Isolation and characterization of the gene encoding bovine adenylate kinase isozyme 2

Isolation and characterization of the gene encoding bovine adenylate kinase isozyme 2

Gene, 93 (1990) 221-227 Elsevier 221 GENE 03650 Isolation and characterization of the gene encoding bovine adenylate kinase isozyme 2 (Mitochondri...

719KB Sizes 3 Downloads 104 Views

Gene, 93 (1990) 221-227

Elsevier

221

GENE 03650

Isolation and characterization of the gene encoding bovine adenylate kinase isozyme 2 (Mitochondrial enzyme; recombinant DNA; pseudogenes; genomic library; phage A vector; alternative splicing; multiple transcription start points; promoter activity)

Hiroshi Tanaka*, Mamoru Yamada*, Fumio Kishi* and Atsushi Nakazawa Department of Biochemistry, Yamaguchi University School of Medicine. Ube, Yamaguchi 755 (Japan) Received by Y. Sakaki: 2 March 1990 Revised: 20 April 1990 Accepted: 30 April and 15 May 1990

SUMMARY

Mitochondrial adenylate kinase isozyme 2 (AK2) exists in two isoforms, AK2A and AK2B, which have the same amino-acid sequence except for the C-terminal portion. We have isolated the gene encoding AK2 from a bovine genomic library. The gene covers about 25 kb and consists of seven exons and six introns. The nucleotide sequences from exon l to the 5' half of exert 6 encode the portion common to AK2A and AK2B, while the sequences of the 3' half of exon 6 and exon 7 direct the unique portions of AK2A and AK2B, respectively. Therefore, an alternative splicing mechanism is suggested in generating two types of mRNA encoding AK2A and AK2B. The 5'-flanking region of the gene lacks a TATA box, but contains three CAAT boxes. The G + C content of this region is high and eight copies of GC box are found. These features of the promoter region resemble those of'housekeeping' genes. S l mapping and primer extension analyses revealed multiple transcription start points. The 58 l-bp region just upstream from the start codon functions as a promoter in the expression of the cat gene in Chinese hamster ovary cells.

INTRODUCTION

Adenylate kinase (AK) is a ubiquitous enzyme which contributes to homeostasis of adenine nucleotide composition in the cell (Atkinson, 1977). In vertebrates, three isozymes (AKI, AK2 and AK3) have been identified. AKI is present in the cytosol of skeletal muscle, brain and erythrocytes (Khoo and Russell, 1972), while AK2 is localized in the mt intermembraneous space of liver, kidney, spleen and heart (Khoo and Russell, 1972). AK3, correctly Correspondenceto: Dr. A. Nakazawa, Dept. of Biochemistry, Yamaguchi University School of Medicine, 1144 Kogushi, Ube, Yamaguchi 755 (Japan) Tel. (0836)22-2214; Fax (0836)22-2315. * Present addresses: (H.T.) Dept. ofOrthopedics, Yamaguchi University School of Medicine, 1144 Kogushi, Ube, Yamaguchi 755 (Japan) Tel. (0836)22-2268; (M.Y.) Dept. of Agricultural Chemistry, Faculty of Agriculture, Yamaguchi University, 1677-1 Yoshida, Yamaguchi, Yamaguchi 753 (Japan)Tel. (0839)22-6111; and (F.K.) Dept. of Pediatrics, Yamaguchi University School of Medicine, 1144 Kogushi, Ube, Yamaguchi 755 (Japan) Tel. (0836)22-2258. 0378-1119/90/$03.50© 1990Elsevier SciencePublishers B.V.(BiomedicalDivision)

called GTP:AMP phosphotransferase, exists in the mt matrix ofliver and heart (Tomasselli et al., 1979). To understand the physiological significance of these isozymes, knowledge of the organization and expression of their genes should be necessary. We previously isolated and characterized cDNAs for chicken AK 1 (Kishi et al., 1986), bovine AK2 (Kishi et al., 1987) and bovine AK3 (Yamada et al., 1989). The genes for chicken AKI (Suminami et al., 1988) and human AKI (Matsuura et al., 1989) were also cloned and characterized. In our study on AK2 eDNA, a new Abbreviations: aa, amino acid(s); AK, adenylate kinase; AK, gone (DNA) encoding AK; AKI, AK isozyme I; AK2, AK isozyme 2; AK3, AK isozyme 3; pGal, #-galactosidase; bp, base pair(s); CAT, Cm acetyltransferase; cat,gone encoding CAT; CHO, Chinese hamster ovary; Cm, chloramphenicol; kb, kilobase(s) or 1000bp; mr, mitochondria(I); nt, nucleotide(s); oligo, oligodeoxyribonucleotide;SDS, sodium dodecyl sulfate; tsp, transcription start point(s).

222

10 m

|

I$ t

20 t

25 I

30 kb !

AK2-III ~,AK2-044 AK2-008 ~,AK2.003 2

exon !

5 6

34

Sm e~n5

~

7

E

_-

Sc P ~

ED ._

D

Ac E

exon7

Fig. I. Structure of the bovine A K 2 gene and the sequencing strategy.High-molecular weight D N A of bovine liverwas partiallydigested with Sau3Al, size-fractionatedthrough a sucrose gradient (Maniatis et al.,1980), ligated with the B a m H l + Sail-digested arms of ~ E M B L 4 (Frischauf et al.,1983) and packaged by using Gigapack (Stratagene).A mixture of the purifiedEcoRI fragments of 2bAK2A-3 and ~bAK2B-5 (Kishi et al.,198"/)was ~2P-labeled and used as a probe. Phages were screened by plaque hybridization as described (Benton and Davis, 1977). The filters were washed at the final stringency of 0.1 × SSC I × SSC: 0.15 M NaCi/0.015 M sodium citrate (pH 7.0) and 0 . 1 ~ SDS at 42°C. Phage D N A was purified from the positive plaque and characterized by restriction enzyme mapping and Southern blot analysis. The nt sequences ofthe exons, exon-intron junctions, and the 5'- and Y-flanking regions were determined by the dideoxy chain-termination method (ganger et el., 1977). The exon/intron structure of the gene is depicted in the middle. Closed and open boxes represent coding and noncoding exons, respectively. The start codon (ATG) and polyadenylation signals (ATTAAA, AATAAA) are indicated by vertical arrows. Inserts of four clones covering the AK2 gene are shown in the upper part. Detailed restriction maps of exons ! - 7 and nt sequencing strategy are shown in the lower part. The abbreviations used for restriction enzymes are as follows: A, Aiul; Ac, Accl; B, BamHl; D, D m l ; E, EcoRI; H, H/ndllI; He, Hincll; K, Kpnl; N, Ncol; P, Pstl; Sa, Sad; SA, Sau3Al; Sc, Seal; Sin, Smal; St, Sml; X, Xbal. The nt sequences determined in this study have been submitted to the Gene BankX~/EMBL Data Bank with accession numbers D90065, D90066, D90067, D90068 and D90069.

isozyme, AK2B, was found which difl'eredfrom the known AK2 (AK2A) at the C-terminai portion as well as in the tissue distribution (Kishi et al., 1987). The aim of present study was (i)to isolate and characterize the bovine AK2 gene, and (ii)to determine whether two types of mRNA encoding AK2A and AK2B isoforms might be generated by alternative splicing mechanisms.

RESULTS AND DISCUSSION

(a) Isolation of the bovine AK2 gene A bovine genomic library was constructed in the ~EMBL4 vector. Aider about 106 recombinant phages of the library were screened by plaque hybridization usingthe AK2A and AK2B cDNA clones as probes, 68 positive

33

34

Giy ?hr Gin EXON 1 (200 bp)

....

AlaFroZys

GGTACCCAGgtgagctgct -- I~RON 1 (12 kb) .... ccctccacag ~A CCC AAA ?5

Gi),L4~rZe# BXON2 (127 bp)

BXON 3 (101

---

BXON2

?6

FalSer Asp

GGGAAGCTG gtaUttcag .... INT~N Z (2.5 kb) ..... tEaettccag GTGAGTGAT --- EXON3

112 113 Jim GiuJet /.euJspJsp bp) --- GCAGAAAT(;gtaastgact ..... INTRON3 (129 bp) .... ctttttacagCTT GATGAC --- EXON4 144

Thr GD, J.'~t'

EgOS 4 (96 bp) . . . . ACl~GGF,AG gtatttgtct --- Ik~RON4 (5 kb) ...... tatcttacq

145 Zeulle & ¢TG ATT .... BXON 5

168 169 Z~r ,fsp Jsp lie ?itrGIr BION 5 (?4 bp) .... AAA GATGAT8tatgtaaat --- INTRON5 (?00 bp) --- ttcttctcag ATCACTGGG .... ElON 6a 233 Ala ?hr O,~ to,sAs# AAA GAC --- UON 6b (170 bp) .... MON ~ (lg6 bp) -- ~ A~ T ~T

234

--- INTRON6 (2080 bp) -- tttatttca8 CC TAGTAA --- EXON? (877 bp) Fig. 2. Exon-intron organization ofthe bovine AK2 gene. Exon sequences are shown in upper-case, whereas intron sequences are presented in lower-case letters. The corresponding aa residues are shown and numbered above the nt sequence. The lengths ofeach exon and intron are indicated. Exon 6b which is spliced out together with intron 6 to generate AK2B mRNA is placed as an intmn.

223 A° - 923

A~CTGTTGCTACTTATTTAA~TCA`~`~`~`~G~TTATTGTTG~AACA~TTAGGATATTGCCTTTGTTGTTG~TG~TATGTG~rT~TAGGTTrGGGGACCG~T~AAGTG

- 823

~TA~A~TAATATTTTTGTT~A~AGTCTTTG~GCTGCTA~GCG~T~G~TGTGTCCGACTCT~TGCG~CCCAT~GACCGCACGCC~CTAG~TCCTCTG

- 723

TCCCT~GGATTCTCC~GGCAAGAATAGTGG~GT~GGTTGCCATTTCCTTCTCCAAT~CACCAA~GTGA~AAGGG~AAGT~A~GTCGCT~"AG~GTGTCC~

- 623

ACTCCT~GCGACCCC~TGGACTGCAGCCCACCAGGCTCCTCCGTCCxTGGGATTTTCCAGGCA~G~GTACT~GGGTGGGTTGCCATTGCCTTCTCC~AAC

- 523

AAGTCTTTACATATATCCTAG~TTATGTCCTTGGATTTCCTGGA~GT~G~TTGCTAGACC~AAGG~TATTTGATAC~TATTAG~[~r~TCTTCTA~

#CO i

- 423

A~CTGATTGCACTCATTTGTACTCCCACGGG~-~TT~TAGGACTTCACTCGGGTTGCCCGTGCCATTAACTGTCAA~AGCCTACCT~GA~CCCGGAGG

- 323

C•GGCCTCTAGGCTTGCTTCACGTTTTCCCGCCCACTCCGC•CACCCGGCGCG•GCCCCGCCCCTTCGTACAT•GTGGCCCGCCCCTTCCT•CTCACTC•

- 223

~CGATTCTGT(~A-1~CAAA~GCAGCTTT~C~CACTCTGGCTG~GTGGTTGGTAAGCTGA~GTTTCCGT~GCG~TCAGGGGCGGG~GG~GCGG~CTGAC~G

- 123

~C~TGGC~CGGACCAGTGTG~GCGG~G~G~GAGGCGTGCGTGGC~TG~GTGGCGCGCTGGCCTGG~AGCG~11;G~1`cTG~GAGGTGTGCGGGCGGGT1`(;

-

23

AG•AGTG•GAGGT•TTGG•GGCCATGG•TC••AA•GTG•CTG•GG••GAA••GGT•••GGAGTCT••TA•AGG•GT••GGG••GTG•TCTTAGGG•CGCC

+

78

~GGAGC~GG~AA~GGTA~AG~t~aRctRctRaRcc~RRcttRtcRRtRRaRctcRtRa~attc~aRRtc~R~ccRaa~tccRRttcRcccca~a

+I

" - - "

B. ATTAAAAAAGCACTTGCTTCATGTacctt t R t e t g t R c R c a a a t c t c a t c t c a t c t e t R t e t R t R t R t e n t R t e t a t U t R c R t R c a c ~ t R c a c a c a c R t R t R t a ~

tataa2tat2t2tacactcttatacttcttaaattRtaRRCRaRact~tt

tacttctttaRcc~tactctttattttcRa

Fig. 3. The nt sequences oftbe 5'-fianking region (A) and the junction region between exon 6 and intron 6 (B) ofthe bovine AK2 gene. (A) The nt sequences in the 5'-upstream region and the exon 1 are shown in upper case, and those in intron 1 are shown in lower ease, letters. The sequence is numbered from the first nt of the AT(} start codon. The asterisks above and below the nt sequence indicate the tsp observed in SI mapping and primer extension, respectively. CAAT boxes are boxed; GC boxes are doubly underlined, Wavy lines indicate the sequences homologous to the AP- I-binding site consensus. The 15-bp and 12-bp direct repeats are shown by horizontal arrows, The first nt ofthe longest eDNA clone is indicated by a vertical arrow, The overlined sequence is the complementary antisense oligo used in S I mapping and primer extension. (B)The nt sequence of exon 6 is shown in upper case, and that ofintron 6 is shown in lower case, letters. The 3'-end portion ofexon 6 was assigned from the 3'-nt sequence of AK2A eDNA (Kishi et al,, 1987). The polyadenylation signal and GT clusters are underlined.

clones were isolated. By detailed restriction mapping, Southern blotting and partial nt sequencing, four overlapping clones which covered about 30 kb altogether were found to contain the bovine AK2 gene (Fig. 1). The gene was split into seven exons. The sequences at the 5' and 3' ends of each intron conformed to the GT-AG rule (Fig. 2).

Ai

IRXON

Genomi

c

8equonco

AK2A

eDNA

AK2B

eDNA

...... - I

6u

...... I ATC ACT . . . . . .

I/e I}r

P

• r/?r

A m o n g the isolated clones, at least six clones seemed to be

pseudogenes which were homologous to AK2 eDNA. Some of the restriction endonuclease fragments ofgenomic D N A hybridized to the AK2 eDNA probe in the previous Southern blot analysis (Kishi etal., 198"/) were not accounted for by the map o f t h e A K 2 gene. T h e s e fragments

BXON

6b

...... " ' " " 1 GCC ACA TGT AAA GAC. . . . . .

EXON

. . . . . . . . . . . . '°1

't

.... " " " 1

I ATTAAA---I

Ale l ~ r C~,slos Asp

I I

Ala Yhr I

B AK2A mhNA

~ l J~,"'l ~ ~ t - - poly (A)

t AK2 gene

AK2B mRNA

ex

~___poly

(A)

Fig. 4. Alternative splicing ofthe bovine AK2 gene. (A) Comparison of two types ofcDNA and their genomic sequences around exon 6 and exon 7. Boxes indicate ~he exons. Polyadenylation signals are underlined. (B) Schematic representation of alternative splicing. Each numbered box shows the corresponding exon. The exons common to both mRNAs are shown by shaded boxes, while alternate exons (exon 6b and exon 7) are shown by open boxes.

224 were ascribable to the restriction fragments derived from a processed AK2 pseudogene (unpublished results). The nt sequence of the 5'-flanking region of the gene is presented in Fig. 3A. (b) Alternative splicing The nt sequence ofexons 1-5 was common to AK2A and AK2B cDNA. The 5' halfof exon 6 (exon 6a)was common to AK2A and AK2B cDNA, while the 3' half (exon 6b) was

A

C

B 1

CTAG

unique to AK2A c D N A and contained the polyadenylation signal. Exon 7 was located about 2 kb downstream from exon 6 and comprised the unique sequence ofAK2B c D N A which ended at the polyadenylation site. Cleavage of the primary transcript and polyadenylation could occur both at the end of exon 6 and at the end of exon 7, giving rise to AK2A and AK2B mRNA, respectively (Fig. 4A). In the latter case, the primary transcript should be spliced between exon 6a and exon 7. In fact, a sequence CAT/GTAAAG at

23

CT A G

{Slm~ng

I

I"~lll

mRNA

r--'-J, probe al

" ..

il¸¸

Stul 1,;319

tm

1

ATG I+1i

intron 1 I -

primer

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

~

,

mRNA

Ill i¸

I primerextension I e= " i ID

.,il

II - 8 illllB

~

""lql B

"

i t

O

lil

m,

e.

O

ill

~

ell,,

'~

,L

Fig. 5. Mapping of transcripts from the bovine AK2 gene. (Panel A) S I mapping: A single-stranded DNA was synthesized as described (Lompreet al., 1984). A synthetic oligo 5 '-GGGAGCCATGGCCGCCAA-Y (I 8-mer)complementary to nt -9 to + 9 ofexon I which contains the first eodon (Fig. 3A) was saP-end.labeled and used as a primer. An Ml3mpl8 clone of the 1.2-kb Kpnl-F,coRi fragment containing exon I and the 5°-flanking region was used as a template. Al~er a DNA polymerizing reaction, the reaction product was digested with Sml and subjected to strand-separation electrophoresis to obtain an end.labeled 328-bp probe. The probe was annealed to poly(A) + RNA (10/Ag) isolated from bovine liver (Chirgwin et al., 1979) at 58°C for 3 h in 80% formamide/400 mM NaCI/I mM EDTA/40 mM PIPES (pH 6.4). The sample was diluted with ice-cold buffer containing 30 mM K. acetate (pH 4.6)/1 mM ZNSO4/250 mM NaCl/20/~g per ml heat-denatured salmon sperm DNA, and digested with SI nuclease for I h at 37°C. The products were denatured and analyzed on an 8 M urea/8% polyacrylamide gel followed by autoradiography. Lane 1, undigested probe; lanes 2 and 3, digested products by S 1 nuclease (10 and 20 units, respectively). The protected fragments are indicated by arrows. The sequencing ladders (panels A and B) were generated by using the same synthetic oligo as the sequencing primer. (Panel B) Primer extension: The end-labeled primer was hybridized to bovine liver ,poly(A) ÷ RNA (10 #g) and subjected to reverse transcriptase reaction (Matsuura et ai., 1989). The products were analyzed on an 8 M urea/8% polyacrylamide gel followed by autoradiography. The positive fragments are indicated by arrows. (C)The outline of the two experiments is shown. A map ofthe 5' end of the bovine AK2 gene is shown in the middle, with the start codon (ATG), the Sml site and the location ofthe synthetic oligo used as a primer (open box). saP-labeled ends are marked by asterisks.

225 the junction of exons 6a and 6b was homologous to the consensus sequence C ( A ) A G / G T A ( G ) A G T for the 5' splice site (Mount, 1982) (Figs. 2 and 4A). Thus two types of m R N A are probably generated from a single gene by an alternative splicing mechanism (Fig. 4B). In our previous Northern blot analysis, the ratios of AK2A to A K 2 B m R N A s in liver and heart were 1: 5 and 5:1, respectively (Kishi etal., 1987). Therefore, tissue-specific m R N A processing should have occurred in generating the AK2 isoforms. (c) The 3'-end regions Both the 3'-noncoding sequences of AK2A and A K 2 B c D N A were rich in A and T. 'ATTTA' motifs, which are suggested to contribute to m R N A stability (Clemens, 1987), were recognized in the 3' sequences; one in exon 6 and three in exon 7. In addition, G T clusters were found downstream from the polyadenylation site of exon 6 (Fig. 3B), which presumably play an important role in the cleavage and polyadenylation of the primary transcript (Birnstiel et al., 1985). Such clusters were not found in the region downstream from exon 7. (d) Characterization of the 5'-flanking region In S 1 mapping analysis to define the tsp, protected D N A fragments were heterogeneous in size (Fig. 5A). The major

fragments corresponded to transcripts that initiated at nt -46, -59 and -113. No protected fragment was detectable when tRNA was used (data not shown), which ruled out the possibility of artificial protection of the probe. In primer extension analysis of the 5' end of m R N A (Fig. 5B), the reaction products were also heterogeneous in size, corresponding to tsp at nt -27, -37 to -40, -54, -102 and -108. The band patterns generated by the two independent analyses did not correlate exactly, but the fragments in S 1 analysis were always about 5 nt longer than corresponding fragments in pr:aner extension, the reason for which is not clear at present. From these results we concluded that multiple initiation occurred in transcription of the bovine AK2 gene. In the upstream region (Fig. 3A), there were three CAAT boxes (nt -212, -392 and -438), but no canonical TATA box was found. The region from nt -300 to + 300 contained about 6 6 ~ of G + C, and eight GC boxes, which are thought as the binding site of transcription factor Spl (Dynan and Tjian, 1985) were recognized (one in intron 1, two in exon I and five in the 5'-flanking region, two overlapped). Furthermore, seven heptamer sequences were found which resemble the consensus binding site T G A C T C A of transcription factor AP-I (Lee et al., 1987), five of which were on the antisense strand. There were two pairs of directly repeated sequences in the 5'-upstream

TABLE I Functional analysis of the promoter region of the bovine ,4K2 gene Plasmida

Cell extract b (~8 protein/assay)

CAT activity~ (% conversion/30min)

pSV00AK2cat

17 33 28 29 26 4 8 15 3O

13.9 34.6 0 0 2.4 32.4 45.9 69.1 95.8

pSV00AK2rcat pSV00cat PAioeat2 pSV2cat

Relative specific activityd 7.5 0 0 1.8 100

" pSV00AK2catand pSV00AK2rcatwere constructed by ligatingan end-filledNcol fragment(nt -578 to + 3, see Fig. 3) with an end-filledHindlll digest ofpSV00cat, a low background expression vector (Araki et al., 1988),in correct and reverse orientations, respectively. PAsocat2contains the SV40early promoter, while pSV2cat contains an enhancer sequence in addition to the promoter (Laimins et ai., 1982). b CHO cells were seeded at 106/30-mmdish and grown for 3 days at 37°C under 5% CO2 in modified Ham's FI2 medium supplemented with 10% fetal calf serum, penicillin (50 units/ml), and streptomycin (100 og/ml). The Ca. phosphate precipitates containing 10 ~g of a CAT plasmid and 10#8 of a monitor plasmid pAc-lacZ which directs the synthesis of #Gai under the control of the p-actin promoter (Miyazaki et ai., 1989)were added to each dish (Graham and Van der Eb, 1973).After 5 h the cells were washed with 0.14 M NaCi/0.75 mM NazHPO4/0.25 M HEPES pH 7.1, treated with 15% glycerol for 3 min (Parker and Stark, 1979), and further incubated for 24 h in the fresh medium. The cells resuspended in 50 pl of 250 mM Tris. HCI (pH 8.0) were broken by five cycles of freezing and thawing. The supernatants (5.2-6.6 mg protein/ml) were obtained by centrifugation at 10000 x g. c The reaction mixture (180 #i), containing 0.48 mM acetylCoA/14 ~M ['4C]Cm (53 mCi/mmol)/0.25 M Tris. HCi pH 8.0 and the cell extract, was incubated for 30 min at 37°C. The reaction was stopped and mixed with 250 #1 of ethyl acetate. The organic phase was subjected to thin-layer chromatography, followedby autoradiography.The percentage ofconversion of Cm into mono- and di.acetylated forms was obtained after densitometric scans of the autoradiograms. d The CAT activity was normalized to the #Gal activity directed by cotransfected pAc-lacZ which was assayed as described (Herbomel et ai., 1984). Specific CAT activity (normalized ~ conversion/30min/mg protein) was calculated and expressed as relative values taking the activity generated by pSV2cat as 100%.

226 region. The latter sequences were highly homologous to G A G G C G T G G C , which is present in the long terminal repeat of the HTLV-III retrovirus (Jones et eL, 1986). The promoter of the bovine AK2 gene has features similar to those of 'housekeeping' genes, such as encoding adenosine deaminase (Valerio et al., 1985), dihydrofolate reductase (Crouse et el., 1982), hypoxanthine phosphoribosyl transferase (Melton et al., 1984) and mt creatine kinase (Haas etal., 1989). The features common to promoters of these genes include absence of a typical TATA box, G + Crich upstream sequences, presence of G C boxes and multiple tsp. (e) Functional analysis of the bovine AK2 gone promoter To examine promoter activity of ~he 5' region of the bovine AK2 gone, an Ncol fragment (nt -578 to + 3) was inserted into a site upstream from the promoterless cot gone in plasmid pSV00cat. The ability of the plasmid to produce CAT enzyme was measured 24 h after transfection into C H O cells. Relative CAT activities after normalizing to /~Gal activity from a cotransfected plasmid are summarized in Table I. The CAT activity directed by the AK2 promoter (pSV00AK2cat) was 7,5~o of that obtained from the SV40 early promoter with an enhancer (pSV2cat). A control piesmid containing only the SV40 early promoter (PAtocat2) showed a little activity. When the AK2 promoter was placed in the reverse orientation (pSV00AK2rcat), or no promoter was present (pSV00cat), essentially no CAT activity was observed. Therefore, we conclude that the Ncol fragment can direct the initiation of transcription in C H O cells. This region contains most of the elements shown in Fig. 3A which may function in the regulation of gene expression.

ACKNOWLEDGEMENTS We thank Dr. Y. Ebina for providing plasmid pSV00cat, Dr. M. Mori for pSV2cat and pAiocat2, and Dr. J. Miyazaki for pAc-lacZ, H.T. is grateful to Prof. S. Kawai for providing the opportunity to complete this work.

REFERENCES

Araki, E., Shimada, F., Shichiri,M,, Mori, M. and Ebina, Y.: pSV00CAT: low background CAT plasmid. Nucleic Acids Res, 16 (1988) 1627. Atkinson, D.E.: Cellular Energy Metabolism and Its Regulation. Academic Press, Orlando, Fla,, 1977, pp. 85-107. Benton, W.D. and Davis, R.W.: Screening ~gt recombinant clones by hybridization to single plaques in site, Science 196 (1977) 180-182, Birnstiel, M.L., Busslinger, M. and Strub, K.: Transcription termination and 3' processing: the end is in sitet Cell 41 (1985) 349-359. Chirgwin, J.M., Przybyla, A.E., MacDonald, RJ. and Rutter, WJ.: Isolation ofbiologicallyactive ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18 (1979) 5294-5299.

Clemens, M.J.: A potential role for RNA transcribed from B2 repeats in the regulation of mRNA stability. Cell 49 (1987) 157-158. Crouse, G.F., Simonsen,C.C.,McEwan, R.N. and Schimke, R.T.: Structure of amplified normal and variant dihydrofolate reductase genes in mouse sarcoma S180 cells. J. Biol. Chem. 257 (1982) 7887-7897. Dynan, W.S. and Tjian, R.: Control of eukaryotic messenger RNA synthesis by sequence-specific DNA-binding proteins. Nature 316 (1985) 774-778. Frischauf, A.M., Lehrach, H., Poustka, A. and Murray, N.: Lambda replacement vectors carrying polylinker sequences. J. Mol. Biol. 170 (1983) 827-842. Gr~aham, F.L. and Van der Eb, AJ.: A new technique for the assay of infectivityof human adenovirus 5 DNA. Virology52 (1973) 456-467. Haas, R.C., Korenfeld, C., Zhang, Z., Perryman, B., Roman, D. and Strauss, A.W.: Isolation and characterization of the gene and eDNA encoding human mitochondrial creatine kinase. J. Biol. Chem, 264 (1989) 2890-2897. Herbomel, P., Bourachot, B. and Yaniv, M.: Two distinct enhancers with dllterent cell specificities coexist in the regulatory region of polyoma. Cell 39 (1984) 653-663. Jones, K.A.,Kadonaga, J.T., Luciw, P.A. and Tjian, R." Activation of the AIDS retrovirus promoter by the cellular transcription factor, Spl. Science 232 (1986) 755-759. Khoo, J.C. and Russell, PJ.: Isoenzymes of adenylate kinase in human tissue. Biochim. Biophys. Acta 268 (1972) 98-101. Kishi, F,, Maruyama,M., Tanizawa, Y. and Nakazawa, A.: Isolation and characterization of eDNA for chicken muscle adenylate kinase. J. Biol. Chem. 261 (1986) 2942-2945. Kishi, F., Tanizawa, Y. and Nakazawa, A.: Isolation and characterization of two types of eDNA for mitochondrial adenylate kinase and their expression in Escherichla coil J. Biol. Chem. 262 (1987) ! 1785-11789. Laimins, L.A., Khoury, G., Gorman, C., Howard, B. and Gruss, P.: Host-specific activation of transcription by tandem repeats from simian virus 40 and Moloneymurine sarcoma virus, Proc. Natl. Acad, Sci. USA 79 (1982) 6453-6457. Lee, W., Haslinger, A., Karin, M. and Tjian, R.: Activation of transcription by two factors that bind promoter and enhancer sequences of the human metailothionein gone and SV40. Nature 325 (1987) 368-372. Lompr6, A.M., Nadai-Ginard, B. and Mahdavi, V.: Expression of the cardiac ventricular ~- and p-myosin heavy chain genes is developmentally and hormonally regulated. J. Biol. Chem. 259 (1984) 6437-6446. Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molec~lar Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982,pp. 270-294. Matsuura, S., Igarashi, M., Tanizawa, Y., Yamada, M., Kishi, F., Kajii, T., Fujii, H., Miwa, S., Sakurai, M. and Nakazawa, A.: Human adenylate kinase deficiencyassociated with hemolyticanemia.J. Biol. Chem. 264 (1989) 10148-10155. Melton, D.W., Konecki, D,S., Brennand, J. and Caskey,C.T.: Structure, expression and mutation of the hypoxanthine phosphoribosyltransferase gone. Proc. Natl. Acad. Sci. USA 81 (1984) 2147-2151. Miyazaki,J.-i.,Takaki, S., Araki, K., Tashiro, F,, Tominaga~A., Takatsu, K. and Yamamura, K.-i.: Expression vector system based on the chicken p-actin promoter directs efficientproduction ofinterleukin-5. Gene 79 (1989) 269-277. Mount, S.M.: A catalogue of splice junction sequences. Nucleic Acids Res. 10 (1982) 459-472. Parker, B. and Stark, G.: Regulation of simian virus 40 transcription: sensitive analysis of the RNA specific present early in infection by virus or viral DNA. J. Virol. 31 (1979) 360-369.

227 Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with chainterminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 5463-5467. Suminami, Y., Kishi, F., Torigoe, T. and lqakazawa, A.: Structure and complete nucleotide sequence ofthe 8ene encoding chicken cytosolic adenylate kinase. J. Biochem. 103 (1988) 611-617. Tomasselli, A.G., Schirmer, R.H. and Noda, L.H.: Mitochondrial GTPAMP phosphotransferase. Eur. J. Biochem. 93 (1979) 257-262.

Valerio, D., Duyvesteyn, M.G.C., Dekkcr, B.M.M., Weeda, G., Berkvens, T.M., Van der Voom, L., Van Ormondt, H. and Van der Eb, AJ.: Adenosine deaminase: characterization and expression ofa sene with a remarkable promoter. EMBO J. 4 (1985) 437-443. Yamada, M., Shahjahan, Md., Tanabe, T., Kishi, F. and Nakazawa, A.: Cloning and characterization of cDNA for mitochondrial GTP: AMP phosphotransferase of bovine liver. J. Biol. Chem. 264 (1989) 19192-19199.