Appl Microbiol Biotechnol (2005) 68: 425435
DOI 10.1007/s00253-005-0003-0
MINI-REVIEW
F. R. Schmidt
Optimization and scale up of industrial fermentation processes
Received: 8 March 2005 / Revised: 12 April 2005 / Accepted: 15 April 2005 / Published online: 7 July 2005
# Springer-Verlag 2005
Abstract To increase product yields and to ensure con-
sistent product quality, key issues of industrial fermen-
tations, process optimization and scale up are aimed at
maintaining optimum and homogenous reaction conditions
minimizing microbial stress exposure and enhancing met-
abolic accuracy. For each individual product, process and
facility, suitable strategies have to be elaborated by a com-
prehensive and detailed process characterization, identifi-
cation of the most relevant process parameters influencing
product yield and quality and their establishment as scale-
up parameters to be kept constant as far as possible. Phys-
ical variables, which can only be restrictedly kept constant
as single parameters, may be combined with other pertinent
parameters to appropriate mathematical groups or dimen-
sionless terms. Process characterization is preferably based
on real-time or near real-time data collected by in situ and
on-line measurements and may be facilitated by supportive
approaches and tools like neural network based chemo-
metric data analysis and modelling, clarification of the
mixing and stream conditions through computational fluid
dynamics and scale-down simulations. However, as fer-
mentation facilities usually are not strictly designed ac-
cording to scale-up criteria and the process conditions in
the culture vessels thus may differ significantly and since
any strategy and model can only insufficiently consider and
reflect the highly complex interdependence and mutual
interaction of fermentation parameters, successful scale up
in most cases is not the result of a conclusive and straight-
lined experimental strategy, but rather will be the outcome
of a separate process development and optimization on
each scale. This article gives an overview on the problems
typically coming along with fermentation process optimi-
zation and scale up, and presents currently applied scale-
F. R. Schmidt (*)
Sanofi-Aventis Deutschland, Biocenter H 780,
Industriepark Hchst,
65926 Frankfurt, Germany
e-mail: Frank-Rainer.Schmidt@sanofi-aventis.com
Tel.: +49-69-3056757
Fax: +49-69-30516371
up strategies while considering future technologies, with
emphasis on Escherichia coli as one of the most com-
monly fermented organisms.
Introduction
The scale up of fermentation processes as a central prob-
lem in biotechnology was first recognized and described
during industrial penicillin production at the beginning
of the 1940s. At present, several aspects of this subject are
supported as joint research projects by European and na-
tional research foundations.
Fermentation scale up is aimed at the manufacture of
larger product quantities, if at all possible, with a simul-
taneous increase or at least consistency of specific yields
and product quality. As will be illustrated in the next sec-
tions, the changed geometric and physical conditions in
larger scales, however, lead to a less favourable mixing
behaviour and to impaired physiological reaction condi-
tions which in turn may lead to a decreased process con-
stance and reproducibility, to reduced specific yields and to
an increase in unwanted side products and thus ultimate-
ly to a diminished batch-to-batch consistency and product
quality, key issues in industrial production processes.
Scale up related problems
Reduced mixing quality and enhanced stress exposure
The problem of reduced mixing quality in larger scales is
aggravated with increasing vessel sizes: the opposite sub-
strate and oxygen gradients along the vessel height which
are formed as a result of the conventional fermenter design
according to which substrate feed usually occurs from the
top and aeration from the bottom, are more pronounced in
larger reactors due to (1) longer distances to be covered
leading to larger substrate and oxygen depletion zones, (2)
larger volumes of culture broth to be stirred and therewith
longer mixing times and (3) stronger hydraulic pressure
426
gradients influencing the oxygen transfer rate (OTR; Eq. 14,
15, Table 1) (discussed in more detail by Larsson et al.
1996; Bylund et al. 1998, 1999, 2000; Enfors et al. 2001).
Cells at the fermenter top are exposed to excess glucose
concentrations and simultaneously suffer from oxygen limi-
tations, whereas those at the bottom are exposed to glucose
starvation. Excess glucose concentrations (threshold value
for Escherichia coli at 30 mg/l) result in acetate overpro-
duction (overflow metabolism), a simultaneous oxygen limi-
tation further induces the formation of ethanol, hydrogen,
formiate, lactate and succinate (mixed acid fermentation,
Bylund et al. 2000; Xu et al. 1999; Castan and Enfors 2001).
The produced acids can become reassimilated in oxygen-
rich zones, but in any case first lead to a temporary acid-
ification of the microenvironment, later eventually also of
larger regions. Combined with a decreased transportation
and elimination of carbon dioxide, detrimental metabolites
and surplus heat generated by agitation and metabolic
processes, resulting in zonal overheating, the lower mixing
rates in large scales thus lead to the formation of zones with
enhanced stress conditions. The subsequent activation of
stress genes (a survey on bacterial stress proteins is given by
Kwint et al. 2003) only partially protects the cell against
detrimental stress effects. For instance, despite activation of
heat stress genes like E. coli dnaK and clpB, reducing
misfolding and aggregation of heat-sensitive proteins (Mogk
et al. 1999), metabolic changes and damages such as trans-
location of proteins and membranes have been observed
upon heat exposure (Umakoshi et al. 1998). The evidently
unavoidable detrimental effects of the repeated and cyclic
passing of different stress zones and the subsequent con-
tinuous activation and shut down of the corresponding
stress genes are believed to lead to a completely altered
physiology (Schweder et al. 1999; Enfors et al. 2001) with
constant metabolic shifts, which ultimately reduce growth
and productivity and increase by-product formation.
Metabolic shifts
One of the most important consequences of stress-induced
metabolic shifts affecting product quality is the misincor-
poration of amino acids into native and recombinant pro-
teins, the reasons for which are complex and not yet fully
clarified: according to current knowledge, stress-triggered
changes in glycolysis, citrate cycle and of pathways in-
volved in anaerobic and mixed acid fermentation lead to
shifts in the coupled amino acid biosynthetic pathways
(Fenton et al. 1997). Even though translation processes
are considered to be generally performed quite precisely
(Weickert and Apostol 1998), the resulting shifts in the
amino acid pool in turn lead, due to the inherent inaccu-
racy of the translation machinery (unspecificity of amino-
acyl-tRNA synthases, wobbling of tRNA), to amino acid
misincorporations in freshly synthesized proteins, partic-
ularly in phases of high protein production rates which are
characterized by amino acid shortages. Muramatsu et al.
(2002) report the incorporation of -methylnorleucin in-
stead of isoleucin into recombinant E. coli hirudin in po-
sitions 29 and 59. A misincorporation of norvalin instead
of leucin and methionine into human recombinant haemo-
globin in E.coli has been published by Apostol et al. (1997)
and Kiick et al. (2001). Amino acid misincorporations are
thus a cause for an increase in by-products at the expense
of the desired main product yields in larger scales.
Plasmid stability
An essential prerequisite for high product yields particu-
larly in larger scales, in which cultures pass a higher num-
ber of generations due to larger culture broth volumes and
longer inoculation chains from the cell bank to the produc-
tion stage, is the stable propagation of plasmids to daugh-
ter cells. Plasmid stability is influenced by the plasmid
properties, including size and nucleotide sequence, by the
genetic background of the host (details discussed by Friehs
2004; Thiry and Cingolani 2002) as well as by process
parameters like temperature, growth rates and substrate
concentrations. Lin and Neubauer (2000) show that rapid
glucose oscillations favour plasmid stability and recom-
binant protein production rate, whereas high glucose con-
centrations diminish plasmid stability (Neubauer et al.
2003). Plasmid stability and plasmid numbers are thus neg-
atively influenced, more difficult to control and less easy to
be maintained in larger scales. A concept to render plasmid
stability and expression rates more independent from such
physiological influences is the application of runaway plas-
mids, the replication of which can be induced separately
from growth in the desired fermentation phase (e.g. Ansorge
and Kula 2000; Nordstrom and Uhlin 1992), enabling copy
numbers of up to 1,000 in the expression phase and, due
to the separation of their replication and expression from
the growth phase, simultaneously a more precise replica-
tion and amino acid incorporation. Further possible mea-
sures to enhance the accuracy of metabolic processes are (1)
a reduction the generation time and generation numbers
throughout the process, (2) a deceleration of the metabolic
speed, e.g. through reduction of temperature, and (3) the
development more stress-resistant microbial strains.
A prioritized goal of process optimization and scale up
thus consists of an appropriate process design which im-
proves the physiological conditions and the metabolic ac-
curacy by minimizing microbial stress exposure.
Process characterization
To identify process-specific stress factors and to understand
the physiological responses to the vessel specific physical
conditions, the mutual influences and interactions of the
various physical and physiological parameters have to be
analysed in detail. An overview on analytical methods
currently applied and yet in development is given in the
following subsections.

Analysis of physiological parameters
Among the physiologically most relevant parameters are,
besides the pH as discussed, biomass, cell viability, the
concentrations of substrates, metabolites and products, the
partial oxygen and carbon dioxide pressures in culture broth
and the composition of the exhaust gas, giving information
about respiratory states as indicated, e.g. by the respiratory
quotient (RQ; Eq. 18, Table 1). Established analytical
methods and tools are colorimetric procedures, chromatog-
raphy (van de Merbel 1997), mass spectroscopy (Buchholz
et al. 2001; Matz and Lennemann 1996), enzymatic and
electrophoretic methods (Klyushnichenko 2004; Nishi et al.
1989, 1990), hybridizing techniques, biochips (Enfors et al.
2001; Schweder et al. 1999; Gabig-Ciminska et al. 2004)
or flow cytometry. Multiparameter flow cytometry permits,
in addition to the analysis of usual metabolites, also the
analysis of the cellular DNA, RNA and protein content, cell
viability, membrane potential, intracellular pH, cell size and
cell development stages (Hewitt and Nebe-von Caron 2001,
2004; Hewitt et al. 2000; Kell et al. 1991). In general, the
data are, if possible, preferably captured as real-time values
by in situ on-line measurements as occurring for physical
parameters (Harms et al. 2002) to avoid falsified or gappy
data due to time differences between sampling and analysis.
Ex situ methods can be designed as real-time procedures
in combination with robust automated sampling and sam-
ple preparation methods like ultrafiltration (Schaefer et al.
1999; Matz and Lennemann1996; van der Merbel 1997)
and flow injection analysis (FIA; Rocha and Ferreira 2002;
Tothill et al. 1997; Schgerl et al. 1991; van de Merbel et
al. 1993; Huang et al. 1995; further discussion and refer-
ences given by Vaidyanathan et al. 1999). An on-line FIA
glucose analysis method of the native culture broth with-
out prior sampling and filtration has been presented by
Arndt and Hitzmann (2004) and Kleist et al. (2003). For
their robustness, in situ measurements of biological param-
eters are preferably performed by optical probes (optodes),
working either on a physical base, e.g. by refractional
or spectroscopic measurements (for further physical sen-
sor principles, see Turner 1994), chemically as by ion ex-
change reactions (Spichiger-Keller 1997) or enzymatically
by gel-embedded biocatalysts, the activity of which can
be measured by means of pH changes or by fluorophores
(chemo- and biosensors; Wolfbeis 2002; Griffiths and Hall
1993). The application of the hitherto employed enzymes
has been restricted mostly to the analysis of sugar com-
pounds (Vaidyanathan et al. 1999), whereby the most fre-
quent applications are reported for glucose (Tothill et al.
1997). However, despite intensive research, only a few con-
cepts practically applicable for industrial purposes could
have been developed. Broader applicable are physical op-
todes enabling the simultaneous measurement and quanti-
fication of several parameters and metabolites through
continuous scans or excitations and subsequent absorption
and scattering measurements within defined wavelength
ranges like (1) 2D fluorescence spectroscopy (Boehl et al.
2003; Solle et al. 2003; Stark et al. 2002), as shown for the
determination of NAD(P)H concentrations through exci-
427
tation at 350 nm and the fluorescence measurements at
450 nm (more details and literature in Vaidyanathan 1999);
(2) near-infrared spectroscopy (NIR; Hall et al. 1996;
Vaccari et al. 1994; Macaloney et al. 1994), which has been
employed as a 700- to 2,500-nm scan in the antibiotic
fermentation at Pfizer (Hammond 1992; Hammond and
Brooks 1992) for the measurement of nutrient and product
as well as side product concentrations; or (3) measure-
ments in infrared (Pollard et al. 2001), ultraviolet, or visible
wavelength spectra (Vaidyanathan et al. 1999). By means
of capacitance measuring electrodes (Aber Instruments
Ltd., UK), viable cell counts can be performed by seizing
the electric capacitance and conductivity of cells with intact
membranes in a generated electric field. To gain a picture
as comprehensive as possible by enhancing the amount of
analysable substances, multi-channel arrays with parallel
employment of various analytical procedures monitoring
several parameters simultaneously have been established
and new concepts like artificial noses (Dickinson et al.
1998) and electronic tongues (Turner et al. 2003) are cur-
rently tried to be integrated into fermentation technology,
allowing the analysis of a large number of analytes by
cross-talking semispecific sensors which act in analogy to
the human olphactory sensoric system. The realization of
this concept will be facilitated by a further miniaturization
of the sensors and chips employed (Moser et al. 2002), as
it has been achieved with biochips developed for the phar-
maceutical natural product screening, onto which the need-
ed number of reactants is fixed in smallest space or are
brought together in microchannels and the analytes are
separated in microcapillars (lab-on-a-chip technology, e.g.
Collins et al. 2004; Neils et al. 2004; Farinas et al. 2001).
For analysis, interpretation and correlation of the ob-
tained signals and data through chemometric modelling
(Boehl et al. 2003; Solle et al. 2003), neural network tools
(e.g. Unscrambler 9.0, Camo, Norway) seem to gain an
increasing important role (Ferreira et al. 2001). Despite
inherent limitations (Chen and Rollins 2000; Trelea et al.
2001), adequately trained neural networks are capable of
recognizing and revealing non-linear, highly complex and
even non-obvious, hidden relations between a multitude
of various physical, biochemical and physiological pro-
cess parameters and to design the according process mod-
els. Hooked up to expert systems, they allow immediate
and short-term reactions to process deviations by anticipat-
ing physiological drifts and their influences on product
yield and quality and to change the according set points and
process profiles. In this way, neural network tools help to
optimize process control and to facilitate the identification
of suitable strategies for process optimization scale up, as
illustrated in the next section.
Process optimization and scale up
Process control
The on-line connection of mass spectrometric, NIR, elec-
trochemical and multi-array gas sensors by the neuronal net
Table 1 Fermentation parameters, coefficients and terms, implicated in mixing, aeration, oxygen and heat transfer, suitable as scale-up variables to be kept constant alone or combined with
each other or other process relevant variables, preferably but not necessarily, as dimensionless groups as described and discussed, e.g. by Hubbard (1987) and Wang and Cooney (1997)
Parameter/coefficient
Power input (P); volumetric
power input (P/V)
Dimensionless power
number (NPo)
Impeller tip speed (vtip)
Reynolds number (Re)
Mixing time (Tm)
Dimensionless mixing time
(Tmn)
Modified dimensionless
power number
Modified dimensionless
mixing time
Aeration rate (volume per
volume per minute, vvm)
Superficial gas velocity (vs)
Gas hold up ()
Fractional gas hold up ()
Gassing number (NQg)
Oxygen transfer rate (OTR) 14
15
Volumetric oxygen mass
transfer
coefficient (kLa)
kLa Scale-dependent
according to
WangCooney
Respiratory quotient (RQ)
428
Mathematical characterisation Symbol explanation


1 P 2
nM NPo n3 dI5 kgm2 s
2 W P=power input; n=stirrer speed; M=momentum; NPo=dimensionless power
number (impeller specific); =density of the medium; n=stirrer speed;
2 NPo P=n3 dI5 dI=impeller diameter
3 vtip 2
n dI m=sec n=stirrer speed; dI=impeller diameter
4 Re ndI2 = Re=Reynolds number; n=stirrer speed; dI=impeller diameter; =density of the
medium; =dynamic viscosity
5 Tm f n;dI ;V =QV = Nfl n dI3 sec Tm=mixing time; n=stirrer speed; dI=impeller diameter; =kinematic viscosity;
V=volume; Q=volumetric flow rate; Nfl=pumping number
6 Tm nf Re; NPo Tm=mixing time; n=stirrer speed; Tmn=dimensionless mixing time; Re=Rey-
nolds number NPo=dimensionless power number
0 NPo=modified dimensionless power number Tmn=modified dimensionless
7 NPo NPo Re3 dF =dI PdF 2 =3
mixing time; NPo=dimensionless power number; Re=Reynolds number;
.h i
8 T0m n nTm dF =dI 2 Re Tm =dF 2  dF=inside vessel diameter; dI=impeller diameter; P=power input; n=stirrer
speed; tm=mixing time; =fluid density; =dynamic viscosity


9 AR FG =VR m3 =m3 min AR=aeration rate; FG=volumetric gas flow rate; VR=fermenter reaction volume
10 vs FG =A m=sec vs=superficial gas velocity; FG=volumetric gas flow rate; A=fermenter cross
section
11  VR =FG =gas hold up; VR=fermenter reaction volume; FG=volumetric gas flow rate;
12 " VG =VR =fractional gas hold up; VG=dispersed gas volume
   
13 NQg f Fr;Re;S; g =0 ; g =0 qg =ndI3 NQg=gassing number; Fr=Froude number; Re=Reynolds number; S=fluid
constant; =density; =dynamic viscosity; g=aerated; 0=not aerated; qg=gas
throughput; n=stirrer speed; dI=impeller diameter


OTR kL aCG
CL kL a LO2 pO2G
pO2L kg O2 =m3 h with OTR=oxygen transfer rate from gas to liquid phase; kL=mass transfer
CG 0; 526pi =36 T mg=1 coefficient; a=specific interfacial surface area; CG=oxygen saturation
concentration in the gas phase; CL=measured oxygen saturation concentration
in the liquid phase; LO2=oxygen solubility in the liquid phase; pO2G=partial
pressure of oxygen in the gas phase; pO2L=partial pressure of oxygen in the
liquid phase; pi=vessel back pressure [bar]; T=temperature [C]
16 kL aa0 P=VR b vcs s
1  kLa=oxygen transfer coefficient; P=power input; VR=fermenter reaction
volume; vs=superficial gas velocity; a, b and c as specific fluid constants to be
determined experimentally
 a
17 kL a k 0 Pg =VR vs b B=60;8 j=dI 0;3 s
1  k, a, b=vessel specific coefficients; Pg=power input in aerated vessel;
V=volume of culture broth; Vs=superficial gas velocity; B=number of stirrers;
j=buffle width; dI=impeller diameter
18 RQ CER=OUR pCO2E
pCO2i =pO2i
pO2E mol CO2 =mol O2  RQ=respiratory quotient; CER=carbon dioxide emission rate; OUR=oxygen
uptake rate; pCO2=partial pressure of carbon dioxide; pO2=partial pressure of
oxygen; i=air feed; E=exhaust air
 429

For instance, the kLa valueas the most frequently applied physical scale-up variablehas been combined with (1) the vessel backpressure p by maintaining the product of kLap constant
through variation of pressure, agitation and aeration for scale up of Bacillus thuringiensis fermentations (Flores et al. 1997), or (2) the aeration rate vvm under variation of power input,
Qt=total energy; Qmet=energy generated by metabolic activities; Qag=energy

approximate metabolic constant (460 kJ/mol O2); Ro=molar oxygen uptake

modul NeurOn-line capable of processing 1,800 signal

generated by agitation; Qaer=energy generated by aeration; Qevap=energy

=heat transfer coefficient; dQ/dt=heat transfer; A=surface area; dT=tem-

simultaneously for the calculation of the process state and

losses through evaporation; Qhxch=energy losses through cooling; Yho=-

adaptive control as well as for the development of appro-
priate process models and its connection to a real-time ex-
pert system based on the Gensym G2 system has been
described by Cimander et al. (2003). Kishimoto and co-
workers, who developed a model for on-line measurements
of cell, glucose, and acetate concentrations as well as of the
specific growth rate on the basis of the carbon dioxide
emission rate (CER; see Eq. 18, Table 1), present a process
control strategy for glutamic acid, -amylase, -galacto-
sidase, and vitamin B12 fermentation based on algorithms
describing and steering fuzzy states and relations (Horiuchi
and Kishimoto 1998). Kolehmainen et al. (2003) identified
five growth phases of Saccharomyces cerevisiae by neu-
rocomputation of exhaust gas data measured by ion mo-
working volume and aeration according to the WangCooney equation (Wong et al. 2002). For details and further explanations, see text

bility spectrometry. As the maximum glucose and oxygen

uptake rates are not constant but depend on growth phases
Symbol explanation

perature difference

and rates and get reduced by induction of product synthe-

sis (Neubauer et al. 2003), Lin et al. (2001) established an
rate (mol O2/s)

integrative kinetic model combining these parameters by

aid of the simulation program SCILAB (Inria, F) enabling
the determination of the maximum glucose and oxygen
uptake rates through determination of the time which passes
between glucose pulses and subsequent changes of partial
oxygen pressure (pO2) and biomass concentration. Such a
model thus allows the alignment of the glucose feed meet-
ing the maximum uptake capacity and therefore avoiding
overflow metabolism. The observed pattern of glucose os-
Qt Qmet Qag Qaer
Qevap
Qhxch kWh kJ with

cillation can even serve as a key parameter condition in later

scale-up stages (Lin and Neubauer 2000). The equilibration
of carbohydrate and oxygen supply thus constitutes a key
component of process control and optimization. Further
strategies aiming at the avoidance of oxygen limitation and
glucose overflow metabolism consist of the employment
of alternative carbon sources like glycerol (Lee 1996) or
galactose (Kim et al. 2003) and of a drastic glucose re-
striction. Wong et al. (2002) report that minimal concen-
trations of glucose and yeast extract yielded the highest
Qmet Yho Ro kWh kJ

concentrations of 0.5 and 1 g/l of a recombinant K99 an-

tigen. Yim et al. (2001) achieved constant yields of 4.4 g/l
Mathematical characterisation

of human granulocyte colony stimulating factor during

 dQ=dt=AdT

scale up from 2.5 to 30 l by limiting glucose feed and

keeping the growth rate at the minimum (=0.116 h1).
Besides the above-mentioned parameters such as the max-
imum uptake capacities, the various different physiological
or directly measurable physical process parameters like
pH, ammonium consumption, pO2, OTR or the growth pa-
rameters can be applied as reference parameter for glucose
19
20

21

feed in a closed loop design. Dantigny et al. (1991) describe

a biomass controlling feed depending on ethanol forma-
tion rate and RQ for S. cerevisiae. To achieve a more uni-
Heat transfer coefficient
Energy balance (Qges)
Parameter/coefficient

form glucose distribution in large-scale tanks, Larsson et al.

Table 1 (continued)

(1996) and Bylund et al. (1998) suggested a glucose feed

at the dynamic zones of the fermenter bottom together
with the injected air. A step ahead in this regard would be
a further enhancement of the mixing quality by a reactor
and process design permitting a multilevel injection of
both, air and substrates into high turbulence zones.
430
Process design
The realization of concepts aimed at the improvement of
mixing quality in larger scales to ensure homogenous re-
action conditions and to reduce both, the size of stress
zones and the zonal residence times, requires the char-
acterization of the large-scale hydrodynamic and reaction
conditions. As supportive approaches, the conditions gov-
erning in large vessels and their relevant zones are simu-
lated and depicted through high-performance computing
[computational fluid dynamics (CFD); e.g. Cant 2002] and
through transfer into small reactors (scale down).
Scale down approach
Scale down designs, like unstirred bypass loops to simulate
and investigate effects of oxygen limitations, excess glu-
cose concentrations or pH fluctuations, are illustrated by
Amanullah (2001); Enfors et al. (2001); Hewitt et al. (2000)
and Onyeaka et al. (2003). The characterization of the
governing stress conditions by tracking of the correspond-
ing stress gene expression (see above) can be performed
on-line by means of bioluminescence, e.g. by the green
fluorescing protein (GFP signaling) and the detection by
appropriate probes (Reischer et al. 2004). The sort and
amount of parameters which can be simulated in small
scales, however, appear to be restricted (see Bylund et al.
2000). On one side, the viability of cells is impaired in
smaller scales (Hewitt et al. 2000; Enfors et al. 2001) and
physiological parameters evidently are not straight-lined
transferable. When comparing E. coli fermentations in dif-
ferent scales with similar and identical courses of biomass,
respiration, cell lysis, mixed acid fermentation, overflow
metabolism, glucose and oxygen concentrations, Bylund
et al. (2000) found that the yields of human growth hor-
mone were 80% higher and the concentration of pro-
teolytically digested side products 30% lower in the small
scale. Evidently, knowledge about the conditions to be
transferred is not yet sufficiently comprehensive. Current
efforts thus try to complete the picture by a detailed char-
acterization of the hydrodynamic conditions, as performed
by Enfors et al. (2001) for investigation and description
of gradients and oscillations of substrate concentrations
through computational fluid dynamics (CFD).
Computational fluid dynamics
Generally, hydrodynamic model development can be per-
formed mathematically abstract as well as data driven on
the basis of experimental flow and circulation measure-
ments (Davidson et al. 2003), during which the trajectory
and velocity of tracer particles (fluorescence dyes, magnet-
ic or radio pills) is pursued and detected optically or by
means of antenna or induction coils wound around the ves-
sel (flow follower techniques). A further optical detection
procedure, the application of which however is restricted
to glass vessels, is the laser Doppler method, which de-
tects the scattered and reflected light pattern (Doppler shift)
generated by floating particles passing a dual laser wave
interference field. For the mathematical calculation and
description of turbulent flows [e.g. through direct numeri-
cal simulation (DNS)], the reaction volume is discretized
and meshed into as many zones and cells as possible and
subsequently, on the basis of given geometric and rhe-
ological data, the physical conditions governing these cells
are described by solving the corresponding NavierStokes
equations under consideration of the conservation laws of
mass, momentum and energy. The generation and distri-
bution of the resulting field variables enable the vectorial
representation of physical conditions at any point and
therewith detailed information with respect to trajectories,
mixing time distributions, mixing rates and profiles, hy-
drodynamic conditions, flow behaviour, mass transports
and shear stress (Bezzo et al. 2003) in dependence of al-
ready existing or still to be developed reactor and stirrer
design (Kelly and Humphrey 1998), properties of the cul-
ture broth like density and viscosity as well as the adjust-
ed process parameters. In this way, also scale-up influences
are predictable. Williams et al. (2002) report a standard-
ization of hydrodynamic and mixing conditions as well
as the avoidance of shearing stress and oxygen limita-
tions even in critical reactor zones for the scale up of car-
tilage cell fermentations. Broadly applied in the field of
numerical flow simulations are the flow modelling software
tools of Fluent Inc. (USA) (Fluent 4, 5, 6 and MixSim). A
simulation of the trajectories and distributions of gas bub-
bles and mass transports enables the determination of zon-
al pO2 values and oxygen transfer rates. The combination
with parameters like substrate concentrations and gradients,
the residence times in the respective zones, population
dynamics and metabolic fluxes (structured metabolic mod-
els) leads to integrative models [integrated fluid dynamics
(IFD)], which permit the prediction of physiological ef-
fects and reactions. Even though it is emphasized that the
integration of CFD and structured biokinetics at present
requires a deeper understanding of the dynamics of meta-
bolic and regulatory networks and cascades of signal trans-
duction triggered by microenvironmental fluctuations and
thus further research (Schmalzriedt et al. 2003), it will
most likely contribute at long terms to a realistic model-
ling of the interplay between physics and physiology and
thus will facilitate the identification of key parameters
influencing product yield and quality the most and there-
with of suitable scale-up parameters and strategies. Prob-
lems related to the scale up of physical parameters are
illustrated in the next section.
Physical scale-up parameters
Suitable for employment as physical scale-up parameters
are all known process parameters and coefficients exerting
known physiological effects, particularly those affecting
oxygen supply, heat transportation and mixing, such as
power input (Eq. 1; Table 1), aeration and agitation rate
(Eq. 3, 9, 10; Alves and Vasconcelos 1996), mixing time

(Eq. 5; Oniscu et al. 2001) and dimensionless mixing time
(Eq. 6), gas hold up and fractional gas hold up (Eq. 11, 12),
pO2, OTR (Eq. 14, 15; Diaz and Acevedo 1999; Russell
et al. 1995), oxygen mass transfer coefficient (kLa, Eq. 16,
17), and biocalorimetric variables like heat fluxes and heat
transfer coefficients (Eq. 19, 20, 21). Biocalorimetric mea-
surements were found to be in definite correlation with
metabolic activities and with OTR and are employable as
growth phase indicators (Trker 2004; Anderson et al.
2002); however, the calculations require a precise measure-
ment and knowledge of heat fluxes and sources, energy
inputs and of the temperature distribution in the vessel. In
most cases, however, it is, depending on the scale-up factor,
not or only restrictedly possible from the physical view-
point to keep physical parameters constant throughout the
scale up. Constance of even a single specific parameter
mostly leads to an uncontrolled and unpredictable change
of other variables into dimensions that are technically not
realizable. Classical examples are (1) the mixing time,
which inevitably increases in larger vessels due to the
larger volumes to be stirred and which cannot be un-
limitedly compensated and kept constant by increasing
stirrer speeds and energy inputs and (2) the volumetric
energy input P/V itself. A constant volumetric power input
has indeed been successfully applied as scale-up parameter
for the early industrial penicillin fermentations (1 hp per
gallon, equivalent to 1.8 kW per 1 m3) and in fermentations
with low energy inputs (see Kim et al. 2003), but is limited
in fermentations requiring high energy inputs, like recom-
binant E. coli cultures.
For this reason, mathematically driven approaches are
pursued both for process and reactor scale up by form-
ing dimensionless coefficients (e.g. by the Buckingham
Pi method as cited by Hubbard 1987), which are kept
constant by an appropriate choice and adjustment of the
relevant influencing process and fermenter parameters.
Well-established examples are, among many others, the
dimensionless power number (Eq. 2), Reynolds number
(Eq. 4), gassing number (Eq. 13) and the modified di-
mensionless power number (Eq. 7) and the modified di-
mensionless mixing time (Eq. 8). The comparison of the
latter two coefficients and of the corresponding curves
enable the identification of appropriate types of stirrers
capable of exerting the desired mixing performance at a
given stirrer speed with a minimum of energy consump-
tion and therewith of compensating the above-mentioned
limitations of volumetric energy inputs at larger scales.
(A quantum leap in stirrer technology in this regard seems
to be the Visco-Jet of Inotec, which possesses cone formed
short tubes instead of blades and is reported to exhibit best
mixing performance with a minimum of power input and
shear stress without foam generation.) The performance of
further various types and designs of stirrers is presented and
discussed by Junker et al. (1998). Stirrer performance co-
efficients and curves are thus crucial scale-up aids.
In addition, any of these parameters and coefficients can
be combined with other variables to set up and create new,
process-specific parameter correlations, coefficients, terms,
431
groups and characteristic curves (examples given, e.g. by
Wang and Cooney 1997).
This relates also to the kLa value (Eq. 16), which is
currently the mostly applied physical scale-up variable
since it includes the relevant parameters influencing oxy-
gen supply like agitation (via energy input) and aeration as
superficial gas velocity (Eq. 10) and which, as a component
of dimensionless terms, is also frequently employed for
reactor scale up (see Yawalkar 2002).
Taking into account the limitation of a kLa-oriented scale
up in the form of a limited power input, OTR and tolerable
shear stress, Flores et al. (1997) integrated the backpressure
p by maintaining the product of kLap constant through
variation of pressure, agitation and aeration for scale up of
Bacillus thuringiensis fermentations. While the specific
growth was reduced and the biomass production and spor-
ulation efficiency remained constant, fermentation time
could be shortened and toxin yields were increased. Wong
et al. (2002) successfully scaled up E. coli fermentations
from 5 to 200 l by keeping constant the product of kLa
and aeration rate (vvm) under variation of power input,
working volume and aeration according to the Wang
Cooney equation (Eq. 17; Wang and Cooney 1997), which
reflects these parameters in dependence of the fermenta-
tion scale. Diaz and Acevedo (1999) argue that the oxygen
transfer capacity, indicated by the kLa value, is not the most
process-relevant parameter, but the effective oxygen trans-
fer rate as a product of kLa and mass transfer potential, i.e.
the difference between the pO2 values in the gas and the
liquid phase and suggest an OTR based scale-up strategy
(Eq. 14, 15).
A common simple and robust method is the maintenance
of a constant pO2 in the culture broth by variation of stirrer
speed and aeration rate. To avoid shear stress and to keep
the energy input (P) at the lowest possible level, the pO2
is steered at the minimum limit. Riesenberg et al. (1990)
employ a stirrer speed and glucose feed steered pO2 of 20%
for interferon E. coli fermentations in a 30- and 450-l
scale. These examples demonstrate that, for each specific
process, appropriate variables and coefficients and there-
with scale-up strategies have to be identified individually.
Conclusion
Despite the central role of the scale-up issue in biotech-
nology and the comparably large body of literature, there
seems to be no common, generally applicable strategy es-
tablished. For each individual product, process and facility,
a suitable scale-up strategy has to be elaborated.
A wholistic scale-up strategy consists of a comprehen-
sive and detailed process characterization to identify key
stress factors and key parameters influencing product yield
and quality the most, and of an appropriate process control
and process design ensuring optimum mixing and reaction
conditions, supported by appropriate knowledge and data
driven models (Berkholz and Guthke 2001; Berkholz et al.
2000) as well as computational tools (Fig. 1).
432

Flow cytometry Analysis

Hybridization
Electrophoresis
Spectroscopy
Chromatography Influence Integrated Fluid Dynamics
On-line / in-situ
measurements
Electric tongues Physiological Physical
Artificial noses parameters parameters
Lab-on-a-chip
technology

Process characterisation

Scale down
Chemometric modelling Model development Computational
Neural networks
Process optimisation Fluid Dynamics

Identification of
Scale up parameters

SCALE UP

Fig. 1 Scheme roughly illustrating the theoretical steps and stages of
fermentation process optimization and scale up, characterised by
gliding transitions. Process development usually starts with a detailed
process characterisation through elucidation of the mutual influences
of physiological parameters (e.g. productivity, metabolite and sub-
strate concentrations, cell mass and viability, respiration) and of phys-
ical process parameters (e.g. aeration, agitation, pressure, temperature,
fluid dynamic conditions), eventually supported by neural networks
and integrated fluid dynamics. Employable analytical tools for phys-
iological parameter characterisation are the established physical and
chemical procedures like flow cytometry, spectroscopy, chromatog-
raphy, etc. as well as techniques yet in development for biotechno-
logical application like electric tongues, artificial noses or the lab-on-a-
chip technology. Process characterisation as well as the subsequent
process optimisation, which aim at the improvement physiological
reaction conditions and reduction of stress exposure and which enable
the identification of suitable scale-up parameters and strategies, are
usually performed on the basis of experimentally developed data
and knowledge driven assumptions and models. Supportive approaches
and tools for model development are scale down experiments, (neu-
ral network based) chemometric modelling and computational fluid
dynamics. Examples of established physical scale-up parameters are
given in Table 1

It should be kept in mind, however, that any approach References

and any model will always be approximative and a com-
promise as the process-related knowledge will always be
gappy and the known mathematical methods and relations
cannot completely reflect the highly complex interactions
and relationships of the physical conditions governing the
fermentation process (Liden 2002).
As a matter of fact, it appears that, in view of the high
complexity of the fermentation parameters influencing
each other and the only rudimentary and fragmentary re-
flection of the reality any model can deliver and finally the
different layouts of vessels and facilities which are rarely
designed according to strict scale-up criteria, successful
scale up in most cases will not be the result of a conclusive
and straight-lined experimental strategy, but rather is the
outcome of an independent optimization on each scale
which highly depends on the experience, skill and last, but
not least, the intuition of the experimentalist.
Alves SS, Vasconcelos JMT (1996) Optimisation of agitation and
aeration in fermenters. Bioprocess Biosyst Eng 14:119123
Apostol I, Levine J, Lippincott J, Leach J, Hess E, Glascock CB,
Weickert MJ, Blackmore R (1997) Incorporation of norvaline at
leucine positions in recombinant human hemoglobin expressed
in Escherichia coli. J Biol Chem 272:2898028988
Amanullah A, McFarlane CM, Emery AM, Nienow AW (2001)
Scale-down model to simulate spatial pH variations in large-
scale bioreactors. Biotechnol Bioeng 73:390399
Anderson RKI, Jayaraman K, Voisard D, Marison IW, von Stockar U
(2002) Heat flux as an on-line indicator of metabolic activity
in pilot scale bioreactor during the production of Bacillus
thuringiensis var. galleriae-based pesticides. Thermochim Acta
386:127138
Ansorge MB, Kula MR (2000) Production of recombinant L-leucine
dehydrogenase from Bacillus cereus in pilot scale using the
runaway replication system E. coli (pIET98). Biotechnol Bioeng
68:557562
Arndt M, Hitzmann B (2004) Kalman filter based glucose control at
small set points during fed-batch cultivation of Saccharomyces
cerevisiae. Biotechnol Prog 1:377383
Berkholz R, Guthke R (2001) Model based sequential experimental
design for bioprocess optimisationan overview. Focus Bio-
technol 4:129141

Berkholz R, Rohlig D, Guthke R (2000) Data and knowledge based
experimental design for fermentation process optimization. En-
zyme Microb Technol 27:784788
Bezzo F, Macchietto S, Pantelides CC (2003) General hybrid
multizonal/CFD approach for bioreactor modeling. AIChE J 49:
21332148
Boehl D, Solle D, Hitzmann B, Scheper T (2003) Chemometric
modelling with two-dimensional fluorescence data for Clavi-
ceps purpurea bioprocess characterization. J Biotechnol 105:
179188
Buchholz A, Takors R, Wandrey C (2001) Quantification of
intracellular metabolites in Escherichia coli K12 using liquid
chromatographic electrospray ionization tandem mass spectro-
metric techniques. Anal Biochem 295:129137
Bylund F, Collet E, Enfors EO, Larsson G (1998) Substrate gradient
formation in the large scale lowers cell yield and increases by-
product formation. Bioprocess Eng 18:171180
Bylund F, Guillard F, Enfors SO, Trgardh C, Larsson G (1999)
Scale down of recombinant production: a comparative study of
scaling performance. Bioprocess Eng 20:289377
Bylund F, Castan A, Mikkola R, Veide A, Larsson G (2000) In-
fluence of scale-up on the quality of recombinant human growth
hormone. Biotechnol Bioeng 69:119128
Cant S (2002) High-performance computing in computational fluid
dynamics: progress and challenges. Philos Trans Math Phys Eng
Sci 360:12111225
Castan A, Enfors SO (2001) Formate accumulation due to DNA
release in aerobic cultivations of Escherichia coli. Biotechnol
Bioeng 77:324328
Chen VCP, Rollins DK (2000) Issues regarding artificial neural
network modeling for reactors and fermenters. Bioprocess Bio-
syst Eng 22:8593
Cimander C, Bachinger T, Mandenius CF (2003) Integration of
distributed multi-analyzer monitoring and control in bioprocess-
ing based on a real-time expert system. J Biotechnol 103:237
248
Collins GE, Wu P, Lu Q, Ramsey JD, Bromund RH (2004) Compact,
high voltage power supply for the lab-on-chip. Lab Chip 4:408
411
Dantigny P, Ninow JL, Lakrori M (1991) A new control strategy for
yeast production based on the L/A approach. Appl Microbiol
Biotechnol 36:352357
Davidson KM, Sushil S, Eggleton CD, Marten MR (2003) Using
computational fluid dynamics software to estimate circulation
time distributions in bioreactors. Biotechnol Progress 19:1480
1486
Diaz A, Acevedo F (1999) Scale-up strategy for bioreactors with
Newtonian and non-Newtonian broths. Bioprocess Eng 21:21
23
Dickinson TA, White J, Kauer JS, Walt DR (1998) Current trends in
artificial nose technology. TIBTECH 16:250258
Enfors SO, Jahic M, Rozkov A, Xu B, Hecker M, Jrgen B, Krger
B, Schweder T, Hamer G, OBeirne D, Noisommit-Rizzi N,
Reuss M, Boone L, Hewitt C, McFarlane C, Nienow A, Kovacs
T, Trgardh C, Fuchs L, Revstedt J, Friberg PC, Hjertager B,
Blomsten G, Skogman H, Hjort S, Hoeks F, Lin HJ, Neubauer P,
van der Lans R, Luyben K, Vrabel P, Manelius A (2001) Phys-
iological responses to mixing in large scale bioreactors. J Bio-
technol 85:175185
Farinas J, Chow AW, Wada HG (2001) A microfluidic device for
measuring cellular membrane potential. Anal Biochem 295:
138142
Fenton D, Lai PH, Lu H, Mann M, Tsai L (1997) Control of norleucine
incorporation into recombinant proteins. US patent 5599690
Ferreira LS, de Souza MB, Folly ROM (2001) Development of an
alcohol fermentation control system based on biosensor mea-
surements interpreted by neural networks. Sens Actuators B
Chem 75:166171
Flores ER, Perez F, de la Torre M (1997) Scale-up of Bacillus
thuringiensis fermentation based on oxygen transfer. J Ferm
Bioeng 83:561564
433
Friehs K (2004) Plasmid copy number and plasmid stability. Adv
Biochem Eng Biotechnol 86:4782
Gabig-Ciminska M, Holmgren A, Andresen H, Bundvik Barken K,
Wumpelmann M, Albers J, Hintsche R, Breitenstein A,
Neubauer P, Los M, Czyz A, Wegrzyn G, Silfversparre G,
Juregn B, Schweder T, Enfors SO (2004) Electric chips for rapid
detection and quantification of nucleic acids. Biosens Bioelec-
tron 15:537546
Griffths D, Hall G (1993) Biosensorswhat real progress is being
made. TibTech 11:122130
Hall JW, McNeil B, Rollins MJ, Draper I, Thompson BG, Macaloney
G (1996) Near-infrared spectroscopic determination of acetate,
ammonium, biomass and glycerol in an industrial Escherichia
coli fermentation. Appl Spectrosc 50:102108
Hammond SV (1992) NIR analysis of antibiotic fermentation. In:
Murray I, Cowe IA (eds) Making light work: advances in near
infrared spectroscopy, vol 1. VCH, New York, pp 584589
Hammond SV, Brookes IK (1992) Near infrared spectroscopya
powerful technique for at-line and on-line analysis of fermenta-
tions. In: Ladisch MR, Bose A (eds) Harnessing biotechnology
for the 21st century: proceedings of the 9th international
symposium and exhibition. Am Chem Soc, Washington DC, pp
325333
Harms P, Kostov Y, Rao G (2002) Bioprocess monitoring. Curr Opin
Biotechnol 13:124127
Hewitt CJ, Nebe-von Caron G (2001) An industrial application of
multiparameter flow cytometry: assessment of cell physiolog-
ical state and its application to the study of microbial fermen-
tations. Cytometry 44:179187
Hewitt CJ, Nebe-Von Caron G (2004) The application of multi-
parameter flow cytometry to monitor individual microbial cell
physiological state. Adv Biochem Eng Biotechnol 89:197223
Hewitt CJ, Nebe-von Caron G, Axelsson B, McFarlane CM, Nienow
AW (2000) Studies related to the scale-up of high-cell-density E.
coli fed-batch fermentations using multiparameter flow cytom-
etry: effect of a changing microenvironment with respect to
glucose and dissolved oxygen concentration. Biotechnol Bioeng
70:381390
Horiuchi JI, Kishimoto M (1998) Fuzzy-aided estimation of biolog-
ical parameters based on material balances. J Ferment Bacteriol
86:111117
Huang YL, Foellmer TJ, Ang KC, Khoo SB, Yap MGS (1995)
Characterization and application of an on-line flow injection
analysis/wall-jet electrode system for glucose monitoring during
fermentation. Anal Chim Acta 317:223232
Hubbard DW (1987) Scaleup strategies for bioreactors containing
non-newtonian broths. Ann N Y Acad Sci 506:600607
Junker BH, Stanik M, Barna C, Salmon P, Buckland BC (1998)
Influence of impeller type on mass transfer in fermentation
vessels. Bioprocess Biosyst Eng 19:403413
Kell DB, Ryder HM, Kaprelyants AS, Westerhoff HV (1991)
Quantifying heterogeneity-flow cytometry of bacterial cultures.
Anton van Leeuwenhoek Int J Gen Mol Microbiol 60:145158
Kelly WJ, Humphrey AE (1998) Computational fluid dynamics
model for predicting flow of viscous fluids in large fermentor
with axial flow impellers and internal cooling coils. Biotechnol
Prog 14:248258
Kiick KL, Weberskirch R, Tirell DA (2001) Identification of an
expanded set of translationally active methionine analogues in
E. coli. FEBS Lett 502:2530
Kim CH, Rao KJ, Youn DJ, Rhee SK (2003) Scale-up of
recombinant hirudin production from Saccharomyces cerevisi-
ae. Biotechnol Bioprocess Eng 8:303305
Kleist S, Miksch G, Hitzmann B, Arndt M, Friehs K, Flaschel E
(2003) Optimization of the extracellular production of bacterial
phytase with Escherichia coli by using different fed-batch fer-
mentation strategies. Appl Microbiol Biotechnol 61:456462
Klyushnichenko V (2004) Capillary electrophoresis in the analysis
and monitoring of biotechnological processes. Methods Mol
Biol 276:77120
434
Kolehmainen M, Rnkk P, Raatikainen O (2003) Monitoring of
yeast fermentation by ion mobility spectrometry measurement
and data visualization with self-organizing maps. Anal Chim
Acta 484:93100
Kwint K, Nachin L, Diez A, Nystrm T (2003) The bacterial
universal stress protein: function and regulation. Curr Opin
Microbiol 6:140145
Larsson G, Trnkvist M, Wernersson ES, Trgard C, Noorman H,
Enfors SO (1996) Substrate gradients in bioreactors: origin and
consequences. Bioprocess Eng 14:281289
Lee SY (1996) High-cell density culture of Escherichia coli. Trends
Biotechnol 14:98105
Liden G (2002) Understanding the bioreactor. Bioprocess Biosyst
Eng 24:273279
Lin HY, Neubauer P (2000) Influence of controlled glucose
oscillations on a fed-batch process of recombinant Escherichia
coli. J Biotechnol 79:2737
Lin HY, Mathiszik B, Xu B, Enfors SO, Neubauer P (2001)
Determinantion of the maximum specific uptake capacities for
glucose and oxygen in glucose-limited fed-batch cultivations of
Escherichia coli. Biotechnol Bioeng 73:347357
Macaloney G, Hall JW, Rollins MJ, Draper I, Thompson BG, McNeil
B (1994) Monitoring biomass and glycerol in an Escherichia
coli fermentation using near-infrared spectroscopy. Biotechnol
Tech 8:281286
Matz G, Lennemann F (1996) On-line monitoring of biotechnolog-
ical processes by gas chromatographic-mass spectrometric anal-
ysis of fermentation suspensions. J Chromatogr 750:141149
Mogk A, Tomoyasu T, Goloubinoff P, Rudiger S, Roder D, Langen
H, Bukau B (1999) Identification of thermolabile Escherichia
coli proteins: prevention and reversion of aggregation by DnaK
and ClpB. EMBO J 18:69346949
Moser I, Jobst G, Urban GA (2002) Biosensor array for simultaneous
measurement of glucose, lactate, glutamate, and glutamine.
Biosens Bioelectron 17:297302
Muramatsu R, Negishi T, Mimoto T, Miura A, Miosawa S, Hayashi
H (2002) Existence of -methylnorleucine in recombinant
hirudin produced by Escherichia coli. J Biotechnol 93:131142
Neils C, Tyree Z, Finlayson B, Folch A (2004) Combinatorial mixing
of microfluidic streams. Lab Chip 4:342350
Neubauer P, Lin HY, Mathiszik B (2003) Metabolic load of
recombinant protein production: inhibition of cellular capacities
for glucose uptake and respiration after induction of a heterol-
ogous gene in Escherichia coli. Biotechnol Bioeng 83:5364
Nishi H, Tsumagari N, Kakimoto T, Terabe S (1989) Separation of -
lactam antibiotics by micellar electrokinetic chromatography. J
Chromatogr 477:259270
Nishi H, Fukuyama T, Matsuo M (1990) Separation and determina-
tion of aspoxicillin in human plasma by micellar electrokinetic
chromatography with direct sample injection. J Chromatogr
515:245255
Nordstrom K, Uhlin BE (1992) Runaway-replication plasmids as
tools to produce large quantities of proteins from cloned genes
in bacteria. Biotechnology 6:661666
Oniscu C, Galaction AI, Cascaval D, Ungureanu F (2001) Modelling
of mixing in stirred bioreactors 1. Mixing time for non-aerated
simulated broths. Roum Biotechnol Lett 6:119129
Onyeaka H, Nienow A, Hewitt CJ (2003) Further studies related to
the scale-up of high cell density Escherichia coli fed-batch
fermentations: the additional effect of changing microenviron-
ment when using aqueous ammonia to control pH. Biotechnol
Bioeng 84:474484
Pollard D, Buccino R, Connors N, Kirschner T, Olewinski R, Saini
K, Slamon P (2001) Real-time analyte monitoring of a fungal
fermentation, at a pilot scale, using in mid-infrared spectrosco-
py. Bioprocess Biosyst Eng 24:1324
Reischer H, Schotola I, Striedner G, Ptschacher F, Bayer K (2004)
Evaluation of the GFP signal and its aptitude for novel on-line
monitoring strategies of recombinant fermentation processes. J
Biotechnol 115125
Riesenberg D, Menzel K, Schulz V, Schumann K, Veith G, Zuber G,
Knorre WA (1990) High cell density fermentation of recombi-
nant Escherichia coli expressing human interferon alpha 1. Appl
Microbiol Biotechnol 34:7782
Rocha I, Ferreira EC (2002) On-line simultaneous monitoring of
glucose and acetate with FIA during high cell density fermen-
tation of recombinant E. coli. Bioelectrochem 56:127129
Russel AB, Thomas CR, Lilly MD (1995) Oxygen transfer mea-
surements during yeast fermentations in a pilot scale airlift
fermenter. Bioprocess Biosyst Eng 12:7179
Schaefer U, Boos W, Takors R, Weuster-Bootz D (1999) Automated
sample device for monitoring intracellular metabolite dynamics.
Anal Biochem 270:8896
Schmalzriedt S, Jenne M, Mauch K, Reuss M (2003) Integration of
physiology and fluid dynamics. Adv Biochem Eng Biotechnol
80:1968
Schgerl K, Brandes T, Dullau K, Holzhauer-Rieger K, Hotop S,
Hbner U, Wu X, Zhou W (1991) Fermentation monitoring and
control by on-line flow injection and liquid chromatography.
Anal Chim Acta 249:87100
Schweder T, Krger E, Xu B, Jrgen B, Blomsten G, Enfors SO,
Hecker M (1999) Monitoring genes that respond to process-
related stress in large-scale bioprocesses. Biotechnol Bioeng
65:151159
Solle D, Geissler D, Stark E, Scheper T, Hitzmann B (2003) Chemo-
metric modelling based on 2D-fluorescence spectra without a
calibration measurement. Bioinformatics 19:173177
Spichiger-Keller UE (1997) Ion- and substrate specific optode
membranes and optical detection modes. Sens Actuators B
Chem B38:6877
Stark E, Hitzmann B, Schgerl K, Scheper T, Fuchs C, Koster D,
Mrkl H (2002) In-situ-fluorescence probes: a useful tool for
non-invasive bioprocess monitoring. Adv Biochem Eng Bio-
technol 74:2138
Thiry M, Cingolani D (2002) Optimizing scale-up fermentation
processes. Trends Biotechnol 20:103105
Tothill IE, Newman JD, White SF, Turner APF (1997) Monitoring of
the glucose concentration during microbial fermentation using a
novel mass-producible biosensor suitable for on-line use. En-
zyme Microb Technol 20:590596
Trelea IC, Titica M, Landaud S, Latrille E, Corrieu G, Cheruy A
(2001) Predictive modelling of brewing fermentation: from
knowledge-based to black box models. Math Comput Simul
56:405424
Trker M (2004) Development of biocalorimetry as a technique for
process monitoring and control in technical scale fermentations.
Thermochim Acta (in press)
Turner AP (1994) Biosensors. Curr Opin Biotechnol 5:4953
Turner C, Rudnitskaya A, Legin A (2003) Monitoring batch fer-
mentations with an electronic tongue. J Biotechnol 103:8791
Umakoshi H, Kuboi R, Komaswa I, Tsuchido T, Matsumura Y
(1998) Heat-induced translocation of cytoplasmic beta-galacto-
sidase across inner membrane of Escherichia coli. Biotechnol
Prog 14:210217
Vaccari G, Dosi E, Campi AL, Gonzalez-Vara A, Matteuzi D,
Mantovani G (1994) A near infrared spectroscopy technique for
the control of fermentation processes: an application to lactic
acid fermentation. Biotechnol Bioeng 43:913917
Vaidyanathan S, Macaloney G, Vaughan J, McNeil B, Harvey LM
(1999) Monitoring of submerged bioprocesses. Crit Rev
Biotechnol 19:277316
van de Merbel NC (1997) The use of ultrafiltration and column liquid
chromatography for an on-line fermentation monitoring. Trends
Anal Chem 16:162173
van de Merbel NC, Lingemann H, Brinkmann UAT (1993)
Automated monitoring of biotechnological processes using
on-line ultrafiltration and column liquid chromatography. Anal
Chim Acta 279:3950
Wang DIC, Cooney CL (1997) Translation of laboratory, pilot, and
plant scale data. In: Wang DIC et al (eds) Fermentation and
enzyme technology. Wiley, New York, pp 194211

Weickert MJ, Apostol I (1998) High-fidelity translation of recom-
binant human hemoglobin in Escherichia coli. Appl Environ
Microbial 64:15891593
Williams KA, Saini S, Wick TM (2002) Computational fluid
dynamics modeling of steady-state momentum and mass trans-
port in a bioreactor for cartilage tissue engineering. Biotechnol
Progress 18:951963
Wolfbeis OS (2002) Fiber-optic chemical sensors and biosensors.
Anal Chem 74:26632678
Wong I, Hernandez A, Garcia MA, Segura R, Rodriguez I (2002)
Fermentation scale up for recombinant K99 antigen production
cloned in Escherichia coli MC1061. Proc Biochem 37:1195
1199
435
Xu B, Jahic M, Bomsten G, Enfors SO (1999) Glucose overflow
metabolism and mixed acid fermentation in aerobic large-scale
fed-batch processes with Echerichia coli. Appl Microbiol Bio-
technol 51:564571
Yawalkar AA, Heesink ABM, Versteeg GF, Vishwas GP (2002) Gas
liquid mass transfer coefficient in stirred tank reactors. Can J
Chem Eng 80:840848
Yim SC, Jeong KJ, Chang HN, Lee SY (2001) High-level secretory
production of human granulocyte-colony stimulating factor by
fed-batch culture of recombinant E.coli. Bioprocess Biosyst Eng
24:249254