Proc. Nadl. Acad. Sci. USAVol. 81, pp. 5956-5960, October 1984Biochemistry

Synthesis of a gene for human growth hormone and its expressionin Escherichia coli

(trp promoter/hybrid plasmid/promoter efficiency/modified Shine-Dalgarno region)

M. IKEHARA*, E. OHTSUKA*, T. TOKUNAGA*, Y. TANIYAMA*, S. IWAI*, K. KITANO*, S. MIYAMOTO*,T. OHGI*, Y. SAKURAGAWA*, K. FUJIYAMA*, T. IKARI*, M. KOBAYASHI*, T. MIYAKE*,S. SHIBAHARA*, A. ONOt, T. UEDAt, T. TANAKAt, H. BABAt, T. MIKIt, A. SAKURAIt,T. OISHI*, 0. CHISAKA§, AND K. MATSUBARA§*Faculty of Pharmaceutical Sciences, Osaka University, Suita, Osaka 565, Japan; tFaculty of Pharmaceutical Sciences, Hokkaido University, Sapporo 060,Japan; lInstitute of Physical and Chemical Research, Wako, Saitama 351, Japan; and §Institute for Molecular and Cellular Biology, Osaka University,Osaka 530, Japan

Communicated by Susumu Ohno, June 15, 1983

ABSTRACT A gene coding for human growth hormone,which consists of 192 amino acids, was chemically synthesized.The synthesis entailed ligating 78 deoxyribooligonucleotides,which had been synthesized on polymer supports by the phos-photriester method with frequently occurring amino acid co-dons of Escherichia coli. The chemically synthesized gene wasinserted into an E. coli plasmid downstream from the E. colithp promoter, with a modified ribosome-binding region car-ried on pBR322. E. coli cells transformed with this recombi-nant plasmid synthesized 2.9 x 106 molecules per cell of hu-man growth hormone upon induction. The induced polypep-tide was identical with natural human growth hormone in sizeand in immunological properties, as well as in biological activi-ty as examined by the tibial test with hypophysectomized rats.

The improvement of techniques in chemical synthesis ofdeoxyribooligonucleotides has made possible the total syn-thesis of genes of 100-500 nucleotides (1-9). By this tech-nique, genes with designed amino acid sequences can inprinciple be synthesized, and the products of their expres-sion in bacteria can be examined. In the present work, a genefor human growth hormone (hGH) (10), which contains 191amino acids and methionine, was totally synthesized bychemical means, using amino acid codons most frequentlyappearing in Escherichia coli (11-13) (Fig. 1). To establish ageneral procedure for efficient expression of genes for rela-tively large peptides, hGH was thought to be a suitable tar-get. Use of the synthetic gene will provide sufficient sites forrestriction enzymes, which should be useful in mutations ofgenes for replacement of protein domains. Rapid synthesisof 78 gene fragments of 7-26 bases was carried out, followedby sequential ligation to construct the gene of 584 base pairs(bp). This is the longest gene so far synthesized chemically.Separately, a modified trp promoter was constructed that al-lows expression of any synthetic genes with an initiation co-don placed downstream of it. A recombinant plasmid carry-ing the promoter and the hGH gene produced the hormone inE. coli at a high level. The induced polypeptide was indistin-guishable from the natural hGH in several features. The puri-fied hormone was as active as the natural growth hormone inbiological tests (14).

MATERIALS AND METHODSDeoxyoligonucleotides. Deoxyoligonucleotides with a

chain length of 7-26 bases were synthesized by the phospho-triester solid-phase method, using 1% cross-linked polysty-rene (15) as the support. Altogether 78 chains were prepared

by addition of dinucleotide blocks to the nucleosides linkedthrough a 3'-succinyl group to the resin. Tetranucleotideblocks prepared by condensation of dimers in the liquidphase were used for the synthesis of the 25-mer (A-UO; Fig.1). Sixteen dinucleotide blocks were prepared with mesity-lenesulfonyl-3-nitro-triazolide (16) by methods essentially asreported (17) and purified mostly by reversed-phase chroma-tography (18) under conditions described (19, 20). A typicalsynthesis of a 15-mer started with treatment of the 5'-dimeth-oxytritylated nucleoside resin (5 ,umol) with benzenesulfonicacid (17) (2%; 2 ml) for 1 min twice in a small vessel with asintered glass plate and a stopcock at the bottom. In the caseof deoxyadenosine resin, either zinc bromide (21) or 3% tri-chloroacetic acid (22) was used. A dimer (20 ,umol) wastreated with MSNT (70 ,umol) in pyridine (200-300 Al) at40'C for 20 min. The resin was treated with acetic anhydride(0.2 ml) in 4-dimethylaminopyridine-pyridine (0.1 M; 1.8 ml)for 3 min. The product was deblocked with 1,1,3,3-tetra-methylguanidium 2-pyridinealdoximate in dioxane-H20 (0.5M, 90% dioxane) (16) at 300C overnight and with concentrat-ed ammonia at 550C for 5 hr. The dimethoxytritylated 15-merwas separated on a column (0.7 x 6 cm) of C-18 silica gel(Waters; 35-100 pum) with a gradient of acetonitrile (10%to-35%; total, 200 ml) in 0.05 M ammonium acetate and treatedwith 80% acetic acid. The product was applied to a column(0.7 x 21 cm) of DEAE-Toyopearl 650S (Toyosoda) with agradient of sodium chloride (0.1-0.3 M; total, 200 ml) in 7 Murea/20 mM Tris HCl, pH 7.5. The 15-mer was checked byhigh-pressure liquid chromatography (HPLC) on C-18 silicagel (Toyosoda, TSK-410 AK) and ion exchange (Toyosoda,TSK gel IEX540K). Impurities were removed by HPLC, andthe purity and sequence were confirmed by mobility-shiftanalysis (23).Enzymes and Proteins. DNA ligase was obtained from Ta-

kara Shuzo and Boehringer Mannheim. Restriction endonu-cleases Cla I and Alu I were purchased from BoehringerMannheim and Bethesda Research Laboratories. Other en-zymes, including restriction endonucleases (Bgl II, BamHI,Sal I, HinfI, and Hpa I) were obtained from Takara Shuzo.Methionyl hGH was a gift from H. Yamamoto (SumitomoChemical).

Construction of the Promoters and the hGH Gene. A 0.5-kilobase-pair DNA fragment carrying the intact trp promoterwas extracted by HinfI digestion from E. coli plasmid DNA,ptrp ED 5-1 (24). It was cloned into the unique EcoRI site ofpBR322, using EcoRI linker. The upstream EcoRI site wascleaved by the enzyme and filled in with T4 DNA polymer-ase. Then, the remaining downstream EcoRI site was

Abbreviations: hGH, human growth hormone; SD sequence, Shine-Dalgarno sequence; bp, base pair(s).

5956

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. NatL Acad Sci. USA 81 (1984) 5957

1 10 20Met Phe Pro Thr Ile Pro Leu Ser Arg Leu Phe Asp Asn Ala Met Leu Arg Ala His Arg Leu His Gln Leu Ala Phe

Cla I UO Hinf IUl U2 V U3 V U4Ica ATG TTC CCA ACT ATT CCA CTG AGTCGC CTG TTC GAT AAC GCG ATG CTG CGT GCG CAT CGT CTG CAC CAA CTG GCT TTCz TAC AAG GGT TGA TAA GGT GAC TCA fLCG GAC AAG CTA TTG C% TAC GAC GCA CGC GTA WA GAC GTG GTT GAC CGA A4G

L0Ot L1 L2 C L3A part

30 40 50Asp Thr Tyr Gin Glu Phe Glu Glu Ala Tyr Ile Pro Lys Glu Gln Lys Tyr Ser Phe Leu Gln Asn Pro Gln ThrGvCATC TTGTU5v U6 U0 Aiu I vU

GAC ACT TC CAG GAG TTC GAA GAA GYA TAC ATC CCG AAA GAA &G AM TAC A TTC Cl4 CAG MC CCA CAG ACCCTG IGA ATG GTC CTC AAG CTT CTT CGT ATG TAG GGC TTT CTT GTC TTT AG It.AAG GAA GTC TTG GGT GTC TGG

L4 L5 L6 " L0 A

B part

60 70Ser Leu Cys Phe Ser Glu Ser Ile Pro Thr Pro Ser Asn Arg Glu Glu Thr Gin Gin Lys Ser Asn Leu Glu Leu

V U2 U3 V U4 US V U6TCG TTG TGT TTC TCT YAA AGT ATC CCG ACC CCT TCT AAC CGC GAA GAGVACC.CAG CAG AAA TCG AAC CTT GAA CTGAGC AAC ACA A~c AGA CTT TCA TAG GGC TGG GGA AGA TTG GCG CTT CTC TGG GTC GTC TTT AGC TTG GAA CTT GACLi " L2 L3 A L4 A L5 A

80 90 100Leu Arg Ile Ser Leu Leu Leu Ile Gln Ser Trp Leu Glu Pro Val Gin Phe Leu Arg Ser Val Phe Ala Asn Ser

V U7 v U8 v U9 V U1 0 V UiiCTT CGT ATC TCG CTG CTT CTC ATT CAG TCG TGG CTG GAG CCA GTA CAG TTC CTG CGT TCG GTT TTC GCA AAC TCAGAA GCA TAG AGC GAC GAA GAG TAA GTC AGC ACC GAC CTC GGT CAT GTC AAG GAC GCA AGC CAA AAG CGT TTG AGTL6 A L7 A L8 A L9 A L10

A

110 120Leu Val Tyr Gly Ala Ser Asp Ser Asn Val Tyr Asp Leu Leu Lys Asp Leu Glu Glu Gly Ile Gln Thr Leu Metv U12 V U13 v U14 v U15 Rn'ILV U16

CTG GTT TAC GGT GCG TCT GAC AGT AAC GTT TAC GAC CTG CTG AAA GAC CTT GAA GAA G rAG ACC CTG ATGGAC CAA ATG CCA CGC AGA CTG TCA TTG CM ATG CTG GAC GAC TTT CTG GAA CTT CTT CC TC TGG GAC TACLii A L12 A L13 A L14 A L15

130 140 150Gly Arg Leu Glu Asp Gly Ser Pro Arg Thr Gly Gln Ile Phe Lys Gin Thr Tyr Ser Lys Phe Asp Thr Asn Serv U17 V U18 V.Ui9Bql II Ui v U2 v U3

GGT CGC CTG GAA GAT GGT TCA CCA CGC ACT GGT Ck ATC TC AAA CAG ACT TAC TCC AAA TTC GAT ACT AAC TCTCCA GCG GACACTT CTA CCA AGT GGT GCG TGA CCA GCTAGAA~ TTT GTC TGA ATG AJG TTT AAG CTA TGA TTG AGAL16 A L17 A L18i Li L2 A

C part160 170

His Asn Asp Asp Ala Leu Leu Lys Asn Tyr Gly Leu Leu Tyr Cys Phe Arg Lys Asp Met Asp Lys Val Glu Thr

v U4 V U5 U6 V U7 v U8CAT AAC GAT GAT GCT CTG CTG AAA AAC TAC GGC CTG CTG TAC TGT TTC CGT AAA GAT ATG GAT AAA GTT GM ACTGTA TTG CTA CTA CGA GAC GAC TTT TTG ATG CCG GAC GAC ATG ACA AAG GCA TTT 9TA TAC CTA TTT CAA CTT TGA

L3 L4 A 15 A 16 Li

180 190Phe Leu Arg Ile Val Gln Cys Arg Ser Val Glu Gly Ser Cys Gly Phe StopStopv U9 V U10 Uli U12 Sal I

TTC CTG CGT ATC GTT CAG TGT CGT TCT GTT GAA &G TCG TGT GGC TTC TAA TAGMG GAC GCA TAG CAA GTC ACA GCA AGA CAA CTT CCC AGC ACA CCG AAG ATT ATC aqct

L8 " L9 " L10 " Li1

FIG. 1. Human growth hormone and its synthetic gene.

cleaved by EcoRI, followed by BAL-31 digestion to elimi-

nate a sequence between the trpE-coding region and its

Shine-Dalgarno (SD) sequence. A chemically synthesizedCla I linker d(A-A-T-C-G-A-T-T) was inserted just down-

stream of the SD sequence. Sequence of the relevant regionin the resulting plasmid pCT-1 is shown in Fig. 2. It has the

trp promoter, the SD sequence for trpL (attenuator), alongwith another SD sequence of trpE in front of the Cla I linker,

d(A-A-T-C-G-A-T-T).Oligonucleotides (0.2 A260 unit; 10 ug; -1.5 nmol) were

phosphorylated by treatment with [y32P]ATP (500 pmol; 1-2

mCi/hmol; 1 Ci = 37 GBq) and polynucleotide kinase (3

units), pH 8.0 at 370C for 15 min, and then with unlabeled

ATP (10 mmol) and polynucleotide kinase (1 unit) at 370C for

90 min. Phosphorylated fragments (5-8 pieces) and nonphos-phorylated ends were mixed as shown in Fig. 2, then an-

nealed at 70'C, in the presence of 66 mM Tris*HCI, pH 7.5/6mM MgCl2/500 tkM ATP, in a volume of 270 ul. The mixture

was cooled to 15°C during 1 hr and treated with T4 DNA

ligase (1.2 units; 4.4 units/mol) together with dithiothreitol

(10 mM) at 15°C for 13-15 hr. The reaction was checked bygel electrophoresis on 10% acrylamide. The products were

precipitated or separated by the same electrophoresis when

needed. For example, in the synthesis of the C part gene, the

larger three fragments (50 pmol) were phosphorylated at pH9.6 and ligated at 20'C for 16 hr, and the mixture was treatedwith Cla I (30 units) at 370C for 6 hr then with Sal 1 (120 units)at 370C for 15 hr. The product (0.5 gg; -2.5 pmol) was isolat-ed by 5% gel electrophoresis. A, B, and C parts of the genewere ligated as shown in Fig. 2, and the duplex (0.5 pmol)was joined to the Cla I/Sal I sites of the expression vector

pCT-1 to yield plasmid phGH-1 (Fig. 3).To improve the efficiency of its expression, the promoter

region of phGH-1 was modified as follows. First, the regionbetween the Hpa I site at -11 and the Cla I site at + 157 was

eliminated to remove the attenuator (trpL) along with the SDsequence of trpE. These were replaced by a synthetic oligo-nucleotide consisting of either 33, 35, or 37 nucleotide pairs,each carrying a SD sequence (A-A-G-G) and with appropri-ate termini. The resulting plasmids, pGH-L9, pGH-L11, andpGH-L13, carry 9, 11, or 13 bases upstream of the initiationcodon (ATG) of the hGH gene (Fig. 3). Plasmid pGH-L8 hasa similar structure to the others, except that the gene was

inserted into the Taq I site just downstream of the trpL SDsequence.

Characterization of the Protein. Radioimmunoassay was

carried out with a Phadebas hGH PRIST kit (26) and an EI-KEN-hGH-I kit for the double-antibody method from Eiken

Biochemistry: Ikehara et aL

5958 Biochemistry: Ikehara etaLP

1 A 15 B-I 2B 35 B-E-4 -4 3 -414 24 34 44

45 B-N53

54 C-1 64 C U

63 7

Clal A A

1) T4ligase2) Clal, Bgilx

A A____ Bgv v v v

8

AB 415b.p

Bgif[ sail

C I 165bp

Clal AB

1) T4IigaseCSol ,2))ClaL Soal

,.SallI CallI A Sall-v- v -

)< C 176b.p.pCT 1/Cial, Sall

Sall

HGH gene 584 bP

FIG. 2. Synthesis of the hGH gene.

Immunochemical. Biological activity of the induced product,which was fractionated by chromatography on DEAE-cellu-lose (Whatman DE-52), was tested on hypophysectomizedrats (14).

RESULTSSynthesis of hGH Gene. The amino acid sequence of hGH

and the synthetic gene fragments for methionyl hGH areshown in Fig. 1. The gene was divided into three parts-theamino terminal (A), the middle part (B), and the carboxylterminal (C). Amino acid codons used in the present genewere chosen mainly from those found in a gene for elonga-tion factor Tu (tufA) of E. coli (13). Homologous or comple-mentary sequences were sought with a microcomputer, andcodons were altered to avoid mismatches in duplex align-ments. The amino acid codons used in the present synthesisare shown in Table 1, together with those found in a humangene (27). As indicated in Fig. 1, sequences for the restric-tion enzymes Cla I, Sal I, Hinf Bgl II were incorporated toclone the fragments into the vector and to facilitate modifica-tions. A single cleavage sequence for each of these enzymeswas designed. The methionine codon and the termination co-dons were also added at the appropriate positions. Dinucleo-tide blocks were mainly used, except for the synthesis of

AUO (25-mer), where tetramer blocks were used. As expect-ed, the use of larger oligonucleotides as condensing unitswas found to be advantageous in obtaining the pure product,although synthesis of protected tetranucleotides was timeconsuming. All fragments were finally examined by HPLCon reversed-phase and mobility shift analyses (23).

Construction of Plasmids Carrying the hGH Gene UnderControl of the trp Promoter. Enzymatic ligation of the 78chemically synthesized oligonucleotides was carried out byjoining 5-10 pieces in the first step. The subfragments thusobtained were further ligated to yield 415-bp (A+B segment:amino-terminus portion) and 176-bp (C segment: carboxyl-terminus portion) polynucleotides. The C-segment was inte-grated into a plasmid, replacing the region between the Cla I

and Sal I sites of pCT1 (Fig. 2) to generate phGHC-1. Afteramplification, the 165-bp C segment was extracted by diges-tion with Bgl II and Sal I, and ligated to the 415-bp A+Bsegment to yield the whole hGH gene, which was insertedinto the plasmid pCT1 as shown in Fig. 2. The resulting plas-mid phGH-1 carries the 584-bp synthetic gene under the con-trol of the trp promoter and the SD sequence for trpE.The attenuator was next removed from phGH-1, and the

synthetic gene was placed at different distances from thetrpL SD sequence A-A-G-G (Fig. 3). Plasmids pGH-L8,

pCT-1pBR322 Ptrp trpL trpL trpE pBR322

Sal A r HpaI S.D. TaqI attenuator S.D. ClaI SallIt T-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~--GAGAAICGA----0 --GTTAAC--1---0 TA +2C7TG--+--TGA ------+-1+-1-

-300 -1 0 +1 4 +27 +71 +1 40 +1 54

replace right ClaI-Sall fragmentwith synthetic HGH gene (584b.p.)

pHG-1~

-

pHGH-1 Ptrp trpL trpL trpESail A r HpaI S.D. TaqI attenuator S.D. Clal HGH gene Sail-CJ-------"---tI- C--1---AAGGGTATCGACAATG---TAG-------GAGAATCGA

--300 -10 +14 +27 +71 +140 +154

remove TaqI(+22)-ClaI(+157) fragmentyielding pGH-L8 or

replace HpaI(-11)-ClaI(+157) fragment with synthetic DNAfragments ,yielding pGH-L9,pGH-Ll l,pGH-Ll3

trpLHpaI S.D. Clal HGH gene

pGH-L8 AACTAGTACGCAAGTTCACGTAAAAAGGGTATCGA1'g -----------------

-10 +1

pGH-L9 AC-------------------------GAAGAM----------------

pGH-L1 AAC-------------------------GTATCATCGAg --------------

pGH-L13 MC-------------------------GTATC G 1-----------

FIG. 3. Construction of plasmids containing the synthetic hGH gene. pCT-1 and phGH-1 were obtained as described in Materials andMethods. Dotted lines are identical to the sequence of the trp operon (25). To replace Hpa I/Cla I fragments, DNA duplexes of 33, 35, and 37 bpwere prepared by joining four synthetic fragments in each duplex.

nc-7

Proc. NatL Acad ScL USA 81 (1984)

I

~~~Proc. NatL Acad. Sci. USA 81 (1984) 5959

Table 1. Codon usage in synthetic (and natural) hGH gene

First Thirdposition Second position position(5' end) U C A G (3' end)U Phe 0 (3) Ser 5(3) Tyr 0 (3) Cys 4 (2) U

Phe 13 (10) Ser 1 (5) Tyr 8 (5) Cys 0 (2) CLeu 0 (1) Ser 2 (3) Term. 1 (0) Term. 0 (0) ALeu 1 (0) Ser 6 (1) Term. 1 (1) Trp 1 (1) G

C Leu 5 (1) Pro 1 (0) His 2 (1) Arg 7 (1) ULeu 1 (8) Pro 0 (5) His 1 (2) Arg 4 (4) CLeu 0 (4) Pro 5 (2) Gln 1 (2) Arg 0 (0) ALeu 19 (12) Pro 2 (1) Gln 12 (11) Arg 0 (1) G

A Ilie 2 (2) Thr 6(1) Asn 0 (0) Ser 3 (1) Ulie 6 (6) Thr 4 (4) Asn 9 (9) Ser 1 (5) Cle 0 (0) Thr 0 (4) Lys 9 (1) Arg 0 (0) AMet 4 (3) Thr 0 (1) Lys 0 (8) Arg 0 (5) G

G Val 5 (0) Ala 2(1) Asp 7 (2) Gly 4 (0) UVal 0 (3) AlaO0(5) Asp 4 (9) Gly 2 (5) CVal 2 (0) Ala 2 (1) Glu 11 (6) Gly 0 (0) AVal 0 (4) Ala 3 (0) Glu 3 (8) Gly 2 (3) G

pGH-L9, pGH-L11, and pGH-L13, respectiveiy, carry the

SD sequence 8, 9, 11, and 13 bases upstream of the methio-

nine codon.

Transformation and Expression of the Gene in E. coli. The

synthetic gene inserted in the plasmid, and the modified re-

gions around the SD sequences were identified by sequence

analyses (28). The plasmids were used to transform E. coli

HB1O1 (29), and the transformants were examined for

expression of the hGH gene upon induction by 3-indolyla-

crylic acid at eariy logarithmic phase (24). The cells were

collected after 23 hr of induction, lysed, and the proteins

were analyzed by electrophoresis in reducing NaDodSO4/

polyacrylamide gels (30). The results are shown in Fig. 4. E.

coli cells carrying pGH-L9 synthesized the product most ef-

ficiently. From radioimmunoassay, the yield of hGH was es-

timated to be 169 ji~g per ml of culture medium, or 2.9 x 106

molecules per cell. As shown in Table 2, pGH-L8, pGH-L11,

and pGH-L13 also yielded hGH at a higher level upon induc-

tion compared to phGH-1. E. coli cells carrying these plas-

mids had to be treated with 6 M guanidine hydrochloride

(31). The amount of hGH in cells without the treatment was

almost one-tenth of that observed after denaturation. This

suggests that some of the induced proteins in cells might be

Mr 1-- AA-- HGH

10O3 GH L8 L9 L13 GH L8 L9Lii1L13 5gg1lOptg

92.5

68~

43I

25.7

18.4 411% -t-l-

1,W. am- ... ME .-12.3 - X _:;.

19 2) 7 16 16 (%)

FIG. 4. Analysis of the induced protein by NaDodSO4/POIY-acrylamide gel electrophoresis. 3-Indolylacrylic acid (IAA) was

used as the inducer. A mixture of proteins containing cytochrome C,

i3-lactoglobulin, a-chymotrypsinogen, ovalbumin, bovine serum

albumin, phosphorylase b, and myosin (H chain) was used as size

markers. hGH, methionyl hGH.

in the coagulated form. Thin-section electron micrographs

(Fig. 5), in fact, show a large mass of uniformly stained mate-

rial.

Analysis of the Induced Protein. The molecular weight of

the product was estimated from the mobility in reducing gel

electrophoresis as 21,000 (Fig. 4). The amino terminus of the

product was analyzed by Edman degradation and found to

be methionine.

Immunological properties of the induced peptide were ex-

amined by two procedures. Table 2 summarizes the results

of radioimmunoassay of hGH in extracts from bacteria con-

taining expression plasmids by the solid method (26).

Biological activity of the induced peptide was examined in

hypophysectomized rats. In tests of weight increase and in-

crease in the width of the proximal epiphyseal cartilage of

the tibia, the synthesized hGH showed equal activity to that

of the natural growth hormone.

DISCUSSION

hGH is known to be responsible for the growth of bones and

is currently used for treatment of hypopituitary. dwarfism

(10). Other biological activities can be studied if hGH is

available in quantity. The present study has. shown that a

totally synthesized hGH gene was efficiently expressed in E.

ccli under the control of the trp promoter. Enzymatically

synthesized cDNA corresponding to the major part of the

hGH gene has also been cloned, together with a synthetic

Table 2. Radioimmunoassay of hGH in E. ccli extracts

milliunits/ MoleculesPlasmid* Induction mit pag/mit per cell

phGH-1 + 67.7 33.9 4.9 x i0- ~~~7.3 3.6 4.9 x 104

pGH-L8 + 214.2 107.1 1.7 x 106- ~~~21.4 10.7 1.6 x i05

pGH-L9 + 337.5 168.7 2.9 x 106- ~~~27.2 13.6 2.0 x i05

pGH-L11 + 217.1 108.5 1.7 x 106- ~~~26.0 13.0 1.8 x 105

pGH-L13 + 234.4 117.2 1.9 X 106- ~~~17.4 8.7 1.3 x i05

pCT-1 + ND ND-- ~~ND ND

ND, not determined.*In E. coli HB1O1.tOne milliliter per culture.

Biochemistry: Ikehara et aL

5960 Biochemistry: Ikehara etaLP

A;' ^~AFIG. 5. Electron micrograph of E. coli. (a) Cells containing

pHG-L9. Cells were stained with glutaraldehyde and osmium tetra-oxide. (b) Cells containing pCT1.

amino-terminal gene, and expressed in E. coli under the con-trol of lac (26) or tac (32, 33) promoters. It was shown thatpGH-L9, which carried the SD sequence of trpL 9 bases up-stream from the initiation codon, produced methionyl hGHmost efficiently. The production of methionyl hGH wasabout 5 times that of phGH-1, which contained the attenua-tor and the SD sequence of trpE. This efficiency is also 10times higher than that reported previously (32, 33). The dis-tance of the synthetic gene from the trpL SD sequence A-A-G-G seemed to affect the efficiency of the expression. Asshown in Table 2, the efficiency for pGH-L9 is significantlyhigher than that for pGH-L8, pGH-L11, or pGH-L13.The usage of amino acid codons most frequently appearing

in E. coli should favor the expression, because it was shownthat there was a good correlation between the frequency ofcodon usage and the amount of corresponding tRNAs.

Expression vectors used in this study can be applied toany genes having proper sequences for restriction enzymesand the initiation and termination codons.Any proteins of -200 amino acids may be prepared by the

recombinant DNA technique in bacteria by use of chemical-ly synthesized genes. Site-directed mutations of amino acidsor domains can be designed by replacing nucleotides in thecorresponding region of the gene fragments. The presenthGH gene was also designed to facilitate modification ofhGH by replacement of deoxyoligonucleotides.

The authors thank Drs. J. Kawamura and T. Terao of the NationalInstitute of Hygiene of Japan for the biological testing of the prod-uct. We thank Dr. A. Matsushiro and T. Wada of the MicrobiologyDisease Institute of Osaka University and Dr. S. Aida of the JEOLCo. for electron microscopy. We are indebted to Dr. H. Yamamotoof the Sumitomo Chemical Co., Ltd., for his generous gift of methio-nyl hGH.

1. Khorana, H. G., Agarwal, K. L., Buchi, H., Caruthers,M. H., Gupta, N. K., Kleppe, K., Kumar, A., Ohtsuka, E.,RajBhandary, U. L., van de Sande, J. H., Sgaramella, V.,Terao, T., Weber, H. & Yamada, T. (1972) J. Mol. Biol. 72,209-217.

2. Khorana, H. G. (1979) Science 203, 614-625.3. Itakura, K., Hirose, T., Crea, R., Riggs, A. D., Heynecker,

H. L., Boliver, F. & Boyer, H. W. (1977) Science 198, 1056-1065.

4. Crea, R., Kraszewski, A., Hirose, T. & Itakura, K. (1978)Proc. Natl. Acad. Sci. USA 75, 5765-5769.

5. Hsing, H. M., Sung, W. L., Broussan, R., Wu, R. & Narang,S. A. (1980) Nucleic Acids Res. 8, 5753-5765.

6. Suzuki, M., Sumi, S., Hasegawa, A., Nishizawa, T., Miyoshi,K., Wakisawa, S., Miyake, T. & Misoka, F. (1982) Proc. Natl.Acad. Sci. USA 79, 2475-2479.

7. Edge, M. D., Greene, A. R., Heathcliffe, G. R., Meacock,P. A., Schuch, W., Scanlon, D. B., Atkinson, T. C., Newton,C. R. & Markham, A. F. (1981) Nature (London) 292, 756-762.

8. Tanaka, S., Shima, T., Ohsuye, K., Ono, T., Mizono, K.,Ueno, A., Nakazato, H., Tsujimoto, M., Higashi, N. & Nogu-chi, T. (1983) Nucleic Acids Res. 11, 1707-1723.

9. Edge, M. D., Heathcliffe, G. R., Moore, V. E., Faulkner,N. J., Camble, R., Petter, N. N., Trueman, P. T., Schuch,W., Hennam, J., Atkinson, T. C., Newton, C. R. & Markham,A. F, (1983) Nucleic Acids Res. 11, 6419-6435.

10. Escamilla, R. E. (1975) in Hormonal Proteins and Peptides,ed. Li, C. H. (Academic, New York), Vol. 3, pp. 147-190.

11. Granthem, R., Gautier, C. & Gouy, M. (1980) Nucleic AcidsRes. 8, 1893-1912.

12. Ikemura, T. (1981) J. Mol. Biol. 151, 389-409.13. Yokota, T., Sugisaki, H., Takanami, M. & Kaziro, Y. (1980)

Gene 12, 25-31.14. Greenspan, F. S., Li, C. H., Simpson, M. E. & Evans, H. M.

(1949) Endocrinology 45, 455-463.15. Ito, H., Ike, Y., Ikuta, S. & Itakura, K. (1982) Nucleic Acids

Res. 10, 1755-1769.16. Reese, C. B., Titmas, R. C. & Yau, L. (1978) Tetrahedron

Lett., 2727-2730.17. Broka, C., Hozumi, T., Arentzen, R. & Itakura, K. (1980) Nu-

cleic Acids Res. 8, 5461-5471.18. Fritz, H.-J., Belagaje, R., Brown, E., Fritz, R. H., Jones,

R. A., Lees, R. G. & Khorana, H. G. (1978) Biochemistry 17,1257-1267.

19. Ohtsuka, E., Fujiyama, K. & Ikehara, M. (1981) Nucleic AcidsRes. 9, 3503-3522.

20. Ohtsuka, E., Taniyama, Y., Marumoto, R., Sato, H., Hirosa-ki, H. & Ikehara, M. (1982) Nucleic Acids Res. 10, 2597-2608.

21. Kierzek, R., Ito, H., Blatt, R. & Itakura, K. (1981) Tetrahe-dron Lett., 3761-3764.

22. Tanaka, T. & Letsinger, R. (1982) Nucleic Acids Res. 10,3249-3260.

23. Sanger, F., Donelson, J. E., Coulson, A. R., Kossel, H. &Fischer, D. (1973) Proc. Natl. Acad. Sci. USA 70, 1209-1213.

24. Hallewell, R. A. & Emtage, S. (1980) Gene 9, 27-47.25. Yanofsky, C., Platt, T., Crawford, I. P., Nichols, B. P., Cris-

tie, G. E., Horowitz, H., VanCleemput, M. & Wu, A. M.(1981) Nucleic Acids Res. 9, 6647-6667.

26. Goeddel, D. V., Heynecker, H. L., Hozumi, T., Arentzen, R.,Itakura, K., Yansura, D. G., Ross, M. J., Miozzari, G., Crea,R. & Seeburg, P. H. (1979) Nature (London) 281, 544-548.

27. Martial, J. A., Hallewell, R. A., Baxter, J. D. & Goodman,H. M. (1979) Science 205, 602-607.

28. Maxam, A. & Gilbert, W. (1980) Methods Enzymol. 65, 499-560.

29. Sadler, J. R., Techlenburg, M. & Betz, J. L. (1980) Gene 8,279-300.

30. Laemmli, U. K. (1970) Nature (London) 227, 680-685.31. Hager, D. A. & Burgess, R. R. (1980) Anal. Biochem. 109, 76-

86.32. de Boer, H. A., Comstock, L. J. & Vasser, M. (1983) Proc.

NatI. Acad. Sci. USA 80, 21-25.33. de Boer, H. A., Comstock, L. J., Yansura, D. G. & Hey-

necker, H. L. (1982) in Promoters: Structure and Function,eds. Rodriguez, R. L. & Chamberlin, M. J. (Praeger, NewYork), pp. 462-481.

Proc. NatL Acad Sci. USA 81 (1984)