of polypeptides A and B was significantly different (Fig. 4). These results did not support the hypothesis that chainsA and B result by simple covalent modification of thesameprecursor. While these experiments were in progress, Lacombe et al. (11)isolated and sequenced the gene encoding for NDP kinase from Dyctiostelium discoideum. The protein showed high structural homology to human Nm23 (26)3 and Drosophila melanogaster Awd proteins (27). Moreover, the amino acid composition of polypeptide B, andespecially of polypeptide A, was strikingly similar to that of human Nm23 protein.Knowing the primary structure of human Nm23 protein as deduced from the nucleotide sequence of the corresponding gene (26), we determined thesequence of subunits A and B of human erythrocyte NDP kinase. CarboxymethylatedA6 and B6isoenzymes were subjected to tryptic digestion and peptides were isolated by reverse-phase HPLC ina pure state and in mostcases with a good yield (Fig. 4). They were then easily identified by amino acidcomposition(A6 form) and manual degradation steps (B6 form).For the latterisoenzyme (Fig. 5), only one overlap with one CNBr peptide (not shown) was necessary for the fragment 40-43 (thetrypticpeptideT7having beenlost during the purification). The tryptic peptides T I 0 (A) and T11 (B) were isolated from the tryptic core (not shown). The tryptic peptides TI (A) and T1 (B) could not be sequenced, indicating that theN terminus of both subunits was blocked. The same observation was recently reported by Kimura et al. (12) for rat nucleoside diphosphate kinase. All other fractions were examined, and no evidence was obtained for any of the tryptic peptides which was incompatible with the reference sequence. Thus, asshown in Fig. 5 , all evidence indicated that the A6 isoenzyme hasthepredictedprimarystructure of human Nm23 protein. ChainB of human NDP kinase,which is itself 98% homologous to rat NDP kinase, differed from chain A by 18 positions (88% homology). It is worthwhile to I I I I I mention that seven of the differing amino acids are "clustered at the C-terminal end of the molecule. 16 18 20 From the primary structureof the two polypeptide chains T ime (min) A and B we established theirmolecular masses as 17,143 and FIG. 6. Isolation by reverse-phase HPLC of the phosphoryl- 17,294, respectively, whereasmeasurement by SDS-polyacrylated peptide resulted from tryptic digestion of human eryth- amide gel electrophoresis systematicallygave values of 20,500 rocyte chain A NDP kinase. Experimental conditions are the same (A) and 19,000 (B), respectively. This anomalous electrophoas in Fig. 4. "'P-Labeled protein (1 mg) was digested with TPCKretic migration seems to be commonly encountered with other trypsin for 3 h at 25 " C ,and the reaction productwas directly applied NDP kinases (ascite Ehrlichcells (5), HeLa cells (28)) where t o a nucleosil (2-18 column and eluted with a linear gradient of 060% acetonitrile. A, part of the chromatogramwhere a peak contain- two bands (21,000 and 18,000) are also generally found. ing radiolabeled peptide is found. Approximately 70% of the applied Identification of the Phosphorylation Site-Earlier studies radioactivity was recovered in the peak labeled by a n asterisk. The (29, 30) haveshown that the reaction of ATP with NDP inorganic phosphate (Pi)did not interact with the column. B, shift of kinase resulted in the reversible transfer of the y-phosphate elution time of this peptide after dephosphorylation.The radioactive of ATP to a histidine residue of the enzyme yielding 1- or 3fraction was lyophilized and then left a t room temperature, redisphosphohistidine. The phosphohistidine linkage is relatively solved, and rechromatographed under the same conditions. unstable, and it was therefore difficult to identify the phosmolecule. Tryptic digestion of ties. Both enzyme preparations were phosphorylated (17, 25) phorylation site in the to an extent close to one phosphate group/subunit (0.9 for A6 labeled A6 isoenzyme allowed isolation of phosphorylated peptide(s) by reverse-phase HPLC. Approximatively 70% of and 0.75 for B6). Antibodies raised against purified chains A and B were used radioactivity was associated with one peak (Fig. 6). Under t o compare their immunoreactivity in Western blot analysis acidic conditions or at room temperature (thiswork), the ['"PI phosphate is released from histidineresidue. The productwas (Fig. 3)and enzyme activityinhibition.Antiseradirected against peptide A recognized both polypeptide chains in im- then analyzed by HPLC. Before treatment, the peptide was munoblot experiments and inhibited the activityof both A6 eluted at the position labeled by an asterisk in Fig. 6. After became and B6 isoenzymes, althoughtodifferentextents. Cross- incubation at room temperature,itselutiontime reactivity was much less visible with antisera raised against ,"Very recently, Stahl et al. ( 3 5 ) described the isolation and sepeptide B. quence of a Nm23-related gene, Nm23-H2. In the cited paper ( 3 5 ) , Amino Acid Composition of Isolated Chains and Analysisof Nm23 is now called Nm23-H1.

Erythrocyte NDP Kinase

8788 10 Nm23-A1 NDPK A NDPK E N m3 2- E 2 Rat Awd D.disc MST M. xan.

20

M .xan.

50

40

I

I

MANCERTFIAIKPDGVQRGLVGEIIKRFEQKGFRLVGLKFMQASEDLLKEH MANCERTFIAIKPDGVQRGLVGEIIKRFEQKGFRLVGLKFMQASEDLLKEE M A N Z ’ E R T F I A I K P D G V Q R G L V G E I I K R F E Q K G F R L V ~ K F ~ A S E E H L ~ H MANLERTFIAIKPDGVQRGLVGEIIKRFEQKGFRLVAMKFLRASEEELKQH M A N L E R T F I A I K P D G V Q R G L V G E I IKRFEQKGFRLVAMXFLRASEEHLKQH KPDGVQRGLVGKIIE

60

Nm23-H1 NDPK A NDPK E Nm23-H2 Rat Awd

30

7 0

80

90

I

100

I

YVDLKDRPFFAGLVKYMHSGPVVAMVWEGLNVVKTGRVMLGETNPADSKPG~ Y V D L K D R P F F A G L V K Y MHSGPVVAMVWEGLNVVKTGRVMLGETNPADSKPG YIDLKDRPFF~GLVKYMNSGPVVAMVWEGLNVVKTGRVMLGETNPADSKPG YIDLKDRPFFPGLVKYMNSGPVVAMVWEGLNVVKTGRVMLGETNPADSKPG YIDLKDRPFFPGLVKYMNSGPVVAMVWEGLNVVKTGRVMLGETNPADSKPG Y A E ~ R P F F P G L V ERPFFGGLVSFITSGPVV D. disc. A.RPFF.K~QF!~JISGPVV 110

I

120

I

130

I

140

150

characteristic of nonradioactivepeptide T14.Aminoacid coded by two different genes. This seems to be also the case analysis indicated that this peptide corresponded to the seg- for D. discoideum enzyme (27). Moreover Steeg and Liotta (34) have identified in humans a second NDP kinase related ment from Asn-115 to Lys-128 (Asn-Ile-Ile-His-His-Gly-SerAsp-Ser-Val-Glu-Ser-Ala-Glu-Lys). Although an amino acid gene which was recently sequenced (35). Thededuced primary sequence analysis was not possible, these results corroborated structure of the protein encoded, Nm23-H2, is identical with previous studies (29, 30) and favor strongly the fact that the the sequence of the human erythrocyte NDP kinase B (Fig. site of phosphorylation in erythrocyte NDP kinase is His- 7). These results confirm and extend recent conclusions (27) endowedwith catalyticfunctions, 118. This conclusion is also indirectly supported by the fact thatNm23proteinis that this residue is the only histidine conserved in all forms namely is a NDP kinase. of NDP kinase for which the primary structure has been Acknowledgments-We thank 0. Birzu for useful suggestions and established. CONCLUSIONS

NDP kinases are ubiquitous enzymes involved in the formation of nucleoside triphosphates from the corresponding diphosphates. As regeneration of nucleoside triphosphates can be performed by other enzymes, such as phosphoglycerate kinase or pyruvate kinase present inlarge amounts in allcells and exhibiting alsoa broad specificityfornucleotide substrates, it was supposed that NDP kinase might have other functions in the cell. The relationship between NDP kinase ( 6 ) ,elongation and GTP-dependent proteins such as tubulin factor 1 (31), and plasma membrane G-proteins linked to adenylate cyclase function (32) suggested that NDP kinase channels the high-energy phosphate from ATP tothese proteins and thus participatesmore in complex cellularfunctions. This hypothesis received strong support from the recent and surprising observation that NDP kinase from D. discoideum (27) is highly homologous to Nm23 and Awd proteins involved in mammalian tumor metastasis and Drosophila development, respectively. Thispromptedustodeterminetheprimary structure of the isolated chains of the human enzyme erythrocyte NDP kinase. From our studies it emerged that these chains are separate entities and most probably they are en-

constructive criticisms, M. VLron, M.-L. Lacombe, and S. Michelson for fruitful discussion and helpful comments, J. Bellalou for kind advice in immmunochemical experiments, and M. Ferrand for excellent secretarial help. REFERENCES 1. Cheng, Y. C., Agarwal, R. P., and Parks, R. E., Jr. (1971) Biochemistry 10,2139-2143 2. Presecan, E., Vonica, A,, and Lascu, I. (1989) FEBS Lett. 250, 629-632 3. Yokoyama, M., Uesaka, H., and Ohtsuki, K. (1986) FEBS Lett. 206,287-291 4. Kimura, N., and Shimada, N. (1988) Biochem. Biophys. Res. Commun. 151, 248-256 5. Koyama, K., Yokoyama, M., Koike, T., Ohtsuki, K., and Ishida, N. (1984) J. Biochem. (Tokyo) 95, 925-935 6. Nickerson, J. A., and Wells, W. W. (1984) J. Biol. Chem. 259, 11297-11304 7. Robinson, J . B., Jr., Brems, D. N., and Stellwagen, E. (1981) J. Bid. Chem. 256,10769-10773 8. Parks, R.E., Jr., and Agarwal, R. P. (1973) in The Enzymes (Boyer, P. D., ed) Vol. 8, pp. 307-333, Academic Press, New York 9. Islam, K., and Burns, R. G. (1984) Anal. Biochem. 137,8-14 10. Muiioz-Dorado, J., Inouye, M., and Inouye, S. (1990) J. Biol. Chem. 265. 2702-2706 11. Lacombe, M.:L., Wallet, V., Troll, H., and VLron, M. (1990) J .

Downloaded from www.jbc.org at CNRS on October 9, 2006

TIRGDFCIQVGRNIIHGSDSV~SAEKEIGLWFHPEELVDYTSCAQNWIYE Nm23-H1 NDPK A TIRGDFCIQVGRNIIEGSD NDPK E Nm2 3-E2 Rat AWd D.disc. M.xan. FIG. 7. Comparison of the sequence of polypeptide chains A and B from human erythrocyte N D P kinase ( N D P K A and NDPK B ) with those of D. discoideum ( l l )M. , xanthus ( l o ) , and rat (12) N P D kinases, human Nm23-H1 and H2 proteins (26, 35), and theD. melanoguster Awd protein (33).

~ ~ t ~ r o cNlIP y t eKinase Biol. Chem. 265,1~12-10018 N.,Shimada, N., Nomura, K., and Watanabe, K. (1990) J. Biol. Chem. 265,15744-15749 13. Lascu, I., Kezdi, M., Goia, I., Jebeleanu, G., Btrzu, O., Pansini, A., Papa, S., and Mantsch, H. H. (1979) Biochemistry 18, 4818-4826 14, Bohme, H.-J., Kopperschlager, G., Schulz, J., and Hoffmann, E. (1972) J. Chro~atogF.69,209-214 15. Waxdal, M. J., Konigsberg, W.H., Henley, W. L., and Edelman, M. G. (1968) B~ochem~stry 7,1959-1964 16. Chang, J. Y., Brauer, D., and Whittmann-Liebold, B. (1978) FEBS Lett. 93,205-214 17. LasW I*, POP, R. Dv Porumb, Hv Presecan, E., and Proinov, 1. (1983) Eur. J. Biochem. 135,497-503 Kezdi, M Y Kiss, L.3 Bojan, o., Pav& T,, and Birzu, 0-(1976) Anal. Biochem. 76, 361-364 19. Bradford, M. M. (1976)Anal. Biochern. 72,248-254 20. Gornall, A. G., Bardawill, C . J., and David, M. M. (1949) J. Biol. Chem. 177,751-766 21. Laemmli, U. K. (1970) Nature 227,680-685 22. Towbin, H., Staehelin, T., and Gordon, J. (1979)Proc. Natl. Acad. Sci. U. S. A. 76,4350-4354 23. Anderson, c. M., Zucker, F. H., and Steitz, T. A. (1979) Science 204,375-380 12. Kimura,

8789

24. Muirhead, H. (1983) Trends Biochem. Sci. 8,326-329 25. Lascu, I., Presecan, E., and Proinov, I. (1986) Eur. J. Biochem. 158,239-243 26. Rosengard, A. M., Krutzch, H. C., Shearn, A., Biggs, J. R., Barker, E., Margulies, I. M. K., Richter King, C., Liotta, L. A., and Steeg, P. S. (1989) Nature 342, 177-180 27. Wallet, v., Mutzel, R., Troll, H., Btrzu, o., Wtlrster, B., VBron, M., and Lacombe, M. L. (1990) J. Nad. Cancer Inst. 82,11991202 28. Ohtsuki, K., Yokoyama, M., Ikeuchi, T., Ishida, N.,Sugita, K., and Satoh, K. (1986) FEBS Lett. 196, 145-1.50 29. Edlund, B., Rask, L., Olsson, P., Walinder, O., Zetterqvist, and Engstrom, L. (1969) Eur. J. Biochem. 9,451-455 30. Colomb, M. e,,cheruy,A., a n d ~ i g n a i sp. , v, (1972)~ i ~ ~ 11,3378-3386 31. Ohtsuki, K., and yokoyama, M. (1987) fjiochem. 3iophys,R ~ ~ . Commun. 148, 300-307 32. Kimura, N., and Shimada, N. (1988) J. Biol. Chem. 263,46474653 33. Biggs, J., Tripoulas, N., Hersperger, E., Dearolf, C., and Shearn, A. (1988) Genes & Deu. 2,1333-1343 34. Steeg, P. S., and Liotta, L. A. (1990) Proc. Am. ASSOC. Cancer Res. 31,504-505 35. Stahl, J. A., Leone, A., Rosengard, A. M., Porter, L., Richter King, C., and Steeg, P. S. (1991) Cancer Res. 51,445-449

o.,

h

Downloaded from www.jbc.org at CNRS on October 9, 2006