Gene, 105 (1991) 31-36 0

1991 Elsevier

GENE

Science

Publishers

B.V. All rights reserved.

31

0378-l Il9/91/$03.50

06028

Nucleoside diphosphate kinase from Escherichia coli; its overproduction and sequence comparison with eukaryotic enzymes (PCR;

Myxoc~ccus

xanthus;

ATP-binding;

Kohara’s

library;

evolution;

nucleotide

metabolism;

recombinant

DNA)

Hiroko Hama, Niva Almaula, Claude G. Lerner, Sumiko Inouye and Masayori Inouye Department of Biochemistry, Robert Wood Johnson Medical School, University of Medicine and Dentistry ofNew Jersey at Rutgers, Piscataway, NJ 08854 (U.S.A.) Received by .I. Messing: 13 May 1991 Revised/Accepted: 9 June/l0 June 1991 Received at publishers: 1 July 1991

SUMMARY

The gene encoding nucleoside diphosphate (NDP) kinase of Escherichia colt' was identified by polymerase chain reaction using oligodeoxyribonucleotide primers synthesized on the basis of consensus sequences from Myxococcus xanthus and various eukaryotic NDP kinases. The gene (ndk), mapped at 54.2 min on the E. colichromosome, was cloned and sequenced. The E. coli NDP kinase was found to consist of 143 amino acid residues that are 51, 45, 45, 42, 43, and 43% identical to the M. xanthus, Dictyostelium discoideum, Drosophila melanogaster, mouse, rat, and human enzymes, respectively. The ndk gene appears to be in a monocistronic operon and, when cloned in a pUC vector, NDP kinase was overproduced at a level of approx. 25 y0 of total cellular proteins. The protein could be labeled with [ Y-~~P]ATP and migrated at a 16.5 kDa when electrophoresed in SDS-polyacrylamide gel, which is in good agreement with the M,. of the purified E. colt’ NDP kinase previously reported.

INTRODUCTION

Nucleoside diphosphate (NDP) kinase is considered to be essential for the maintenance of functional concentrations of all NTPs and dNTPs in a cell, in both prokaryotic and eukaryotic organisms (Parks and Agarwal, 1973;

Correspondence

to: Dr.

UMDNJ-Robert

Wood

Piscataway,

M.

Inouye,

Johnson

Department

Medical

School,

of 675

Biochemistry, Hoes

Lane,

NJ 08854 (U.S.A.)

Tel. (908)463-4115; Abbreviations:

Fax (908)463-4783.

aa, amino

acid(s);

bp, base pair(s);

A’-2-hydroxyethylpiperazine-N’-2-ethanesulfonic 1000 bp; NDK, nucleotide diphosphate NDK; NDP, nucleotide

diphosphate;

d, deoxy;

Hepes,

acid; kb, kilobase or kinase; ndk, gene encoding

nt, nucleotide(s);

NTP, nucleotide

triphosphate; oligo, oligodeoxyribonucleotide; ORF, open reading frame; PCR, polymerase chain reaction; SD, Shine-Dalgarno (sequence); SDS, sodium

dodecyl

sulfate.

Ginther and Ingraham, 1974). Originally described in 1953 (Berg and Joklik, 1953; Krebs and Hems, 1953) it has been extensively studied in various systems (Ginther and Ingraham, 1974; Ratliff et al., 1964; Mourad and Parks, 1966; Nakamura and Sugino, 1966; Glaze and Wadkins, 1967; Saeki et al., 1974; Rodriguez and Ingraham, 1983; Ohtsuki et al., 1984). In spite of these biochemical and genetic studies, the gene for the enzyme had not been cloned and characterized until recently. The gene (ndk) for NDP kinase was first cloned and sequenced from a Gram - bacterium M. xanthus by Mufioz-Dorado et al. (1990a, b), who crystallized this enzyme. Since then, the NDK-encoding genes from D. discoideum (Lacombe et al., 1990) and from rat (Kimura et al., 1990) have been cloned and characterized. Interestingly, two previously identified genes, the nm23 gene of human and mouse (Rosengard et al., 1989) and the awd (abnormal wing disc) gene of Drosophila (Dearolf et al., 1988; Biggs et al.. 1988) were found to

32 Drosophila, mouse, rat, and human, respectively. The gene product was overproduced to approx. 25 y0 of total cellular proteins and shown to be phosphorylated with [ y-“P]ATP.

encode polypeptides that show extensive sequence similarity to NDKs (Wallet et al., 1990). The nm23 gene was identified as a gene that is poorly transcribed in tumor cells of high metastatic potential (Steeg et al., 1988; Bevilacqua et al., 1989) and thus may be a metastasis suppressor gene (Steeg et al., 1988; Bevilacqua et al., 1989). The awd gene was identified as a gene essential for differentiation in Drosophila; mutations in awd cause abnormal tissue morphology and necrosis and widespread abnormal differentiation (Dearolf et al., 1988). The Nm23 polypeptide was

RESULTS ANDDISCUSSION

(a) Identification of the Escherichia coli ndk gene The aa sequence of the M. xunthus NDK shows high similarity to those from various eukaryotic organisms (Biggs et al., 1990; Williams et al., 1991); when aligned, there are a few highly conserved sequences (Fig. 4). On the basis of these conserved sequences, two oligos, A and B, were synthesized; oligo A corresponds to the aa sequence of the M. xunthus NDK from aa 72-79, while oligo B is complementary to the nt sequence corresponding to the aa sequence from aa 117-121 (Fig. 1A). Oligos A and B are 3 1- and 22-mers and a mixture of 32- and 64-mers, respectively. The mixed oligos were used as primers for PCR on phage 2 DNAs encompassing the region at 54 min of the

found to be 78% identical to the Awd polypeptide (Rosengard et al., 1989), and recently the Awd polypeptide was indeed shown to be an NDP kinase that is associated with microtubules (Biggs et al., 1990). Here, we identify the ndk gene of E. coli by PCR using oligo primers which were synthesized on the basis of consensus sequences among NDKs characterized so far. The gene located at 54.2 min on the E. coli chromosome was subsequently cloned and sequenced. It codes for a polypeptide of 143 aa, that is 57, 45, 45, 42, 43, and 43% identical to NDK from M. xunthus, D. discoideum,

A 3’

5’ Oligo

A

CTGGATCCGTTGTTGCTATGGTTTGGGAAGG BZHIIHI

Oligo

B

G

G

G

G

G

3' 5' CAGGATCCGAATCTGATCCATG G GCTG G BZUllHI

Fig. 1. Amplification of the ndk gene by PCR. The nt sequences of primers used for PCR are shown in part A. Amplified DNA fragments after 25 cycles of PCR were analyzed by 5% polyacrylamide gel electrophoresis (panel B). The template DNAs used are from phage L8E3 (phage 427; lane l), 2D5 (phage 428; lane 2). 7F8 (phage 429; lane 3) 5ElO (phage 430; lane 4), 6FlO (phage 431; lane S), and 8El2 (phage 432; lane 6). These phages are from a phage I library for the E. coli genome (Kohara et al., 1987) and were obtained from Dr. A. Ishihama (National Institute of Genetics, Mishima, Shizuoka Ken, Japan). Unique bands around 160 bp are indicated oligos A (31-mer) and B (22-mer) were used as template

by arrows marked and primers.

a and b. On a 50-~1 scale reaction,

50 ng of phage

DNA and 100 pmol of

33 E. coli chromosome. This region was chosen because the ndk gene of Salmonella typhimurium has been suggested to be located near the hiss gene (Rodriguez and Ingraham, 1983). The hiss gene of E. coli is mapped at 54.1 min on the chromosome (Bachmann, 1990). The following six phages were used from a phage i library for the E. coli genome (Kohara et al., 1987); 8E3 (phage 427), 2D5 (phage 428), 7F8 (phage 429), 5ElO (phage 430), 6FlO (phage 431), and 8E12 (phage 432). These phages contain DNA fragments of 14-18 kb from the E. co/i genome, and the DNA fragments overlap by 2.5-10.6 kb with each other encompassing the region from 2627-2692 kb on the map of Kohara et al. (1987). PCR amplified a few DNA fragments from each phage 2 DNA template as shown in Fig. 1B. Oligos A and B are expected to amplify a DNA fragment of approx. 160 bp if the ndk gene is present. The phages 428, 429, and 432 produced unique bands at approx. 160 bp as indicated by arrows in Fig. 1B. These bands were isolated and sequenced. The aa sequence deduced from the nt sequence of band a matched well with the aa sequence of the M. xanthus NDK; between the two primers there are 37 aa for both E. coli and M. xanthus sequences, of which 20 are identical aa (Fig. 3). (b) Sequence of the Escherichia coli ndk gene The result described above indicates that phages 428 and 429 most likely contain the ndk gene. Subsequently, Southern-blot hybridization of various restriction digests of phage 428 and 429 DNA with the PCR product as a probe was carried out (data not shown). From this analysis, a restriction map around the positive hybridization region was constructed (Fig. 2) which revealed that the gene was located at 2646-2646.5 kb on the E. coli chromosome map of Kohara et al. (1987), corresponding to 54.2 min on the E. coli genetic map (Hama et al., 1991). Originally a 2.7-kb EcoRI fragment from phage 429 (from the leftmost end to site E(a) in Fig. 2) was cloned using the

PCR product as probe into the EcoRI site of pUC9 (Vieira and Messing, 1982) and the resulting plasmid was designated pKT9EE. However, it was found that the 2.7-kb EcoRI fragment did not contain the entire ndk gene. Subsequently the 3. l-kb PvuII fragment from phage 428 (from P(a) to P(b) in Fig. 2) was cloned into the SmaI site of pUC9Sma1, resulting in construction of pKT8P3. The nt sequence adjacent to and containing the PCR-amplified region was determined using pKT9EE and pKT8P3 DNA. Fig. 3 shows 724 nt of sequence thus determined, revealing the existence of an ORF of 143 codons starting from the start codon (nt 181-183) to the stop codon (nt 610-612). The N-terminal sequence of 46 aa completely agreed with the sequence directly determined on the purified E. coli NDK (Ray and Mathews, 1991). There is an SD sequence, GAGG, 5 nt upstream from the start codon. There are inverted repeats 60 nt downstream from the stop codon. Therefore, a transcript from this region could form a stable stem-loop structure with dG of -16 kcal calculated according to Devereux et al. (1984). Since this structure contains a stretch of U at the 3’ end, it probably serves as the signal for Rho-independent transcription termination (Rosenberg and Court, 1979). In Fig. 3, the sequences used as primers for PCR are boxed. (c) Comparison with other NDKs The aa sequence determined above was compared with all the known NDKs from various sources (Fig. 4). The

60 CTCATTTTATTTTAAAAAAA

TGTTACCTTCATCAATAGTCRACGGCCCTGTTGCCCTGTTGCTCAT~

TAATCCGCGCCATCTCGTACGCTGGTACAGACAACAACAG

i’(a) 429

/

E(a)

P(b)

__--

**

*,--

w I

***-

*-.a

_m

--__

PS I

300 40

CACCTGACCGTTGAACAGGCACGTGGCTTTTATGCTGAACCCGTTCTTT H LTVEQARGFl'AEHDGKPFF

360

--__ K I

0.2 kb

E(a)Ps II

_ -

as blackened

map of the 3.1-kb PvuII fragment carried

from the phage

vector

(not shown)

enzymes used are: E,EcoRI; one site occurs,

the ndk

are flanked by two EcoRI sites

(Kohara

K, KpnI; P,PvuII;

(a) and (b) are added.

containing

420 80 480 100

ACGGTTCTGAT H

G

s

D

540 120 600

SVESAAREIAYFFGEGEVCP

140

CGCACCCGTTAATAATTTCGTAAATGCCGCGTGCAAACGTGGC~TCCGTGCGCC~G~TT

660

R

T

R

TGTACAATGCAGCGCCCCCGGACG

by the phage clones are indicated

bars. The genomic fragments

GA.9

P(b)

ndk

DNA fragments

TCGTGGTTTCCGTGCTGGAAG

~GTCGAATCTGCCGCTCGCGAAATCGCTTATTTCTTTGGCGAAGGCGAGTGTG~CCG

--._ I

Fig. 2. Restriction

60

GGTACTCTGCGCGCTGATTACGCTGACAGCCTGACCGAAACGGTAC

-

gene. Genomic

20

GNIFARFEAAGFKIVGTKML

GTLRADYADSLTENGT

--__

240

GGTAATATCTTTGCGCGCTTTGAAGCTGCTGCAGGGTTC-TTGTTGGCACC-TGCTG

AACGCCGTTCAGCGTCACCGCGATCTGCTGGGCGCGACCACGCACTGGCT NAVQRHRDLLGATNPANALR

2 kb 428

AAAAAACGTCATT

E(b)

" -

phage

180 SD

-10

ATGGCTATTGAACGTACTTTTTccATc*TcAAAccG~cGcGG**Gc IrKPNA”?.KN”I MAIERTFS

GATGGTCTGGTTGAATTCATGACCTCTGGCCC D GLVEFMTSGPIVVSVLEGE

ptlage

120

-35

TCCA

GCAGCCG~TCACCGGGGCGTTTCTTTTTTCAACCC

720

724

Fig. 3. The nt sequence of the ndk gene and the deduced Putative -35 and -10 regions and the SD are underlined. are underlined

with facing arrows.

The sequences

aa sequence. The inverted

et al., 1987). Restriction

repeats

Ps, PstI. When more than

to primers for PCR are boxed. The nt sequence was determined dideoxy chain-termination method (Sanger et al., 1977).

corresponding by the

34 000000

E. coli M. xanthus Dictyostelium Drosophila Mouse Rat Human

Fig. 4. Comparison

00000

0

000

0

000

000

00

00

00

50 50 55 52 51 51 51

M?SERTFSIIKPNAVAKNVIGNIFARFEAAGFKIVGTKMLHLTVEQARGF MAIERTLSIIKPDGLEKGVIGKIISRFE~KGLKPVAIRLQHLSQAQAEGF MSTNKVNKERTFLAVKPDGVARGLVGEIIARYEKKGFVLVGLKQLVPTKDLAESH MAANKERTFIMVKPDGVQRGLVGKIIERFEQKGFKLVALKFTWASKELLEKH MANSERTFIAIKPDGVQRGLVGEIIKRFEQKGFRLVGLKFLQASEDLLKEH MANLERTFIAIKPDGVQRGLVGEIIKRFEQKGFRLVAMKFLRASEEHLKQH MANCERTFIAIKPDGVQRGLVGEIIKRFEQKGFRLVGLKFMQASEDLLKEH *** 00 l b 00 0 l

E. coli M. xanthus Dictyostelium Drosophila Mouse Rat Human

00 0 000 00 00 000 0 00000000 00 000000 0 Y~HDGKPFFDGLVE~MTSGPIWSVLEGENAVQRHRDLLGATNP~ALA YAVHAARPFFKDLVQFMISGPWLMVLEGENAVLANRDIMGATNPAQAAE YAEHKERPFFGGLVSFITSGPWAMVFEGKGWASARLMIGVTNPLASAP YADLSARPFFPGLVNYMNSGPVVPMVWEGLNVVKTGRQMLGATNPADSLP YTDLKDRPFFTGLVKYMHSGPWAMVWEGLNVVKTGRVMLGETNPADSKP YIDLKDRPFFPGLVKYMNSGPWAMVWEGLNKTGRVMLGETNPADSKP ~DLKD~PFFAGL~YMHSGPW~EGLN~TGR~LGETNPADSKP l ** l e l l e* l *e 0 l l *e le

100 100 105 102 101 101 101

E. coli M. xanthus Dictyostelium Drosophila Mouse Rat Human

000000 0 00000 0 0 000000 0 0 GTLRADYADSLTENGTHGSDSVESAAREIAYFFGEGEVCPRTR GTIRKDFATSIDKNTVHGSDSLENAKIEIAYFFRETEIHSYPYQK GSIRGDFGVDVGRNIIHGSDSVESANREIALWFKPEELLTE~PNPN-LYE GTIRGDFCIQVGRNIIHGSDAVESAEKEIALWFNEKELVTWTPAAKf)WIYE GTIRGDFCIQVGRNIIHGSDSVKSAEKEISLWFQPEELVEYKSCAQNWIYE GTIRGDFCIQVGRNIIHGSDS~SAEKEIGL~FKPEELIDYKSCAHD~Y GTIRGDFCIQVGRNIIHGSDSVESAEKEIGLWFHPEELVDYTSCAQNWIYE l 00 l 0 l l *** l a*

143 145 156 153 152 151 152

of the E.coli NDK with other NDKs.

M. xanthtts (Mufloz-Dorado

The deduced

et al., 1990a), D. discoideltm (Lacombe

(Kimura et al., 19901, and human (Rosengard et al., 1989). Identical all NDKs are indicated by blackened circles

aa sequence

of the ndk gene product

et al., 1990), Drosophila

of E. coli was compared

(Biggs et al., lPPO), mouse

aa between E. coli and 1%. xanthus are indicated

E. coli sequence can be well aligned with that of the M. ~~~~~~~ NDK without any gaps except that the M, xa&zthus NDK has two extra aa at the C-terminal end. There are 8 1 identical aa between the two sequences (57 “/, identity). Similarly, the E. coli sequence can be well aligned with other NDKs again without any internal gaps; there are 65 identical aa (45% identity) with D. di.~co~deu~~65 aa (459.~) with ~roso~~iZu, 60 aa (427~) with mouse, 61 residues (43 %) with rat, and 6 1 residues (43 % ) with the human NDK. There are a total of 43 aa (30%) conserved in all NDKs. This result strongly demonstrates that the ORF encodes NDK. (d) Identification of the gene product E. coli JM83[pKT8P3] was found to overproduce a 16.5-kDa protein, when examined by SDS-polyacrylamide gel electrophoresis (lane 2, Fig. 5A). The A4,. agrees well with the deduced iw, (15 462) of the E. coli NDK from Fig. 3 as well as the M, of the purified NDK from E. coli (16 kDa; Ohtsuki et al., 1984). Since stop codons are found in all three reading frames in the upstream region shown in Fig. 3, putative -35 and -10 regions can be identified in this region; TAGTCA (nt 96-101 in Fig. 3) for the -35 region and TATAAT (nt 119-124) for the -10 region. Since the ndk gene has its

(Rosengard

with NDKs

from

et al., 1989), rat

by open circles. Conserved

aa among

own transcription termination signal, the E. coli ndk gene appears to consist of a monocistronic operon. Cells were able to overproduce NDK at a level of approx. 25 % of total cellular proteins at stationary phase of growth (data not shown). Since the ndk gene is cloned opposite to the transcriptional direction from the IQC promoter in the vector, this expression came from the original promoter. The M. xnnthus NDK was also overproduced under the fat promoter in E. coli up to 13% of the total protein in the soluble fraction (Mufioz-Dorado, 1990b). In both cases, cell growth was somewhat slowed by the overproduction. It was also found that the overproduced band was readily labeled with [ y-32P]ATP (lane 2, Fig. 5B). It is interesting to note that NDKs are highly conserved throughout evolution having 43% identity between the E. coli and the human enzymes. As human and mouse Nm23 function as metastasis suppressors (Steeg et al.. 1988; Bevilacqua et al., 1989) and ~~os~~~i~a Awd is essential for differentiation by associating with microtubule (Biggs et al., 1990), NDK is likely to play an important role not only in maintaining functional concentrations of NTPs and dNTPs in the cell but also in providing GTP to various GTP-binding proteins required for essential cellular functions. Although NDK of E. coil was first described in 1953 (Berg and Joklik, 1953; Krebs and Hems, 1953) and

35 bacterial and eukaryotic enzymes. The gene (ndk), mapped at 54.2 min on the E. coli chromosome, was cloned and sequenced. (2) The E. coli NDK consists of 143 aa and shows 57, 45,45,42,43, and 43 y0 identity to NDKs from M. xanthus, D. discoideum, Drosophila, mouse, rat and human, respectively. (3) The overproduction of the E. coli NDP kinase was achieved at a level of approx. 25 “/, of total cellular proteins.

ACKNOWLEDGEMENTS

This work was supported by US Public Health Service grants GM26843 (to S.I.) and GM 19043 (to M.I.) from the National Institutes of Health and a grant from Takara Shuzo Co., Ltd. (to M.I.). One of the authors (C.G.L.) was supported by Postdoctoral Fellowship GM 12446 from the National Institutes. of Health.

REFERENCES Fig. 5. Expression product.

of the ndk gene and phosphorylation

Of total cellular

[y-3*P]ATP stained

40 pg were incubated

with 0.5 PM of

and loaded on a 20% SDS-polyacrylamide

with

Coomassie

(panel B). Lanes:

blue

(panel A) and

1, E. coli JM83

pUC9; 2, JM83 carrying the positions

protein

of the ndk gene

pKT8P3.

of size markers.

(Vieira

exposed

and Messing,

Arrowheads

Plasmid

gel. The gel was to x-ray

film

1982) carrying

on the left margin indicate

pKT8P3

contains

the 3.1-kb PvuII

Bachmann,

B.J.: Linkage

Berg, P. and Joklik, Bevilacqua,

from phage 428 DNA that was cloned into the unique SmaI site

carcinomas cal indicators

NDP

kinase

vested

containing extracts

was labeled

at midlog

phase,

and SmaI

with [y-32P]ATP washed

with

obtained

by sonication

13000 x g for 10 min and

as follows.

Cells were har-

10 mM Tris’ HCl buffer pH 7.7

1 mM EDTA and resuspended were

sites were reversed.

in the same buffer. Crude cell followed

the supernatant

by centrifugation

was used

enzyme. 40 pg protein was mixed with 0.5 PM [y-“PJATP in

the

buffer

EDTA/lOO

containing

20 mM

mM NaCl/O. 1 X Lubrol

ice for 15 min and the reaction gel reducing

buffer. Samples

SDS-polyacrylamide

Hepes

buffer

PX. The mixture

stopped

by adding

as a source

at of

(200 Ci/mmol) pH

8.0/l mM

was incubated

on

SDS-polyacrylamide

were boiled for 5 min before loading on 20:~

gel.

8. Micro-

between

nucleoside

172 (1953) 1008-1009.

G., Sobel, M.E., Liotta, L.A. and Steeg, P.S.: Association

SmaI. Plasmid pUC9 SmaI was constructed of the BumHI

Nature

low nm23 RNA levels in human

fragment

a way that the positions

W.K.: Transphosphorylation

polyphosphates.

ofpUC9

from pUC9 in such

coli K-12, Edition

map of.&cherichiu

biol. Rev. 54 (1990) 130-197.

primary

infiltrating

with lymph node involvement of high metastatic

ductal

of

breast

and other histopathologi-

potential.

Cancer

Res. 49 (1989)

5185-5190. Biggs, J., Tripoulas, Analysis

N., Hersperger,

mutations

of Drosophila. Genes

Biggs, J., Hersperger,

E., Dearolf, Develop.

metastasis

L.A. and Shearn,

to a mammalian

codes for a nucleoside

induced Devereux,

of awd”‘, a cell-autonomous by hybrid

dysgenesis.

J., Haeberli,

sequence

analysis

Develop.

P. and Smithies, programs

A.: A

gene associated

diphosphate

Cell 63 (1990) 933-940. Dearolf, C.R.. Tripoulas, N., Biggs, J. and Shearn, sequences

A.:

2 (1988) 1333-1343.

E., Steeg, P.S., Liotta,

Drosophila gene that is homologous with tumour

C. and Shearn,

between the prune and Killer ofprune

ofthe lethal interaction

kinase.

A.: Molecular

Biol. 129 (1988)

169-178.

0.: A comprehensive

for the VAX. Nucleic

con-

of Drosophila

lethal mutation

Acids

set of Res.

12

(1984) 387-395.

purified in 1984 (Ohtsuki et al., 1984), surprisingly mutations in the ndk gene have never been isolated as dna, fts or any other mutations. This may be due to pleiotropic effects of mutations in the ndk gene caused by the association of the E. coli NDK with various cellular proteins. Investigations to examine whether NDK is associated with any cellular proteins in E. coli, are now in progress in our laboratory.

Ginther, CL. and Ingraham, J.L.: Nucleoside diphosphokinase Sulmonellu typhimurium. J. Biol. Chem. 249 (1974) 3406-3411. Glaze,

R.P. and Wadkins,

triphosphate-adenosine

Properties

of a nucleoside

and its relationship

diphosphate

exchange

diphospho-

to the adenosine reaction.

J. Biol.

Chem. 242 (1967) 2139-2150. Hama,

H., Lerner,

phate

kinase

C., Inouye,

S. and Inouye,

M.: Nucleoside

from Escherichiu co/i. J. Bacterial.

diphos-

(1991) in press.

Kimura, N., Shimada, N., Nomura, K. and Watanabe, K.: Isolation and characterization of a cDNA clone encoding rat nucleoside diphosphate

(e) Conclusions (1) The ndk gene of E. coli was identified by PCR using oligo primers on the basis of consensus sequences from

CL.:

kinase from liver mitochondria

of

Kohara,

kinase.

J. Biol. Chem. 265 (1990) 15744-15749.

Y., Akiyama,

K. and Isono, K.: The physical

map of the whole

E. coli chromosome: application of a new strategy for rapid analysis and sorting of a large genomic library. Cell 50 (1987) 495-508.

36 Krebs,

H.A. and Hems,

phosphates

R.: Some reactions

in animal

tissues.

of adenosine

Biochim.

Biophys.

and inosine

Acta

12 (1953)

172-l 80. Lacombe,

of a nucleoside

cloning

kinase from Dicfyostelium discoideum. J.

diphosphate

Biol. Chem. 265 (1990) 10012-10018. N. and Parks

kinase.

J., Inouye,

M. and Inouye,

nucleoside

diphospho-

gene. J. Biol. Chem. 265 (1990a) Mufioz-Dorado,

J., Inouye,

diphosphate

and sequencing

M.: Nucleoside

xanthus, II. Biochemical

of the

diphosphate

Y.: Metabolism

diphosphate reaction

characterization.

J.

of deoxyribonucleotides.

J.

Parks

T. and

Ishida,

N.: Nucleoside

kinase in Escherichiu co/i: its polypeptide

intermediate.

Biochem.

Jr., R.E. and Agarwal,

structure

and

Int. 8 (1984) 715-723.

R.P.: Nucleoside

Boyer, P.D. (Ed.), The Enzymes.

Academic

R.H.,

Lardy,

triphosphate-nucleoside

diphosphokinases.

In:

Press, New York,

1973,

Ray, N.B. and Mathews, link between

H.A. and Kuby,

diphosphate

oside diphosphokinase).

S.A.: Nucleoside

transphosphorylase

(nucle-

C.K.: Nucleoside

intermediary

S.B. and Ingraham,

murium chromosome kinase

in the

metabolism

diphosphokinase: and nucleic

J.L.: Location

of the gene encoding

(ndk). J. Bacterial.

metastasis

Altered mental

a functional acid synthesis.

nucleoside

153 (1983) 1101-l 103.

Cancer

J. Biochem. 76 (1974)

A.R.: DNA sequencing Nat].

Acad.

Sci.

R., Liotta,

adenovirus

with chain-

USA

74 (1977)

L.A. and Sobel, M.E.:

associated 2 Ela

with low tumor

inhibition

of experi-

Res. 48 (1988) 6550-6554.

J.: The pUC plasmids,

mutagenesis

an M13mp7-derived

and sequencing

with synthetic

R., Troll, H., Barzu, O., Wurster,

M.-L.: Dictyostelium nucleoside

homologous

studies on the inhibition

sys-

universal

19 (1982) 259-268.

Wallet, V., Mutzel,

to Nm23

tumor metastasis Williams,

R.L.,

Inouye, diffraction

on the Salmonella

during

Vieira, J. and Messing, Gene

Proc.

G., Pozzatti,

potential,

metastasis.

primers.

H.: Kinetic

of NM23, a gene

expression

E.,

Drosophilu

and aberrant

kinase by desdanine.

S. and Coulson,

inhibitors.

5463-5467. Steeg, P.S., Bevilacqua,

A., Biggs, J.R., Barker,

L.A. and Steeg, P.S.: Reduced

B., Veron, M. and

diphosphate

and Awd proteins

involved

and Drosophila development.

kinase highly in mammalian

J. Natl. Cancer

Inst.

82 (1990) 1199-1202.

J. Biol. Chem. 239 (1964) 301-309.

Curr. Top. Cell. Regul. 32 (1991) in press. Rodriguez,

involved

Annu. Rev. Genet.

342 (1989) 177-180.

diphosphate

623-629. Sanger, F., Nicklen,

Lacombe,

R.L., Weaver,

in tumour

Nature

tem for insertion

pp. 307-333. Ratliff,

protein

metastatic

M., Koike,

H.C., Shearn,

I.M.K., King, C.R., Liotta,

terminating

Biol. Chem. 241 (1966) 4917-4922. K., Yokoyama,

A.M., Krutzsch,

Margulies,

of nucleoside

Biol. Chem. 265 (1990b) 2707-2712. H. and Sugino,

Rosengard,

development.

2702-2706.

S. and Inouye,

kinase from Myxococcus

Ohtsuki,

sequences

of RNA transcription.

Saeki, T., Hori, M. and Umezawa,

S.: Nucleoside

xanthus, 1. Cloning

from My~ococcus

Nakamura,

D.: Regulatory

and termination

Nm23/Awd

Jr., R.E.: Erythrocytic

J. Biol. Chem. 241 (1966) 3838-3844.

Muiioz-Dorado, kinase

M. and Court,

promotion

13 (1979) 319-353.

M.-L., Wallet, V., Troll, H. and Veron, M.: Functional

Mourad,

Rosenberg,

typhi-

diphospho-

Mufioz-Dorado,

M. and Arnold, analysis

J., Jacobo-Molina,

E.: Crystallization

of nucleoside

diphosphate

coccus xanfhus. J. Mol. Biol. 220 (1991) 5-7.

A., Inouye,

and preliminary kinase

S.,

X-ray

from My.xo-