vol ume 17 number 10 october 1995
`
`b-otechnology research and reviews
`I
`
`~11:~"
`
`~ ~ ·-
`, '
`_,_.
`'r 1
`I ~ -....... ._
`'~
`"" ...... ._
`Ill
`--
`"';;:;: L J J
`,...
`'roo.
`-.....J....
`· .. : ?;~
`~ T-..
`~ ---.....J......J.
`-·
`
`. I
`
`I
`
`r---..
`
`I""
`
`:-..
`
`1 3
`
`J. ..J.ll
`L
`''-J,.
`~ ......
`,...,....
`I ~
`-C
`._,.._,_
`............
`-,.....
`t- t-
`J..-1-o' t- r-
`L r- r-
`"""'-.J.,.
`~~ 'J......
`J _I ~ .... }
`
`- r--
`
`l
`_(~
`
`~
`~ --l..J
`I ~ .............
`~ ..
`''
`
`j_
`
`:~1 ( --
`
`-
`
`-
`
`'1...J. I
`,...,__
`
`- ~ I
`-----
`~ I I
`
`--'-'. I
`~ ~ ...............
`-~ '
`
`I....,
`
`., .'.
`
`1-,.....
`
`C)
`
`I ~
`
`Butterworth-Heinemann
`
`-
`-
`l::"h.~
`t-
`.. _:.{,
`r--
`~
`r-r-,..
`-.L,.L..,
`~~~ ~j
`rr ~
`~ c-.--
`
`BEQ 1010
`Page 1
`
`

`
`Enzyme and Microbial Technology is an
`interna.tional monthly biotechnology journal.
`Authors are urged to study th e notes on th e
`inside back cover.
`
`Enzyme and Microbial Technology is
`published monthly by Elsevier Science Inc.,
`655 Avenue of the Americas, New York, NY
`10010. Institutional subscription rate : $825.00.
`Outside the United States, Please add $100.00
`for postage. Individual subscription rate:
`$130.00. Outside the United States, please add
`$40.00 for postage. Claims for missing issues
`can be honored only up to three months for
`domestic addresses, six months for foreign
`addresses. Duplicate copies will not be sent to
`replace one undelivered through failure to
`notify Elsevier of change of address. Single
`copy and back volume information available
`upon request.
`
`This journal is indexed in Biological Abstracts,·
`CABS; Derwent Biotechnology Abstracts;
`Chemical Abstracts; Current Contents;
`Engineering Index Monthly; Scisearch;
`Biotechnology Citation Index; Science Citation
`Index; Research Alert; and Excerpta Medica.
`
`POSTMASTER: Send address changes to
`Enzyme and Microbial Technology, Elsevier
`Science Inc., 655 Avenu e of the America s,
`New York, NY 10010.
`
`Plea se direct inquiries regarding the placing
`of advertising in this journal to Art Carlucci,
`Elsevier Science Inc., 655 Avenue of the
`Americas, New York, NY, 10010. Telephone:
`(21 2) 633-3806; FAX: (212) 633-3820.
`
`No responsibility is assumed by the Publi sher
`for any injury and/or damage to person s or
`property as a matter of product liability,
`negligence or otherwise, or from any use or
`operation of any methods, products,
`instructions or ideas contained in the material
`herein . Because of rapid advances in the
`medical sciences, the Publisher recommends
`that independent verification of diagnosis and
`drug dosages should be made.
`
`© Elsevier Science Inc., 1995. All rights
`reserved. Second class postage paid at New
`York, NY and at additional mailing offices.
`This journal has been registered with the
`Copyright Clearance Center Inc. Consent is
`given for copying of articles for personal or
`internal use, or for the personal or internal
`use of specific clients. This consent is given
`on the condition that the copier pay through
`the Center the per-copy fee stated in the cost
`on the first page of each article for copying
`beyond that permitted by the U.S. Copyright
`Law. If no code appears on an article, the
`author has not given broad consent to copy
`and permission to copy must be obtained
`directly from the author. This consent does
`not extend to other kinds of copying; such as
`for general distribution, resale, advertising
`and promotional purposes, or for creating
`new collective works.
`
`§"·~his paper meets the requirements of
`
`ANSI Standard Z39.48-1992 (Permanence of
`Paper).
`
`All inquiries regarding copyrighted materials
`from this publication, other than reproduction
`through the Copyright Cl earance Center,
`should be directed to: Rights and Permissions
`Department, Elsevier Science Inc., 655 Avenue
`of the Americas, New York, NY 10010. Fax:
`212 633-3977.
`
`IBRAR
`
`866
`
`biotechnology research and reviews
`nr.K
`volume 17, number 10, october 1995, pages 865-960
`Editors: Sheldon W. May and Raymond E. Spier
`L T t 3
`Papers
`Lipasecmediated hydrolysis and esterification
`M. Zarevucka, Z. Zalska, M . Rejzek, L. Streinz, Z. Wimmer,
`M. Mackova, and K. Demnerova
`870 Regul.9 tion and kin.~tic modeling of galactose oxidase secretion
`Z. B. Ogel and M. Ozilgen
`877 Chemical derivatives of Pseudomonas aeruginosa elastase showing
`increased stability
`C. Besson, G. Favre-Bonvin, C. O'Fagain, and J. Wallach
`882 Studies on papain action in the synthesis of Gly-Phe in
`two-liquid -phase media
`J. A. Feliu, C. de Mas, and J. L6pez-Santin
`888 Low-level endoglucanase contamination in a Trichoderma reesei
`cellobiohydrolase II preparation affects its enzymatic activity
`on [3-glucan
`T. Reinikainen, K. Henriksson, M . Siika-aho, 0. Teleman, and
`K. Poutanen
`893 Effect of temperature on scleroglucan synthesis and organic acid
`production by Sclerotium glucanicum
`Y. Wang and B. McNeil
`900 Purification and characterization of an intracellular [3-glucosidase
`from Botrytis cinerea
`Y. Gueguen, P. Chemardin, A. Arnaud, and P. Galzy
`907 Analysis of plant harvest indices for bioregenerative life
`support systems
`A. Velayudhan, K. L. Kohlmann, P. J. Westgate, and M. R. Ladisch
`911 Study of the loss of streptokinase activity during cross-flow
`microfiltration: I. Immunologic approach
`I. Hernandez-Pinzon, F. Millan, and J. Bautista
`915 Fluidized bed reactors operating with immobilized enzyme systems:
`Design model and its experimental verification
`A. B6dalo, J. L. Gomez, E. Gomez, J. Bastida, and M . F. Maximo
`923 Analysis and optimization of recombinant protein production in
`Escherichia coli using the inducible pho A promoter of the E. coli
`alkaline phosphatase
`C. Li.ibke, W. Boidol, and T. Petri
`929. Characterization of sucrose synthase from rice grains for the
`enzymatic synthes is of UDP and TDP glucose
`L. Elling and M. -R. Kula
`935 Substrate use and production of o:-L-arabinofuranosidase during
`solid -state culture of Trichoderma reesei on sugar beet pulp
`N. Roche, P. Berna, C. Desgranges, and A. Durand
`942 Enhanced cellulase production by a mutant of Trichoderma reesei
`N. J. Gadgil, H. F. Daginawala, T. Chakrabarti, and P. Khanna
`947 Comparison of temperature- and isopropyl -[3-o-thiogalacto(cid:173)
`pyranoside-induced synthesis of basic fibroblast growth factor in
`high-cell -density cultures of recombinant Escherichia coli
`A. Seeger, B. Schneppe, J. E. G. McCarthy, W. -D. Deckwer, and
`U. Rinas
`954 Differential and synergistic action of Streptomyces endoxylanases
`in prebleaching of kraft pulps
`G. Elegir, M. Sykes, and T. W. Jeffries
`
`BEQ 1010
`Page 2
`
`

`
`(!IUTTERWOR TH
`E I NEMANN
`
`Comparison of temperature- and
`isopropyl-(3-n-thiogalacto-pyranoside-induced
`synthesis of basic fibroblast growth
`factor in high-cell-density cultures of
`recombinant Escherichia coli
`
`Anke Seeger,* Bernard Schneppe,t John E. G. McCarthy,t Wolf-Dieter Deckwer,*
`and Ursula Rinas*
`
`*GBF National Research Center for Biotechnology, Biochemical Engineering Division, and
`tDepartmenr of Gene Expression, Braunschweig, Germany
`
`Two different expression systems 1vere de veloped for expression of the eDNA encoding human basic fibroblast
`growth factor (bFGF) using Escherichia coli TGJ as host organism. Th e bFGF structural gene was cloned into
`two vectors differing only with respect to the promoter, 11·hich was either the bacteriophage 'A PRPL- or the E.
`coli lac-promoter. The resulting expression systems 1vere studied in high-cell density cultures . Cells were grown
`in a fed-bat ch procedure with a predetermined exponential f eedin g rate based on mass balances and kinetic
`equations to ensure constant specific growth rates. Prior to induction , cells were grown at 30°C. Product
`formation was induced by either a temperature shift from 30 to 4rc or by the addition of isopropyl-f3-o -thio(cid:173)
`galactopyranoside (IPTG). Under comparable culture conditions-induction of bFGF expression at41 to 45 g
`t- 1 dl )' eel/weight and the addition of feed medium with identical rates-bFGF accumulated to 4 .9 and 1.1 g
`1- 1 using the temperature- and /PTG-inducible expression systems, respectively . Th e final biomass concentra(cid:173)
`tions obtained using temperature- and IPTG-inducible expression systems were 61 and 135 g / - 1 dt)' eel/weight,
`whereas the specific product concentrations were 80 and 8 .1 mg bFGF g - 1 dry eel/weight, re:,pecti vely . When
`a temperature shift was used for product induction , 30% of bFGF was recovered as inclusion bodies in the
`insoluble cell ji·action. IPTG-dependent induction yielded exclusively soluble bFGF.
`
`Keywords: Basic fibroblast growth fac tor; comparison of different ex pression systems; high-cell-densit y cultivation ;
`recombinant Escherichia coli
`
`Introduction
`
`Economic processes for the production of recombinant
`pharmaceutic proteins require high specific product yields
`and the development of simple high-cell-density cultivation
`techniques that comply with cGMP requirements . The ac(cid:173)
`cumulation of growth-inhibiting metabolic by-products
`such as acetic acid in response to oxygen limitation, excess
`
`Address reprint requests to Dr. Rinas, GBF National Researc h Center for
`Biotechnology, Biochemical Engineering Division , Mascheroder Weg I ,
`38124 Braunschweig, Germany
`Received 22 September 1994; accepted 28 Nove mber 1994
`
`carbon, or cellular stress caused by recombinant protein
`expression represent the main obstacles to obtaining high
`cell densities and high product concentrations in the biore(cid:173)
`actor . The highest specific product concentrations are gen(cid:173)
`erally obtained by placing the structural gene downstream
`of strong promoters such as the temperature shift-inducible
`bacteriophage 't...P L or 't...P R promoters. However, high tem(cid:173)
`peratures during recombinant protein synthesis can increase
`the probability of inclusion body formation, 1 which need to
`be subjected to solubilization and refolding steps to allow
`generation of functional proteins .
`Fed-batch procedures have proved to be the most effec(cid:173)
`tive means of maximizing cell mass concentration in the
`bioreactor. Several different feeding strategies have been
`
`Enzyme and Microbial Technology 17 :947-953, 1995
`© 1995 by Elsevier Science Inc.
`655 Avenue of the Americas, New York, NY 10010
`
`0141-0229/95/$10.00
`SSDI 0141 -0229(94)00123-9
`
`BEQ 1010
`Page 3
`
`

`
`Papers
`
`developed to maintain carbon-limited growth using defined
`medium without accumulation of toxic levels of acetic acid.
`For example, the highest cell concentrations of nonrecom(cid:173)
`binant Escherichia coli were obtained in fed-batch pro(cid:173)
`cesses by feeding solid glucose via a special apparatus into
`a pressurized fermentor ( 134 g 1- 1 DCW) , 2 by removing
`inhibitory substances such as acetic acid by dialysis (174 g
`1- 1 DCW), 3 and by using feedback control strategies for
`carbon source feeding ( 110-125 g I - 1 DCW). 4-6
`In contrast to high-cell-density cultures of nonrecombi(cid:173)
`nant E. coli, cell concentrations of E. coli synthesizing
`recombinant proteins have not exceeded 100 g 1- 1 DCW.
`The application of high-cell-density fed-batch cultures of E.
`coli for recombinant protein production has been summa(cid:173)
`rized recently, 7.8 with the highest cell density (77 g 1- 1
`DCW) obtained during temperature-induced synthesis of a
`SpA-f3-galactosidase fusion protein. 9 Higher cell concen(cid:173)
`trations (92-97 g 1- 1 DCW) of E. coli synthesizing recom(cid:173)
`binant proteins under the control of IPTG-inducible promot(cid:173)
`ers using predetermined exponential feeding profiles have
`been reported subsequently . 10·11
`In this study we compare the performance of tempera(cid:173)
`ture- and IPTG-inducible expression systems in high-cell(cid:173)
`density cultures of recombinant E. coli synthesizing basic
`fibroblast growth factor (bFGF) as a heterologous protein of
`pharmaceutic importance in wound healing, tissue grafting,
`and nerve regeneration . 12 To ensure comparability of the
`recombinant constructs, the bFGF structural gene was
`cloned into the pBR322-derived expression vector pCY(cid:173)
`TEXP1'3 ·14 under the control of either the X.PRPc or the
`lac-promoter. Except for the application of different induc(cid:173)
`tion procedures, the cultivation conditions were identical
`during fed-batch cultivation of both expression systems.
`
`Materials and methods
`Bacterial strains and plasmid construction
`
`Escherichia coli K 12 strains TG I and TG2 have been described
`previously .15 General recombinant DNA methods were used for
`plasmid construction. 15
`
`Shaken flask experiments and pulse labeling
`
`Pulse labeling was performed essentially as described previ(cid:173)
`ously .16 Cultures were shaken at 30°C in defined medium supple(cid:173)
`mented with amino acids (50 J.Lmol each, except methionine). At
`OD550 = 0.5 , bFGF synthesis was induced by either temperature
`shift to 42°C or by addition of IPTG (final concentration 2 mM).
`After 20 min, cells were pulse labeled with 35S-methionine for I
`min .
`
`High-cell-density cultivation: Medium and
`inoculum preparations
`
`· 11
`The medium was prepared essentially as described previously 6
`The compositions of the batch and feed media are given in Table
`1. For preparation of 2.5 I batch medium , (NH 4 hHP04 , KH 2P04 ,
`citric acid, EDTA , and trace elements were dissolved in 2.3 I
`distilled water in the bioreactor, pH was adjusted to 6.3 with 5N
`NaOH, and the solution was sterilized for 30 min at 12l °C. Stock
`solutions of MgS0 4 and glucose were each sterilized separately for
`30 min at l2! °C. Thiamine and ampicillin were steri lized sepa-
`
`Table 1 Medium composition
`
`Components
`
`Glucose· H20 "
`KH 2 P0 4
`(NH 4 ) 2 HP0 4
`MgS0 4 · 7H 20
`Citric acid
`EDTA
`CoCI 2 · 6H 20
`MnCI 2 · 4H 20
`CuCI 2 · 2H 20
`H3 B03
`Na 2 Mo04 · 2H 20
`Zn(CH 3C00b · 2H 20
`FE(III)citrate
`Thiamine · HCI
`Antifoam SP1
`Ampicillin
`
`Batch
`medium
`
`27.5gl - 1
`13.3 g 1- 1
`4.0 g 1- 1
`1.2 g 1- 1
`1.7 g 1- 1
`14.1 mg 1- 1
`2.5mgl - 1
`15.0 mg 1- 1
`1.5 mg 1- 1
`3.0mgl - 1
`2.1 mg 1- 1
`33.8 mg 1- 1
`100.8 mg 1- 1
`4.5 mg 1- 1
`0.1 mll - 1
`50mgl - 1
`
`Feeding
`solution
`
`875gl - 1
`
`20.0 g 1- 1
`
`13.0 mg 1- 1
`4.0mg l- 1
`23.5mgl - 1
`2.3mgl - 1
`4.7 mg 1- 1
`4.0mgl - 1
`16.0 mg 1- 1
`40.0 mg 1- 1
`4.5 mg 1- 1
`
`90 mg 1- 1
`
`"Initia l concentration of glucose for precultures was 12.0 g 1- 1
`
`rately by filtration. After cooling , all solut ions including antifoam
`reagent SPI (Th. Goldschmidt AG, Essen , Germany) were com(cid:173)
`bined, and the pH was adjusted to 6. 7 with aqueous NH3 (25 % w
`w - 1) prior to inoculation. All components of the feed medium
`were sterilized separately before being mixed . The following stock
`(909 g 1- 1) ,
`solutions were prepared: g lucose · H 2 0
`MgS04 · 7H 20 (1.0 g ml - 1), EDTA (8.4 mg ml - 1), and trace
`· 6H 20 (2 .5 mg ml - 1), MnCl 2 • 4H20 (15.0 mg
`elements: CoCI 2
`ml - 1), CuCI 2
`· 2H 20 (1.5 mg ml - 1), H3B0 3 (3.0 mg ml - 1),
`· 2H 20 (2.5 mg ml - 1), Zn(CH3C00h · 2H 20 (13 .0
`Na 2Mo0 4
`mg m1 - 1) and Fe(lll)citrate ( 12 .5 mg ml - 1). Glucose and MgS04
`were sterilized for 30 min at 121 °C . EDTA and trace elements
`were sterilized by filtration. Stock solutions of thiamine (I 0 mg
`ml - 1) and ampicillin (50 mg ml - 1) were prepared directly prior to
`use (sterilized by filtration). Antifoam reagent SPI was dissolved
`in ethanol (50% v v - 1).
`The first preculture (10-ml batch medium in 100-ml shake
`flask) was inoculated with a single colony from M9 minimal agar
`plates 15 and incubated on a rotary shaker at 30°C for 10-1 4 h.
`Following this, two funher precultures (each 100-ml batch me(cid:173)
`dium in I ,000-ml shake fl asks) were inoculated with the first
`preculture (I % v v - 1), incubated on a rotary shaker at 30°C for
`10-12 h, and used for inoculation of the bioreactor.
`
`High-cell-density cultivation: Bioreactor culture,
`controls, and calculations
`
`The main cultivations were carried out at 30°C in a 5-1 bioreactor
`(Type Biostat MD; B. Braun Diessel Biotech GmbH, Melsungen,
`Germany) equipped with an additional cooling device (B. Braun
`Diessel Biotech GmbH). The initial batch culture conditions were
`as follows: initial culture volume = 2.5 I, air flow = 2 .5 I min - I,
`stirrer speed = 500 min - 1 • Thermal mass flow meters (Brooks
`Instruments B. Y., Yeenendaal, The Netherlands) were used for
`mixing air and oxygen. The dissolved oxygen concentration was
`maintained at 40% of air saturation by increasing the stirrer speed.
`The pH was maintained at pH 6. 9 by addition of aqueous ammonia
`(25% w w- 1). Control of pH and dissolved oxygen was carried
`out by the digital control unit (DCU) of the bioreactor. The con(cid:173)
`centrations of oxygen and carbon dioxide in the ex haust gas were
`determined by paramagnetic and infrared gas analysis systems,
`respectively (Mai hak , Hamburg , Germany).
`
`948 Enzyme Microb. Techno!., 1995, vol. 17, October
`
`BEQ 1010
`Page 4
`
`

`
`bFGF synthesis by recombinant E. Coli: A. Seeger et al.
`
`resuspended in distilled water, centrifuged again, and dried at
`40°C under vacuum until constancy of weight.
`Glucose and acetic acid were analysed by HPLC using an
`Aminex HXP-87H column (BioRad) for separation (T = 25°C)
`and ultraviolet and refractive index detectors (Techlab, Germany)
`for detection. Sulphuric acid (0.005 M) was used as mobile phase
`(flo w rate = 0.5 ml min - 1). In addition , a glucose analyser
`(Yellow Springs, OH) was employed to determine the concentra(cid:173)
`tion of glucose. A polarographic electrode (Ingold, Germany) was
`used to analyze the concentration of dissolved oxygen. Ammo(cid:173)
`nium concentrations were analysed by an ammonium electrode
`(type Orion 95-12; Colora, Lorch , Germany). The concentration
`of phosphate was analysed according to a modified procedure
`described in the German Standard Methods. 17
`Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
`(SDS-PAGE) was performed according to the method of Laem(cid:173)
`mli.1 8 Samples were mixed with sample buffer (1 :3) consisting of
`7% (w/v) SDS , 35% (w/v) glycerol , !50 mM DTT (dithiothreitol) ,
`250 mM Tris/HCl (pH 6.8), boiled for 10 min , and e lectrophoresed
`immediately using precast gels (ExcelGel 8-18; Pharn1acia). Gels
`were stained with Coomassie Brilliant Blue R250 . Quantification
`was done by densitometry (Hirschmann elscript 400 , Germany)
`using bovine serum albumin (A2 153 , Sigma) for calibration . For
`immunodetection of bFGF samples were run on 9 to 16% SDS(cid:173)
`PAGE gels at constant current (40 mA per gel , 160
`· 200 · 1.5
`mm 3) and electroblotted to Immobilon PYDF membranes (Milli(cid:173)
`pore) according to the method described by Towbin eta!. 19 Rabbit
`anti -bFGF (S igma F3393 ; 1-24 amino acid peptide) antiserum ,
`biotinylated goat-anti-rabbit-IgG (Gibco, BRL), streptavidin(cid:173)
`alkaline phosphatase (Gibco BRL) , and an alkaline phosphatase
`substrate (Gibco , BRL, cat. no. 8280SA) were used for immun(cid:173)
`odetection of bFGF. Immunostaining was performed according to
`the manufacturers ' instructions.
`
`Results and discussion
`Construction of plasmids for temperature- and
`IPTG-inducible expression of the human bFGF gene
`
`The synthetic structural gene encoding full -length human
`bFGF (155 amino acids) was isolated as an Ndei/Sa!I(cid:173)
`fragment from pJLA503-hbFGF. 13 The entire sequence of
`the gene was verified to ensure expression of authentic hu(cid:173)
`man bFGF. For temperature-inducible production of bFGF
`the expression vector pCYTEXP 114 was digested with Ndel
`and Saii and ligated with the Ndei/Sall-fragment containing
`the bFGF gene . The resulting plasmid pAFGFB carried the
`bFGF gene under the control of the AP Rp cpromoter. For
`efficient repression of bFGF expression at permissive tem(cid:173)
`peratures the A cl857 repressor was also encoded on plas(cid:173)
`mid pAFGFB . To construct the IPTG-inducible expression
`vector, pCYTEXP 1 was digested with Notl and X hoi to
`remove the fragment containing the A c/857 repressor and
`P Rp cpromoter. Klenow treatment and religation resulted in
`plasmid pCTY tlc!PRPL This plasmid was digested with
`Kpni and ligated with the Kpnl!Kpnl fragment containing
`the lac-promoter from plasmid pBI. 20 The resulting plasmid
`with the lac-promoter in the correct orientation relative to
`the pCYTEX-polylinker was named pCYTiacPro. Diges(cid:173)
`tion of pCYTiacPro with Ndel and Sail and ligation with the
`Ndei!Sall-fragment containing the bFGF gene resulted in
`the expression vector pLACFGFB. All cloning experiments
`were verified by subsequent sequencing . Figure 1 shows
`the two expression vector constructs used in the present
`
`After consumption of the initial glucose, as indicated by an
`increase of the dissolved oxygen concentration , the fed-batch
`phase was started. Feeding was canied out essentially as described
`previously. 11 To maintain a constant specific growth rate , the
`time-dependent substrate mass flow rate m,(t) (g h - 1) of glucose
`during the fed-batch process with a desired specific growth rate
`f.Lser (h - 1) , starting at cultivation time IF (h) , and cell density XF
`(g 1- 1) was calculated as follows 11 :
`
`· f.Lset + Ill] · XF · VF · e f'-'<d t- tr)
`
`1-
`
`Ills(!) = [-
`Yxts
`
`where Yx 1s (g g- 1) is the yield coefficient for glucose (0.5), 111 (g
`g- 1 h - I) the maintenance coefficient (0 .025) , and V F (I) the vol(cid:173)
`ume in the bioreactor at the onset of the fed-batch procedure.
`The feeding solution and aqueous ammonia (25 % w w - 1) were
`each placed on a balance to allow the determination of the time(cid:173)
`dependent consumption of glucose (concentration of glucose was
`0.61 g g- 1) and to calculate the time-dependent volume of the
`culture broth (densities of glucose feeding solution and aqueous
`NH 3 were 1.3 and 0 .91 g cm- 3 , respectively) . Changes in culture
`volume caused by sampling were corrected manually by subtract(cid:173)
`ing sample volume from the volume of the culture broth. The
`dissolved oxygen concentration was maintained at 40% of air sat(cid:173)
`uration by increasing the stirrer speed and/or blending air with
`pure oxygen. Foam was suppressed when necessary by the addi(cid:173)
`tion of anti foam reagent SP I . DCU , mass flow meters, balances,
`feeding pumps , and exhaust gas analysis systems were interfaced
`to a VME-bus microcomputer using Universal Bio-Process Con(cid:173)
`trol System (UBICON) software (esd , Hannover , Germany). In
`addition to control functions carried out by the DCU (temperature,
`pH , dissolved oxygen concentration by changing stiner speed),
`UBICON was used to control the mass flow meters (control of
`dissolved oxygen concentration by blending air with pure oxygen)
`and the substrate feeding pump. UBICON was also used to cal(cid:173)
`culate the time-dependent consumption of glucose and the time(cid:173)
`dependent volume of the culture broth .
`Yield coefficients Y x ts (biomass/glucose) or Y Pts (bFGF/
`glucose) were determined considering the total amount of glucose
`consumed and biomass or product produced during a particular
`phase of the cultivation process . Specific growth rates (fL,.eat) dur(cid:173)
`ing the production phase (phase 2) were determined without taking
`into account the final part of the production phase (decreasing
`growth rate with final cessation of cell growth).
`
`Preparation of total cell lysates, as well as soluble
`and insoluble cell fractions
`
`All procedures were carried out at 4°C unless otherwise indicated.
`Cells were harvested by centrifugat ion and resuspended in 50 mM
`sodium phosphate buffer (pH 7) to an OD600 of 3 (temperature
`induction) or 10 (IPTG induction). Total celllysates were prepared
`by sonication on ice . Soluble lysates were obtained as supernatants
`after centrifugation of total celllysates at 15 ,000 g for 45 min. The
`remaining pellet was collected as insoluble cell lysate (cell debris
`and inclusion bodies) and resuspended in 50 mM sodium phosphate
`buffer (pH 7) in one-fourth of the original volume. All samples
`were stored at -70°C until further analysis.
`
`Analytical methods
`
`Cell growth during high-cell-density cultivation was followed by
`measurement of the optical density at a wavelength of 600 nm
`(Novaspec II ; Pharmacia LKB , Freiburg, Germany) . In addition,
`dry cell weights (DCW) were determined from 1-ml aliquots of
`culture broth collected in balanced 1.5 ml centrifugation tubes .
`Cell pellets were collected by centrifugation for 3 min at 3 ,300 g,
`
`Enzyme Microb. Technol., 1995, val. 17, October 949
`
`BEQ 1010
`Page 5
`
`

`
`Papers
`
`NOl l
`
`Kpn l/
`
`X hoi
`'-
`
`Nde l
`/
`
`Sail
`
`pHGfB
`
`30
`
`100
`
`6
`
`5
`
`TIS-E
`
`fd term
`
`X hoi
`Kpnl
`K pn t '-.
`
`Ndcl
`/
`
`pLACI'GFB
`
`.::::.
`~
`3
`u..
`(.!)
`u..
`
`4
`
`2
`
`_[]
`
`•
`
`'=' '
`2 0
`~
`.,
`"
`0
`u
`-=
`(.!)
`\1
`
`10
`
`10
`
`.::::.
`~
`3:
`()
`a
`0
`
`TI S-E
`
`fd term
`
`0
`
`0
`
`b a lch
`
`fed - balch
`
`2
`
`A \0°
`I'
`
`I
`
`\ 'V
`
`10
`
`20
`
`30
`
`0.1
`
`0
`
`0.7
`
`0 .6
`
`.::::. 0.5
`~
`"0 0 . 4
`u
`0
`u
`;;
`u
`<(
`
`0.3
`
`0.2
`
`Figure 1 Stru cture and unique restriction sites of plasmids
`pHGFB and pLACFGFB
`
`work. The synthetic bFGF gene was preceded by a synthetic
`copy of the intercistronic sequence normally located up(cid:173)
`stream of the E. coli atpE gene [Translational Initiation
`Sequence-E (TIS-E)].2 1 Transcription (Figure 1, see ar(cid:173)
`rows) was driven by either the bacteriophage major pro(cid:173)
`moter region A. P Rp L (p'A.FGFB ) or by the E. coli lac(cid:173)
`promoter (pLACFGFB), and was terminated by the bacte(cid:173)
`riophage fd terminator (fd term). The positions of a number
`of the unique sites that can be used to exchange the com(cid:173)
`ponents of these modular vectors are indicated (Figure 1).
`
`-
`
`-
`
`1 2
`
`3 4
`
`5 6
`
`7 8
`
`Figure 2 Pulse-label experim ents w ith temperature- and IPTG (cid:173)
`indu cible expression systems using Escherichia coli TG2 as host
`organism. Cells w ere pulse labeled prior and after product in(cid:173)
`duction . The temperature-inducibl e expressi on system with
`pCYTEXP 1 as contro l plasmid (lanes 1 and 3) and pA.FGFB as
`bFGF express ion vecto r (lanes 2 and 4) were pulse labeled prior
`to (lanes 1 and 2) and after temperature shift (lanes 3 and 4)
`from 30 to 42°C. The IPTG-inducible express ion system with
`pCYT/acPro as co ntrol pl asmi d (lanes 5 and 7) and pLACFGFB as
`bFGF expression vector (lanes 6 and 8) were pul se labeled prio r
`(lanes 5 and 6) and afte r addition of IPTG (l anes 7 and 8). The
`arrows indicate the position of bFGF
`
`950 Enzym e Microb. Tec hno!., 1995, val. 17, October
`
`3.5
`
`3.0
`
`'=' 2.5
`'
`~
`E 2.0
`::1
`c:
`1.5
`0
`E
`E 1.0
`<(
`0
`
`0 .5
`
`15
`
`.::::.
`10
`~
`2
`0
`.,
`.<:
`c._
`•
`
`0
`.<:
`(L
`
`5
`
`0.0
`
`0
`
`A 0.1
`
`0.0
`0
`
`10
`
`20
`
`30
`
`time [h)
`
`Figure 3 Temperature-induced production of bFGF in high-cell(cid:173)
`density cultures . After unlimited growth during batch mode
`(f-1-max = 0.51 h - 1 ), fed-batch mode was started with a desired
`specific growth rate of f-l. set = 0.12 h - 1 (phase 1 ). Product for(cid:173)
`mation (phase 2) was induced at 45 g 1- 1 DCW by temperature
`shift from 30 to 42°C. Feeding was reduced in phase 2 of the
`fed-batch process such that a specific growth rate of f-1-set = 0.08
`h - 1 would have bee n maintained at 30°C. (A) Time course of
`bFGF formation, glucose consumption, and cell growth . (B)
`3 - ), and ammonium
`Time course of acetic acid , phosphate (P0 4
`(NH 4 +)concentrations (g 1- 1 ) in the cell fre e culture broth. Each
`arrow indicates the addition of 16.6 g KH 2 P0 4 and 5 g
`(NH 4 ) 2HP04 to the culture broth
`
`Temperature- and IPTG-induced expression of
`human bFGF-shakenflask experiments and
`pulse labeling
`
`Both expression strains were grown in shake flasks on de(cid:173)
`fined medium supplemented with am ino acids and exam(cid:173)
`ined for overexpression of bFGF . After induction of bFGF,
`cells were pulse labeled with radioactive methionine as de(cid:173)
`scribed in Materials and methods. Expression from the
`pCYTEXPl construct p'A.FGFB was at least 10 times greater
`than that of the equivalent P1a c construct (Figure 2) . It is not
`known whether the stability of the recombinant bFGF was
`different in the two expression strains. However, there is no
`reason to assume that there was increased proteolytic deg(cid:173)
`radation of bFGF in the strain carrying the IPTG- inducible
`express ion vector pLACFGFB . Indeed, proteolytic degra(cid:173)
`dation of heterologous or abnormal proteins not protected
`within inclusion bodies is in general more prominent at
`higher temperatures 22
`·23 Therefore, proteolytic degrada(cid:173)
`tion of bFGF would be expected to be more pronounced
`during temperature-induced expression (42°C versus 30°C
`
`BEQ 1010
`Page 6
`
`

`
`bFGF synthesis by recombinant E. Co li: A. Seeger et al.
`
`30
`
`b a l c h
`100 A
`
`fed- b alch
`
`6
`
`5
`
`~ 4
`
`"-
`'-"
`"-
`
`' "' ~ 3
`.D 2 •
`
`10
`
`.,
`
`Ill
`0
`u
`::>
`'-"
`\I
`
`10
`
`~
`u
`0
`0
`
`Cultivation started as a batch process with cells growing
`at 30°C with maximal specific growth rates (TG 1 :pt..FGFB:
`, TGl:pLACFGFB: IJ!max = 0 .56 h - 1
`f!- 11wx = 0. 5 1 h - 1
`) .
`The cell mass concentrations at the end of the batch process
`reac hed 11. 2 and 12.4 g 1- 1 dry cell weight using temper(cid:173)
`ature- and IPTG-inducible ex pression systems , respec(cid:173)
`tively . The highest concentrations of acetic acid accumu(cid:173)
`lated during unlimited growth in the batch phase of E . coli
`TG I can-ying the plasm ids pt..FGFB and pLACFGFB were
`0 .23 and 0.5 g 1- 1, respectively (see also Figures 3b and
`4b). Feeding was started after glucose was consumed (Fig(cid:173)
`ures 3a and 4a) and the exponential feeding rate was ad(cid:173)
`justed in such a way that carbon-limited growth continued
`with a constant specific growth rate of f!-set = 0 .12 h - 1
`(phase I ; see Table 2) . This growth rate was chosen because
`accumulation of acetic acid was not observed at this growth
`rate (f!-set < f!-cr;1). [Using the described cultured conditions
`and nonrecombinant E. coli TG 1 the critical growth rate
`(f!-cr;1) that did not cause accumulation of acetic acid was
`approximately 0.17 h - 1• However, when recombinant E .
`coli TG 1 was employed , a reduction in the critical growth
`rate (depending on the type of plasmid present) was ob(cid:173)
`served.] During this phase of the fed-batch process , the
`actual specific growth rate (f!-rea t) was identical to the de(cid:173)
`sired specific growth rate ( fl. set) . The temperature was main-
`tained at 30°C during fed-batch phase 1 prior to induction of
`bFGF expression.
`Expression of bFGF was induced when the cell mass
`concentrations reached 45 and 41 g 1- 1 dry cell weight
`using the temperature- and IPTG-inducible expression sys(cid:173)
`tems , respectively. Product formation (phase 2; Figures 3
`and 4) was induced either by temperature shift from 30°C to
`42°C (TG 1 :pt..FGFB) or by addition of IPTG (final concen(cid:173)
`tration I mM, TGL:pLACFGFB) without a temperature
`shift.
`
`1 2 3 4 s 6 7 8 9 10 11 12
`
`Figure 5
`lmmunoblots of total cell lysates (lanes 1, 4, 7, and
`10). insoluble fractions (lanes 2, 5, 8, and 11). and soluble frac(cid:173)
`tions (lan es 3, 6, 9, and 12). Samples of TG1 :pt..FGFB (lanes 1-6)
`and TG1 :pLACFGFB (lanes 7- 12) were collected directly prior to
`product induction (lanes 1-3 and 7-9) and at the end of the
`cultivations (lanes 4- 6 and 10--12). Samples of TG1 :pLACFGFB
`(lanes 7-12) w ere concentrated twice relative to samples of
`TG1 :pAFGFB (lanes 1-6). Insoluble fractions were concentrated
`four t imes relative to the respective soluble fractions . The arrow
`indicates the position of bFGF
`
`Enzyme Microb. Techno!., 1995, val. 17, October 951
`
`0
`
`0
`
`3.5
`
`3.0
`
`~ 2.5
`~
`E 2.0
`" ·c 1.s
`0
`E
`E 1.0
`<(
`0 0.5
`
`0.0
`
`15
`
`";;; 10
`
`2
`0
`.<:
`"(cid:173)
`Ill
`0
`.<:
`
`0.. •
`
`5
`
`0
`
`0. 7 , . - - - - . - - - - - - - - - - - - .
`2
`B
`
`0.6
`
`0 .5
`
`0 . 4
`
`0.3
`
`"
`"'
`
`-u
`u
`0
`u
`
`e o.z
`<
`""
`
`0 . 1
`
`10
`
`20
`
`30
`
`time (h]
`
`Figure 4
`IPTG -induced production of bFGF in high-cell-density
`cultures. After unl imited growth during batch mode (JJ.max =
`0.56). fed-batch mode was started with a desired spec if ic growth
`rate of JJ.ser = 0.12 h - 1 (phase 1). Product formation (phase 2)
`was induced at 41 g 1- 1 DCW by adding IPTG to the bioreactor
`at a final concentration of 1 mM. Feeding was reduced in phase
`2 of the fed-batch process to mainta in a specific growth rate of
`JJ.s et = 0.08 h - 1 at 30°C. (A) Time course of bFGF formation,
`glucose