Review

Science-based bioprocess design for

filamentous fungi
Andreas E. Posch, Christoph Herwig, and Oliver Spadiut
Vienna University of Technology, Institute of Chemical Engineering, Research Area Biochemical Engineering, Vienna, Austria

Industrial bioprocesses are commonly based on empiricism rather than scientific process understanding. In this
review, we summarize current strategies for sciencebased bioprocess design and control for filamentous
fungi aiming at reducing development times and increasing process economics. We discuss recent developments and trends regarding three crucial aspects
throughout the bioprocess life cycle of filamentous fungi, namely (i) strain and inoculum characterization, (ii)
morphology, and (iii) rheology, as well as their effects on
process performance. Complex interconnections between strain, inoculum, morphology, rheology, and process design are outlined and discussed. Only combining
different hard type sensors with soft sensor technology
and the development of simplified mechanistic models
can enable science-based bioprocess design for filamentous fungi.
Industrial application of filamentous fungi
Economic, reliable, and controllable bioprocesses for filamentous organisms, in particular filamentous fungi, are of
utmost importance for the large scale production of a wide
range of value-added products including organic acids,
enzymes, and antibiotics. Due to complex interactions
between process technology, filamentous morphology
(Figure 1), and overall process performance, traditional
bioprocess design is still commonly carried out stepwise by
time-consuming, empirical strategies (Figure 2).
In contrast to biopharmaceuticals, the majority of products from filamentous fungi are industrial (white) biotechnological bulk products and thus not subject to tight
regulatory demands. Consequently, manufacturers can
continuously optimize their production strains and industrial processes to ensure competitiveness. Common empirical approaches, however, lack scientific insight into
process technology and key process parameters (KPPs)
(see Glossary) inevitably leading to sub-optimally designed
bioprocesses and high process failure rates. Therefore,
bioprocess engineers pursue two overall goals summarized
as two-times 50%, which means twice the productivity and
reducing bioprocess development time to 50%. In order to
meet these economic requirements, it is necessary to understand and control the biological system used.
In this review, we give an overview of recently developed measurement and control strategies for major
Corresponding author: Spadiut, O. (oliver.spadiut@tuwien.ac.at)
Keywords: bioprocess development; bioprocess characterization; filamentous fungi;
strain screening; morphology; rheology.

obstacles throughout the bioprocess life cycle of filamentous fungi that significantly affect the overall process
performance. Points that impact the process performance
are (i) strain and inoculum, (ii) morphology, and (iii)
rheology. Furthermore, we provide future perspectives
targeting a holistic understanding of complex interdependencies in science-based process design for filamentous fungi.
Strain and inoculum
Screening for strain characteristics
The first step in bioprocess design is the identification of
promising candidate strains (Figure 2). To achieve automation and high-throughput, strains are commonly identified in micro titer plate (MTP) systems. Besides
identification of highly productive strains, initial screening should also consider morphological characteristics,
such as complex growth morphology or adherent wall
growth, which significantly affect reproducibility of
results and the subsequent scale-up procedure. Missing
these important factors, selected clones may fail at delivering productivities observed at small scale. Reducing the
headspace was proven to be a useful strategy for preventing wall growth and consequent heterogeneities [13].
Another study compared physiological and morphological
process performance in MTPs and 1-liter benchtop reactors. Addition of glass beads promoted mycelial growth

Glossary
Carbon dioxide evolution rate and oxygen uptake rate (CER/OUR): respiratory
rates reflecting physiological culture activity.
Critical quality attribute (CQA): a physical, chemical, biological, or microbiological characteristic that should be within an appropriate limit, range, or
distribution to ensure desired product quality.
Critical process parameter (CPP): a process parameter whose variability has an
impact on a critical quality attribute and therefore should be monitored or
controlled to ensure that the process produces the desired quality.
Design of Experiments (DoE): a structured, organized method for determining
the relationship between factors affecting a process and the output of that
process.
Key process parameter (KPP): an adjustable parameter (variable) of the process
that, when maintained within a narrow range, ensures optimum process
performance. A key process parameter does not meaningfully affect critical
product quality attributes. Ranges for KPPs are established during process
development, and changes to operating ranges will be managed within the
quality system.
Process analytical technology (PAT): a system for designing, analyzing, and
controlling manufacturing through timely measurements of critical quality and
performance attributes of raw and in-process materials and processes with the
goal of ensuring final product quality.
Quality by Design (QbD): a systematic approach to development that begins with
predefined objectives and emphasizes product and process understanding and
process control, based on sound science and quality risk management.

0167-7799/$ see front matter 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tibtech.2012.10.008 Trends in Biotechnology, January 2013, Vol. 31, No. 1

37

Review

[(Figure_1)TD$IG]

Trends in Biotechnology January 2013, Vol. 31, No. 1

Pellet
morphology

Disperse
morphology
Developing
hyphae
Spores

TRENDS in Biotechnology

Figure 1. Morphological life cycle of filamentous fungi. Independent of capitalizing Aspergillus spp. for the production of enzymes or organic acids, Trichoderma spp. for
cellulases, or Penicillium spp. for antibiotics, these filamentous industrial workhorses all share a unique and complex growth morphology. Fungal bioprocesses are
commonly inoculated from spores. After successful germination: branching, extending, and entanglement of single hyphae introduce disperse hyphal networks. Over
process duration, pellets may develop via growth from single spores or by agglomeration. High shear forces and nutrient limitation cause pellet disintegration.

High-throughput screening for

highly producing strains.

Screening of process morphology

and rheology at this stage
demands veried, mechanisc
scale-up models.

2. Process
development

1. Strain screening

[(Figure_2)TD$IG]

Characterization of the inoculum

Spore inoculum concentration, quality, and viability severely affect fungal morphology, physiology, and productivity [7]. To date however, the amount of inoculum and the
setting of process parameters are still empirically optimized with each inoculum batch. Consequently, initial
processes performed with a fresh inoculum batch only
gradually reach an empiric optimum after several test
runs. This procedure completely disagrees with the overall
two-times 50% target in industrial (white) biotechnology.
In this respect, parallel microbioreactor devices and
MTP cultivation systems [1,4,5] represent useful tools
for the characterization of germination efficiency and spore
viability. Moreover, fluorescence microscopy [8] and Fourier transform infrared (FTIR) spectromicroscopy [9,10]
provide spatial resolution of the intracellular biochemical
composition of filamentous fungi. Although these methodologies are still in their infancy, results show promise for a
means to quantify spore inoculum quality in the future.
The correct transfer time point from the seed reactor to
the production reactor after successful spore germination
is as crucial as the optimal spore inoculum quality.

Measurement and control of

KPP kinecs at limited
throughput in liter scale.
Inial invesgaon of process
morphology and rheology
based on hydraulic
scale-down.

3. Implementaon
at producon scale

and concomitantly decreased wall growth [4]. Using this

approach, physiological performance parameters identified during MTP screening were similar to benchtop reactors. The authors also studied the influence of different
cultivation regimes on morphological process behavior by
measuring particle size distributions. Although physiological screening results were similar at miniaturized and
lab-scale, growth morphology was different, probably due
to differing shear forces.
A disposable microbioreactor system for high-throughput screening of germination efficiency was introduced
recently [5]. Reactor chambers of 100 ml volume enabled
parallelized characterization of growth morphology during
spore germination as well as quantification of inoculum
quality. Although such miniaturized reactors do not allow
for complete characterization of growth morphology, it was
shown that morphological parameters, such as the hyphal
growth unit length, could be correlated to pellet formation
kinetics and pellet size [6].
Another crucial factor that has to be considered when
analyzing strain characteristics is the growth medium
(Box 1).

Process control at manufacture

scale based on KPP ranges
idened in step 2.
Smooth integraon of novel
clones into the producon is
oen hampered by the lack of
mechanisc understanding of
up-scaling eects.

TRENDS in Biotechnology

Figure 2. Traditional bioprocess design for filamentous fungi. The development strategy starts with qualitative screening of suitable clones, followed by quantitative
bioprocess development in lab-scale bioreactors. Thereby, key process parameters (KPPs) that ensure that the biological system is steered to target product quantity are
identified. Subsequent piloting aims at scalability of the process, analyzing adaptation of the process parameters, and scale-up effects [53].

38

Review
Box 1. Media development
To date, most of the established production processes with
filamentous fungi are performed in chemically undefined media.
The major advantages of complex media are that they contain a rich
variety of compounds allowing unimpaired growth of microorganisms and that they are made from inexpensive sources. However,
significant lot-to-lot variability of different components in complex
media results in large variations in the production efficiency.
Currently, there are two strategies to tackle this problem: (i)
characterizing complex raw materials and taking corresponding
actions depending on the variations, and (ii) switching production
processes from complex to defined media.
For the characterization of complex media components, spectroscopic techniques, including near infrared (NIR), mid infrared (MIR),
and attenuated total reflectance mid infrared (ATR-MIR) spectroscopy are commonly employed [54]. A comprehensive overview of
most of these methods, containing valuable suggestions for
possible applications of the respective technique, was given in
2004 [55]. Major drawbacks of spectroscopic tools are sensor fouling
during inline application, extensive calibration procedures for
reliable quantitative chemometric models as well as the demand
for complex multivariate data analysis [54,56]. Hence, for the sake of
simplicity and controllability, avoiding raw material variability by
switching to defined cultivation media is preferable.
Fast strain characterization of filamentous fungi in batch cultivations on combined complex/defined media allows identification of
the most promising candidate strain for switching the bioprocess
from complex to fully defined media [3]. This methodology is based
on minimal analytical needs and the indirect determination of
specific uptake rates of complex media components by applying
statistically verified redundant mass balancing techniques. This
approach has not only enabled the switch from complex to defined
media but also a twofold increase of the overall penicillin space time
yield and a threefold increase in the maximum specific penicillin
formation rate [3].

Measurement of the carbon dioxide evolution rate (CER),

multi-wavelength fluorescence spectroscopy, turbidity
measurement, and near-infrared (NIR) spectroscopy are
used to determine this time point online [11,12]. Spectroscopic approaches do not require off-gas analysis and thus
also apply to MTPs and shake flasks.
Miniaturized and parallel screening approaches provide
fast quantification of strain and inoculum characteristics.
Screening for relevant physiological and morphological
characteristics at this initial stage of bioprocess design
will facilitate prediction of large scale process behavior.
Morphology
Coagulating versus non-coagulating species
The mechanism of pellet formation depends on the fungal
strain. A fungal strain can be coagulating or non-coagulating, a description of the phenomena of conidial aggregation. Several studies have analyzed the event of conidial
aggregation, which is further influenced by the media,
mechanical forces, as well as the inoculum concentration
and viability [7,13]. Aggregation occurs in two steps: (i)
immediately after inoculation, only affected by the pH
value and the osmolality of the medium [14]; (ii) at germination, affected by pH, osmolality, agitation, and aeration
[13]. Conidial aggregation can be nicely monitored by 2D
fluorescence spectroscopy [11] and has been described in a
reliable process model [7].
For strains of non-coagulating fungi, pellets either form
from a single spore that germinates or from entanglements

Trends in Biotechnology January 2013, Vol. 31, No. 1

of branched hyphae and disperse mycelial clumps. The

spore inoculum concentration is a KPP for pellet formation.
In fact, pellet formation accelerates at reduced spore inoculum concentrations [15]. In order to reliably describe the
behavior of filamentous fungi over the production process,
knowledge of inoculum and its coagulation characteristics
should be integrated into a global morphological population balance as well as predictive multivariate models [13].
Different morphologies during production processes
Most filamentous fungi processes are multi-stage cultivations during which filamentous fungi exhibit different
morphologies. The predominant morphology during cultivations depends on several factors like the strain, the spore
inoculum concentration and quality, and agitation [15].
Nutrient limitation and oxygen diffusion gradients affecting pellet micromorphology [16] as well as effects of dO2
and dCO2 concentrations in the bioreactor on fungal
macromorphology have all been quantified recently [17
19]. This is especially important for process scalability
because hydrostatic backpressures at the bottom of large
vessels may create high residual dCO2 concentrations
[17,18,20]. Other studies have investigated morphological
changes during prolonged cultivation periods [21], during
which substrate concentration gradients across fungal
pellets affect pellet micromorphology and biomass activity
[16].
Morphology significantly influences several other KPPs
including broth viscosity, mixing power demands, and
substrate diffusion, but also strain-specific parameters like
productivity and yields [22]. Morphology and rheology
interdependencies can be investigated by digital image
analysis [23] or particle size analysis [24]. Certain growth
morphologies are directly related to the production of
industrially relevant compounds, for example, pellets
are preferred for citric acid production [15], whereas mycelial growth enhances formation of fructofuranosidase
[14]. Hence, reliable morphology analysis and control are
crucial for optimized bioprocesses.
Morphology analysis and control
Growth morphology is increasingly targeted as a KPP
during process development because of the undisputable
link between fungal morphology and process performance
[15,22,25]. A prerequisite for morphological process control
is the detailed understanding of a fungus morphological
response to process parameters. Such knowledge, however,
can only be gathered if accurate tools for morphological
analysis are available. An excellent, recently published
review discusses methods to analyze fungal morphology
including atomic force microscopy, flow cytometry, and
real-time applicable techniques [25]. Although for some
of these techniques applicability in the production environment still needs to be proven, the review showcases
existing approaches to elucidate relevant, mechanistic
morphological information for optimized bioprocess development.
Reliable analytical techniques should allow detailed
investigation of the effects of changes in process parameters, including spore inoculum, media composition, and
mixing kinetics, on fungal morphology [14,15,22,26,27]. So
39

Review
far, however, these analyses have mostly been based on a
limited number of manually recorded microscopic images
using digital image analysis techniques introduced more
than a decade ago [27]. Because manual image recording
prevents high-throughput applicability, automatable lowlevel morphological characterization approaches based on
particle size distributions should be used for industrial
routine applications [4,24].
For closer insights, intermediate-level morphological
characterization approaches, giving access to mechanistic
morphological information, can be used. Fractal dimensions can be employed as morphological descriptors [28]
and also describe branching behavior and pellet morphology [29]. Fractal dimensions measure the degree of complexity as well as mass filling properties, and a large
number of morphological parameters can be effectively
replaced by these indices [28]. Moreover, another morphological descriptor, the morphology number, is closely correlated with productivity for Aspergillus niger [14].
The optimal strategy, however, is a high-throughput,
high-level morphological characterization approach describing fungal morphology based on morphological classification and frequency distributions for several
morphological parameters [6]. Such approaches allow
modeling and understanding of heterogeneities that affect
the overall process performance based on population balances [30]. Promising, structured morphological modeling
strategies for predicting morphological process behavior
and the link to productivity were already introduced in the
90s, but still have not been implemented for industrial
process development [31,32]. To date, the lack of reliable
tools for automated high-throughput analysis of fungal
morphology has prevented the usage of morphological
analysis as a routine analysis procedure. Devices for online, inline, and real-time microscopic measurements have
been developed in recent years, but their applicability as
automated online measurement tools is yet to be demonstrated [33]. Despite the availability of computational
workflows minimizing manual intervention during morphological analyses [34,35], there is still a need for image
analysis systems with fully automated image acquisition
and analysis [14,36]. A recent high-throughput method for
fast morphological characterization of filamentous fungi
interlinks fully automated recording of thousands of microscopic images with concomitant statistical verification
for complete, accurate quantification of fungal morphology
[6].
Pellet micromorphology and substrate diffusion limitations affecting pellet growth, erosion, and breakage can be
quantified by confocal laser scanning microscopy [16].
Recent, novel tools for micromorphological analysis include (i) FTIR spectromicroscopic imaging, which allows
3D spatial resolution of pellet micromorphology by investigating the biochemical composition of pellet microtome
sections, and (ii) MALDI-TOF intact cell mass spectrometry (ICMS), which provides micromorphological analysis
that facilitates identification and quantification of protein
patterns in different physiological or morphological process states (A.E. Posch et al., unpublished). Such highlevel techniques represent an important step towards
mechanistic modeling of interdependencies between
40

Trends in Biotechnology January 2013, Vol. 31, No. 1

process parameters, physiology, mass transfer, and micromorphology.

To date, morphological control strategies for industrial
production include the selection of a candidate strain that
exhibits the desired morphology during process development [6] as well as adjustment of spore inoculum concentration, media composition, and agitation regime [15].
Moreover, microparticles can be used to customize fungal
morphology [37]. Similarly, control of osmolality facilitates
control of disperse mycelial morphology [14].
We believe that high-throughput morphological characterization approaches ranging from light microscopy to
flow cytometry [6,24] will facilitate morphology- and thus
bioprocess-control in the future. Updating mechanistic,
structured morphological models with quantitative morphological data will advance approaches for predictive 3D
morphological modeling of pellet and disperse morphology
[38]. Sound combination of sensors for biomass volume
[39], for example, dielectric (DE) spectroscopy, and redundant elemental mass balancing for estimation of biomass
production [3] may be a promising path to facilitate realtime monitoring of biomass characteristics by soft sensors.
Rheology
Rheology and morphology
There are complex interconnections between morphology,
biomass concentration, and rheology [23,40]. Here, we
highlight the dependency of the rheology on fungal morphology. Freely dispersed fungal growth causes high broth
viscosity and impairs oxygen transfer. Unless there is a
direct linkage between morphology and productivity, pellet
morphology is preferred in production processes due to
lower viscosity. However, pellet growth limits substrate
diffusion to the pellet core at critical pellet diameters,
which causes lysis of pellet core cells [6]. This event has
been spatially resolved by confocal laser scanning microscopy [16].
Hydrodynamic conditions in the bioreactor influence the
morphological development of the fungus and thus the
rheology. High agitation rates, which are required for
optimum mass transfer, lead to high shear stress. Consequently, mycelial networks are fragmented, increasing the
amount of free filaments and hence the viscosity. High
viscosity causes insufficient mass transfer and oxygen
limitation [41]. Modeling fungal fragmentation requires
calculation of the hyphal strength and the mechanical
forces exerted on the fungus by the turbulent flow. In this
respect, it is crucial to know the energy dissipation by
impellers and aeration. Higher energy dissipation causes
strong erosion and changes in pellet morphology as shown
for A. niger [41]. In general, hydrodynamic conditions affect
pellet formation on three levels: (i) interaction of pellets
and eddies; (ii) impact between pellets and stirrers/baffles;
and (iii) collision among pellets [42]. These interactions
may strip off hyphae or completely rupture whole pellets.
To reliably describe a fungal bioprocess, the influence of
hydrodynamic shear stress on morphology, rheology, and
productivity has to be integrated into models. The average
energy input, an easily accessible parameter, can be used
to characterize fluid dynamics and rheology. This parameter, however, depends on the experimental set-up and thus

Review

Trends in Biotechnology January 2013, Vol. 31, No. 1

Box 2. Predicting rheology

Projected power requirements from lab to large scale often diminish
process economics or even exceed power availability. Hence,
bioprocesses with filamentous fungi need to be adapted during
piloting to meet available power at manufacturing scale, for example,
by media adaptation and introduction of dilution steps [20]. To date,
scale-up of filamentous bioprocesses is commonly targeted using
empirical rule of thumbs for hydraulic scale-up including the
maximum shear rate or impeller tip speed, kla value, oxygen
saturation, volumetric power input, and mixing time [5759].
Science-based process design applying morphological rheological
modeling for predictive morphological scale-up would represent a
major step towards increasing the process success rate and thus
contribute at reducing process development times and increasing
process economics (Figure I).
Unfortunately, it is still very complex and difficult to predict the
rheology of a fermentation broth. Important formulas to describe
different aspects influencing the rheology are the well-known
formulas for the apparent viscosity, the Metzner and Otto method
power law, and the Herschel-bulkley model. However, a single
formula cannot describe the complex processes contributing
to the viscosity in a bioreactor. Complementing conventional
low-throughput image analysis for rheological modeling [23],

[(Box_2)TD$FIG]

multivariate methods, and high-throughput accessible size distribution data can be used to predict rheological properties in a
bioreactor [24]. The most important parameters to describe broth
rheology can be successfully predicted from size distribution,
biomass concentration, and feed mode. Current approaches combine particle size distribution data and biomass concentration and
analyze these data using different chemometric tools, such as
partial least squares (PLS) models. These multivariate approaches
model rheological properties of filamentous fermentation broths
with equal or greater accuracy than traditional power-law type
relationships based on population average data from image
analysis [24]. However, the calibrated model is specific for a certain
fungal strain, the scale and the operating conditions and thus only
valid in a limited range.
We believe that superior particle characterization methods (highthroughput, statistically verified image analysis) will significantly
increase the predictive power of rheology models in the future by
providing mechanistic modeling approaches [23,38] with highdimensional quantitative morphological population data [6]. Thus,
we suggest predictive morphological scale-up along with scalable,
nonlinear control strategies for reliable extrapolation of production
process performance from screening results.

Science-based process design

Rheology

Miniaturizaon and
parallelizaon
High-throughput analysis
and data vericaon
Online monitoring
and modeling

Morphology

Strain and inoculum

TRENDS in Biotechnology

Figure I. Recommended strategies for science-based process design targeting reduced process development times and increased process economics. Over the coming
years, we will see the implementation of recently introduced advanced sensors for strain and inoculum characterization, morphology, rheology, and process
performance, as well as their concomitant improvement. This will allow more accurate recording of KPPs and give access to a much greater number of (online) process
data. Extraction of relevant information from such accessible multidimensional data by multivariate analysis and integration into simplified, mechanistic process
models will be necessary to increase process success rate from miniaturized screening to manufacturing scale. Combination of orthogonal, complementing
measurement techniques by sound soft sensors already enables resolving the interplay of highly cross-linked KPPs and thus describes a promising approach towards
science-based bioprocess design for filamentous fungi.

cannot be transferred across models. A holistic model has

to use transferable parameters and describe all processes
independent of the operating conditions. Thus, the interaction between eddies and the pellet surfaces needs to be
related to shear forces. Shear forces are unequally distributed within the reactor, so the maximum shear rate rather
than average parameters should be used to describe the
rheology in a bioreactor [41].
Rheology measurement tools
Reliable, automatable approaches for timely determination of rheological broth characteristics are still needed.
Traditional techniques can be grouped into three categories: (i) capillary techniques, (ii) rotational techniques, and
(iii) falling ball techniques [43]. However, these techniques

cannot be used for fungal fermentation broths because they

are considered to be non-Newtonian liquids. Fungal fermentation broths are a complex system: at the beginning of
a process the broth basically describes a Newtonian liquid,
but due to biomass growth and changing fungal morphology, the broth becomes heterogeneous exhibiting non-Newtonian characteristics. Recent developments in instrument
technology for rheometry, especially with regard to nonNewtonian liquids, have been achieved [43,44]. Despite the
availability of inline viscometers from the chemical industry, rheology is commonly not measured in real-time for
filamentous bioprocesses. Novel, miniaturized devices, including ultrasound spectroscopic tools, electromagnetic
acoustic resonators, and vibrating bridge devices are still
in their infancy but describe a promising online tool for
41

Review

Trends in Biotechnology January 2013, Vol. 31, No. 1

Table 1. Methods for real-time determination of filamentous biomassa

Method

Cross-sensitivity

Gas
bubbles

Microscopy
R
Turbidimetry

Flow cytometry
Vibrational spectroscopy R

Applicability

Refs

Solid media
components
+
R
R
R

Salt
content

High cell
density

R
R

Morphology Sensor
fouling
+
R
R
R
+

R/+
R

Screening Development Production

A
A
NA
A

A
A
A
A

R
R
A
R

Dielectric spectroscopy

NA

Fluorescence
spectroscopy
Calorimetry
Mass balancing from
pH/base addition
Mass balancing from
pO2/OUR from pO2
Off-gas measurement

R/+

NA
A

A
A

A
A

[33,47]
[46,47]
[24,47]
[12,46,
54]
[39,46,
47]
[11,39,
46,47]
[49,50]
[51]

[51]

NA

Soft sensors

[3,11,
12,39]
[3,46,51]

Abbreviations: +, quantifiable cross-sensitivity; , no cross-sensitivity; R, case-specific risk assessment necessary due to non-quantifiable cross-sensitivity; A, applicable;
NA, not applicable.

measuring filamentous broth viscosity. Hence, current

efforts are focused on reliable models that predict broth
rheology based on measurements of more easily accessible
parameters including biomass concentration, particle size
distribution [24], and energy input [45]. Accurate measurement of broth rheology as well as its prediction during
scale-up is crucial for successful bioprocess design (Box 2).
Effects on process performance
To quantify process performance of filamentous fungi, the
biomass concentration, viability as well as its physiological
state should be monitored.
Biomass concentration and viability
Volumetric productivity, the target parameter in industrial process design, strongly depends on the biomass
concentration and the amount of viable, actively producing cells. Also KPPs, such as the rheology [23], are influenced by the biomass concentration. Reliable online
determination of the biomass concentration and the
amount of active biomass would represent a major step
towards effective bioprocess control. To monitor biomass
characteristics, different strategies, like turbidity measurements, IR spectroscopy, 2D spectroscopy, dielectric
(DE) spectroscopy, and different soft sensors have been
developed [46,47].
Broth viscosity can be used as an indicator for the
biomass concentration of the filamentous model organism
Streptomyces spp.. The viscosity is proportional to the
biomass concentration up to the end of the exponential
growth phase [48]. However, during later stages of the
process, viscosity does not indicate biomass concentration
due to changes in the rheological properties of the fermentation broth arising from changes in morphology [48].
Biocalorimetry is used to monitor heat production during bioprocesses online. The metabolic heat production
correlates with growth, morphology, and physiology
42

[49,50]. Investigating heat evolution over process time

by so-called power-time curves provides information on
biomass growth and physiology. Furthermore, heat release
patterns also reflect changes in fungal morphology, which
is why this online strategy could be a useful tool for noninvasive morphology monitoring [49,50].
Soft sensors can be used to accurately estimate biomass
by combining easy-to-monitor data, like the carbon dioxide
evolution rate (CER), the oxygen uptake rate (OUR), or the
base consumption, with hard type sensors, which are
closely related to biomass growth [51]. Hence, soft sensors
are a powerful approach for tracking KPPs, which would
only be accessible for direct measurement by costly equipment and manpower. Provided that the relevant components have been identified and can be measured online,
first principle stoichiometric mass balancing allows online
estimation of biomass concentration even for cultivations
on complex, solid media [3]. A comparison of common
online biomass sensors has shown that DE spectroscopy,
turbidity measurement, and CER gave the most accurate
estimates. A combination of sensors also reveals information on the morphology of the fungus, changing yields, and
the biomass viability [39].
Currently, a large variety of different measurement
tools to determine the fungal biomass concentration is
available. Table 1 gives an overview of these technologies,
shows cross-sensitivities, and describes the stages of the
fungal bioprocess life cycle, where the respective technology can be used. In summary, throughout all stages of
bioprocess design, the combination of complementary measurement principles with first principle stoichiometric
mass balancing in soft sensors is highly recommended.
Biomass viability, concentration, morphology, and activity
can be estimated to ultimately facilitate predictive process
modeling and control. In addition to such representative
biomass measurements, the physiological process state
should be reliably monitored (Box 3).

Review

Box 3. Monitoring physiological process performance

Different strategies can be applied and combined for in situ
monitoring of the physiological process state. 2D fluorescence
spectroscopy can be employed to monitor the productivity and the
metabolic activity of a recombinant A. niger strain during fed-batch
cultivations [11]. Glucose in Streptomyces coelicolor cultivations
has been accurately monitored by NIR spectroscopy [60]. Performance of a production process on complex media for a recombinant
animal cell line has been nicely predicted by combining NIR and 2D
fluorescence spectroscopy [61]. We believe that this strategy also
describes a potential tool for bioprocesses with filamentous fungi.
Unfortunately, it is difficult to develop reliable models for spectroscopic methods. Biomass and broth absorbance may hide absorbance of other analytes, and peaks in the spectra can overlap; filtrate
samples have to be analyzed separately in order to formulate a
model for a certain key analyte [54,60].
The relationships between strain, inoculum, morphology, rheology, and process design with physiology and productivity are
complex. To date, reliable and verified process analytical technology (PAT) tools for in situ monitoring of bioprocesses of filamentous fungi are still scarce. Online sampling strategies allowing
automated real-time at-line parameter measurement have been
suggested for other microbial systems within recent years and
illustrate the ongoing research in this field (e.g., [62]). However,
sampling techniques need to be validated for each specific
application [63]. Moreover, PAT tools that have been developed
for unicellular systems may not be transferred easily to filamentous
fungi due to several fungal specific characteristics, such as the
complex growth morphology or adherent wall growth affecting
probe fouling.
Simplified, mechanistic modeling presents an alternative strategy
to predict and control productivity, rheology, and morphology [52].
In this context, a mechanistic approach to model enzyme production
in A. oryzae has been based on common hydraulic parameters
including agitation and aeration regimes but also 1st principle
relationships for reaction stoichiometry, mass transfer, and broth
viscosity [45]. The model accurately infers KPPs including productivity, broth viscosity, and oxygen transfer, as well as the specific
growth rate at varying cultivation regimes.

Concluding remarks and future trends

Challenges in bioprocess design for filamentous fungi are
heavily interdependent (see Figure I in Box 2). Systematic
investigation of interdependencies between KPPs requires a
Design of Experiments (DoE) approach. Relevant process
information should be extracted from experimental raw
data by statistically verified macroscopic balancing and
morphological analysis [3,6]. Furthermore, underlying
physiological information should be examined with mechanistic modeling [45,52]. Although not yet a regulatory
threat, we believe that following Quality by Design (QbD)
guidelines [ICH (2009) Q8, Pharmaceutical Development
(R2); see http://www.ich.org] would be also beneficial in
the case of industrial (white) biotechnological bulk products.
Science-based process development based on initial risk
assessment and identification of critical quality attributes
(CQAs), critical process parameters (CPPs), and KPPs
followed by systematic resolution of their interdependencies
has been demonstrated for the production of biopharmaceuticals [CMC-Biotech-Working-Group (2009) A-Mab: a Case
Study in Bioprocess Development; http://www.casss.
org/associations/9165/files/A-Mab_Case_Study_Version_
2-1.pdf, 1-278]. Implementation of such an automated
workflow will also enable fast integration of novel, improved strains into the production, decrease development

Trends in Biotechnology January 2013, Vol. 31, No. 1

times, and increase process economics for the production

of value-added products from filamentous fungi.
References
1 Siebenberg, S. et al. (2010) Reducing the variability of antibiotic
production in Streptomyces by cultivation in 24-square deepwell
plates. J. Biosci. Bioeng. 109, 230234
2 Meyer, V. et al. (2010) Genetics, genetic manipulation, and approaches
to strain improvement of filamentous fungi, In Manual of Industrial
Microbiology and Biotechnology, (3rd edn), American Society for
Microbiology, pp. 318329
3 Posch, A.E. et al. (2012) Switching industrial production processes from
complex to defined media: method development and case study using
the example of Penicillium chrysogenum. Microb. Cell Fact. 11, 88
4 Sohoni, S.V. et al. (2012) Robust, small-scale cultivation platform for
Streptomyces coelicolor. Microb. Cell Fact. 11, 9
5 Demming, S. et al. (2011) Disposable parallel poly(dimethylsiloxane)
microbioreactor with integrated readout grid for germination screening
of Aspergillus ochraceus. Biomicrofluidics 5, 14104
6 Posch, A.E. et al. (2012) A novel method for fast and statistically
verified morphological characterization of filamentous fungi. Fungal
Genet. Biol. 49, 499510
7 Grimm, L.H. et al. (2004) Kinetic studies on the aggregation of
Aspergillus niger conidia. Biotechnol. Bioeng. 87, 213218
8 Herbrich, S. et al. (2012) Label-free spatial analysis of free and enzymebound NAD(P)H in the presence of high concentrations of melanin. J.
Fluoresc. 22, 349355
9 Kaminskyj, S. et al. (2008) High spatial resolution analysis of fungal
cell biochemistrybridging the analytical gap using synchrotron FTIR
spectromicroscopy. FEMS Microbiol. Lett. 284, 18
10 Szeghalmi, A. et al. (2007) A synchrotron FTIR microspectroscopy
investigation of fungal hyphae grown under optimal and stressed
conditions. Anal. Bioanal. Chem. 387, 17791789
11 Ganzlin, M. et al. (2007) In situ multi-wavelength fluorescence
spectroscopy as effective tool to simultaneously monitor spore
germination, metabolic activity and quantitative protein production
in recombinant Aspergillus niger fed-batch cultures. J. Biotechnol. 132,
461468
12 Nordon, A. et al. (2008) In situ monitoring of the seed stage of a
fermentation process using non-invasive NIR spectrometry. Analyst
133, 660666
13 Grimm, L.H. et al. (2005) Influence of mechanical stress and surface
interaction on the aggregation of Aspergillus niger conidia. Biotechnol.
Bioeng. 92, 879888
14 Wucherpfennig, T. et al. (2011) Morphology engineeringosmolality
and its effect on Aspergillus niger morphology and productivity.
Microb. Cell Fact. 10, 58
15 Papagianni, M. (2004) Fungal morphology and metabolite production
in submerged mycelial processes. Biotechnol. Adv. 22, 189259
16 Hille, A. et al. (2005) Oxygen profiles and biomass distribution in
biopellets of Aspergillus niger. Biotechnol. Bioeng. 92, 614623
17 El-Sabbagh, N. et al. (2006) Dissolved carbon dioxide effects on growth,
nutrient consumption, penicillin synthesis and morphology in batch
cultures of Penicillium chrysogenum. Enzyme Microb. Technol. 39,
185190
18 El-Sabbagh, N. et al. (2008) Effects of dissolved carbon dioxide on
growth, nutrient consumption, cephalosporin C synthesis and
morphology of Acremonium chrysogenum in batch cultures. Enzyme
Microb. Technol. 42, 315324
19 Li, Q. et al. (2008) The effects of bioprocess parameters on extracellular
proteases in a recombinant Aspergillus niger B1-D. Appl. Microbiol.
Biotechnol. 78, 333341
20 Junker, B. et al. (2009) Pilot-scale process development and scale up for
antifungal production. Bioprocess Biosyst. Eng. 32, 443458
21 Haack, M.B. et al. (2006) Change in hyphal morphology of Aspergillus
oryzae during fed-batch cultivation. Appl. Microbiol. Biotechnol. 70,
482487
22 Wucherpfennig, T. et al. (2010) Morphology and rheology in
filamentous cultivations. Adv. Appl. Microbiol. 72, 89136
23 Riley, G.L. and Thomas, C.R. (2010) Applicability of Penicillium
chrysogenum rheological correlations to broths of other fungal
strains. Biotechnol. Lett. 32, 16231629
43

Review
24 Petersen, N. et al. (2008) Multivariate models for prediction of
rheological characteristics of filamentous fermentation broth from
the size distribution. Biotechnol. Bioeng. 100, 6171
25 Krull, R. et al. (2012) Characterization and control of fungal
morphology for improved production performance in biotechnology.
J. Biotechnol. http://dx.doi.org/10.1016/j.jbiotec.2012.06.024
26 Ahamed, A. and Vermette, P. (2009) Effect of culture medium
composition on Trichoderma reeseis morphology and cellulase
production. Bioresour. Technol. 100, 59795987
27 Paul, G.C. and Thomas, C.R. (1998) Characterisation of mycelial
morphology using image analysis. Adv. Biochem. Eng. Biotechnol.
60, 159
28 Papagianni, M. (2006) Quantification of the fractal nature of mycelial
aggregation in Aspergillus niger submerged cultures. Microb. Cell Fact.
5, 5
29 Barry, D. et al. (2009) Relating fractal dimension to branching
behaviour in filamentous microorganisms. ISAST Trans. Electron.
Signal Process. 4, 7176
30 Muller, S. et al. (2010) Origin and analysis of microbial population
heterogeneity in bioprocesses. Curr. Opin. Biotechnol. 21, 100113
31 Paul, G.C. and Thomas, C.R. (1996) A structured model for hyphal
differentiation and penicillin production using Penicillium
chrysogenum. Biotechnol. Bioeng. 51, 558572
32 Paul, G.C. et al. (1998) A structured model for penicillin production on
mixed substrates. Biochem. Eng. J. 2, 1121
33 Bluma, A. et al. (2010) In-situ imaging sensors for bioprocess
monitoring: state of the art. Anal. Bioanal. Chem. 398, 24292438
34 Barry, D.J. et al. (2009) Morphological quantification of filamentous
fungal development using membrane immobilization and automatic
image analysis. J. Ind. Microbiol. 36, 787800
35 Lecault, V. et al. (2007) Morphological characterization and viability
assessment of Trichoderma reesei by image analysis. Biotechnol. Prog.
23, 734740
36 Barry, D.J. and Williams, G.A. (2011) Microscopic characterisation of
filamentous microbes: towards fully automated morphological
quantification through image analysis. J. Microsc. 244, 120
37 Driouch, H. et al. (2011) Improved enzyme production by bio-pellets of
Aspergillus niger: targeted morphology engineering using titanate
microparticles. Biotechnol. Bioeng. 109, 462471
38 Celler, K. et al. (2012) Structured morphological modeling as a
framework for rational strain design of Streptomyces species.
Antonie Van Leeuwenhoek 102, 409423
39 Ronnest, N.P. et al. (2011) Introducing process analytical technology
(PAT) in filamentous cultivation process development: comparison of
advanced online sensors for biomass measurement. J. Ind. Microbiol.
Biotechnol. 38, 16791690
40 Riley, G.L. et al. (2000) Effect of biomass concentration and mycelial
morphology on fermentation broth rheology. Biotechnol. Bioeng. 68,
160172
41 Kelly, S. et al. (2006) Effects of fluid dynamic induced shear stress on
fungal growth and morphology. Process Biochem. 41, 21132117
42 Kelly, S. et al. (2004) Agitation effects on submerged growth and
product formation of Aspergillus niger. Bioprocess Biosyst. Eng. 26,
315323
43 Hou, Y.Y. and Kassim, H.O. (2005) Instrument techniques for
rheometry. Rev. Sci. Instrum. 76, 101101101119
44 Jakoby, B. et al. (2010) Miniaturized sensors for the viscosity and
density of liquids-performance and issues. IEEE Trans. Ultrason.
Ferroelectr. Freq. Control 57, 111120

44

Trends in Biotechnology January 2013, Vol. 31, No. 1

45 Albaek, M.O. et al. (2011) Modeling enzyme production with

Aspergillus oryzae in pilot scale vessels with different agitation,
aeration, and agitator types. Biotechnol. Bioeng. 108, 1828
1840
46 Kiviharju, K. et al. (2008) Biomass measurement online: the
performance of in situ measurements and software sensors. J. Ind.
Microbiol. Biotechnol. 35, 657665
47 Madrid, R.E. and Felice, C.J. (2005) Microbial biomass estimation.
Crit. Rev. Biotechnol. 25, 97112
48 Ferreira, A.P. et al. (2005) Evaluation of a new annular capacitance
probe for biomass monitoring in industrial pilot-scale fermentations. J.
Biotechnol. 116, 403409
49 Dhandapani, B. et al. (2012) Impact of aeration and agitation on
metabolic heat and protease secretion of Aspergillus tamarii in a
real-time biological reaction calorimeter. Appl. Microbiol. Biotechnol.
94, 15331542
50 Dhandapani, B. et al. (2012) Energetics of growth of Aspergillus tamarii
in a biological real-time reaction calorimeter. Appl. Microbiol.
Biotechnol. 93, 19271936
51 Luttmann, R. et al. (2012) Soft sensors in bioprocessing: a status report
and recommendations. Biotechnol. J. 7, 10401048
52 Gernaey, K.V. et al. (2010) Application of mechanistic models to
fermentation and biocatalysis for next-generation processes. Trends
Biotechnol. 28, 346354
53 Herwig, C. (2010) Process analytical technology in biotechnology.
Chem. Ing. Tech. 82, 405414
54 Roychoudhury, P. et al. (2006) At-line monitoring of ammonium,
glucose, methyl oleate and biomass in a complex antibiotic
fermentation process using attenuated total reflectance-mid-infrared
(ATR-MIR) spectroscopy. Anal. Chim. Acta 561, 218224
55 Pons, M.N. et al. (2004) Spectral analysis and fingerprinting for
biomedia characterisation. J. Biotechnol. 113, 211230
56 Bakeev, K. (2010) . In Process Analytical Technology: Spectroscopic
Tools and Implementation Strategies for the Chemical and
Pharmaceutical Industries (Bakeev, K.A., ed.), John Wiley
& Sons
57 Antonio, R-V.J. et al. (2006) From shake flasks to stirred fermentors:
scale-up of an extractive fermentation process for 6-pentyl-a-pyrone
production by Trichoderma harzianum using volumetric power input.
Process Biochem. 41, 13471352
58 Pollard, D.J. et al. (2006) Scale up of a viscous fungal fermentation:
application of scale-up criteria with regime analysis and operating
boundary conditions. Biotechnol. Bioeng. 96, 307317
59 Junker, B.H. et al. (2004) Early phase process scale-up challenges for
fungal and filamentous bacterial cultures. Appl. Biochem. Biotechnol.
119, 241277
60 Petersen, N. et al. (2010) In situ near infrared spectroscopy for analytespecific monitoring of glucose and ammonium in Streptomyces
coelicolor fermentations. Biotechnol. Prog. 26, 263271
61 Jose, G.E. et al. (2011) Predicting Mab product yields from cultivation
media components, using near-infrared and 2D-fluorescence
spectroscopies. Biotechnol. Prog. 27, 13391346
62 Dietzsch, C. et al. (2012) On-line multiple component analysis for
efficient quantitative bioprocess development. J. Biotechnol. http://
dx.doi.org/10.1016/j.jbiotec.2012.03.010
63 Spadiut, O. et al. (2012) Evaluating online sampling probes for
substrate concentration and protein production by a Design of
Experiments screening approach. Eng. Life Sci. 12, 507513