`gene
`‘/\
`
`an international journal focusing on gene
`cloning and gene structure and function
`
`\
`
`ELSEVIER
`
`1987
`
`Completing this volume
`
`volume 55 nos. 2 and 3
`BEQ 1041
`
`BEQ 1041
`Page 1
`
`
`
`© 1987, Elsevier Science Publishers B. V. (Biomedical Division)
`All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any
`means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the Publisher, Elsevier Science Publishers
`B.V. (Biomedical Division), P.O. Box 1527, 1000 BM Amsterdam (Molenwerf 1, 1014 AG Amsterdam) The Netherlands.
`
`No responsibility is assumed by the Publisher for any injury and/or damage to persons of property as a matter of products liability, negligence or
`otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of the rapid advances
`in the medical sciences, the Publisher recommends that independent verification of diagnoses and drug dosages should be made.
`
`Although all advertising material is expected to conform to ethical (medical) standards, inclusion in this publication does not constitute a guarantee
`or endorsement of the quality or value of such product or of the claims made of it by its manufacturer.
`
`the author(s) will be asked
`journal,
`Special regulations for authors. Upon acceptance of an article by the
`the article to the Publisher. This transfer will ensure the widest possible dissemination of information.
`
`to
`
`transfer copyright of
`
`Submission of a paper to this journal entails the author's irrevocable and exclusive authorization of the publisher to collect any sums
`or considerations for copying or reproduction payable by third parties (as mentioned in article 17 paragraph 2 of the Dutch Copyright
`Act of 1912 and in the Royal Dutch Decree of June 20, 1974 (S. 351) pursuant to article 16b of the Dutch Copyright Act of 1912)
`and/or to act in or out of Court in connection therewith .
`
`Inc. Consent
`the Copyright Clearance Center,
`registered with
`journal has been
`in the U.S.A. This
`Special regulations for readers
`is given for copying of articles for personal or internal use, or for the personal use of specific clients. This consent is given on
`the
`condition that the copier pays through the Center the per-copy fee stated in
`the code on
`the first page of each article for copying
`beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. The appropriate fee should be forwarded with a copy of
`the first page of the article to the Copyright Clearance Center, Inc. , 27 Congress Street, Salem, MA 01970, U.S.A. If no code appears
`in an article, the author has not given broad consent to copy and permission to copy must be obtained directly from the author. All
`articles published prior to 1980 may be copied for a per-copy fee of US $2.25, also payable through the Center. This consen t does
`not extend to other kinds of copying, such as for general distribution, resale, advertising and promotion purposes, or for creating neW
`collective works. Special written permission must be obtained from the Publisher for such copying .
`
`Printed in Belgium
`
`BEQ 1041
`Page 2
`
`
`
`Gene. 55 ( 1987) 189-196
`Elsevier
`
`GEN 02039
`
`189
`
`High-level secretion of human growth hormone by Escherichia coli
`
`(Recombinant DNA; alkaline phosphatase promoter; heat-stable enterotoxin signal; periplasmic secretion)
`
`Chung Nan Chang, Michael Rey *, Barry Bochner, Herbert Heyneker * and Gregory Gray*
`
`Department of Cell Genetics, Genentech. Inc ., South San Francisco, CA 94080 (U.S.A.) Tel. (415) 266-1000 and* Genencor,
`Inc .. South San Francisco, CA 94080 (U.S.A.) Tel. (415) 747-7500
`
`Received I March 1987
`Revised 2 April 1987
`Accepted 6 April 1987
`
`SUMMA RY
`
`A gene encoding the mature form of human growth hormone (hGH) was fused to the secretion signal coding
`sequence of the Escherichia coli heat-stable enterotoxin II (STII). This hybrid gene was preceded by two
`Shine-Dalgarno sequences derived from the fi]J and STII-coding genes and was expressed in E. coli under the
`transcriptional control of the E. coli alkaline phosphatase (phoA) promoter. In low-phosphate growth media,
`cells synthesized about 15 to 25 Jlg of hG H/ml/ 1 A 550 unit of cells. This represents 6 to 10 % of total cellular
`protein. The majority of the hGH produced (more than 90 %) was processed precisely and secreted into the
`periplasmic space. These results demonstrate that E. coli cells are able to synthesize and secrete high levels of
`this human protein using a prokaryotic signal sequence.
`
`INTRODUCTION
`
`Most bacterial secretory proteins, like mammalian
`secretory proteins, are synthesized as larger pre-
`
`Correspondence to: Dr. C.N. Chang, Protein Design Labs, Inc.,
`3181
`Porter Drive, Palo Alto, CA
`94304, U.S.A.,
`Tel.(4 15) 424-0111.
`
`Abbreviations : aa, ami no acid(s); Ap, ampicillin; bp, base
`pair(s); HPLC, high-press ure liquid chromatography; hGH ,
`hum an growth hormone; LB, Luria broth; P, promoter; PA,
`polyacrylamide; phoA, alkaline phosphatase gene; Pollk,
`Klenow (large) fragment of E . coli DNA polymerase I ; PTH ,
`5 ,
`3-phenyl-2-thiohydantoin; R,
`resistant;
`sensitive; SD,
`Shine- Dalga rno seq uence; SDS, sodium dodecyl sulfate; STII,
`heat-sta ble enterotoxin II ; STII, gene coding for STII; Tc,
`tetracycline; 11p, tryptophan gene; [],designates plasmid-carrier
`state.
`
`cursor molecules that contain a signal sequence of 15
`to 30 aa at the N terminus (Blobel eta!., 1979;
`Silhavy et a!. , 1983). All signal sequences are charac(cid:173)
`terized by several charged amino acids at the N
`terminus, a stretch of hydrophobic amino acids in
`the central region, and an amino acid with a short
`side chain at the cleavage site. It has been demon(cid:173)
`strated that the signal sequence functions to facilitate
`the translocation of secretory proteins across the
`inner membrane (Inouye eta!., 1984). The mem(cid:173)
`brane-bound signal peptidase then cleaves off the
`signal sequence to release the mature protein into the
`periplasmic space located between the inner and
`outer membranes of E. coli.
`A number of mammalian secretory proteins have
`been correctly processed and secreted by E . coli cells
`
`0378-1119/87/S03.50 © 1987 Elsevier Science Publi shers B.V. (Biomedical Division )
`
`BEQ 1041
`Page 3
`
`
`
`190
`
`using either a natural mammalian or an £. coli signal
`sequence. These include rat proinsulin (Talmadge
`et at., 1980a,b ), human proinsutin (Chan et at.,
`1981 ), human immunoglobulin light chain (Zemel(cid:173)
`Dreasen et at., 1984), hGH (Gray et at., 1985), and
`human epidermal growth factor (Oka et at., 1985).
`However, the levels of these secreted proteins are
`generally low. For instance, both proinsulin and
`immunoglobulin were detectable only by using either
`a radioactive tracer or an immunoblot. Human
`growth hormone and epidermal growth factor were
`synthesized at higher levels using a more efficient
`system but the levels were less than 0.5% of total
`cellular protein.
`In this paper we describe the high-level synthesis,
`correct processing and secretion of hGH using the
`signal sequence of the E. coli STII (Picken et at.,
`1983). This secretion system allows easier extraction
`and purification of the natural form ofhGH from the
`periplasmic space. A preliminary account of this
`work has been published (Chang et at., 1984).
`
`MATERIALS AND METHODS
`
`(a) Enzymes
`
`All restriction enzymes, Pollk, T4 polynucleotide
`kinase and T4 DNA ligase were obtained from either
`New England Biolabs or Bethesda Research Labo(cid:173)
`ratories, and used according to the manufacturers'
`recommendations.
`
`(b) Bacterial strains, plasm ids and growth condi(cid:173)
`tions
`
`E. coli strain 294 (endA 1 thi-1 hsdR F - sup£44;
`ATCC31446) (Backman et at., 1976) was generally
`used for transformations and analysis of plasmid
`DNA; E. coli
`strain
`JMlOl
`(L11ac-pro
`sup£
`thi[F'traD36 proAB + lacJQ ZL1Ml5]) (Messing,
`1979) for oligodeoxynucleotide-directed M 13 muta(cid:173)
`and £. coli
`genesis;
`strain W311 0
`(tonA;
`ATCC27325) for analysis of hGH expression and
`secretion .
`Plasmids phGH207- l (De Boer et at., 1983),
`pAPH-1 (Gray et at. , 1985), and pWM501 (Picken
`et at., 1983) have been described elsewhere.
`
`For plasmid analysis, £. coli cells harboring
`plasmid were grown overnight at 37 °C in LB
`medium supplemented with 50 J.Lg Ap/ml or 5 ~tg
`Tc jmt.
`For hGH expression experiments, cells were
`in a
`low-phosphate
`grown overnight at 37 °C
`medium which contained 150 mM
`trieth anol(cid:173)
`amine· HCl, pH 7.4, 10 mM KCI, 1.6 mM MgS 0 4 ,
`60 mM NaCl, 20 mM NH4 Cl and 0.5 mM
`NaH 2 P04 supplemented with 0.15 % glucose, 50 11g
`per ml each of 20 aa (or 0.30 % NZAmine YT) and
`appropriate antibiotic.
`
`(c) Transformation and analysis of DNA
`
`E. coli was transformed and plasmid DNA was
`prepared according to the method of Birnboim and
`Doly ( 1979). Agarose and PA gel electrophoreses
`were performed for analysis of DNAs as descri bed
`by Sharp et at. (1973) and Maniatis et at. (1 975),
`respectively.
`
`(d) Preparation of Escherichia coli extracts and
`hGH assays
`
`E. coli cells were grown overnight at 37 oc in low(cid:173)
`phosphate medium and 1 A 550 unit/ml of cell culture
`was pelleted in the microcentrifuge for 2 mjn. The
`resulting cell pellets were resuspended and sonicated
`for 20 s at 4 o C in 200 J.Ll of lysis buffer which
`contained 30 mM Tris · HCl, pH 8.0 and 2 % SDS.
`Cell extracts were diluted serially in horse serum
`(Hybritech Inc.) and the hGH levels were then deter(cid:173)
`mined using a radioimmunoassay as previously de(cid:173)
`scribed (Goedde! et at., 1979).
`For analysis of periplasmic hGH, the cells were
`extracted by the osmotic-shock method as previous(cid:173)
`ly described (Koshland and Botstein, 1980) and the
`resulting shock fluid was diluted with horse serum
`and assayed as above.
`
`(e) Purification and N-terminal sequencing of hGH
`
`The shock fluid obtained from the periplasm of
`E. coli cells was passed through an immunoaffinity
`chromatography column containing monocl onal
`anti-hGH coupled
`to Affigel 10 (Bio-Rad) as
`described previously (Gray et at. , 1985). Fractio ns
`containing hGH were pooled and small portions
`
`BEQ 1041
`Page 4
`
`
`
`191
`
`Xbo I, Ktenow Poll + 4dNTP's
`BomHL
`isolate vector fraQmenr
`._ _______________
`4
`
`Eco Rl, Bom HI
`isolate 560bp fragment
`Mnll
`isolole 183bp fragment
`
`T4 DNA liQOSe
`
`Xbo L. Klenow Pall + 4dNTP's
`BomHI
`isolare hGH fragment
`
`Bglll, Klenow Poll t 4dNTP's
`BomHI
`'--- - - - - - -,--- - - - - -_ J ISola te veclor fragmenr
`
`EcoRI
`
`""'
`
`Synthetic Oligonucleotide
`5 ·ICAAATGCC TATGCATTCCCAACTATACC -~·
`
`""'
`
`""'
`
`EcoRl
`
`Psi[
`
`lA A TTCCC TGccdTGCA
`GGGACGGG
`I
`
`EcoRI/Pstlcul
`p8R322 large fragment
`I
`1 T4 DNA ligase
`
`EcoRI
`
`Psi!
`A oR
`
`Psi I
`
`Synthetic End of hGH Gene
`
`H1nd!J
`
`EroRI
`
`CTGTGGCT TCT AGAIAGC T TCCG T ACGI
`GACACCGAAGATCT TCGA AGGCATGCTT AA
`I
`I
`Xool
`Pw[[
`
`PwU
`
`BomH{
`
`EcoRL
`par rial Pvu II
`isolate 1080bp frog men!
`
`EcoR I
`alkaline phosphalose
`
`Fig. I. Construction of secretion plas mids. For a detailed explanation , consult RESULTS , section a. R and L relate to the clockwise
`and counterclockwise orientations of transcription in the expression and secretion unit ofplasmids phGH4R and phGH4L, respectively.
`Single lines indicate pBR322 nucleotide sequence. Blackened boxes represent TcR and ApR genes. Open boxes are inserted DNAs.
`Hatched boxes indicate DNAs encoding STI I signal sequence. The phoA signal-coding seq uence is included in the stippled region. Four
`dNTP represent four deoxynucleotide triphosphates dA TP, dGTP, dCTP and dTTP.
`
`BEQ 1041
`Page 5
`
`
`
`192
`
`were analyzed on an SDS-12.5% PAgel as pre(cid:173)
`viously described (Laemmli, 1970; Blobel and
`Dobberstein, 197 5).
`For N-terminal sequence analysis, the purified
`hGH was subjected
`to five cycles of Edman
`degradations (Edman and Begg, 1967). The resulting
`thiazolinones were converted to PTH-amino acids.
`Each PTH-amino acid was then identified by com(cid:173)
`parison to the retention times of a mixture of stan(cid:173)
`dard PTH-arnino acids in a reverse-phase HPLC.
`
`RESULTS
`
`(a) Plasmid construction
`
`The construction of the final plasmids phGH4L
`and phGH4R for expression and secretion of hGH
`is detailed in Fig. 1.
`The first stage of the construction was to achieve
`the fusion of the signal-peptide-coding sequence
`from STIJ gene to the mature coding region of hGH
`gene, i.e. , without its natural signal-peptide-coding
`sequence (Goedde! et al., 1979). The STU-signal(cid:173)
`coding sequence was isolated as a 550-bp Rsal frag(cid:173)
`ment from plasmid pWM501 (Picken et al., 1983)
`and was cloned in the Sma I site in the polylinker of
`M 13mp8. The STII fragment was then reisolated
`from the above construction by EcoRI and Pst i
`digestion and subsequently cloned
`in pBR322
`resulting in plasmid pBR322STU (Fig. 1). Plasmid
`pTrpSTU was constructed by ligating a 183-bp
`Mn/1-Bam HI fragment containing the STU-signal(cid:173)
`coding sequence from pBR322STU into phGH207-
`1 vector (De Boer et al. , 1983) which was previously
`digested with Xba I, followed by a fill-in reaction with
`Po Ilk enzyme and subsequently cut with Bam HI.
`The next intermediate plasmid, pTrpSTUhGH(cid:173)
`fusion, was constructed to bring the hGH structural
`gene in the vicinity of the trp promoter and STU(cid:173)
`signal-coding sequence as indicated in Fig. 1. M 13
`mutagenesis (Zoller and Smith, 1982) was then
`utilized to delete the extraneous DNA separating the
`end of the STU-signal-coding sequence and the start
`of the mature hGH gene in pTrpSTUhGH-fusion.
`The resulting plasmid, pTrpSTIIhGH, contains the
`tip promoter functionally linked to the STII signal(cid:173)
`peptide-coding sequence (Picken et al., 1983) which
`
`in turn was directly fused to the 5' end of th e mature
`hGH gene.
`The second stage was to replace the lip promoter
`of plasmid pTrpSTUhGH with the phoA promoter.
`For this purpose, the 260-bp EcoRI-Hpai fragment
`containing the tip promoter was replaced by a 380-bp
`fragment
`from plasmid pAPHI
`EcoRI-Rsai
`containing the phoA promoter. The resulting plasmid
`was designated as phGHl.
`Finally, novel restriction sites were introduced
`into the 3' end of the hGH gene by ligating a
`synthetic oligodeoxynucleotide to a 1080-bp EcoRI(cid:173)
`PvuU DNA segment ofphGH1 . TheresultingEcoRI
`fragment was then ligated into the E coRI site of
`pBR322-2 in both orientations, resulting in plasmids
`phGH4L and phGH4R. Plasmid pBR322-2 is a
`pBR322 derivate in which the 750-bp Eco R I-Ps!I
`DNA fragment ofpBR322 (Bolivar et al. , 1977) was
`replaced with a short synthetic linker. Since the ApR
`gene was partially deleted by such a replacement,
`cells that contain plasmid phGH4L (or phG H4R)
`will not synthesize the secretory penicillinase (en(cid:173)
`coded by the ApR gene) which might compete with
`hGH for secretion into the periplasmic space. Thus,
`the final plasmids phGH4L and phGH4R contain
`expression and secretion units flanked by E co RI
`sites consisting of the phoA promoter, the STII
`signal sequence and the entire mature hGH gene.
`
`(b) Regulated synthesis of hGH
`
`Plasmid pAPH1 (Fig. 1) was previously con(cid:173)
`structed by the fusion of the phoA promoter and
`phoA secretion-signal-coding sequence to the mature
`form of the hGH gene (Gray et al., 1985). In the
`present study this plasmid was included as a control
`to study the phosphate regulation of hGH produc(cid:173)
`tion fr~m transformants
`contammg plasmid
`phGH4R or phGH4L. Table I shows the levels of
`hGH synthesized in cells with the various construc(cid:173)
`tions using different phosphate concentrations in the
`growth media. For high-phosphate medium, no
`hGH was produced in either E. coli strain 294 or
`strain W3110 having plasmid pAPH l. Human
`growth hormone was produced only in growth media
`containing a low level of phosphate ion . These re(cid:173)
`sults are in agreement with our previous observations
`(Gray et al. , 1985). However, E. coli transformants
`having plasmid phGH4L or phGH4R produced
`
`BEQ 1041
`Page 6
`
`
`
`TAB LE I
`
`Regul ~ ted synthesis of hGH in Escherichia coli
`
`Strain [plasmid] "
`
`Amount of hGH synthesized b
`(pg fml/A 55o)
`
`294[pA PHI]
`294[phG H4R]
`294[phG H4L]
`
`W311 0[pAPHI]
`W311 0[phGH4R]
`W311 0[ phGH4L]
`
`H
`
`0
`0.7
`2.7
`
`0
`0.9
`4.0
`
`L
`
`3.4
`3.6
`5.8
`
`3.0
`7.9
`15.4
`
`in MATERIALS AND
`strain s are desc ribed
`• E. coli
`METHO DS , section band plasmids are described in RESULTS ,
`section a.
`b Tran sformed cells were grown overnight in the high- or low(cid:173)
`phospha te medium and I A 550 unit of cells was collected and
`processed for hG H radioimmunoassay as described in MATE(cid:173)
`RIA LS AND METHODS , sections b and d. The yield of hGH
`was ex pressed as p g ofhGH produced/ml/ 1 A 550 unit of ce lls. H
`and L relate to the high- (15mM) and low- (0.5 mM) phosphate
`concent rations, respecti vely, in th e growth media.
`
`substantial amounts of hGH in high-phosphate(cid:173)
`containing media. For plasmid phGH4L, strain
`W3110 produced 26 % as much hGH in high-phos(cid:173)
`phate-containing medium when compared with the
`low-phosphate-containing medium; whereas strain
`294 produced 46 % as much. With plasmid
`phGH4R, E. coli strain W3110 produced 11 % as
`
`TAB LE II
`
`Cell ul ar location of hGH in Escherichia coli cells"
`
`193
`
`much hGH in high-phosphate-containing medium as
`compared
`to
`that obtained
`in
`low-phosphate(cid:173)
`containing medium; whereas strain 294 produced
`about 19 % as much. These results indicate that the
`E. coli strain W311 0 exhibits a tighter phosphate
`regulation than E . coli strain 294 and that transfor(cid:173)
`mants containing phGH4L appear to have less
`phosphate regulation than those harboring plasmid
`phGH4R.
`The hG H level from the cells also varied with
`different plasmids and strains irrespective of phos(cid:173)
`phate concentration in the media. Strain W3110 cells
`containing phGH4L yielded the highest amount of
`hGH, i.e., 15.4ttg of hGH/ml/ 1 A 550 unit of
`stationary cells. In strain 294, cells produced only
`one-third of that amount. Cells harboring either
`plasmid phGH4R or pAPH 1 produced the same
`amount of hGH; however, twice as much hGH was
`produced in strain W3110 compared to strain 294.
`We conclude that the production and secretion of
`hGH by E. coli strain W3110 is superior over strain
`294 and that the counterclockwise orientation of the
`phoA promoter in phGH4L is preferred for optimal
`hGH synthesis.
`Analysis of plasmid DNA by agarose gel electro(cid:173)
`phoresis has revealed that cells containing plasmid
`phGH4L have two- to three-fold more plasmid
`DNA than those containing plasmid phGH4R (not
`shown). The plasmid copy number could account for
`the differences in hGH level and phosphate regu(cid:173)
`lation.
`
`Strain [plasmid]
`
`Amount of hGH
`synthesized
`(pg fmlfA 55o)
`
`Cellular di stribution of hGH b
`(% oftotal)
`
`Medium
`
`Peri plasm
`
`Cell-associated
`hGH after osmotic(cid:173)
`shock method
`
`W3 11 O[phGH4R]
`W3110[phGH4L]
`W3!10[phGH207-I]
`
`7.9
`15.4
`21.1
`
`2
`
`< I
`
`94
`90
`4
`
`4
`9
`>95
`
`a Cell growth a nd hGH assays were done as described in Table I.
`b One A 550 unit of cells was fraction a ted into the peri plasmic fraction and cell-associated fr action acco rding to the standard osmotic(cid:173)
`shock procedure (Koshland and Botstein , 1980). Growth medi a after removal of cells were diluted with horse serum and assayed as
`descri bed in MATERIALS AND METHODS , section d. To cells that contain plasmid phGH207-l was added 50 11g of3 {3-indole acrylic
`acid at the log phase to induce the synthesis of intracellular methionine-hGH.
`
`BEQ 1041
`Page 7
`
`
`
`194
`
`The plasmid pTrpSTIIhGH (Fig. 1) has also been
`studied for the biosynthesis and secretion of hGH
`(not shown). Transformants having this plasmid
`yielded the same amount ofhGH as those containing
`phGH4R in the same amount of cells; however,
`E. coli containing this plasmid cannot be grown to a
`very high cell density in the fermentor. The easily
`controllable, tightly regulated phoA promoter might
`explain the difference in growth response.
`
`(c) Periplasmic secretion of hGH
`
`To investigate whether hGH produced in the cells
`harboring plasmid phGH4L or phGH4R is secreted,
`the cells were extracted by a standard osmotic-shock
`procedure (Koshland and Botstein, 1980) and the
`levels ofhGH were measured in the various fractions
`as shown in Table II. A majority of the hGH (90 to
`95 % of total) produced was found in the peri plasm
`and only a small fraction of hGH was still cell(cid:173)
`associated after
`the osmotic-shock procedure.
`Fractions from cells containing plasmid phGH4L
`were further analyzed by SDS-12.5 % PAgel elec(cid:173)
`trophoresis and Coomassie-blue staining. Human
`growth hormone accounts for 14 % of total cellular
`protein and about 40 % of peri plasmic protein
`(Fig. 2, lanes 1 and 3, respectively) as estimated by
`the densitometric measurements of the gel. These
`results indicate that hGH is the most abundant
`protein among the periplasmic or total cellular pro(cid:173)
`teins. Thus, the STII signal sequence appeared to
`promote efficient transport of hGH from the site of
`its synthesis in the cytoplasm of the periplasmic
`space.
`transformed with plasmid phGH207-1
`Cells
`(Fig. 1) containing the mature hGH gene without
`signal coding sequence, synthesized only intra(cid:173)
`cellular methionine-hGH (methionine residue is the
`translational initiator of mature hGH). These cells
`were used as experimental controls for the osmotic(cid:173)
`shock procedure. There is a trace amount of intra(cid:173)
`cellular methionine-hG H ( 4 % of total) released from
`these cells (Table II).
`
`(d) Processing of periplasmic hGH
`
`To study whether periplasmic hGH was properly
`processed, hGH was extracted from
`the cells
`containing phGH4L and isolated as described in
`
`1234
`
`Fig. 2. Purification ofhGH from the periplasmic fraction . E. coli
`cells containing plasmid phGH4L were grown in the low(cid:173)
`phosphate medium as described in Table I. The peri plasmic pro(cid:173)
`teins were extracted by the osmotic-shock method (Koshland
`and Botstein, 1980) and purified by the antibody affinity chroma(cid:173)
`tography (Gray et al., 1985 ). Protein samples were resolved on
`an SDS-12.5 % PA gel (Laemmli , 1970; Blobel and Dobberstein,
`1975). Lanes: I, proteins derived from th e whole cells ; 2, cell(cid:173)
`associated proteins after the osmotic-shock procedure; 3, peri(cid:173)
`plas mi c pro teins, and 4, hGH purified from the periplasmic
`proteins by antibody affinity chro matography. The arrowhead
`indicates the position of hGH.
`
`BEQ 1041
`Page 8
`
`
`
`rv!ATERIALS AND METHODS, section d. Analysis of
`purified hGH on an SDS-12.5~<, PAgel showed a
`single band (Fig. 2, lane 4). TheN-terminal sequence
`of this purified hGH was identified as Phe-Pro-Thr(cid:173)
`IIe-Pro, indicating that the cleavage had occurred
`correctly to generate the mature protein. No alterna(cid:173)
`tive cleavages were detected.
`
`DIS CUSSION
`
`The results presented here demonstrate that a high
`level of synthesis and secretion of hGH has been
`achieved using E. coli cells transformed with plasmid
`phGH4L. This plasmid contains a strong phoA pro(cid:173)
`moter and an efficient STII signal sequence.
`It is
`interesting that cells harboring plasmid
`phGH4L synthesize more hGH than those harbor(cid:173)
`ing plasmid phGH4R since they differ only in the
`orientation of the DNA fragment for expression and
`secretion. One possible explanation for this result is
`that the phoA promoter is a rather strong one and
`that transcriptional readthrough may affect the
`region of copy-number control near the origin of
`replication. Therefore, the difference in hGH yield
`may be due to difference in copy number of plasmid
`inside the cell.
`The amount of hGH synthesized in E. coli strain
`W31 10[phGH4L] is 15.4 ~Lgjml/A 550 unit of sta(cid:173)
`tion ary cells in a shake-flask experiment (Table I).
`In
`the highly controlled environment of
`the
`ferm enter, the same transformant has consistently
`synthesized 20 to 25 ~g hGH/ml/A 550 unit of cells
`as determined by radioimmunoassays. Since it is
`known that 1 ml of E. coli cells at 1 A 550 unit has
`approx. 250 pg of protein, 15 to 25 pg of hGH
`produced represents 6 to 10 % of total cellular pro(cid:173)
`tein in E. coli. This number is smaller than the 14%
`estimated from the densitometric scanning of a
`Coomassie-blue stained gel (Fig. 2, lane 1 ).
`The E. coli secretion machinery is capable of
`han dling the high level of the hGH synthesized since
`the vast majority (more than 90 % ) of the hormone
`is transported, resulting in hGH being the most
`abundant protein in the periplasm (Fig. 2, lane 3).
`Extraction and purification of hGH from the peri(cid:173)
`plasm (Koshland and Botstein, 1980; Nossal and
`Heppe!, 1966; Ames et al., 1984) is easier than the
`current isolation procedure in which methionine-
`
`195
`
`hGH is obtained from intracellular inclusion bodies
`(Goedde! et al., 1979).
`While this paper was in preparation, Hsiung et al.
`( 1986) reported that the ompA signal was also effec(cid:173)
`tive in the synthesis of hGH. They estimated that
`hGH is about 6 % of total cellular protein by the
`densitometric scanning of a Coomassie-blue-stained
`gel. Since different promoters were used in both
`studies, it is difficult to assess the relative efficiencies
`of the ompA signal versus the STII signal.
`
`ACKNOWLEDGEMENTS
`
`The authors would like to thank Herman de Boer,
`Ron Hitzeman, Anna Hui, and Amy Chovnick for
`helpful discussions, Bill Kohr for N-terminal se(cid:173)
`quencing, and Socorro Cuisia for help in manuscript
`preparation .
`
`REFERENCES
`
`Ames, G.F-L., Prody, C. and Kustu , S.: Simple, rapid, and quan(cid:173)
`titative rele ase ofperiplasmic proteins by chloroform. J. Bac(cid:173)
`terial. 160 (1984) 1181-1183.
`Backman, K., Ptashne, M. and Gilbert, W.: Construction of
`plasm ids carrying the ci gene of bacteriophage J.. Proc. Nat!.
`A cad. Sci. USA 73 ( 1976) 4174-4178.
`Birnboim , H.C. and Doly, J.: A rapid alkaline ex traction proce(cid:173)
`dure for sc reening recombinant plasmid DNA. Nucl. Acids
`Res. 7 (1979) 1513-1523.
`Blobel , G. , Walter, P., Chang, C.N., Goldman , B.M. , Erickson ,
`A.H. and Lingappa, V.R.: Translocation of proteins across
`membranes: the signal hypothesis and beyond. Symp. Soc.
`Exp. Bioi. 33 ( 1979) 9-36.
`Blobel , G. and Dobberstein, B.: Transfer of proteins across
`membranes, I. Presence ofproteolytically processed and non(cid:173)
`processed nasce nt immunoglobulin light chains on mem(cid:173)
`brane-bound ribosomes of murine myeloma. J. Cell. Bioi. 67
`(1975) 835-851.
`Bolivar, F. , Rodriguez, R.L., Greene, P.J. , Betlach, M.C. ,
`Heyneker , H.L. , Boyer, H.W., Crosa, J.H. and Falkow, S.:
`Construction and characterization of new cloning vehicles,
`II. A multipurpose cloning system . Gene 2 ( 1977) 95-113.
`Chan, S.J. , Weiss, J. , Konrad , M., White, T., Bah!, C., Yu , S.-D.,
`Marks , D. and Steiner, D.F.: Biosynthesis and periplasmic
`segregation of human proinsulin in Escherichia coli. Proc.
`Nat!. Acad. Sci. USA 78 (1981) 5401-5405.
`Chang, C.N., Gray, G.L., Heyneker, H.L. and Rey, M.: Secretion
`of hete rologous proteins in Escherichia coli. U.S. Pate nt
`Application (US 658342, Oct. 5, 1984).
`
`BEQ 1041
`Page 9
`
`
`
`196
`
`De Boer, H., Comstock, L.J. and Vasser, M.: The toe promoter:
`a functional hybrid derived from the lip and lac promoters.
`Proc. Nat!. Acad. Sci. USA 80 (1983) 2 1-25.
`Edman, P. and Begg, G.: A protein sequencer. Eur. J. Biochem.
`I ( 1967) 80- 91.
`Goedde!, D .V., Heyneker, H.L. , Hozumi, T. , Arentzen, R.,
`Itakura, K., Yansura, D.G., Ross, M.J., Miozzari , G., Crea,
`R. and Seeburg, P.H. : Direct expression in Escherichia coli of
`a DNA sequence coding for human growth hormone. Nature
`281 ( 1979) 544-548.
`Gray, G.L. , Baldridge, J.S., McKeown , K.S ., Heyneker, H .L. and
`Chang, C.N.: Peri plasmic production of correctly processed
`human growth hormone in Escherichia coli: natural and
`bacterial signal sequences are interchangeable. Gene 39
`(1985) 247-254.
`Hsiung, H.M., Mayne, N.G. and Becker, G.W. : High level
`expression, efficient secretion and folding of human growth
`hormone in Escherichia coli. Biotechnology 4 ( 1986) 991-995.
`Inouye, S., Vlasuk , G.P., Hsiung, H. and Inouye, M.: Effects of
`mutations at glycine residues in the hydrophobic region of the
`Escherichia coli prolipoprotein signal peptide on the secretion
`across the membrane. J. Bioi. Chem. 259 (1984) 3729-3733.
`Koshland , D. and Botstein , D.: Secretion of beta-lactamase
`requires the carboxyl end of the protein. Cell 20 ( 1980)
`749-760.
`Laemmli, U.K.: Cleavage of structural proteins during the
`assembly of the head of bacteriophage T4. Nature 227 ( 1970)
`680-685.
`Maniatis, T., Jeffrey, A. and Van de Sande, H.: Chain length
`determination of small double- and single-stranded DNA
`molecules by polyacrylamide gel electrophoresis. Bio(cid:173)
`chemistry 14 (1975) 3787-3794.
`Messing, J.: A multipurpose cloning sys tem based on the single(cid:173)
`stranded DNA bacteriophage M 13. Recombinant DNA
`
`Technical Bulletin, NIH Publication No. 79-99, 2 (t 979)
`43-48.
`Nossal , N.G. and Heppe!, L.A.: The release of enzymes by
`osmotic shock from Escherichia coli in exponential phase. J.
`Bioi. Chem. 241 (1966) 3055-3062.
`Oka, T. , Sakamoto, S., Miyoshi, K-1. , Fuwa, T., Yoda, K.,
`Yamasaki , M. , Tamura, G. and Miyake, T.: Synthesis and
`secretion of human epidermal growth factor by Escherichia
`coli. Proc. Nat!. Acad. Sci. USA 82 (1985) 7212-721 6.
`Picken , R.N. , Mazaitis, A.J., Maas, W.K., Rey, M. and Heyneker,
`H.L.: Nucleotide sequence of the gene for heat-sta ble
`enterotoxin II of Escherichia coli. Infect. Immun. 42 ( 1983)
`269-275.
`Sharp, P.A., Sugden, B. and Sambrook, J.: Detection of two
`restrict ion
`endonuclease
`activities
`in Ha em ophilus
`parainfluenzae using ana lytical agarose-ethidium bromide
`electrophoresis. Biochemistry 12 ( 1973) 3055-3063.
`Si lhavy, T.J. , Benson, S.A. and Emr, S.D. Mechanisms of protein
`localization . Microbial. Rev. 47 (1983) 313-344.
`Talmadge, K., Stahl , S. and Gilbert, W.: Eukaryotic signal
`sequence transports insulin antigen in Escherichia coli. Proc.
`Nat!. A cad. Sci. USA 77 ( 1980a) 3369-3373.
`Talmadge, K., Kaufman, J. and Gilbert, W.: Bacteria mature
`preproinsulin to proinsulin. Proc. Nat!. Acad. Sci. U SA 77
`( 1980b) 3988-3992.
`Zemel-Dreasen, 0. and Zamir, A.: Secretion and processing of
`an immunoglobulin ligh t chain in Escherichia coli. G ene 27
`( 1984) 315-322.
`Zoller, M.J. and Smith, M.: Oligonucleotide-directed muta(cid:173)
`genesis using M 13-derived vectors: an efficient and gene ral
`procedure for the production of point mutations in any frag(cid:173)
`ment of DNA. Nucl. Acids Res. 10 ( 1982) 6487-6500.
`
`Communicated by F. Bolivar.
`
`BEQ 1041
`Page 10