Cellular responses to the induction
`of recombinant genes in
`Escherichia coli fed-batch cultures
`
`Dissertation
`
`in fulfillment of the requirements for
`the degree of
`
`doctor rerum naturalium (Dr. rer. nat.)
`
`submitted to the
`
`Martin-Luther-Universität Halle-Wittenberg
`Faculty of Mathematics and Natural Sciences
`Department of Biochemistry and Biotechnology
`
`by
`
`Hongying LIN
`
`
`
`
`
`
`
`
`
`February 2000
`
`BEQ 1021
`Page 1
`
`

`
`Referees:
`
`1.
`
`2.
`
`3.
`
`Prof. Dr. Rainer Rudolph
`Institut für Biotechnologie
`Martin-Luther-Universität Halle-Wittenberg, Germany
`
`Prof. Dr. Sven-Olof Enfors
`Department of Biotechnology
`Royal Institute of Technology, Sweden
`
`Prof. Dr. Michael Hecker
`Institut für Mikrobiologie und Molekularbiologie
`Ernst-Moritz-Arndt-Universität Greifswald, Germany
`
`Halle (Saale), 28.02.2000
`
`BEQ 1021
`Page 2
`
`

`
`Lin, Hongying (1999). Cellular responses to the induction of recombinant genes in Escherichia coli fed-batch
`cultures. Department of Biochemistry and Biotechnology, Martin-Luther-Universität Halle-Wittenberg, Halle
`(Saale), Germany
`
`Abstract
`This thesis concerns the investigation of cell growth, plasmid stability and amplification, recombinant protein
`overproduction, cellular metabolism, and several stress responses after induction in glucose limited fed-batch
`recombinant cultures. The extensive study was specifically focused on one process for production of a yeast
`a-glucosidase in E. coli RB791 by derepression of the Ptac promoter with IPTG. This investigation was also
`compared to other model systems, including CRIMI (creatinine imino hydrolase) in a small scale cultivation and
`ZZ protein (a modified domain B of staphylococcal protein A) in a large-scale process.
`Induction of a-glucosidase formation led to an inhibition of cell growth and glucose uptake. The growth
`inhibition was connected to a decrease of the colony forming ability of the cells, which declined to approximately
`5 % within 4 h after induction in the strain with coexpression of argU-tRNA. The non-culturable cells were shown
`to have not lost all metabolic activities, and even succeed to maintain some glucose uptake and respiratory
`ability. The ability of these cells for replication is apparently not only impaired by competition of the synthesis of
`the recombinant product to the formation of cellular house-keeping proteins, but specifically by continued
`damage of the chromosomal DNA or loss of superhelicity. The cells are unable to induce the SOS response, as
`the product formation occupies a large part of the protein synthesis machinery, and consequently the cells loose
`their ability to recover irrevocably. This model of the inability of cells to have not the opportunity to respond to
`DNA damage is new in view of recombinant protein production.
`Furthermore, up-growth of plasmid-free cells was observed in the a-glucosidase process. The maximal glucose
`uptake capacity decreased to only about 25 % of the q smax of the batch phase. Reduction of q smax may be a serious
`problem in recombinant fed-batch processes, because it results in overfeeding of substrate which in a turn
`supports the up-growth of plasmid-free cells and therefore lowers the productivity.
`After induction, the recombinant plasmid pKK177glucC was amplified by a factor of three to five. The plasmid
`copy number increased from about 100 to 300-400 per cell within a period of six hours in glucose-limited fed-batch
`cultivations. In contrast, no amplification occurred if product formation was not induced. Cultures with the same
`E. coli strain, but other recombinant ampicilline based plasmids, were also overgrown by plasmid-free cells, when
`the growth was inhibited by overexpression of the recombinant genes, but showed no up-growth of plasmid-free
`cells and no plasmid amplification when product formation did not inhibit growth.
`Glucose limited fed-batch cultivations of Escherichia coli cells are characterized by a transient increase of the
`stringent response regulator ppGpp (guanosine-3’,5’-bisphosphate), a higher concentration of the general stress
`response regulator s S and an accumulation of extracellular cAMP during the shift from unlimited to limited
`growth. The influence of the overexpression of recombinant genes on the concentration of these stress regulators
`was compared in different expression systems. It has been shown that the response can be different in
`dependence on the recombinant gene. In case of the a-glucosidase process, no general stress response was
`induced, and the concentration of the three regulators (ppGpp, s S and cAMP) decreased to very low levels. In
`contrast, induction of the recombinant CRIMI caused a strong increase of s S and continuous accumulation of
`cAMP in the cultivation medium. Although the different products were accumulated to similar levels in these
`various systems, significant differences were also detected in connection to the influence of the recombinant
`production on the cellular growth and cell survival. The results suggest that induction strength on the
`transcriptional level and the strength of the ribosome binding site, but specifically the gene codon usage of a
`recombinant gene influence the behavior of stress signals.
`The small scale process of a-glucosidase was also investigated by a down-scale procedure, where glucose
`oscillations were created by an on/off feeding mode in either short cycles (1 min) or long cycles (4 min). The
`influence of repeated short-term glucose starvation on the cell death rate, product stability, and up-growth of
`plasmid-free cells was concluded from investigation of a number of general and specific process parameters.
`Although the glucose uptake capacity was inhibited in all cultures performed, the up-growth of plasmid-free cells
`in the culture was strongly inhibited by fast oscillations. In connection to product formation the initial
`a-glucosidase accumulation was the same in all cultures, but the stability of the product was significantly lower in
`the cultivation with long cycles, possibly because of a higher stress level.
`Finally, a study of cell growth kinetics and physiology during large-scale (12 m3 / 30 m3) fermentation of E. coli
`W3110 including a recombinant ZZ protein process was performed within a EU network project. The data
`obtained from the large-scale processes demonstrate the existence of gradients for glucose and oxygen and show
`the effect of mixing on cell growth and product formation.
`Keywords: E. coli, recombinant protein, cell segregation, cell viability, plasmid stability, plasmid amplification, glucose
`uptake, stress response, cAMP, ppGpp, sigma S, energy charge, nucleotide, glucose oscillations, fed-batch
`fermentation, scale-down, large-scale, a-glucosidase, creatinine imino hydrolase, ZZ
`
`BEQ 1021
`Page 3
`
`

`
`Lin, Hongying (1999). Zelluläre Reaktionen auf die Induktion rekombinanter Gene in Fed-
`batch Prozessen mit Escherichia coli. Institut für Biotechnologie der Martin-Luther-Universität
`Halle-Wittenberg, Deutschland
`
`Zusammenfassung
`
`Ziel dieser Arbeit ist die Untersuchung zellulärer Reaktionen auf die Induktion rekombinanter Gene in
`
`Fed-batch Prozessen mit Escherichia coli, wobei auch Einflüsse der Maßstab-vergrößerung
`
`biotechnologischer Prozesse mit rekombinanten Mikroorganismen auf die Produktivität und die
`
`Zellphysiologie berücksichtigt werden. Als Modellsystem wurde im Rahmen dieser Arbeit ein
`rekombinanter Fed-Batch-Prozeß zur Produktion von a
`wurde in Bezug auf Wachstum, Zellsegregation, Plasmidstabilität, und Produktbildung charakterisiert.
`
`-Glucosidase ausgewählt. Dieser Prozeß
`
`Darüber hinaus wurden jedoch auch Veränderungen der Substrataufnahme, des Nukleotidpools, des
`
`Proteinmusters, sowie der Einfluß der Induktion auf die Expression verschiedener mRNA's
`untersucht. Die am a
`rekombinanten Prozessen verglichen (Creatinin-Iminohydrolase, ZZ-Protein), um Faktoren zu
`
`-Glucosidase-Prozeß gewonnenen Ergebnisse wurden mit zwei weiteren
`
`evaluieren, die verschiedene rekombinante Prozesse voneinander unterscheiden.
`
`Aus grundlagenorientierter Sicht hat die Arbeit folgende wichtige Nachweise geliefert:
`
`1) Nach Induktion rekombinanter a
`zellulärer Prozesse. Als Folge davon kommt es zu einer Deregulation der Plasmidreplikation, mit
`
`-Glucosidase kommt es zu einer Inhibition verschiedener
`
`der Folge einer 3-6-fachen Plasmidamplifikation. Das Phänomen der Plasmidamplifikation nach
`Induktion ist nicht auf den a
`bei denen die Produktbildung mit einer starken Inhibition des Wachstums einhergeht.
`
`-Glucosidase-Prozeß beschränkt, sondern tritt in allen Systemen auf,
`
`2) Die Entstehung nicht-teilungsfähiger Zellen nach Induktion der a -Glucosidase ist eine Folge der
`Last der Produktsynthese auf den Syntheseapparat der Zelle. Es konnte mittels
`
`Elektronenmikroskopie gezeigt werden, daß die nicht-teilungsfähigen Zellen durch eine
`
`Ausdehnung des Chromosoms gekennzeichnet sind. Diese Zellen stellen eine metabolisch
`
`absterbende Population dar, die jedoch über einen längeren Zeitraum nicht lysiert und in der über
`
`längere Zeit noch bestimmte metabolische Aktivitäten nachgewiesen werden können. In diesem
`
`Zusammenhang wird diese Population als viable but non-culturable (VBNC-Status) diskutiert.
`
`starke
`
`Induktion
`
`rekombinanter Proteine
`
`3) Die
`Glucoseaufnahmekapazität der Zellen. Diese Eigenschaft kann in Abhängigkeit von den
`
`führt
`
`zu
`
`einer
`
`Inhibition
`
`der
`
`Prozeßbedingungen problematisch in industriellen Prozessen sein, da die im Wachstumsmedium
`
`BEQ 1021
`Page 4
`
`

`
`akkumulierende nichtmetabolisierte Glucose das Überwachsen der Kultur durch plasmidfreie
`
`nichtproduzierende Zellen begünstigt. Im Rahmen der Arbeit wurde eine on-line nutzbare
`
`Methode zur schnellen Bestimmung der Glucoseaufnahmekapazität entwickelt, auf deren Basis
`
`eine optimale Regelung des Glucose-Feedings möglich ist.
`
`4) Im Rahmen der Arbeit wurden die zellulären Alarmone ppGpp, s S und cAMP in Abhängigkeit
`von der Stärke der Glucoselimitation und der Wachstumsrate gemessen. Es konnte gezeigt
`werden, daß nach Induktion der a
`-Glucosidase im Fed-Batch-Prozeß die Konzentrationen der
`Regulatoren der zellulären Adaptationssysteme an Glucoselimitation (ppGpp und s S) reduziert
`sind, im Vergleich zu Kulturen ohne Induktion. Nach Überexpression der a
`-Glucosidase kommt
`zu einem Abfall der zellulären Konzentrationen von ppGpp, s S und der extrazellulären cAMP-
`Konzentration. Untersuchungen, die in diesem Zusammenhang mit verschiedenen Mutanten
`
`durchgeführt wurden, lassen vermuten, daß das Absterben der Zellen nach Induktion mit der
`
`fehlenden Adaptation an die Streßbedingungen in Zusammenhang steht.
`
`5) Im Unterschied zum a
`Produktion von Creatinin-Iminohydrolase (CRIMI) keine Wachstumsinhibition und keine
`
`-Glucosidase-Prozeß wurde im zweiten untersuchten System zur
`
`Plasmidamplifikation nach Induktion beobachtet, obwohl das Produkt in höherer Konzentration
`(>30% vom Gesamtzellprotein) als die a
`wurde. Weiterhin kommt es nach Überexpression von CRIMI zu einem starken Anstieg der
`zellulären s S-Konzentration und zu einer kontinuierlichen Akkumulation von cAMP im
`Kulturmedium. Obwohl im Rahmen der Dissertation die molekulare Basis der unterschiedlichen
`
`-Glucosidase mit höherer spezifischer Rate gebildet
`
`Reaktion beider Systeme nicht experimentell geklärt wurde, werden im Diskussionsteil der Arbeit
`
`Hypothesen im Zusammenhang mit der Konkurrenzsituation auf Ebene von Transkription und
`
`Translation zwischen Produktsynthese einerseits und zellulären Synthesen andererseits, diskutiert.
`
`6) Mittels einer Scale-down Strategie wurden Zonen mit unterschiedlicher Konzentration von
`Glucose imitiert, die beim Up-scaling von bakteriellen Fermentationsprozessen entstehen und
`
`untersucht, in wieweit sich Oszillationen der Kohlenstoffquelle auf die mikrobielle Physiologie und
`
`die Produktbildung auswirken. Die Ergebnisse belegen, daß oszillierender Glucosehunger die
`
`Produktbildung in rekombinanten Prozessen beeinflußt. Dies betrifft sowohl die Produktausbeute,
`
`als auch die Physiologie der kultivierten Zellen. Ein unerwartetes Ergebnis der Untersuchung war,
`
`daß sich oszillierender Glucosehunger wahrscheinlich vorrangig positiv auf den Prozeßverlauf
`
`auswirkt. Möglicherweise stellen regelmäßige Oszillationen ein schwaches Streßsignal dar, das
`
`BEQ 1021
`Page 5
`
`

`
`eine entsprechende Adaptation der Zellen auslöst und sie resistenter macht gegen den starken
`
`Streß, den die Produktion des rekombinanten Produktes darstellt.
`
`7) Im Rahmen dieser Arbeit wurden Prozeßfaktoren evaluiert, die die Maßstabsvergrößerung
`rekombinanter biotechnologischer Prozesse beeinflussen. Diese Studien wurden im Rahmen des
`
`Europrojektes “Bioprocess Scale-up Strategy — based on Intergration of Microbial
`
`Physiology and Fluid Dynamics” durchgeführt, das neben der biologischen Charakterisierung
`
`auch die Entwicklung entsprechender Simulationsprogramme auf der Grundlage von
`
`Kompartimenten und Fluid Dynamics in oszillierenden Umgebungen, sowie die Large-Scale-
`
`Verifizierung im nicht rekombinanten Wildtyp E. coli W3110 und in einem rekombinanten Prozeß
`
`(ZZT2-Protein) beinhaltet. Es konnte durch umfassende Analysen gezeigt werden, daß die
`
`Vermischung im Großreaktor, insbesondere auftretende Glucose- und Sauerstoff-Gradienten die
`
`Zellphysiologie und Produktbildung beeinflussen.
`
`BEQ 1021
`Page 6
`
`

`
`CONTENTS
`
`1
`
`INTRODUCTION .............................................................................................................. 1
`
`1.1
`
`1.2
`
`1.3
`
`1.4
`
`Principle aspects of recombinant gene expression in E. coli......................................2
`
`Fed-batch as the cultivation strategy...........................................................................4
`
`The scale of production.................................................................................................6
`
`Objectives......................................................................................................................9
`
`2 MATERIALS AND METHODS.....................................................................................11
`
`2.1
`2.1.1
`2.1.2
`
`Strains and Plasmids ...................................................................................................11
`Strains.....................................................................................................................11
`Plasmids..................................................................................................................11
`
`2.2
`2.2.1
`2.2.2
`2.2.3
`2.2.4
`
`Cultivation media and conditions................................................................................12
`Cultivation medium..................................................................................................12
`Shake flask cultivation.............................................................................................13
`Laboratory scale cultivation.....................................................................................13
`Industrial scale cultivation.........................................................................................14
`
`Analytical methods ......................................................................................................15
`2.3
`Cell concentration....................................................................................................15
`2.3.1
`Analysis of medium compounds...............................................................................16
`2.3.2
`2.3.2.1 Glucose concentration..........................................................................................16
`2.3.2.2 Acetate concentration...........................................................................................16
`2.3.2.3 Ammonia concentration........................................................................................17
`2.3.3
`Enzyme assays ........................................................................................................17
`2.3.3.1
`a-glucosidase activity ...........................................................................................17
`2.3.3.2
`Creatinine imino hydrolase activity........................................................................17
`2.3.4
`Preparation and quantification of DNA and RNA.....................................................18
`2.3.4.1 DNA agarose gel electrophoresis .........................................................................18
`2.3.4.2
`Plasmid purification and quantification...................................................................19
`2.3.4.3
`Cell transformation...............................................................................................19
`2.3.4.4 mRNA analysis ....................................................................................................20
`2.3.5
`Protein preparation and quantification.......................................................................21
`2.3.5.1
`Cell disruption in a cell mill....................................................................................21
`2.3.5.2
`Preparation of inclusion bodies (IB’s) ...................................................................21
`2.3.5.3 Quantification of protein on SDS-gels ...................................................................21
`2.3.5.4
`Protein quantification according to Bradford..........................................................22
`2.3.6
`Protein analysis by immunoblot ................................................................................23
`2.3.6.1 Analysis of s S concentration..................................................................................23
`2.3.6.2 Analysis of H-NS concentration...........................................................................23
`2.3.6.3 Analysis of LexA concentration............................................................................24
`2.3.6.4 Analysis of ribosomal protein S8 concentration.....................................................24
`2.3.7
`Determination of nucleotide concentration by HPLC.................................................24
`
`BEQ 1021
`Page 7
`
`

`
`2.3.7.1 HPLC configuration.............................................................................................24
`2.3.7.2 Nucleotide (AXP) analysis ...................................................................................24
`2.3.7.3
`ppGpp analysis ....................................................................................................25
`2.3.7.4
`cAMP analysis.....................................................................................................26
`2.3.8
`Flow cytometry.......................................................................................................26
`2.3.9
`Rate determination of replication, transcription, and translation..................................26
`2.3.10
`Transmission electron microscopy of cell samples.....................................................27
`
`2.4
`2.4.1
`2.4.2
`2.4.3
`2.4.4
`
`On-line measurements and calculation......................................................................27
`On-line measurements .............................................................................................27
`Kinetic parameters ..................................................................................................28
`Glucose uptake capacity..........................................................................................28
`Respiration data ......................................................................................................28
`
`3 RESULTS........................................................................................................................30
`
`3.1
`3.1.1
`3.1.2
`
`Cell growth and segregation in recombinant bioprocesses.......................................30
`Cell growth in recombinant E. coli fed-batch cultivations..........................................30
`Cell segregation and plasmid stability after IPTG induction........................................34
`
`3.2
`3.2.1
`3.2.2
`3.2.3
`3.2.4
`3.2.5
`3.2.6
`
`Cellular responses after strong induction of recombinant aa-glucosidase ................39
`Activity of replication, transcription and translation....................................................39
`Plasmid amplification after induction.........................................................................41
`Influence of a-glucosidase production on the chromosomal DNA supercoiling...........43
`Analysis of DNA binding protein (H-NS) and LexA protein after induction...............45
`The energy situation following induction of a-glucosid ase ..........................................47
`Inhibition of glucose uptake rate after overexpression of recombinant genes ..............49
`
`3.3
`3.3.1
`3.3.2
`3.3.3
`3.3.4
`
`Stress responses after induction of recombinant gene expression...........................52
`Level of the stringent response regulator ppGpp.......................................................52
`The s S- related general stress response.....................................................................54
`Comparison of mRNA levels of genes controlled by different s factors......................58
`Level of cAMP in fed-batch fermentations of E. coli ................................................61
`
`3.4
`3.4.1
`3.4.2
`3.4.3
`
`Cell segregation and stress responses in large-scale cultivations ...........................64
`Large-scale cultivations of E. coli W3110................................................................64
`Large-scale cultivations of recombinant E. coli W3110 pRIT44T2...........................69
`Cell lysis and cAMP level in large-scale processes ...................................................72
`
`3.5
`
`3.5.1
`3.5.2
`3.5.3
`
`Influence of glucose oscillations on the aa-glucosidase process by using a scale-
`down technique ............................................................................................................75
`Effect of glucose oscillations on cell growth and a-glucosidase formation...................75
`Effect of controlled glucose oscillations on cell segregation and maintenance..............77
`Effect of controlled glucose oscillations on cell lysis and cellular responses.................81
`
`4 DISCUSSION ..................................................................................................................85
`
`4.1
`
`4.2
`
`Influence of recombinant gene overexpression on cell growth.................................85
`
`Cell segregation after induction.................................................................................89
`
`BEQ 1021
`Page 8
`
`

`
`4.2.1
`4.2.2
`
`Cell segregation into viable but non-culturable cells...................................................89
`Cell segregation into plasmid-free cells .....................................................................91
`
`4.3
`
`Plasmid content after induction..................................................................................94
`
`4.4
`4.4.1
`4.4.2
`
`Stress responses during fed-batch cultures of recombinant E. coli .........................94
`Stress responses to glucose limitation/starvation........................................................95
`Stress responses after the induction of recombinant genes.........................................98
`
`4.5
`
`4.6
`
`Cell physiology in large-scale bioprocesses ............................................................101
`
`Influence of substrate oscillations ............................................................................102
`
`5 CONCLUSION ..............................................................................................................104
`
`6 ACKNOWLEDGEMENTS ..........................................................................................105
`
`7 REFERENCES.............................................................................................................107
`
`BEQ 1021
`Page 9
`
`

`
`Abbreviations
`
`ADP
`AMP
`ATP
`BSA
`cAMP
`CRIMI
`CRP
`EC
`HPLC
`IAA
`IPTG
`kbp
`OD500
`ONPG
`NADHP
`PAGE
`PFR
`p-NPG
`ppGpp
`RBS
`RNAP
`SDS
`STR
`VBNC
`ZZ
`
`Nomenclature
`
`CCO2in
`CCO2out
`CER
`cfu
`CO2in
`CO2out
`DCW
`DOT
`Fs
`H
`kD
`m
`OUR
`P
`Q
`qCO2
`qO2
`qp
`qS
`R
`RQ
`S
`t
`T
`V
`X
`Yx/s
`
`µ
`
`3´,5´-adenosine diphosphate
`3´,5´-adenosine monophosphate
`3´,5´-adenosine triphosphate
`bovine serum albumin
`cyclic 3´,5´-adenosine monophosphate
`creatinine imino hydrolase (EC 3.5.4.21, 45 kDa)
`cAMP receptor protein, also called CAP (catabolite activator protein)
`energy charge
`high performance liquid chromatography
`b-indole acrylic acid
`isopropyl-b-D-thiogalactopyranoside
`kilo base pairs
`optical cell density at 500 nm
`o-nitrophenyl- b-D-galactopyranoside
`nicotinamide-adenine dinucleutide
`polyacrylamid gel electrophoresis
`plug flow reactor
`p-nitrophenyl-a-D-glucopyranoside
`guanosine 5´-diphosphate 3´-diphosphate
`ribosome binding site, also called Shine-Dalgarno-Sequence
`RNA polymerase
`sodium dodecyl sulfate
`stirred-tank reactor
`viable but non-culturable cell population (also as VNC)
`a modified domain B of staphylococcal protein A (17.7 kDa)
`
`carbon dioxide concentration in outlet gas
`carbon dioxide concentration in inlet gas
`volumetric carbon dioxide evaluation rate [mmol L-1h-1]
`colony forming units [mL-1]
`oxygen concentration in inlet gas (in %)
`oxygen concentration in outlet gas
`dry cell weight [g L-1]
`dissolved oxygen tension [%]
`substrate feed rate [g h -1]
`Henry constant
`specific death rate [h -1]
`maintenance coefficient [g substrate g -1 biomass h-1]
`volumetric oxygen uptake rate [mmol L-1 h-1]
`product concentration [g L-1]
`outlet gas flow rate [L h-1]
`specific carbon dioxide evaluation rate [mmol g -1 biomass h-1]
`specific oxygen uptake rate [mmol g -1h-1]
`specific product formation rate [g product g -1h-1]
`specific substrate consumption rate [g substrate g -1 biomass h-1]
`standard gas constant; = 22.4 [L mol-1]
`respiratory quotient [mol CO2 mol-1 O2]
`substrate concentration [g L-1]
`cultivation time [h]
`temperature [°C]
`culture volume [L]
`cell mass; dry cell weight [g L-1]
`yield coefficient for biomass per substrate [g biomass g -1 biomass]
`
`specific growth rate [h -1]
`
`BEQ 1021
`Page 10
`
`

`
`INTRODUCTION
`
`1 Introduction
`
`1
`
`The recent progress of genetic engineering allows the enrichment of high value therapeutics and other
`
`recombinant proteins in bacteria up to very high levels of the cell protein. However, for successful
`
`production of a protein the thoughtful integration of information from bacterial genetics, physiology,
`
`nucleic acid and protein chemistry, and biochemical engineering is required (Georgiou, 1996).
`
`An effective industrial process is characterized by high product concentrations at a high cell mass
`
`(Riesenberg & Guttke, 1999). By the common way high cell densities are obtained with a fed-batch
`
`procedure. During the feed phase one defined medium component is continuously added to the
`
`fermenter in a growth-limiting amount in order to control the growth conditions, such as overflow
`
`metabolism, accumulation of toxic compounds and oxygen availability (Yamané and Shimizu, 1984).
`
`As the growth rate in a fed-batch culture is generally lower than the maximum growth rate of the
`
`organism, a cell which is cultivated under fed-batch conditions to high cell densities has a very
`
`different physiological and metabolic status than a cell which is grown at low density in nutrient broth
`
`in shake flasks. This difference surely influences synthesis rates, protein stability, and protein folding,
`
`and therefore it can be suggested that the process also has a major influence on the down stream
`
`purification process.
`
`A further important parameter in the industrial production is the scale of process. Large conventional
`
`bioreactors are commonly inhomogenous systems with respect to nutrient concentrations, gas
`
`distribution, and pH profile, mainly due to mixing and mass transfer limitations caused by a realistic
`
`power input. Recent studies indicated that microorganisms react to gradients in the reactor by a
`
`short-term response which finally can influence the process (Larsson & Enfors, 1988; Neubauer et
`
`al., 1995a,b; Larsson et al., 1996; Bylund et al., 1998; Xu et al., 1999). However, the overall effect
`
`of such inhomogeneity on the process performance is still not well investigated.
`
`The aim of this thesis was to study extensively the effects of recombinant protein production on the
`
`host cell physiology in context to the cultivation method and the production scale for one model
`protein, a heterologous a -glucosidase from Saccharomyces cerevisiae. Thereby, we concentrated
`
`on specific parameters, such as cell growth, viability, plasmid stability, product formation, and some
`
`connected cellular responses. In the case of scale effects the study was focused on the question how
`
`repeated short term glucose starvation influences the production.
`
`BEQ 1021
`Page 11
`
`

`
`INTRODUCTION
`
`2
`
`The comprehensive study on this specific model protein, a -glucosidase, was for some specific
`
`questions extended to processes with two other model proteins. Although the investigations with
`both other proteins are comparable to the a -glucosidase process only in a limited number of
`
`parameters, the study with the three different proteins was important for the critical discussion of the
`
`influence which is caused by the specific product characteristics.
`
`1.1 Principle aspects of recombinant gene expression in E. coli
`The overexpression of heterologous genes is influenced by several factors like plasmid stability,
`
`plasmid copy number, strength of promoter, stability of mRNA, availability of ribosomes,
`
`transcription and translation efficiency, post-translational modification, the stability and solubility of
`
`the recombinant protein itself, as well as host cell and culture conditions (Sawers & Jarsch, 1996).
`
`Recombinant processes aiming for a high amount of heterologous protein are often based on the use
`
`of strong expression systems which are regulated at the level of transcription (Swartz, 1996; Vicente
`et al., 1999). Therefore, strong inducible promoter are used, such as Plac, l PL, and l PR, or the
`promoter of the T7 RNA polymerase (Remaut et al.,1981; DeBoer et al., 1983; Studier and
`
`Moffatt, 1986). Such systems are commonly used for transient production of the recombinant
`
`protein, which is induced after a growth phase during which product formation is low. In many cases,
`
`after performing the inducing signal the specific production rate increases to a maximum only within a
`
`short time and product synthesis continues for one to four hours. Although in most cases it is
`
`sufficient to increase the product to a high part of the cell protein, mistranslation, aborted translation,
`
`product modification, product aggregation and degradation are consequences, which could be
`
`suggested to negatively influence the down-stream purification process. Whereas optimization is
`
`mostly performed by random screening procedures, a more comprehensive knowledge about the
`
`cellular processes and regulations in inducible recombinant systems is necessary for a knowledge
`
`based optimization.
`
`Several cellular processes have been investigated in different expression systems in connection to the
`
`question how they are influenced following induction. So high synthesis of heterologous proteins often
`
`effects the central carbon metabolism, which sometimes results in an elevated accumulation of
`
`acetate (Shimizu et al., 1988; Seeger et al., 1995). Also the respiratory activity has been described
`
`to increase after IPTG addition (Bhattacharya & Dubey, 1997), however, the interconnection
`
`between the change of the carbon metabolism and respiration has not been analyzed in detail yet.
`
`BEQ 1021
`Page 12
`
`

`
`INTRODUCTION
`
`3
`
`Although one should assume a drastic change of the protein synthesis pattern after induction when
`
`the synthesis of the recombinant product occupies most of the total protein generating system, only a
`
`few articles were looking on this fact (Bailey, 1993) and a comprehensive analysis is yet missing.
`
`However, it is obvious that transcription as well as translation of the product compete with the
`
`synthesis of house-keeping proteins and decrease their synthesis within minutes after induction (Vind
`
`et al., 1993; Rinas, 1996; Dong et al., 1995). Interestingly, all three groups, although using different
`
`systems and procedures for production of their recombinant product, found a strong reduction of the
`
`synthesis rate and the concentration of ribosomal proteins.
`
`Aside from this reduction of house-keeping proteins, recombinant protein production often also
`
`causes a heat shock like response which is possibly triggered by incorrectly folded intermediates of
`
`the product (Goff & Goldberg, 1985, 1987; Kosinski & Bailey, 1991; Kosinski et al., 1992b).
`
`Possibly the appearance of incorrectly folded intermediates is the cause that