INDUSTRIAL
`
`]OURNAL OF
`
`
`
`STOCKTON
`
`VOLUME 16 - NUMBER 3 - MARCH 1996
`
`BEga1g0e0Z
`
`BEQ 1007
`Page 1
`
`

`
`~
`STOCKTON
`
`Journal of Industrial /vlicrobiology
`is published by Stockton
`Press, a division of Macmillan Press Ltd, on behalf of the Society
`for Industrial Microbiology.
`
`lvlicrobiology
`Industrial
`Journal of
`an
`is
`Scope The
`international journal which publi shes original research letters,
`short communications , critical reviews and peer-reviewed letters
`to
`the Editor
`in
`the field of biotechnology , fermentation,
`environmental microbiology, biodegradation , biodeterioration ,
`quality control and other areas of applied microbiology .
`
`This journal is covered by Current Advances in Biochemi stry ,
`Current Advances in Ecological Sciences, Current Advances in
`Genetics
`and Molecular Biology, Current Biotechnology
`Abstracts and Excerpta Medica.
`
`Editorial Submitted manuscripts will be reviewed by at least
`two members of the Journal ' s Editorial Board. Manuscripts which
`are not concise or do not conform to the conventions and
`standards of the Journal will be returned to the authors for
`revision. Submission of an article to the Journal implies that it
`has not been publi shed previously and that it
`is not being
`considered for publication elsewhere.
`
`Manuscripts and all editorial correspondence should be sent to:
`Dr JJ Cooney
`Environmental Science Program
`University of Massachusetts
`I 00 Morrissey Bl vd
`Boston MA 02 125-3393, USA
`Tel: + I 617 287 7446
`Fax : + I 617 287 7474
`
`Advertisements Enquiries concern ing adverti sements should be
`addressed to:
`Michael Rowley
`Hasler House
`High Street
`Great Dunmow
`Essex
`CM6 lAP, UK
`Tel: +44 (0) 1371 874613
`Fax: +44 (0) 137 1 872273
`Publisher All business correspondence, supplement enquiries
`and reprint requests should be addressed to Joumal of Industrial
`Microbiology,
`Stockton Press, Houndmills, Basingstoke,
`Hampshire RG21 6XS , UK. Tel : +44 (0) 1256 29242. Fax: +44
`(0) 1256 810526.
`Publisher: Jayne Marks
`Production Controller: Sharon Clark
`
`Subscriptions - EU/Rest of World Subscription price per
`annum (two volumes, six issues) £491 , rest of the world £491
`(surface), £585 (airmail) or equivalent in any other currency.
`Orders must be accompanied by re mittance. Cheques should be
`to Macmillan Maga zines and sent
`made paya ble
`to: The
`Subscription Department, Macmillan Press Ltd, Houndmills,
`Basingstoke, Hampshire RG2 1 6XS , UK. Where appropriate,
`subscribers may make payments
`into UK Post Office Giro
`Account No. 519 2455 . Full details must accompany the payment.
`Subscribers from EU territories should add sales tax at the local
`rate.
`
`Subscriptions - USA USA subscribers call toll free 1-800-22 1
`2 123 or send check/money order/credit card details to: Stockton
`Press, 345 Park A venue South, lOth Floor, New York, NY I 00 I 0-
`1707. Tel: + I 212 689 9200. Fax: + 1 2 12 689 9711. USA annual
`subscription rates (institutional/corporate): $730 (surface), $930
`(airmail).
`
`Reprints of any article in thi s journal are available from Stockton
`Press, Houndmill s, Basingstoke, Hampshire RG 2 1 6XS , UK. Tel:
`+44 (0) 1256 29242. Fax: +44 (0) 1256 810526.
`
`Copyright © 1996 Society for Industrial Microbiology
`ISS N 0169-4 146
`
`All rights of reproduction are reserved in respect of all papers,
`articles, illustrations, etc., published in this journal in all countries
`of the world.
`
`All material publi shed in thi s journal is protected by copyri ght,
`which covers exclusive rights to reproduce and distribute the
`material. No material published in thi s journal may be reproduced
`or stored on microfilm or in electronic, optical or magnetic fo rm
`without the written authorization of the Publisher.
`
`Authorization to photocopy items for internal or personal use of
`specific clients, is granted by Stockton Press, for libraries and
`other users registered with the Copyright Clearance Center (CCC)
`Transaction Reporting Service, provided that the base fee of
`$ 12 .00 per copy is paid directl y to CCC, 2 1 Congress St, Salem,
`MA 01970, USA 0169-4146/96 $ 12.00 + $0.00.
`
`Apart from any fair dealing for the purposes of research or pri vate
`study, or criticism or review, as permitted under the Copyri ght,
`Designs and Patent Act 1988, this publication may be reproduced,
`stored or transmitted, in any form or by any means, only with the
`prior permission in writing of the publi shers, or in the case of
`reprographic reproduction ,
`in accordance with
`the
`terms of
`licences issued by the Copyright Licensing Agency.
`
`\:Vhilst e very effon is made by the publishers and ed itorial board to see that no inaccurate or mi slead ing data. opinion o r statement appears in thi s Jo urnal , they wish
`to make it clear that the data and opinio ns appenring in the artic les and advertise ments here in are the responsibility of the contributor or advertiser concerned.
`Accordingly. the publi shers, the editorial commit!ee and their respecti ve employees, officers and agents accept no liability whatsoever for the conseq uences of any
`such inaccurate or misleadi ng data, opinio n or statement.
`
`Typeset by Photo-graphi cs, Honiton, Devon, UK
`Printed on acid-free paper, effective with Volume 12, Issue I, 1993
`
`BEQ 1007
`Page 2
`
`

`
`Journal of Industrial Microbiology (1996} 16, 145- !54
`© !996 Society for Industrial Microbiology 0! 69-4!46/96/$! 2.00
`
`I Review: Optimizing inducer and culture conditions for
`1 expression of foreign proteins under the control of the lac
`I promoter
`I RS Donovan 1
`2 CW Robinson 1
`2 and BR Glick2
`•
`•
`'
`I Departments of 1Chemical Engineering; 2 Biology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1
`
`I
`
`1
`
`•
`
`I This review examines factors which influence the expression of foreign proteins in Escherichia coli under the tran(cid:173)
`scri ptional control of the lac and tac promoters, and discusses conditions for maximizing the production of a foreign
`! protein using this system. Specifically, the influence of IPTG (isopropyl-{3-o-thiogalactoside) concentration, tempera(cid:173)
`ture, composition of the growth medium, the point in the growth curve at which cells are induced with either IPTG
`or lactose, and the duration of the induction phase are discussed.
`
`I Keywords: lac promoter; lac promoter; recombinant DNA; protein overexpression; fermentation strategies; IPTG ; lactose
`
`I
`d
`.
`· lntro uctton
`. During the past two decades, research utilizing recombinant
`1 DNA tec hnology has led to the development of a variety
`/ of products for use in fields as di verse as cancer treatment
`and pulp bleaching [29]. The majority of these products
`have been produced using Escherichia coli [ 17,3 1 ,36] since
`\
`, this bacterium has been well characteri zed at the molec ular
`level and the introd ucti on of foreign genes into it is easil y
`accomp lished. In addition, E. coli can be grown rapidly to
`. a high cell density using inex pensive media and readil y
`{available fermentation eq uipment.
`I A wide vari ety of commerciall y important proteins has
`been expressed in recombinant E. coli, including xy lose
`1 i omerase [8], xy lanase [63] T4 DNA ligase [86] , CD4
`[70] , and a ran ge of different human immunomod ulators,
`
`I growth factors, hormones, blood proteins and viral antigens
`
`· for use as vaccines [34]. The expression of these proteins
`\ in E. coli is generall y accomplished by placing the foreign
`I gene in to a multi-copy plasmid vector under the transcrip-
`tional control of either a consti tuti ve or regul atable strong
`I promoter. Constitutive promoter systems provide a simple
`means of overproducing a foreign protein in a bacterial cell
`,, since the target protein is continuall y expressed . Unfortu(cid:173)
`natel y, continuous hi gh level expression often causes a
`J drain or metabolic burden on the cell 's energy resources
`leading
`to a reduction or
`inhibition of cell growth
`1 [12,30,80]. Since the overall yield of the target protein is
`a fun ction of both the cell yield (g of cells L - t) and the yield
`J of product per cell , the reduced cell growth th at generall y
`accompani es constitutive expression leads to lowe r yields
`of the target protein than is possible when using regulatable
`I promoter systems.
`Regulatable promoter systems provide the abi lity to 'turn
`1, on' the expression of a foreign gene by varying an environ(cid:173)
`! mental factor such as temperature [66,86,89], dissolved
`
`I -------------------------------------
`1 Correspondence: BR Glick. Dept of Biology. University of Waterloo.
`IVaterloo. Ontario, Canada N2L 3G I
`( ~cceived II September 1995 ; accepted 6 December 1995
`I
`
`oxygen tension [4 1 ,6 1] or the concentration of a particular
`component in the growth medium [ 18,59]. For exam ple,
`the bacteriophage lambda pL and pR promoters in conjunc(cid:173)
`tion with a temperature sensiti ve cl 857 repressor gene can
`be thermally induced when the temperature of the fermen(cid:173)
`tation broth is raised from the growth-optimum temperature
`of 30-35°C to 42°C [56,86]. The E. coli trp and phoA pro(cid:173)
`moters are induced fo llowing depletion of tryptophan and
`phosphates, respectively, from the medium [ 18,70] , while
`the lac promoter is induced in the presence of lactose or
`isopropy l-{3-o-th iogalactoside (IPTG) [59). The ability to
`induce foreign gene expression at will allows for the separ(cid:173)
`ation of cell growth from the product synthesis or induction
`phase of the fermentation . After obtaining a hi gh concen(cid:173)
`tration of cells in a fermenter during the growth phase, the
`foreign gene can then be induced resulting in higher total
`yields of the recombinant protein than otherwise could be
`achieved using constitutive express ion . In some cases, the
`level of transcription of the foreign gene can be regulated
`by using an appropri ate level of the inducing stimulus. This
`in turn can further improve yie lds by allowing the level of
`expression to be optimi zed during the induction phase. In
`order to take advantage of the flexibility of inducible pro(cid:173)
`moter systems, the optimal point in the fermentation for
`inducing expression as well as the optimal level of inducer
`that needs to be used in the process must be determined.
`The lac promoter from the E. coli lactose operon is one
`of the promoters most commonly used to regu late the
`ex press ion of recombinant genes in bacteria. It is the most
`we ll understood of all bacterial promoter and ha been
`extensively characterized at the molecular level [6,69] ;
`however, only a small number of studies have optimized
`the production of recombinant proteins using this system.
`This article rev iews factors that influence foreign protein
`expression from the lac (or tac) promoter(s) in E. coli, and
`thus provides a starting point for researchers attempting to
`optimize the condi ti ons for expressing a target protein using
`this system.
`
`BEQ 1007
`Page 3
`
`

`
`146
`
`Transcriptional regulation of the lac promoter
`
`Co n ~;!;; Hirjllion or
`Glucose
`Lactose
`
`cA ~W
`
`Inc Promoter-Op..e.r.a1.QL_Rcg inn
`
`Level of
`Tri]nscripllon
`
`Expression of foreign proteins in E. coli
`RS Donovan et at
`
`l
`
`The lactose (lac) operon in wild-type E. coli consists of
`three genes (lacZ, lacY and lacA) which encode proteins
`responsible for metabolism of the sugar lactose. The lacZ
`protein, {3-galactosidase, converts lactose to glucose and
`gal actose ; the lacY gene codes for lactose permease which
`provides the active transport of lactose across the cytoplas(cid:173)
`mic membrane, while the lacA gene product has transace(cid:173)
`tylase activity and may function to detoxify lactose analogs
`which are harmful to the cell [91].
`Transcription from the lac promoter is regulated by the
`lac repressor, the product of the lacl gene [9,69]. In the
`absence of inducer (ie lactose or IPTG), the lac repressor
`inhibits transcription from the lac promoter by binding to
`the operator reg ion of the lac operon. When the repressor
`is bound to the operator, its presence prevents the proper
`binding of RNA polymerase to the promoter region so that
`transcription of the lac genes does not occur. However,
`since a chemical equilibrium exists between bound and
`unbound repressor molecules , the operator site is not con(cid:173)
`tinuously occupied by the lac repressor and thus there is
`generally a low basa l level of transcription of the lac genes.
`When the cell encounters lactose or other ga lactosides,
`transcription of the lac genes may increase up to 1000-
`fold [40]. The increased transcription occurs because, once
`in side the cell , a small portion of the lactose is converted,
`by the basal level of {3-galactosidase, to an intermediate
`compound known as allolactose, which then binds to the
`repressor protein. This binding causes a conformational
`change in the repressor protein reducing its affinity for the
`lac operator. With no repressor bound at the operator
`region, sign ificantly increased transcription of the lac genes
`occurs and express ion is induced.
`Transcription from the lac promoter is also regulated by
`the binding of the catabo lite activator prote in (CAP) to the
`promoter region [9 ,69]. When CAP bi nels to the promoter,
`it increases the affinity of the promoter for RNA poly(cid:173)
`merase, thereby increasi ng transcription of the lac genes.
`The affinity of CAP for the promoter is enhanced by its
`association with cyclic adenosine monophosphate (cAMP).
`Cellular levels of cAMP are highest when the amount of
`glucose in the medium is lowest. Thus , provided that no
`repressor is bound to the operator, hi gh intracellular con(cid:173)
`centrations of cAMP lead to high levels of cAMP-CAP
`complex at the promoter si te and cause a hi gh leve l of tran(cid:173)
`scription of the lac genes . These interaction s are illustrated
`schematicall y in Figure I [I].
`Catabolite repress ion allows E. coli to metaboli ze glu(cid:173)
`cose preferentially prior to lactose (and a variety of other
`sugars) when a mi xture of sugars is present in the growth
`medium. When the cell encounters hi gh concentrations of
`glucose, transport of glucose into the cell reduces the intra(cid:173)
`cellular level of cAMP. Thus , even in the presence of
`indu cer, a hi gh concentration of g lucose leads to a low level
`of transcription from the lac promoter (Figure I). In small
`sca le culture, the effects of catabolite repress ion however,
`may be reduced by adding cAMP to the growth medium
`lac promoter [ 10]. Glucose al so
`when
`inducing
`the
`indirectly modulates express ion from the lac promoter by a
`process known as carbohydrate-mediated inducer exclusion
`
`Low
`
`Low
`
`H igh
`
`low
`
`High
`
`Low
`
`Low
`
`Pro moter
`
`lacZ
`
`Low
`
`High
`
`High
`
`Low
`
`Promoter
`
`O~ c r<lto r
`
`/rtr Z
`
`low
`
`cAM P
`
`Low
`
`:-ligh
`
`High
`
`larZ
`
`High
`
`Figure 1 Effect or glucose and lactose concentrati on in the growth
`medium on the level of transcripti on from the lac promoter in £. coli.
`Adapted from [I].
`
`[65]. Glucose (and certain other sugars , such as fru ctose
`
`and mannitol) are transported across the cytoplasmic mem- !
`port of glucose into the cell results in dephosphorylation of I
`·l
`
`brane via the phosphotransferase sys tem (PTS). The trans(cid:173)
`
`the glucose specific-PTS enzyme liiGJc, which in its non(cid:173)
`phosphorylated form interacts with lactose permease and
`inhibits the transport of lactose, and hence the lac inducer,
`into the cell .
`IPTG is common ly used for inducing expression from
`the lac promoter and offers a number of distinct advan(cid:173)
`tages, espec iall y in sma ll sca le experiments. Unlike lactose
`and other galactosides, IPTG is a metabolic-free, or gratu(cid:173)
`itous inducer, because it is not metaboli zed by the cell. This
`ensures that the level of induction remains constant follow(cid:173)
`ing the addition of IPTG to the growth medium. IPTG is
`transported into the cell by methods other than lac perme(cid:173)
`ase making this compound less susceptible to
`inducer
`exclus ion by glucose and suga rs whose assimilati on is
`mediated by the PTS system.
`A number of variations on the lac promoter have been
`reported. For example, the lacUV5 promoter contains a
`mutation in the consensus region of the lac promoter that
`increases the promoter strength [69], and therefore has been
`used in place of the wild type lac promoter in some high
`express ion plasmid vectors [28,82]. The lac prom oter, a
`tryptophan (flp) and lacUYS promoters
`hybrid of the
`[4,23], is reported to be 5-10 times stronger th an the
`lacUVS promoter [23 ,24] and is induced using IPTG and
`other ga lactosides while not being subject to catabo lite
`repression.
`
`•
`
`Expression of foreign proteins
`
`There are a wide variety of factors which influ ence the
`express ion of foreign proteins in E. coli. As mentioned pre(cid:173)
`viously, the presence of reco mbinant plasmid DNA and the
`expression of a recombinant protein generally im poses a
`metabolic drain on the cells' energy resources [ L3.30]. This
`added metabolic burden often reduces cellular growth rates,
`causes segregationa l and structu ral plasmid in stability and
`causes metabolic, genetic and physiological changes in the
`host cell which, in turn , may substantiall y affect the total
`and fun ctional yi eld s of a target product. While in some
`
`BEQ 1007
`Page 4
`
`

`
`- - - - - - -------
`
`- ----..,
`
`147
`
`cases problems of metabolic burden and low yields of
`recombinant protein have been solved by altering codon
`usage [38] or host cell metaboli sm through genetic manipu(cid:173)
`lation [21 ], functi onal yields ha ve also been improved by
`modifyi ng culture conditions and the level of recombinant
`protein ex press ion during th e induction phase [73 ,86].
`Thus, it is important to understand the influence of con(cid:173)
`dit ions such as temperature, growth medium composition,
`ind ucer concentration as well as the point of induction and
`the duration of the induction phase on the ex pression of the
`target protein in E. coli.
`
`The effect of IPTG concentration
`Despite the popularity of the lac expression system for pro-
`! ducing recombinant products, few studi es in the literature
`have reported optimizi ng the IPTG concentration for
`ind ucing recombinant gene express ion , and in most cases
`the reasons for using a particul ar level of IPTG are rarely
`stated. While a wide range of IPTG concentrati ons, ranging
`from 0.005
`to 5 mmol L - I ,
`have been
`reported
`I (19 ,32,62,82,87], ex pression is commonly induced using
`· I mmol L- 1 IPTG.
`As a result of metabolic burden (so metimes called meta(cid:173)
`bol ic load), th e hi gh concentrations of inducer that are often
`used in an effort to fully induce the lac promoter do not
`necessarily lead to maximal expression of a target protein
`[30]. In many cases. the optimal inducer co ncentration is
`I chosen to bal ance the decreasing yields of recombinant
`1 cells following induction with increasing cellular leve ls of
`target protein [ 12]. For example, in studying the dynamics
`of chlorampheni col-acetyl-transferase (CAT) expression
`· using the tac promoter, Bentley er a/ [ 12] showed that the
`\ increased specific CAT producti vity (g CAT sy nthesized
`I g- 1 biomass), following mid-log phase induction with IPTG
`reduced the specific growth rate of the cell s in batch cui(cid:173)
`lure. Moreover, the severity of the reduction in speci fie
`growth rate increased with increas ing levels of CAT protein
`. expression. At hi gh inducer concentrati ons (3.4 mmol IPTG
`L- 1
`), the reduction in specific growth rate was so severe (ie
`1 80%) that it led to the early onset of the stationary phase
`and cell death, while at slightl y lower levels (2. 1 mmol
`1 L - I), the cells partially recovered from the initial shock fol(cid:173)
`lowing inducti on and es tablished a constant specific growth
`rate that was 60% lower than pre-induction levels. Thus, the
`4 overall level of CAT ex pression in batch culture involved a
`tradeoff between increas ing product ex pression in the cell
`and decreasing cell yields resulting from a lower specific
`growth rate followin g induction with !PTG. As a result, use
`of an intermediate IPTG concentration of approximately
`I 0.8- 1.0 mmol IPTG L - I balanced the opposing phenomena
`. of increasing target protein expression and decreasing cell
`\ Yields to max imi ze the overall expression of the recombi-
`1 nant product [ 12].
`translational
`transcriptional and
`the
`To understand
`t responses to varying levels of induction with IPTG , Wood
`and Peretti [88] used continuous culture to study the steady(cid:173)
`ltate responses of cells to increasing levels of metabolic
`load when producing recombinant {3-galactosidase. The {3-
`lgalactos idase was under the transcriptional control of the
`lac promoter and the metabolic load was modulated by
`I inducing with 0.0 I to 7.5 mmol IPTG L - I. Under these con-
`
`Expression of foreign proteins in E. coli
`RS Donovan et a/
`
`ditions, the syn thesis rate of {3-galactosidase mRNA and
`the steady state activity of this enzyme increased linearl y
`• Further
`with lPTG co ncentration up to I mmol IPTG L - 1
`increases in IPTG concentration did not lead to stoichio(cid:173)
`metric increases in /a cZ transcription. While it was possible
`that most of the lac promoters were full y induced above
`I mmol IPTG L- 1 (due to the complete titration of repressor
`molecul es by IPTG), in thi s case it appears that degradation
`of the {3-galactosidase mRNA played an important role in
`limiting protein expression. The degradati on rate of the {3-
`galactosidase mRNA increased with increasing levels of
`induction such that the 42-fold increase in the {3-galactoside
`mRNA sy nthesis rate when the inducer concentration was
`increased over the range studied led to only a 4-fold
`increase in the steady state level of {3-galactosidase mRNA.
`The fact that rai sing the IPTG concentrati on from I to
`7.5 mmol L- 1 increased the /a cZ transc ription rate by 75 %,
`but increased the steady state levels of {3-galactosidase
`rnRNA and activity by only 32 % and 22%, respectivel y,
`suggested that rnRNA stability rather than transcription
`limited product expression .
`The stud y of Wood and Peretti [88] also suggested ways
`in whi ch the inducti on of recombinant mRNA sy nthesis can
`alter the cellular processes and redirect energy resources.
`At low levels of inducer (< 0.1 mrnol IPTG L- 1
`), increased
`levels of transcription red uced the sy nthesis of ribosomal
`RNA (rRNA). However, when IPTG concentrations greater
`than 0. I mmol L - I were used, the cells responded to induc(cid:173)
`tion by increasing the sy nthesis rate and steady state levels
`of ribosomal RNA. It was theorized that induction of the
`tac promoters with IPTG diverted RNA pol ymerase away
`from ribosome operons, lead ing to the observed decrease
`in rRNA synthesis. At hi gh levels of induction, the
`increased steady state level s of mRNA sy nthesis may have
`severely depleted the pool of free non-translating ribosomes
`in the cell. Under these conditions, the recombinant cells
`may have di verted energy into producing rRNA to increase
`translationa l capaci ty in order to deal with increased trans(cid:173)
`lational demands. Thus, under transient conditions, this
`diversion of energy (metabolic load) to
`increase trans(cid:173)
`lational capacity observed by Wood and Peretti [88], may
`account for at least a temporary reduction in growth rate
`and plasmid stability as the cell attempts to keep pace with
`the added burden of the recombinant gene expression .
`Induction with IPTG also appears to cause a variety of
`stress res ponses in the cell. In non-transform ed strains with
`a single lac promoter on the chromosome, increased syn(cid:173)
`thesis of the heat shock proteins DnaK, GroEL, Gro ES was
`observed 30 min after the addition of 0.5 mmol IPTG L- 1
`[ 45]. A temporary inhibition of H35 protein- normall y pre(cid:173)
`sent onl y during ex ponential growth-following inducti on
`was also encountered, uggest ing that cells must ada pt to
`even low levels of expression induced with IPTG . As well ,
`proteolytic degradati on of abnormal proteins can be influ(cid:173)
`enced by the level of induction with IPTG, with degra(cid:173)
`dation by some pathways increasing or decreasing with
`increasing IPTG levels [44]. High level recombinant pro(cid:173)
`tein ex pression usin g IPTG has also been shown to induce
`the ex pression of a variety of proteases in medi a where the
`concentrations of certain amino acids may be limiting [32].
`The location of the lac repressor as well as the quantity
`
`BEQ 1007
`Page 5
`
`

`
`148
`
`Expression of foreign proteins in E. coli
`RS Donovan et a/
`
`of lac repressor molecules produced in the cell wi ll greatly
`affect the level of IPTG needed to full y induce express ion
`from the lac promoter. For example, in cases with hi gh
`plasmid
`copy number, plasmid-borne
`lac or
`lac
`promoter/operators may partially or completely
`titrate
`repressor molecules expressed from the chromosomal lac!
`gene. As a re ult, fu ll inducti on may be achieved with a ·
`
`little as 0.001 - 0.1 mmol IPTG L- 1, depend ing on plasmid
`copy number during the induction phase [48]. In add ition,
`in those instances in which the region of the plasmid DNA
`that controls the plasmid copy number is just downs tream
`from a lac or lac promoter, and there is no strong transcrip(cid:173)
`tion terminator between these elements, induction of the
`promoter may artificially increase the plasmid copy number
`(ie cause a phenomenon know n as runaway replication).
`Thi s may result in a higher level of gene expression as a
`consequence of the increased gene copy number [84].
`While this strategy mi ght seem to be an attractive way to
`increase the level of forei gn gene ex pression, the potential
`for plasmid instability and other metabolic alterations as a
`consequence is li kely to be problematic especially in large
`scale ferme ntati ons.
`In other cases, where the gene fo r /ad is present on the
`recombinant plasmid to prevent promoter leakiness (ie hi gh
`levels of non-induced background expression), the level of
`lac repressor present wi ll depend on the plasmid copy num(cid:173)
`ber in the cell , and IPTG concentratio ns required for a parti(cid:173)
`cular level of induction wi ll vary accord ingly [88]. Thus,
`the optimal concentration of IPTG appears to be hi ghl y sys(cid:173)
`tem-dependent, and th e types of metabolic changes found
`at the particular IPTG concentrat ions used by Wood and
`Peretti [88] may occur at different concentrations of IPTG
`in other sys tems. Thi s may explain why the optimal induc(cid:173)
`tion of glutathione-S-transferase using the lac promoter in
`the /ad-containing strain DH5a req uired onl y 0.062 mmol
`IPTG L - I [87]. The mutant /ad" gene which produces I 0-
`folcl more repressor than /ad, is often used in recombinant
`strains to provide better repression of recombinant gene
`express ion in the absence of inducer. Thus, the /adq pheno(cid:173)
`type is generall y optimall y induced using greater concen(cid:173)
`trations of IPTG compared to wild type /ad strains.
`Recombinant protein expression
`in strains containing
`either /ad" or plasmid borne lac! may be lower than th at
`of wild-type /ad strains, even when fu ll y induced with high
`levels of IPTG [ 48]. While plas mid-borne /ad genes pro(cid:173)
`vide more complete repression of the lac promoter, the hi gh
`level of repressor protein adds to the metabolic burden of
`recombinant cells and may affect the cells' translational
`capacity [88]. Mathematical models have been developed
`to evaluate the effects of different promoter/repressor con(cid:173)
`fi gurations on recombinant gene ex pression [48,52,53] and
`translational capacity [49] in E. coli . These models, how(cid:173)
`ever, do not account for situati ons such as inclusion body
`formation or for other limitations such as protein fo lding
`and sec retion to the periplasm.
`The characteristics of a particu lar protei n, its desired
`fo rm and the cellular location of the ex pressed protein may
`also signifi cantl y influence the level of IPTG u eel for opti (cid:173)
`mal inducti on. For example, the act ivity of secreted pro(cid:173)
`teins is often enh anced when exp ression is on ly parti all y
`induced. Reducing IPTG co ncentration from I to 0.005
`
`mmol L- 1 greatly enhanced the acti vity of subtilisi n E
`secreted to the peripl asm, vvhile reducing cell lysis [82].
`Replacing the lac promoter with the stronger /acUV5 pro(cid:173)
`moter to increase the ex press ion of secreted Fab fragments,
`provided a 5- to I 0-folcl increase in target protei n yield , but
`gave little to no improvement in the yield of function al Fab
`mo lecul es [76]. Simi larly, a 2- to 10-folcl increase in the
`yield of functional secreted Fab fragments was observed
`when the lac promoter was induced with 0.01 - 0.1 mmol
`IPTG L- 1
`, rather than 1.0 mmol L - I [73]. Effic ient pro(cid:173)
`duction of periplasmic {3-lactamase and human epidermal
`growth factor from the wc promoter also utili zed 0.1 mmol
`IPTG L - I
`[ 19]. In all of these cases, it appears that the
`solubility and the proper foldin g of secreted proteins ben(cid:173)
`efited greatl y from the lower transc ription rates when less
`than the 'standard ' inducer co ncentration of l mmol IPTG
`L - I was used.
`In contrast to the abo ve examples, hi gher levels of tran(cid:173)
`scription have been used to enhance the expression an d/or
`locali zation of some secreted proteins. For example, high
`level ex pres ion from a lac promoter, in overnight cultures,
`was used to improve the ex pression of periplasmic {3-lacta(cid:173)
`mase activity and enhance outer membrane permeabi lity
`so th at 90% of the {3-lactamase activity was released to the
`cul tu re medium [28]. Leakage of the periplasmic proteins,
`however, requ ired hi gh levels of target protein ex pression,
`since leakage of {3-lactamase acti vi ty to the mediu m was
`not observed when a weaker /acll mutant promoter was
`used. This periplasmic leakage, at hi gh levels of {3-lactam(cid:173)
`ase ex pression, may occur because the secretion of the
`forei gn protein interferes with the ex pression of the outer
`membrane proteins and, therefore, weakens the integrity of
`the outer membrane [28].
`Differences in the level of periplasmic expression with
`increasing promoter strength may be clue in part to the nat(cid:173)
`ure of the recombinant product itself. {3-l ac tamase is a
`native E. coli protein with a very effi cient secretion signal
`seq uence. High transcription and secretion rates for this
`protein may, therefore, be possible since evoluti on has
`li ke ly 'optimi zed' its structure for secretion into the per(cid:173)
`iplasm. {3-Iactamase may also fold more efficientl y than
`many forei gn proteins in the periplasm, so that expression
`limitations of acti vity clue to proper fo lding and/or aggre(cid:173)
`gation may be less significant. In the case of antibody frag(cid:173)
`ments, the solubil ity and level of secretion of single chain
`F"s may be significantl y affected by the order in which the
`V H and V L domains are secreted [5]. For Fab and F,. frag(cid:173)
`ments, the individual light and heavy chains must fold and
`associate properly to form functional proteins. Thus, the
`expression of Fnb and F" fragments may become li mited at
`hi gher induction levels by the fo lding of each protein chain
`as we ll as the homocl imeri zation or incorrect chain a soci(cid:173)
`ations leading to aggregation [42]. Since point mutations
`in Fnb and F'" fragments have been found to signi ficantly
`improve their fo lding and soluble ex pression in the per(cid:173)
`iplasm [43]. the amino acid seq uence of a protein can sig(cid:173)
`nificantly influence its soluble and functiona l expression
`when secreted to th e periplasmi c space.
`The level of inducer required for optima l expression
`depends on the strength of th e promoter, the presence or
`absence of repressor genes on a pl asmid , the cellular
`
`BEQ 1007
`Page 6
`
`

`
`location of product express ion, the response of the cell to
`recombin ant prote in express ion, the so lubility of the target
`protein and the characteristics of the prote in itself. For
`inducing the expression of an intercellul ar recombinant pro(cid:173)
`tein , the use of l mmol IPTG L - I is a reasonable starting
`point since max im al induction is pred icted to occ ur for both
`tacT and /acl'~ strai ns at thi s level [48]. For secreted pro-
`1 tein s, however, IPTG concent