Review

Advances and perspectives in in vitro

human gut fermentation modeling
Amanda N. Payne*, Annina Zihler*, Christophe Chassard and Christophe Lacroix
Institute of Food, Nutrition and Health, ETH Zurich, 8092 Zurich, Switzerland

The gut microbiota is a highly specialized organ containing host-specific assemblages of microbes whereby metabolic activity directly impacts human health and disease.
In vitro gut fermentation models present an unmatched
opportunity of performing studies frequently challenged
in humans and animals owing to ethical concerns. Multidisciplinary systems biology analyses supported by omics platforms remain widely neglected in the field of
in vitro gut fermentation modeling but are key to advancing the significance of these models. Model-driven experimentation using a combination of in vitro gut
fermentation and in vitro human cell models represent
an advanced approach in identifying complex hostmicrobe interactions and niches central to gut fermentation
processes. The aim of this review is to highlight the
advances and challenges exhibited by in vitro human
gut fermentation modeling.
Introduction
For decades the contributions of gut microbiota to human
health and disease were widely underappreciated. Acknowledgement of the gut microbiota, described as an
independent organ with immunostimulatory properties,
was first noted in 1992 by Bocci [1]. A decade later Eckburg
et al. identified a suite of gut microbial contributions to the
human host such as providing nourishment, regulating
epithelial cell development and modulating innate immune responses, yet acknowledged that many basic functions of the gut microbiota remain unknown [2]. Since 2005
the majority of studies have focused on species identification and cataloguing, a process that has been hampered by
the inability to culture most gut microbes. Most species are
recognized solely by their 16S rRNA sequence and identification relies heavily on high-throughput technologies
such as second and third generation sequencing platforms
[3]. Success and advancement of these technologies has
demonstrated a link between various microbial species and
multiple pathologies such as obesity, Crohns disease and
irritable bowel syndrome [48].
16S rRNA-based cataloguing of gut microbiota via
high-throughput sequencing platforms does not, however,
provide information on the functionality of any species
identified. Unraveling the complexity of microbemicrobe
interactions and identification of niches central to gut
fermentation are, however, dependent on functional
studies. In vitro gut fermentation models represent an

Corresponding author: Lacroix, C. (christophe.lacroix@ilw.agrl.ethz.ch)

These authors contributed equally to this manuscript.

innovative technological platform consisting of multiple

model designs which permit investigations on both the
existence of gut microbial species and their related functionality. At their core, in vitro gut fermentation models are
each characterized by inoculation of single or multiple
chemostats with fecal microbiota and operated under physiological temperature, pH and anaerobic conditions [912].
The complexity of in vitro gut fermentation models increases
with the number of simultaneously operated chemostats,
creating a technological platform with the potential for
either single colon regions (e.g. proximal colon) or the entire
colon to be modeled (Figure 1). Experimentation within
these models is virtually unlimited and can include assessment of single or multiple dietary components for their
impact on individual gut microbial populations and subsequent changes to their metabolisms as a function of dietary
modulation, evaluation of the impact of potentially new
probiotic strains on gut microbiota function and survival
or mechanistic studies on commensalpathogen interactions [1316]. Other types of applications include assessment studies where the potential lethality restricts in vivo
testing such as the biotransformation of drugs or toxic
compounds (e.g. polyaromatic hydrocarbons, PAHs) or the
impact of radioactivity on the gut microbiota [1721].
Results obtained using in vitro gut fermentation models
describe the functional, yet uncoupled response of the gut
microbiota as the effects of host community modulation as a
function of immune response mechanisms are not simulated
in these models.
The use of in vitro human intestinal cell models is widely
accepted for evaluating the mechanistic effects of probiotics
or drug absorption and transport [22,23], suggesting their
application could be extended to include hostgut microbe
functional studies. In vitro human cell models are typically
composed of polarized and differentiated monolayers of
single or coculture cell lines and are distinguished based
on their physiological properties with absorptive Caco-2 and
mucus secreting HT29-MTX cell lines predominating
[24,25]. The unique physiological properties of each human
intestinal cell line require that their selection be entirely
based on the intended in vitro application. The combination
of in vitro human intestinal cell models with in vitro gut
fermentation models creates an advanced in vitro model
system where samples obtained from host response lacking
in vitro gut models can be directly applied to monolayer cell
models for host function assessment [2628].
Despite advances in in vitro model development, in vivo
studies resist absolute replacement. Animal studies persist as a substitute to human studies, yet the relevance of

0167-7799/$ see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.tibtech.2011.06.011 Trends in Biotechnology, January 2012, Vol. 30, No. 1

17

(Figure_1)TD$IG][ Review

Trends in Biotechnology January 2012, Vol. 30, No. 1

Batch
Key:
Nutrients

Concentration

Filling

Emptying

Metabolites
Planktonic
gut microbes

pH 5.6 5.9

Time (hours)

Inoculation

Continuous

pH 5.6 5.9
proximal

R2

pH 6.1 6.4
transverse

R3

R1

Concentration

R1

pH 6.6 6.9
distal

R2

R3

Steady-state

Time (days)

Inoculation

Continuous with immobilized feces

pH 5.6 5.9
proximal

R2

pH 6.1 6.4
transverse

R3

R1
Concentration

R1

pH 6.6 6.9
distal

R3

Steady-state

Inoculation

Released bacteria

(a)

R2

Time (weeks)

(b)

1-2 mm

1 m

TRENDS in Biotechnology

Figure 1. Types and process characteristics of in vitro colonic fermentation models simulating proximal (R1), transverse (R2) and distal (R3) colon regions, operated at
physiological section-specific constant pH [41], temperature (37 8C) and under strictly anaerobic conditions (e.g. through continuous CO2 or N2 flushing of the headspace).
Initial (R1) and steady-state (R1, R2 and R3) fermentation profiles. (a) Picture detail of a proximal colon reactor containing polysaccharide beads with immobilized fecal
microbiota. (b) Electron microscope image of microbes embedded and attached to the surface of an intestinal bead (the authors acknowledge support of Stefan Handschin
from EMEZ Electron Microscopy at ETH Zurich, Switzerland).

data obtained through animal studies is often questioned

due to physiological differences between animal and human metabolisms [29,30]. Functional studies in humans
remain difficult to perform owing to social and ethical
considerations governing invasive medical procedures
necessary in accessing the human large intestine, rendering human studies primarily limited to fecal sample analyses [2,31]. An ultimate approach to investigating gut
microbiota functionality must therefore include a combination of in vitro and in vivo models, each yielding complementary results which not only strengthen the overall
validity of each individual model but also distinguish
between the functionality of gut microbial and human
processes (Figure 2). The use of -omics technologies
18

and multidisciplinary systems biology analyses must be

applied to this combinatorial approach to fully realize the
potential of obtaining novel results with in vitro models
(Figure 2). Systems biology approaches using -omics
technologies are commonly used in in vivo and in situ
studies but remain elusive in the field of in vitro human
gut fermentation modeling. The aim of this review is to
highlight the potential and perspective advances exhibited by in vitro gut fermentation models and to discuss the
limitations and challenges facing in vitro technologies.
Diversity and functionality of human gut microbiota
An estimated 1014 intestinal microbes belonging to >1000
different species [32] are distributed along the human

(Figure_2)TD$IG][ Review

Trends in Biotechnology January 2012, Vol. 30, No. 1

Human gut systems biology

In vitro gut
fermentation
modeling

In vitro human
intestinal cell
model

In vivo animal
studies

In vivo human
studies

Traditional systems biology

Metagenomics
Proteomics
Metabolomics
Transcriptomics
TRENDS in Biotechnology

Figure 2. Human gut systems biology approach for functional analysis of gut microbiota.

intestinal tract with the highest densities reached in the

colon. Environmental and physicochemical conditions in
addition to gender and age dictate the considerable variations in both abundance and composition of the gut microbiota along the gastrointestinal (GI) tract [33,34]. Bacterial
counts in the stomach are usually less than 103 per gram
due to acidic pH. Numbers increase to 104107 per gram in
the small intestine where rapid transit and the secretion of
bile and pancreatic juice limit growth. The large intestine
is the area most populated due to favorable conditions
including slow gut transit times, nutrient availability
and a favorable pH [35]. Up to 1012 microbes per gram
inhabit the colon, accounting for a total of 12 kg of our
body weight. Over 50 bacterial phyla have been identified,
but the human-associated microbiota is dominated by only
four main phyla: Firmicutes, Bacteroidetes, Actinobacteria
and Proteobacteria [36,37]. The inclusion of Clostridia
within the Firmicutes phylum makes it the most abundant
phylum, constituting 4658% of total bacteria [38,39]. Other
major inhabitants include the BacteroidesPrevotella group
(1030%), Bifidobacterium (4.44.8%), Enterobacteriaceae
(0.10.2%) and Lactobacillus and Enterococcus (both
0.11.8%) [39,40]. Despite interhost differences in population compositions, the existence of a core gut microbiota with
presumably pivotal functions in maintaining gut homeostasis was reported recently [41]. Two different core phylotypes
belonging to the family Lachnospiraceae (core 1) and Dorea
(core 2) of the Clostridium cluster XIVa were found to
predominate in 16S rRNA sequences from a total of 231
individuals from different ethnic backgrounds, consuming
distinctly different diets [4245].
Intestinal metabolism involves a variety of enzymes
and biochemical pathways encoded by specialized members of the microbial community and are distinct from host
supplied resources. This metabolic capacity is estimated
to be encoded on a total of 24 million microbial genes

which is 70140 times greater than that of the host [46,47].

Initial digestion processes in the human intestine
require specific microbial glycosidases encoded by the
hydrolytic gut microbiota. These microbes hydrolyze
the otherwise indigestible polysaccharides (e.g. starch,
resistant starch and both soluble and insoluble dietary
fiber) and proteins (Figure 3) or host-innate mucins (glycosylated high molecular weight proteins) reaching the
large intestine. Subsequent microbial fermentation of
hydrolyzed polysaccharides in the large intestine results
in production of short chain fatty acids (SCFAs), such as
acetate, propionate and butyrate, branched chain fatty
acids, as well as lactate, formate, ethanol and mixed gases
(e.g. CH4, CO2 and H2) [48]. Dissimilatory metabolism of
proteinaceous material produces ammonia, amines, mercaptans and H2S, as well as some toxic indolic and phenolic
compounds [49]. SCFA production resulting from microbial fermentation of hydrolyzed dietary starches, fibers
and sugars provides an estimated 10% daily dietary energy to the host [5052].
Resident commensal bacteria are further responsible for
creating a powerful line of resistance to colonization by
exogenous microbes. Adherent intestinal microbes constantly compete with exogenous microbes for attachment
sites in the brush border of intestinal epithelial cells, preventing pathogenic invasion and translocation into colonic
tissue [53]. Colonization resistance requires a delicate ecological balance to avert the overgrowth of acquired opportunistic bacteria such as Clostridium difficile and prevent
the pathogenesis of these strains [54]. Optimized nutrient
consumption and metabolism by resident gut microbiota
further attenuates pathogen proliferation as resident
microbes are capable of adjusting their nutrient requirements and metabolic activities causing pathogen starvation
[55]. Direct inhibition of pathogens (e.g. Listeria monocytogenes) might also be mediated by antimicrobial compounds
19

(Figure_3)TD$IG][ Review

Trends in Biotechnology January 2012, Vol. 30, No. 1

Transverse colon
1011-1012 cells g-1
Depletion of substrates
Reduction in bacterial activity
Decreasing SCFA production
pH 6.1 - 6.4

Proximal colon
1010-1011 cells g-1

Stomach
<103
cells g-1

Distal colon
1012 cells g-1

Small intestine
104-107 cells g-1

Low substrate availability

Slow bacterial growth

High substrate concentration

Active bacterial growth

Protein fermentation
Low SCFA production

Carbohydrate fermentation

pH 6.1 - 6.9

High SCFA production

pH 5.4 - 5.9

TRENDS in Biotechnology

Figure 3. Schematic representation of the human gastrointestinal tract: the amount of bacteria per gram of intestinal contents, nutrient availability and bacterial
fermentative activity typically found in different sections of healthy individuals.

such as bacteriocin Abp118 produced by the lactic acid

bacterium Lactobacillus salivarius UCC118 [56].
In vitro fermentation models
The design and complexity of in vitro fermentation systems
has broadened from simple batch cultures to single- or
multistage continuous flow models using a variety of different fecal inoculation techniques [10,17,57,58]. In vitro
gut fermentation models enable the stable cultivation of a
complete intestinal microbiota for a defined and modelspecific period of time. Selection of the appropriate model
requires careful evaluation of the study objectives given
the advantages and limitations exhibited by each type of
system (Table 1).
Batch fermentation models
Batch fermentation describes the growth of a pure or mixed
bacterial suspension in a carefully selected medium without the further addition of nutrients. These models are
generally closed systems using sealed bottles or reactors
containing suspensions of fecal material which are maintained under anaerobic conditions. Dietary components

such as inulin-type fructans or resistant starch and other

complex carbohydrates have been investigated for their
potential use as prebiotics using batch fermentation models [59,60]. Batch fermentations are particularly useful for
investigating metabolic profiles of SCFAs arising from
active metabolism of dietary compounds by gut microbiota
as these systems are generally convenient and inexpensive, allowing large numbers of substrates and/or fecal
samples to be tested quickly [17,61]. Microbial fermentation resulting in production of SCFAs results in continuous
changes to pH and redox potential as most batch fermentations proceed in the absence of pH control. Microbial
growth in these systems is dependent on the inoculation
density and substrate depletion rate. Systems inoculated
with low cell densities demonstrate typical S-shaped
growth curve characteristics owing to an initial abundant
nutrient supply, which as it becomes depleted over time
and together with the accumulation of toxic products,
results in arrested growth (Figure 1). By contrast, systems
inoculated with cell densities similar to the very high
cell densities of the colon result in limited growth. Complex
experimentation such as the evaluation of microbial

Table 1. Advantages and limitations of in vitro fermentation models

Models
Batch culture

Advantages
Easy to set up, useful for fermentation studies and especially
substrate digestion assessment

Continuous culture

Continuous flow mimicking conditions found in vivo.

Environmental parameters are well controlled

Multistage continuous culture

Continuous flow into several vessels mimicking conditions

found in portions of the digestive tract

Immobilized continuous culture

High-cell density and long-term stability of continuous

fermentation system with immobilized fecal microbiota
Continuous flow with metabolites and water exchange
mimicking conditions found in vivo

Artificial digestive system

20

Limitations
Short-term fermentation
studies and weakness in
microbiological control
No host functionality and
experiments are time
limited (days or weeks)
No host functionality and
experiments are time
limited (days or weeks)
No host functionality

Refs
[59,61]

No immune and
neuroendocrine response
and experiments are
limited to few days time

[71,81]

[95,97]

[58,95]

[13,14]

Review
community and microbial metabolic modulation requires
continuous substrate replenishment as substrate depletion restricts the operational time of batch fermentations
to several hours and prevents the establishment of steadystate conditions in vitro.
Continuous fermentation models
Continuous culture fermentation models exist as either
single- or multistage systems and are necessary for performing long-term studies as substrate replenishment and
toxic product removal are facilitated. Single-stage continuous fermentation models are often used to elucidate
proximal colon function and metabolic activity as the
mixing of digesta from both the caecum and ascending
colon is well simulated in these models [17]. Single-stage
continuous fermentation models have been used to investigate the mechanism of Salmonella infection in children
as well as colonization of the infant gut [13,57].
Human colonic function occurs, however, over the span
of the ascending, transverse and descending colonic
regions and differences in metabolic activity and microbial
communities have been horizontally demonstrated for
each colonic region [62]. A major advancement of in vitro
gut fermentation models was the development of multistage continuous fermentation models which enable the
simulation of horizontal colon processes [9]. This design
facilitates the spatial, temporal, nutritional and physicochemical properties of the gut microbiota by combining
three chemostats connected in series, replicating the proximal (R1), transverse (R2) and distal (R3) colon regions
(Figure 1) [9]. Adaptation, survival and proliferation of an
in vivo acquired human gut microbiota to in vitro continuous fermentation models are dependent on environmental
parameters such as pH, retention time, temperature, anaerobiosis and flow rate of the medium. Strict control of
these factors facilitates establishment of steady-state conditions regarding both microbial composition and metabolic activity, with these steady-state conditions being
required to create a reproducible system (Figure 1). A wide
range of studies and novel results on gut microbial community modulation and metabolic function have been
obtained using the various multistage continuous fermentation model designs, evidence of the dynamic capacity of
this technology [9,10,14,58]. Probiotic and prebiotic modulation and impact investigations represent the most commonly performed experimentations using three-stage
continuous in vitro fermentation models [63].
Perhaps the major discriminating factor between the
different in vitro continuous fermentation systems is the
technique used for fecal inoculation. Operation of most in
vitro systems uses a liquid fecal suspension as inoculum,
resulting in several limitations due to the free-cell state of
the bacterial populations [57]. Systems with liquid fecal
inocula generally experience a rapid washout of less competitive bacteria and are consequently limited in operational time to less than 4 weeks [9,64,65]. These systems
also struggle in reproducing both the planktonic (free-cell)
and sessile (biofilm-associated) states of bacterial populations in the colon [17,66]. To address problems associated
with inoculum washout, a process for the immobilization of
fecal microbiota has been developed [10,14,57]. Here, fecal

Trends in Biotechnology January 2012, Vol. 30, No. 1

microbiota are suspended within a porous polysaccharide

matrix resulting in the formation of fecal beads, which are
transferred to the growth medium in R1 of a multistage
continuous fermentation model (Figure 1). Limitations on
substrate and toxic product diffusion within beads result in
formation of a high-cell density peripheral layer, where cell
release occurs spontaneously as a result of active cell
growth [57,6769]. The released cells are transported to
R2 and then R3, resulting in a self-contained continuous
fermentation system of very high-cell density and population stability, close to the human GI tract (Figure 1). The
operational time of systems for studies using immobilized
cells varies and has been functionally demonstrated for
time frames of 29, 54 and up to 71 days [10,57,70].
Artificial digestive systems
The inclusion of host digestive functions in vitro coupled
with multistage continuous fermentation modeling represents the most advanced attempt thus far at simulating
interdependent physiological functions within the human
gut, stomach lumen and small intestine (Table 1) [71].
Human digestive functions that are reproduced in the
TIM-1 small intestinal model include bile secretion, motility, pH and absorption capacities of the upper intestine
[72]. Proximal colon simulator models such as TIM-2 include other host functions such as peristaltic mixing, water
and metabolite absorption [11]. The combination of both
TIM-1 and TIM-2 models have led to the creation of an
artificial digestive system which has been applied to pharmaceutical investigations of drug delivery as well as for
advanced nutritional studies [71,73,74]. The SHIME model (simulator of the human intestinal microbial ecosystem)
comprises a series of five fermentation vessels in series,
operated in sequential batch mode, with reactors 1 and 2
simulating the process of ingestion and digestion occurring
in the duodenum/jejunum and ileum and are connected to
the three-stage large intestinal model [12]. The role of
microbial metabolism in increasing the bioavailability of
various toxic compounds such as PAHs or arsenic has been
demonstrated using the SHIME model [20,75]. Host function is, however, only partially reproduced as immunomodulatory and neuroendocrine responses are still lacking.
The preparation of fecal samples used for inoculation
varies considerably in these models and should be thoroughly considered during the experimental design. For
example, an attempt to standardize the inoculum by mixing fecal samples from multiple donors resulted in propagation of a microbiota in TIM-2 with a microbial profile
distinct from that normally witnessed in healthy adults,
demonstrating the difficulty in establishing a representative human gut microbiota in vitro [76].
Challenges of in vitro gut fermentation models
Inoculation and colonization of in vitro gut fermentation
models
Despite the many different designs and combinations of in
vitro gut fermentation models currently in existence, several challenges and limitations are exhibited by this technology. The reproducibility and functional stability of
human gut microbiota in in vitro gut fermentation models
is frequently challenged and the subject of much criticism.
21

Review
Box 1. Challenges and perspectives of in vitro fermentation
models
Permanent challenges
Enhance and evaluate microbial ecosystem stability.
Carefully select models depending on their advantages and
limitations related to the research question.
Make the right compromise between technical complexity and
biological significance.
Assess reproducibility and define repeatability needs.
Perspectives
Assess functional stability of the in vitro microbiome.
Application of omics technologies to in vitro gut fermentation
modeling.
Modeling of diseased microbiota.
Combination of in vitro models.

The robustness of each model depends on the certainty that

any observed effects on microbial composition and metabolic activity are due solely to the applied experimental
treatment and not to gut microbial adaptation to the
simulated gut environment (Box 1). This presumes that
a compositional stability of the in vitro gut ecosystem is in
place prior to applying any experimental treatment and
that this pseudo-steady state can reproducibly generate
identical microbial communities which are the basis for
comparison with other treatments [58].
Inoculation and colonization of the system, however,
results in a newly balanced gut microbiota which is a
result of both environmental factors and the initial qualitative diversity, but not initial quantitative balance of the
fecal inoculum. Ratios of bacterial populations in model
samples are frequently different from those tested in the
fecal inoculum, as recently demonstrated for the TIM-2
and SHIME models using the Human Intestinal Tract
Chip (HITChip) [58,76]. Assuring a strict repetitive process
of fecal sample collection and processing is essential in
obtaining quality fecal inocula as the freshness and quality
of the sample ultimately dictate the colonization efficiency.
These changes in population ratios reflect applied fermentation conditions (e.g. retention time, culture medium, pH,
etc.) which can never be an absolute simulation of the
conditions encountered in the host intestine. A frequent
misconception of in vitro gut fermentation models is that
they aim to be exact 1:1 replicates of the in vivo intestinal
environment; however, the inability to simulate major host
functions remains a major limitation of these models. As
such, the differences in population ratios should not be
perceived as experimental bias as the main objective of in
vitro gut fermentation modeling is to stably cultivate a
complex intestinal microbiota under controlled environmental conditions for experimental modulation.
Biological replication during in vitro gut fermentation
modeling
The lack of true biological replication in in vitro gut fermentation modeling is a justified concern. True biological
replication of complex continuous models (e.g. TIM-2,
SHIME, multistage models with immobilized fecal microbiota) would involve consecutive testing to account for all
error sources (e.g. medium preparation, time, etc.) as
opposed to operating two parallel systems inoculated with
22

Trends in Biotechnology January 2012, Vol. 30, No. 1

the same microbiota. Realization of this concept is, however,

extremely difficult as the microbiota is subject to temporal
modification during consecutive sampling. Replication can
also be performed with multiple donors tested in separate
fermentation systems to assess host-specific microbiota
effects but requires significant human and monetary
resources to perform under continuous fermentation conditions which require long stabilization periods. Fecal samples pooled from different donors might also suffice the
concerns regarding biological replication but could lead to
intermicrobe interactions, favoring growth of certain microbial populations and establishing an unrepresentative human microbial ecosystem as previously demonstrated [76].
The design of an in vitro fermentation model to adeptly
facilitate the demanded degree of biological replication
remains a major challenge in this field.
Perspectives and advances of in vitro gut fermentation
technology
-Omics analyses of functionally stable in vitro gut
fermentation models
Top-down and bottom-up systems biology approaches have
successfully deciphered complex keystone interactions and
metabolic pathways both in vivo and in situ but remain
elusive to in vitro gut fermentation studies [7780]. Application of -omics technologies to multistage or other advanced in vitro gut fermentation modeling has, to date,
been restricted to phylogenetic analyses using microarray
technology or RNA-based stable isotope probing (SIP)
[58,76,81]. SIP is emerging as a useful technology for tracking metabolic fluxes of 13C-labeled substrates as they
become integrated in metabolically active microbes
[82,83]. Metagenomic studies performed in vivo have catalogued the genetic potential of select human gut microbiota
under various experimental conditions, demonstrating the
capacity for -omics technologies in human gut microbiota
research [84,85]. Although metagenomic analyses provide
inventories of genetic potential, it cannot directly link
microbial coding capacity to microbial functionality. The
revolution in metatranscriptomics in environmental microbial ecology has increased the ability to identify functional
changes in gene expression [78,79,86]. The half-life of
microbial mRNAs, relatively low mRNA yield in comparison
to more abundant and stable rRNAs and requirement of
orthologous primer sequences have challenged reverse
transcription PCR (RT-PCR) analyses of cDNA libraries
[86,87]. Direct application of high-throughput sequencing
technology to microbial transcripts has circumvented many
of these problems and transformed the field of microbial
metatranscriptomics [88]. A predictable relationship
between mRNA abundance and protein activity cannot be
unequivocally demonstrated by metatranscriptomics, as
factors such as post-translational regulation remain unaccounted for. Protein extraction, separation, purification
and identification from complex biological samples remain
impedance to unraveling the metaproteome associated
with human gut microbiota function; nevertheless, studies
identifying the relationship and distribution status of the
predicted metaproteome are becoming more common
[89,90]. Metabolic and cellular processes, giving rise to a
multitude of metabolites which might be assigned to the

Review
functionality of certain populations of gut microbiota or
assigned to a particular function of interest, might be elucidated by integrating metabolomic technologies [91,92].
Polyphenol metabolism by human gut microbiota is
just one example of the enormous promise top-down and
bottom-up systems biology approaches hold in discerning
the complexity of microbiomemetabolomic interactions
[80]. Given the recent advances in meta technologies, perhaps the greatest challenge rests with the biocomputational
and bioinformatic community in developing correctly annotated gene databases, facilitating faster, easier analysis of
large datasets arising from high-throughput sequencing
and metabolomic profiling [93].
Perspective applications of in vitro model systems
Elucidating the functionality of the coding capacity of
human gut microbiota modulated during in vitro gut fermentation modeling and the metabolic profiles generated
as a result requires the application of -omics based technologies. The use of a multidisciplinary systems biology
approach in combination with -omics platforms as outlined in Figure 2 will facilitate the most advanced system
for unraveling the complex microbial and host factors
governing human gut microbiota functionality. As mentioned in the previous section, the inability to sufficiently
simulate host responses in in vitro gut fermentation models remains a major limitation of this technology. However,
host response factors can still be assessed in vitro through
a combinatorial model system using both in vitro gut and in
vitro human intestinal cell models. A recent study assessing adhesion and cytokine expression of Caco-2 cells using
bacterial cultures obtained from a three-stage continuous
fermentation model demonstrates this concept [27]. In
another approach, an in vitro small intestinal model was
combined with Caco-2 cells [26]. However, small intestinal
in vitro models are devoid of intestinal microbes, designed
to replicate only digestion and absorption processes and
are thus incapable of providing information on intestinal
fermentation processes. Another type of combined model
approach utilized supernatants obtained from a batch
fermentation of inulin for application onto human colonic
biopsies [94]. The difficulty and ethical barriers involved in
regularly obtaining biopsies for experimentation makes
the use of in vitro cultivated human cells a more reliable
approach. Results using this combined in vitro approach
indicate that the application of fermentation effluents to in
vitro human intestinal cell models is a feasible platform for
assessing epithelial responses as a result of in vitro fermentative processes.
Future in vitro gut fermentation model systems include
but are not limited to the establishment and maintenance
of diseased microbiota. The development of such models
would allow investigations on the functional role of gut
microbiota in inflammatory bowel diseases such as Crohns
disease and ulcerative colitis. Investigation of the etiopathologies of these diseases has been, to date, heavily
centered on immune system function based studies owing
to the complexity and lack of knowledge regarding additional contributing factors [95,96]. Functional studies
performed in vitro could potentially advance the understanding of these diseases beyond the immune system.

Trends in Biotechnology January 2012, Vol. 30, No. 1

Although the cause versus consequence of the gut microbiota in many of the aforementioned diseases remains
heavily debated, in vivo studies alone have, to date, been
unsuccessful at fully elucidating the mechanisms of these
diseases and make a strong case for using a combined
human gut systems biology approach.
Concluding remarks
In vitro fermentation models are an innovative technological platform where the greatest advantages are exhibited
by the virtually limitless experimental capacity as experimentation is not restricted by ethical concerns. Host intestinal function is only partially simulated in some model
designs (e.g. TIM-1 and TIM-2) and together with microbial population balancing remains a major challenge of in
vitro gut fermentation modeling. Host functionality, however, does not need to be completely ignored during such
experimentation. The use of a combined in vitro model
system comprising both in vitro gut and human intestinal
cell models has the ability to provide novel results on host
responses to in vitro fermentation processes. As in vivo
results remain the most coveted, an ultimate approach to
investigating gut microbiota functionality must include a
combination of complimentary in vitro and in vivo models,
each yielding complementary results which not only
strengthen the overall validity of each individual model
but also distinguish between the functionality of gut microbial and human processes.
References
1 Bocci, V. (1992) The neglected organ bacterial-flora has a crucial
immunostimulatory role. Perspect. Biol. Med. 35, 251260
2 Eckburg, P.B. et al. (2005) Diversity of the human intestinal microbial
flora. Science 308, 16351638
3 Zoetendal, E.G. et al. (2008) High-throughput diversity and
functionality analysis of the gastrointestinal tract microbiota. Gut
57, 16051615
4 Ley, R.E. et al. (2005) Obesity alters gut microbial ecology. Proc. Natl.
Acad. Sci. U.S.A. 102, 1107011075
5 Turnbaugh, P.J. et al. (2006) An obesity-associated gut microbiome
with increased capacity for energy harvest. Nature 444, 1027
1031
6 Manichanh, C. et al. (2006) Reduced diversity of faecal microbiota
in Crohns disease revealed by a metagenomic approach. Gut 55,
205211
7 Malinen, E. et al. (2005) Analysis of the fecal microbiota of irritable
bowel syndrome patients and healthy controls with real-time PCR. Am.
J. Gastroenterol. 100, 373382
8 Bischoff, S.C. (2011) Gut health: a new objective in medicine? BMC
Med. 9, 24
9 Macfarlane, G.T. et al. (1998) Validation of a three-stage compound
continuous culture system for investigating the effect of retention time
on the ecology and metabolism of bacteria in the human colon. Microb.
Ecol. 35, 180187
10 Cinquin, C. et al. (2006) New three-stage in vitro model for infant
colonic fermentation with immobilized fecal microbiota. FEMS
Microbiol. Ecol. 57, 324336
11 Minekus, M. et al. (1999) A computer-controlled system to simulate
conditions of the large intestine with peristaltic mixing, water
absorption and absorption of fermentation products. Appl. Microbiol.
Biotechnol. 53, 108114
12 Molly, K. et al. (1993) Development of a 5-step multi-chamber reactor
as a simulation of the human intestinal microbial ecosystem. Appl.
Microbiol. Biotechnol. 39, 254258
13 Le Blay, G. et al. (2009) New in vitro colonic fermentation model for
Salmonella infection in the child gut. FEMS Microbiol. Ecol. 67, 198
207
23

Review
14 Zihler, A. et al. (2010) Unexpected consequences of administering
bacteriocinogenic probiotic strains for Salmonella populations,
revealed by an in vitro colonic model of the child gut. Microbiology
156, 33423353
15 Kheadr, E. et al. (2010) Study of the physicochemical and biological
stability of pediocin PA-1 in the upper gastrointestinal tract conditions
using a dynamic in vitro model. J. Appl. Microbiol. 109, 5464
16 Gagnon, M. et al. (2011) Ecology of probiotics and enteric protection, In
Probiotic and Enteric Infections Cryoprotection by Probiotic Bacteria (1st
edn) (Malago, J.J. et al., eds), pp. 6585, Springer Science+Business
Media BV
17 Macfarlane, G.T. and Macfarlane, S. (2007) Models for intestinal
fermentation: association between food components, delivery
systems, bioavailability and functional interactions in the gut. Curr.
Opin. Biotechnol. 18, 156162
18 Coutts, T.M. et al. (1987) Use of continuous culture to study the
gastric microflora of a hypochlorhydric patient. Toxicol. in vitro 1,
1721
19 Bearne, C.A. et al. (1990) Continuous culture of human fecal bacteria as
an in vitro model for the colonic microflora. Toxicol. in vitro 4, 522525
20 Van de Wiele, T. et al. (2003) Oral exposure to PAH: bioactivation
processes in the human gut. Commun. Agric. Appl. Biol. Sci. 68, 36
21 Crawford, P. and Gordon, J.I. (2005) Microbial regulation of intestinal
radiosensitivity. Proc. Natl. Acad. Sci. U.S.A. 102, 1325413259
22 Mennigen, R. and Bruewer, M. (2009) Effect of probiotics on intestinal
barrier function. Ann. N.Y. Acad. Sci. 1165, 183189
23 Meunier, V. et al. (1995) The human intestinal epithelial cell line Caco2 pharmacological and pharmacokinetic applications. Cell Biol.
Toxicol. 11, 187194
24 Mathias, A. et al. (2010) Potentiation of polarized intestinal Caco-2 cell
responsiveness to probiotics complexed with secretory IgA. J. Biol.
Chem. 285, 3390633913
25 Vieira, M.A.M. et al. (2010) Two atypical enteropathogenic Escherichia
coli strains induce the production of secreted and membrane-bound
mucins to benefit their own growth at the apical surface of human
mucin-secreting intestinal HT29-MTX cells. Infect. Immun. 78, 927
938
26 Deat, E. et al. (2009) Combining the dynamic TNO-gastrointestinal
tract system with a Caco-2 cell culture model: application to the
assessment of lycopene and alpha-tocopherol bioavailability from a
whole food. J. Agric. Food Chem. 57, 1131411320
27 Bahrami, B. et al. (2011) Adherence and cytokine induction in
Caco-2 cells by bacterial populations from a three-stage
continuous-culture model of the large intestine. Appl. Environ.
Microbiol. 77, 29342942
28 Zihler, A. (2010) In vitro assessment of bacteriocinogenic probiotics for
prevention and treatment of Salmonella in children using novel in vitro
continuous colonic fermentation and cellular models, Dissertation,
ETH Zurich, 19059
29 Hegsted, D.M. (1975) Relevance of animal studies to human disease.
Cancer Res. 35, 35373539
30 Cohen, S.M. (1995) Human relevance of animal carcinogenicity
studies. Regul. Toxicol. Pharm. 21, 7580
31 Zoetendal, E.G. et al. (1998) Temperature-gradient gel electrophoresis
analysis of 16S rRNA from human fecal samples reveals stable and
host-specific communities of active bacteria. Appl. Environ. Microb. 64,
38543859
32 Rajilic-Stojanovic, M. et al. (2007) Diversity of the human
gastrointestinal tract microbiota revisited. Environ. Microbiol. 9,
21252136
33 Leser, T.D. and Molbak, L. (2009) Better living through microbial
action: the benefits of the mammalian gastrointestinal microbiota on
the host. Environ. Microbiol. 11, 21942206
34 Mueller, S. et al. (2006) Differences in fecal microbiota in different
European study populations in relation to age, gender, and country: a
cross-sectional study. Appl. Environ. Microb. 72, 10271033
35 Gibson, G.R. and Roberfroid, M.B. (1995) Dietary modulation of the
human colonic microbiota introducing the concept of prebiotics. J.
Nutr. 125, 14011412
36 Dethlefsen, L. et al. (2006) Assembly of the human intestinal
microbiota. Trends Ecol. Evol. 21, 517523
37 Dethlefsen, L. et al. (2007) An ecological and evolutionary perspective
on human-microbe mutualism and disease. Nature 449, 811818

24

Trends in Biotechnology January 2012, Vol. 30, No. 1

38 Harmsen, H.J.M. et al. (2002) Extensive set of 16S rRNA-based probes
for detection of bacteria in human feces. Appl. Environ. Microb. 68,
29822990
39 Lay, C. et al. (2005) Colonic microbiota signatures across five northern
European countries. Appl. Environ. Microbiol. 71, 41534155
40 Zoetendal, E.G. et al. (2006) A microbial world within us. Mol.
Microbiol. 59, 16391650
41 Turnbaugh, P.J. et al. (2010) Organismal, genetic, and transcriptional
variation in the deeply sequenced gut microbiomes of identical twins.
Proc. Natl. Acad. Sci. U.S.A. 107, 75037508
42 Rajilic-Stojanovic, M. et al. (2009) Development and application of the
human intestinal tract chip, a phylogenetic microarray: analysis of
universally conserved phylotypes in the abundant microbiota of young
and elderly adults. Environ. Microbiol. 11, 17361751
43 Claesson, M.J. et al. (2009) Comparative analysis of pyrosequencing
and a phylogenetic microarray for exploring microbial community
structures in the human distal intestine. PLoS ONE 4, e6669
44 Tap, J. et al. (2009) Towards the human intestinal microbiota
phylogenetic core. Environ. Microbiol. 11, 25742584
45 Sekelja, M. et al. (2011) Unveiling an abundant core microbiota in the
human adult colon by a phylogroup-independent searching approach.
ISME J. 5, 519531
46 Possemiers, S. et al. (2011) The intestinal microbiome: a separate organ
inside the body with the metabolic potential to influence the bioactivity
of botanicals. Fitoterapia 82, 5366
47 Qin, J. et al. (2010) A human gut microbial gene catalogue established
by metagenomic sequencing. Nature 464, 5965
48 Robert, C. and Bernalier-Donadille, A. (2003) The cellulolytic microflora
of the human colon: evidence of microcrystalline cellulose-degrading
bacteria in methane-excreting subjects. FEMS Microbiol. Ecol. 46, 8189
49 Smith, E.A. and Macfarlane, G.T. (1996) Enumeration of human
colonic bacteria producing phenolic and indolic compounds: effects of
pH, carbohydrate availability and retention time on dissimilatory
aromatic amino acid metabolism. J. Appl. Bacteriol. 81, 288302
50 McNeil, N.I. (1984) The contribution of the large intestine to energy
supplies in man. Am. J. Clin. Nutr. 39, 338342
51 Scheppach, W. (1994) Effects of short chain fatty acids on gut
morphology and function. Gut 35, S35S38
52 Backhed, F. et al. (2004) The gut microbiota as an environmental factor
that regulates fat storage. Proc. Natl. Acad. Sci. U.S.A. 101, 15718
15723
53 Bernet, M.F. et al. (1994) Lactobacillus acidophilus LA 1 binds to
cultured human intestinal cell lines and inhibits cell attachment
and cell invasion by enterovirulent bacteria. Gut 35, 483489
54 Poutanen, S.M. and Simor, A.E. (2004) Clostridium difficile-associated
diarrhea in adults. CMAJ 171, 5158
55 Hooper, L.V. et al. (1999) A molecular sensor that allows a gut
commensal to control its nutrient foundation in a competitive
ecosystem. Proc. Natl. Acad. Sci. U.S.A. 96, 98339838
56 Corr, S.C. et al. (2007) Bacteriocin production as a mechanism for the
antiinfective activity of Lactobacillus salivarius UCC118. Proc. Natl.
Acad. Sci. U.S.A. 104, 76177621
57 Cinquin, C. et al. (2004) Immobilization of infant fecal microbiota and
utilization in an in vitro colonic fermentation model. Microb. Ecol. 48,
128138
58 Van den Abbeele, P. et al. (2010) Microbial community development in
a dynamic gut model is reproducible, colon region specific, and selective
for Bacteroidetes and Clostridium cluster IX. Appl. Environ. Microbiol.
76, 52375246
59 Pompei, A. et al. (2008) In vitro comparison of the prebiotic effects of
two inulin-type fructans. Anaerobe 14, 280286
60 Lesmes, U. et al. (2008) Effects of resistant starch type III polymorphs
on human colon microbiota and short chain fatty acids in human gut
models. J. Agric. Food Chem. 56, 54155421
61 Gumienna, M. et al. (2011) Bioconversion of grape and chokeberry wine
polyphenols during simulated gastrointestinal in vitro digestion. Int. J.
Food Sci. Nutr. 62, 226233
62 Macfarlane, G.T. et al. (1992) Comparison of fermentation reactions in
different regions of the human colon. J. Appl. Bacteriol. 72, 5764
63 McBain, A.J. and Macfarlane, G.T. (1997) Investigations of
bifidobacterial ecology and oligosaccharide metabolism in a threestage compound continuous culture system. Scand. J. Gastroenterol.
Suppl. 222, 3240

Review
64 Sghir, A. et al. (1998) Continuous culture selection of bifidobacteria and
lactobacilli from human faecal samples using fructooligosaccharide as
selective substrate. J. Appl. Microbiol. 85, 769777
65 De Boever, P. et al. (2001) Development of a six-stage culture system for
simulating the gastrointestinal microbiota of weaned infants. Microb.
Ecol. Health Dis. 13, 111123
66 Macfarlane, S. and Dillon, J.F. (2007) Microbial biofilms in the human
gastrointestinal tract. J. Appl. Microbiol. 102, 11871196
67 Doleyres, Y. et al. (2002) Quantitative determination of the spatial
distribution of pure- and mixed-strain immobilized cells in gel beads by
immunofluorescence. Appl. Microbiol. Biotechnol. 59, 297302
68 Doleyres, Y. et al. (2002) Bifidobacterium longum ATCC 15707 cell
production during free- and immobilized-cell cultures in MRS-whey
permeate medium. Appl. Microbiol. Biotechnol. 60, 168173
69 Lacroix, C. et al. (2003) Immobilized cell technology for the
dairy industry. In Cell Imobilization Biotechnology (Nevidovic, V.
and Willaert, R., eds), Kluwer Academic Press
70 Le Blay, G. et al. (2010) Set up of a new in vitro model to study dietary
fructans fermentation in formula-fed babies. Br. J. Nutr. 103, 403411
71 Blanquet-Diot, S. et al. (2009) Digestive stability of xanthophylls
exceeds that of carotenes as studied in a dynamic in vitro
gastrointestinal system. J. Nutr. 139, 876883
72 Minekus, M. et al. (1995) A multi compartmental dynamic computercontrolled model simulating the stomach and small intestine. Altern.
Lab. Anim. (ALTA) 23, 197209
73 Souliman, S. et al. (2007) Investigation of the biopharmaceutical
behavior of theophylline hydrophilic matrix tablets using USP
methods and an artificial digestive system. Drug Dev. Ind. Pharm.
33, 475483
74 Souliman, S. et al. (2006) A level A in vitro/in vivo correlation in fasted
and fed states using different methods: applied to solid immediate
release oral dosage form. Eur. J. Pharm. Sci. 27, 7279
75 Laird, B.D. et al. (2007) Gastrointestinal microbes increase arsenic
bioaccessibility of ingested mine tailings using the simulator of the
human intestinal microbial ecosystem. Environ. Sci. Technol. 41,
55425547
76 Rajilic-Stojanovic, M. et al. (2010) Evaluating the microbial diversity of
an in vitro model of the human large intestine by phylogenetic
microarray analysis. Microbiology 156, 32703281
77 Waldram, A. et al. (2009) Top-down systems biology modeling of host
metabotype-microbiome associations in obese rodents. J. Proteome Res.
8, 23612375
78 Urich, T. et al. (2008) Simultaneous assessment of soil microbial
community structure and function through analysis of the metatranscriptome. PLoS ONE 3, e2527
79 Frias-Lopez, J. et al. (2008) Microbial community gene expression in
ocean surface waters. Proc. Natl. Acad. Sci. U.S.A. 105, 38053810

Trends in Biotechnology January 2012, Vol. 30, No. 1

80 van Duynhoven, J. et al. (2011) Metabolic fate of polyphenols in the
human superorganism. Proc. Natl. Acad. Sci. U.S.A. DOI: 10.1073/
pnas.1000098107
81 Kovatcheva-Datchary, P. et al. (2009) Linking phylogenetic identities
of bacteria to starch fermentation in an in vitro model of the large
intestine by RNA-based stable isotope probing. Environ. Microbiol. 11,
914926
82 Neufeld, J.D. et al. (2007) Who eats what, where and when? Isotopelabelling experiments are coming of age. ISME J. 1, 103110
83 de Graaf, A.A. et al. (2010) Profiling human gut bacterial metabolism and
its kinetics using [U-13C] glucose and NMR. NMR Biomed. 23, 212
84 Egert, M. et al. (2006) Beyond diversity: functional microbiomics of the
human colon. Trends Microbiol. 14, 8691
85 Backhed, F. et al. (2005) Host-bacterial mutualism in the human
intestine. Science 307, 19151920
86 Moran, M.A. (2009) Metatranscriptomics: eavesdropping on complex
microbial communities. Microbe 4, 329334
87 Botero, L.M. et al. (2005) Poly(A) polymerase modification and reverse
transcriptase PCR amplification of environmental RNA. Appl.
Environ. Microb. 71, 12671275
88 Poretsky, R.S. et al. (2009) Analyzing gene expression from marine
microbial communities using environmental transcriptomics. J. Vis.
Exp. DOI: 10.3791/1086
89 Rooijers, K. et al. (2011) An iterative workflow for mining the human
intestinal metaproteome. BMC Genomics 12, 6
90 Verberkmoes, N.C. et al. (2009) Shotgun metaproteomics of the human
distal gut microbiota. ISME J. 3, 179189
91 Turnbaugh, P.J. and Gordon, J.I. (2008) An invitation to the marriage
of metagenomics and metabolomics. Cell 134, 708713
92 Li, M. et al. (2008) Symbiotic gut microbes modulate human metabolic
phenotypes. Proc. Natl. Acad. Sci. U.S.A. 105, 21172122
93 Martins dos Santos, V.A.P. and Damborsky, J. (2010) Systems biology
at work. Curr. Opin. Biotechnol. 21, 498501
94 Sauer, J. et al. (2007) Products formed during fermentation of the
prebiotic inulin with human gut flora enhance expression of
biotransformation genes in human primary colon cells. Br. J. Nutr.
97, 928937
95 Maccaferri, S. et al. (2010) Rifaximin modulates the colonic microbiota
of patients with Crohns disease: an in vitro approach using a
continuous culture colonic model system. J. Antimicrob. Chemother.
65, 25562565
96 Schirbel, A. and Fiocchi, C. (2010) Inflammatory bowel disease:
established and evolving considerations on its etiopathogenesis and
therapy. J. Digest. Dis. 11, 266276
97 Duncan, S.H. et al. (2009) The role of pH in determining the species
composition of the human colonic microbiota. Environ. Microbiol. 11,
21122122

25