COBIOT-1104; NO. OF PAGES 9
Available online at www.sciencedirect.com
A systems biology approach to studying the role of microbes in
human health
Ines Thiele1,2, Almut Heinken1 and Ronan MT Fleming1,3
Host–microbe interactions play a crucial role in human health
and disease. Of the various systems biology approaches,
reconstruction of genome-scale metabolic networks combined
with constraint-based modeling has been particularly
successful at in silico predicting the phenotypic characteristics
of single organisms. Here, we summarize recent studies, which
have applied this approach to investigate microbe–microbe
and host–microbe metabolic interactions. This approach can
be also expanded to investigate the properties of an entire
microbial community, as well as single organisms within the
community. We illustrate that the constraint-based modeling
approach is suitable to model host–microbe interactions at
molecular resolution and will enable systematic investigation of
metabolic links between the human host and its microbes.
Such host–microbe models, combined with experimental data,
will ultimately further our understanding of how microbes
influence human health.
Addresses
1
Center for Systems Biology, University of Iceland, Reykjavik, Iceland
2
Faculty of Industrial Engineering, Mechanical Engineering & Computer
Science University of Iceland, Reykjavik, Iceland
3
Dept. Biochemistry & Molecular Biology, Faculty of Medicine,
University of Iceland, Reykjavik, Iceland
Corresponding author: Thiele, Ines ([email protected])
Current Opinion in Biotechnology 2012, 24:xx–yy
This review comes from a themed issue on Analytical Biotechnology
Edited by Karsten Hiller and Susann Müller
0958-1669/$ – see front matter, # 2012 Elsevier Ltd. All rights
reserved.
In this review, we focus on existing and potential applications of the COBRA approach to study systematically
microbe–microbe and host–microbe interactions, with
emphasis on the human body as host. First, we will
describe types of host–microbe interactions and then
illustrate experimental methods to functionally determine these interactions, by giving the example of the
human gut microbiota (i.e. totality of microbes in the
human gut). Secondly, we will briefly introduce the
COBRA approach and its application to the human host
and its microbes. Thirdly, we will discuss existing efforts
to model microbial communities and host–microbe interactions. Finally, we will present some challenges ahead
for host–microbe modeling.
Types of host–microbe interactions
One can distinguish three types of species–species interaction (Figure 1). (1) Neutralism is when two organisms
do not depend on each other for growth. However, this
may result in competitive interaction if shared resources
become limiting. (2) In commensalism, one organism
depends on another for growth. For example, the host
not only provides essential nutrients to commensal
microbes, but also a sheltered environment allowing
specifically adapted microbes to thrive. If the host pays
a fitness price for supplying ‘‘room and board’’ to its
microbial inhabitants, this is considered parasitism, or
host–pathogen interaction. (3) Species–species interactions, in which both partners benefit each other, are
known as mutualism, syntrophy, or symbiosis. For
example, many animal, plant, or fungal microbial inhabitants are not only profiting from their hosts, but have coevolved with the host, which in turn provides reciprocal
benefits.
http://dx.doi.org/10.1016/j.copbio.2012.10.001
Introduction
Experimental methods to functionally
determine host–microbe interactions – the gut
microbiota as an example
Trillions of microbes populate the human body, many of
which are beneficial, some even essential for our health,
while others cause infectious diseases. A systems biology
approach is well suited for investigating the relationship
between a host and its microbes as it permits one to study
the interaction between parts of complex biological systems in a holistic context. One possible computational
modeling approach is constraint-based reconstruction and
analysis (COBRA), in which biochemical transformations
are described based on reaction stoichiometry and physico-chemical properties obtained from genome annotation, biochemical, and physiological data [1].
The gut microbiota consists of an estimated 1014 commensal bacteria from thousands of bacterial and archaeal
phylotypes and plays an important role in human health
[2,3]. Numerous model systems exist to study the links
between gut microbiota and host metabolism, including
in vitro cell culture models [4], in vitro gut models [5], ex
vivo organ models [6], animal models [7,8], human in
patient studies [9], and cohort studies (e.g. [10,11]). To
identify microbes responsible for fermentation of certain
dietary carbon sources, 13C isotopomer labeling has been
recently combined with 16S rRNA-based stable isotope
probing [12]. These data are crucial for constructing
www.sciencedirect.com
Current Opinion in Biotechnology 2012, 24:1–9
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
2 Analytical Biotechnology
Figure 1
Possible forms of interactions
Commensalism
Time
Time
Microbe-microbe interactions
Environment
Microbe 1
Mutualism
Growth
Growth
Growth
Neutralism
Microbe 2
Host-microbe interactions
Intracellular interactions
Host cell Intracellular microbe
Environment
Organism 1
Organism 2
Time
Biofilm-associated microbiota
-O2
-Redox potential
-Nutrient supply
varies
-pH varies
Epithelial cell layer
Extracellular interactions
Host cell Extracellular microbe
Environment
Current Opinion in Biotechnology
Overview of possible interactions between two or more organisms.
comprehensive host–microbe interaction models as well
as for benchmarking the models (see next section).
necessary to ensure similar phenotypic properties of
the metabolic reconstruction and the target organism.
Reconstruction process for biochemical
reaction networks
The COBRA approach in a nutshell
Genome-scale, manually curated metabolic reconstructions serve as knowledge-bases as they summarize existing knowledge about cellular pathways in a target
organism in a well-structured, mathematical manner [1]
(Figure 2). The process to assemble these metabolic
reconstructions in a bottom-up manner has been well
established [13] and implemented in an open-source
reconstruction tool [14]. Recently, a semi-automated,
web-based reconstruction tool [15] has been published
that permits the rapid creation of draft reconstructions for
prokaryotes, amenable to the COBRA approach. Importantly, the automated generation of a draft reconstruction
is not a replacement for manual curation. Comparison
with organism-specific physiological data, such as growth
capabilities and medium requirements, and the incorporation of species-specific pathways, such as the mucindegradation pathway in Akkermansia muciniphila, is
Current Opinion in Biotechnology 2012, 24:1–9
There are many detailed reviews on the COBRA
approach (e.g. [13,16,17]), which we briefly summarize
here. The conversion of a metabolic reconstruction into a
condition-specific model includes the transformation of
the biochemical reaction list into a computable, mathematical matrix format (Figure 2). It also requires the
addition of physico-chemical constraints (e.g. mass conservation) and systems boundaries [1]. The COBRA
approach assumes steady-state concentrations, that is,
the change in metabolite concentration (dx/dt) over time
is zero. This is represented by a system of mass balance
equations dx/dt = Sv 0, where S is the stoichiometric
matrix of the metabolic network (S 2 Zm,n), which lists the
m metabolites as rows and n reactions as columns, with the
flux vector v 2 Rn representing the rate of net flux value
for n reactions. At a steady state, the sum of reaction rates
producing a metabolite i is equal to the stoichiometrically
scaled sum of reactions consuming the metabolite i. The
www.sciencedirect.com
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
The role of microbes in human health Thiele, Heinken and Fleming 3
Figure 2
Reconstruction and in
silico models
(a)
Genome
Bibliome
Example methods
(b)
Maximize:
z = cT·v
Such that
S ·v = 0
LB ≤ v ≤ UB
V2
FBA
Metabolic reconstruction
V1
V3
HT data
Condition-specific constraints
(c)
FVA
Max/min:
vk
Such that
cT·v=opt
S ·v = 0
LB ≤ v ≤ UB
vk
V1
V3
(d)
Maximize:
z = cT·v
Such that
S ·v = 0
LB ≤ v ≤ UB
vk = 0
Mathematical formulation
S=
metabolites
reactions
-1
1
0
-1
0
1
0
-1
0
1
0
0
0
0
-1
LB ≤ v ≤ UB
S ·v = 0
V2
1 2 3
Gene
deletion
V2
V1
V3
Current Opinion in Biotechnology
Schematic overview of the constraint-based reconstruction and analysis (COBRA) approach. The reconstruction and model formulation is depicted on
the left side, and some sample methods are illustrated on the right side. While a reconstruction is unique for each organism (as its genome and
annotation is unique), one reconstruction can give raise to many different models, depending on the environmental and genetic constraints applied to
the model (e.g. one model representing anaerobic, rich medium conditions versus a model representing aerobic, rich medium conditions). FBA – flux
balance analysis. FVA – flux variability analysis.
S matrix gives rise to an underdetermined system of linear
equations, that is, there are less equations (mass-balances)
than variables (reaction fluxes) and thus a polyhedral
convex steady-state solution space contains all feasible
steady-state solutions. By adding further constraints (e.g.
nutrient uptake rates, maximal enzyme reaction rates) to
the model, one restricts the solution space toward solutions
that are biologically relevant in a particular condition.
Thus, despite of having incomplete knowledge about
www.sciencedirect.com
many reaction rates, kinetic parameters, metabolite and
enzyme concentrations, the COBRA approach permits the
computation of phenotypic and physiological properties of
the reconstructed networks [18,19].
Computing functional states
Many mathematical modeling tools, used to compute
functional states of metabolic network properties in silico,
rely on linear programming (LP) to solve one or more flux
Current Opinion in Biotechnology 2012, 24:1–9
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
4 Analytical Biotechnology
balance analysis (FBA) problems [20]. FBA based methods
require the statement of an objective function, which an
LP solver optimizes and is guaranteed to find at least one
optimal flux vector v, if a steady state exists in the current
configuration of the model (Figure 2). Identification of
appropriate cellular objective functions has been subject to
several in silico and in vitro experiments [21,22]. Moreover,
multi-objective programming permits the simultaneous
optimization for multiple objectives [23] to investigate
the trade-off between several competing objectives. Alternatively, one may choose methods that allow the characterization of a representative set of the steady-state
solution space, rather than a set of optimal solutions. These
methods include sampling [24], extreme pathway and
elementary mode analysis [25], as well as flux variability
analysis [17]. The COBRA methods have been described
elsewhere in detail [20] and many have been implemented
in the openCOBRA toolbox [26].
Host metabolic reconstructions
In 2007, the first genome-scale, manually curated reconstruction of human metabolism (Recon 1) was published
[27] describing many known metabolic functions occurring in any human cell. Recon 1 accounts for 1496
metabolic genes, 3311 metabolic and transport reactions,
and 1496 unique metabolites distributed over eight cellular compartments. The human metabolic reconstruction has been used as a starting point for creating tissue
specific reconstructions [28]. Moreover, draft reconstructions for other mammals have been generated based on
Recon 1 [29] and manually curated mouse metabolic
reconstructions have been assembled [29,30].
Metabolic reconstructions of human microbes
Genome-scale metabolic reconstructions have been
assembled for a growing number of organisms, including
numerous human microbes (Figure 3). Depending on the
amount of experimental data available, as well as the
exact reconstruction protocol employed, the content,
coverage and predictive sensitivity and specificity of
these reconstructions will differ. The nutrient supply,
oxygen availability, and pH vary depending on the
location within the host. While one can easily model
changes in nutrient and oxygen availability, the consequences of changes in pH on growth and metabolic
capabilities have so far been only addressed in a few in
silico studies [31,32].
Microbe–microbe interaction models
Numerous studies have been published that model
microbe–microbe interactions using the COBRA
approach (see also Figure 1). For instance, novel insight
into cross-feeding between a sulfate-reducing microbe,
Desulfovibrio vulgaris, and a methanogen, Methanococcus
maripaludis, has been gained by assembling the central
metabolism of these two species and modeling metabolic
activities at different growth stages [33]. In an extension
Current Opinion in Biotechnology 2012, 24:1–9
of this work, Klitgord and Segré [34] placed pairs of
seven previously published reconstructions into a joint in
silico environment and computed cross-feeding under
various medium compositions, which induced distinct
microbial interspecies interactions (neutral, commensal,
or mutualistic). Surprisingly, even pathogenic species,
such as Helicobacter pylori, were predicted to engage in
cooperative interactions on certain media. Freilich et al.
[35] analyzed interactions between 118 microbes, using
automatically generated metabolic reconstructions. The
authors predicted neutral, positive, and negative species–
species interactions, some of which were validated
through laboratory experiments, and identified ‘‘winning’’ and ‘‘losing’’ species. Interestingly, many of the
observed cooperative interactions were unidirectional,
meaning that one species profited from the interaction
(taker) while the other one was unaffected (giver) [35].
Zhuang et al. [36] investigated two dissimilatory metalreducing microbes, whose competition may affect bioremediation of uranium-contaminated groundwater. A
multi-objective optimization approach was proposed for
simulation of two-species communites allowing prediction of more clearly defined interactions, including parasitism, that is, host–pathogen relationships [37]. These
studies underline the potential of the COBRA approach
to further our understanding of the metabolic interactions
between the community members as well as the
microbial community interaction with host beyond topological characteristics.
Metabolic networks of microbial communities
Steps toward assembling more realistic microbial communities have been recently done by compiling metabolic networks of the human microbiota found at different
body sites [38,39]. However, these networks are currently
not amenable to the COBRA approach and thus topological analysis was performed to investigate associations
between metabolic network structure of the microbiota
and disease, for example, obesity and inflammatory bowel
disease [38] and to investigate functional diversity associated with organ-specific microbiota [39]. Boundaries between species as well as the origins of genes on the
species level were ignored in these networks [40]. Constructing a multi-species model on an organism-resolved
level would require mapping each enzyme, and by extension the associated reactions, specifically to the species
they are found in. Furthermore, defined boundaries between species would be required, allowing multiple
species to exchange metabolites across a well-defined
boundary. Such a multi-species model would allow us
to move toward network reconstruction of the microbiota
as a whole and allow simulation of microbial species–
species interaction such as competition and cross-feeding.
Modeling microbial biofilms
Biofilms are matrix-enclosed microbial communities
attached to a surface (e.g. epithelial cell layer,
www.sciencedirect.com
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
The role of microbes in human health Thiele, Heinken and Fleming 5
Figure 3
Porphyromonas
gingivalis W83
G: 478
M: 564
R: 679
Leishmania
major Friedlin
G: 560
M: 657
R: 1112
Neisseria meningitis
serogroup B.
G: 555
M: 471
R: 496
Haemophilus
influenzae Rd
G: 400
M: 367
R: 461
Staphylococcus
aureus N315
G: 619/ 551/ 546
M: 571/ 604/ 1431
R: 640/ 712/ 1493
Francisella
tularensis LVS
G: 683
M: 586
R: 605
Pseudomonas
aeruginosa PAO1
G: 1056
M: 760
R: 883
Yersinia
pestis 91001
G: 818
M: 825
R: 1020
Acinetobacter
baumanni AYE
G: 650
M: 778
R: 891
Mycobacterium
tuberculosis H37Rv
G: 661/ 721/ 663
M: 740/ 739/ 742
R: 939/ 849/ 1049
Klebsiella pneumoniae
MGH 78578
G: 1228
M: 1055
R: 1970
Burkholderia
cenocepacia J2315
G: 1028
M: 748
R: 859
Plasmodium
falciparum 3D7
G: 579/ 366
M: 1622/ 616
R: 1375/ 1001
Helicobacter
pylori 26695
G: 341
M: 411
R: 476
Vibrio vulnificus
CMCP6
G: 673
M: 765
R: 943
Streptococcus
thermophilus LMG18311
G: 429
M: NR
R: 522
Cryptosporidium
hominis
G: 213
M: NR
R: 540
Lactobacillus
plantarum WCFS1
G: 721
M: 554
R: 761
Salmonella enterica ssp.
typhimurium LT-2
G: 945/ 1270
M: 1036/ 1119
R: 1964/ 2201
Mycoplasma
genitalium G-37
G: 187
M: 276
R: 264
Escherichia coli
W (ATC 9637)
G: 1273
M: 1111
R: 2477
Escherichia coli
K12 MG-1655
G: 1366
M: 1136
R: 2251
Bacteroides thetaoiotaomicron VPI-5482
G: 853
M: 914
R: 1305
Lactococcus lactis
ssp. lactis IL1403
G: 358
M: 422
R: 621
Current Opinion in Biotechnology
List of human microbes for which genome-scale metabolic reconstructions have been published and their predominant body sites. In red are
highlighted pathogens, orange are opportunistic pathogens, green are commensals, and blue are probiotic bacteria. Dotted lines represent skin as
body site. White background represents prokaryotic organisms, while eukaryotes are shaded gray. Other body sites than represented here may be
infected. G: number of genes included in the metabolic reconstruction. R: number of reactions. M: number of metabolites. NR: not reported in the
original reconstruction paper. For references, please refer to the supplemental table.
www.sciencedirect.com
Current Opinion in Biotechnology 2012, 24:1–9
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
6 Analytical Biotechnology
Figure 1) and play a key role in chronic infections (e.g.
cystic fibrosis). Because of the different microhabitats in a
biofilm, the bacteria have distinct metabolic profiles
owing to varied nutrient and oxygen supply. Furthermore, the metabolic characteristics of biofilm bacteria are
distinct from those of planktonic (i.e. free-living) bacteria.
First attempts for modeling consequences of microenvironments on the metabolic capabilities of the human
pathogen Pseudomonas aeruginosa have been recently
published [41,42].
Host–microbe interaction models
A first constraint-based model of host–pathogen interaction between Mycobacterium tuberculosis and human
alveolar macrophages has been assembled, by placing
the microbial reconstruction into the cytosol compartment of the macrophage reconstruction [43] (see also
Figure 1). This host–pathogen model was subsequently
used to simulate intracellular infection with M. tuberculosis
cells residing in phagosomes. Furthermore, host–
pathogen interaction in malaria was simulated by embedding a reconstruction of the malaria pathogen Plasmodium
falciparum into a human erythrocyte reconstruction [44]
(see also Figure 1). The COBRA approach was also
applied to study the metabolic dependency between a
commensal gut microbe, Bacteroides thetaiotaomicron, and
a murine host [45]. This host–microbe interaction model
successfully captured mutually beneficial cross-feeding
and competitive interactions as a function of diet. Moreover, it was employed to identify essential metabolic links
between the host and its commensal microbe.
Mathematical representation of host–microbe
interactions
Metabolic host–microbe models are a conjunction of the
individual metabolic models (Figure 4), but certain
additional features are recommended. For instance, it
is convenient to maintain each organism’s extracellular
compartment and combine the two reconstructions by
adding an additional environment compartment where
nutrients are supplied (e.g. dietary component uptake
into the lumen) and secretion products are removed
[34,45]. The organisms within this model can take up
or secrete metabolites into this environmental compartment. One key advantage of maintaining the extracellular
compartment is that one can easily trace in silico, which
organism removed or contributed which metabolite, as
there is only one reaction per metabolite crossing the
environment-extracellular compartment boundary but
there may be more than one organism-specific transporters for a particular metabolite. Naive conjunction of two
models can lead to the computation of biologically
implausible steady states, for example, where microbial
(or host) reactions are active even though the microbe (or
host) itself may not generate ATP or produce biomass.
Additional constraints are necessary to couple reaction
Figure 4
Mathematical representation of host-microbe interactions
EH1
RH1
EH2
RH2
RH3
RH4
Microbe
RH5
H
E3
Eμ1
Rμ1
Eμ2
Rμ2
Rμ3
Rμ5
RH1 RH2 RH3 RH4 RH5 EH1 EH2 EH3 E1 E2 E3 E4 E5 Rμ1 Rμ2 Rμ3 Rμ4 Rμ5 Eμ1 Eμ2 Eμ3
Host - Microbe
Eμ3
E1
E4
Rμ4
E2
Rμ1 Rμ2 Rμ3 Rμ4 Rμ5 Eμ1 Eμ2 Eμ3
RH1 RH2 RH3 RH4 RH5 EH1 EH2 EH3
SH=
-1
0
0
1
0
0
0
0
-1
0
0
1
0
0
0
0
0
-1
-1
1
0
0
0
0
0
0
-1
1
0
0
1
0
0
0
-1
1
0
0
0
0
0
0
0 0
1 0
0 -1
0 0
0 0
0 0
0 0
Sμ=
-1
0
0
1
0
0
0
-1
0
0
1
0
0
0
0
-1
0
1
0
0
0
0
-1
1
0
0
1
0
0
-1
1
0
0
0
0
0
0 0
1 0
0 -1
0 0
0 0
0 0
E3
m Metabolites
Host cell
E5
SHμ=
0 0 0 0 0 0 0
0 1 0 0 0 0 0 0 0 0 0 -1 0 0
0 0 0 0 0 -1 0
0 0 1 0 0 0 0 0 0 0 0 0 -1 0
0 0 0 0 0 0 -1 0 0 0 1 -1 0 0 0 0 0 0 0 0 1
0 0 0 0 0 0 0
1 0 0 0 0 -1 0 0 0 0 0 0 0 0
-1 0 0 0 0 1 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 -1 0 0 0 0 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 1 0 0 -1 0 0 0 0 0 0 0 0 0 0 0 0 0
1 0 -1 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 1 -1 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 1 -1 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 1 -1 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 -1 0 0 0 0 1 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0 -1 0 0 0 0 1 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 1 0 0 -1
0 0 0 0 0 0 0
0 0 0 0 0 0 1 0 -1 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0 1 0 -1 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 1 1 -1 0 0 0
n Reactions
Current Opinion in Biotechnology
Mathematical representation of the host and microbe metabolic networks and their representation as stoichiometric matrices. SH – host stoichiometric
matrix, Sm – microbial stoichiometric matrix, SHm – host–microbe stoichiometric matrix. By definition, all substrates have negative entries in the S
matrix, while all products of a reaction have positive entries. Each cell contains the stoichiometric coefficient of the respective metabolite that
participates in a reaction. A common environment compartment joins the two organisms. If more than two organisms are modeled in one modeling
framework than the corresponding metabolic model is added to the environment compartment and the S matrix is extended accordingly. E: exchange
reaction. R: reaction (transport or intracellular). H: host. m: microbe.
Current Opinion in Biotechnology 2012, 24:1–9
www.sciencedirect.com
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
The role of microbes in human health Thiele, Heinken and Fleming 7
rates in host–microbe models [45], similar to coupling
constraints previously applied to couple utilization and
synthesis of macromolecules in integrated models of
metabolic and macromolecular networks in a single
organism [46].
References and recommended reading
Conclusion
Appendix A. Supplementary data
In this review, we highlighted recent advances in the
constraint-based modeling community, which will pave
the way for the systematic investigation of metabolic
interactions between the human host and its microbes,
and how these interactions may contribute to health and
disease states. Moreover, these computational models
could be used to propose novel drug targets against
intracellular or extracellular pathogens [47].
While we focused on the human host, other host–microbe
interactions can be also modeled, such as plant–microbe
or insect–microbe interactions. For instance, nitrogen
fixation in Rhizobium etli bacteroids at the root nodules
of the bean plant has recently been modeled [48]. Moreover, Buchnera aphidicola, intracellular endosymbiont of
pea aphids, and Sodalis glossinidius, endosymbiont of
tsetse flies, have been recently reconstructed [49,50].
As host–microbe metabolic models will become more
complex, for example, by considering many species
within a microbial community or by capturing more
cellular processes (e.g. macromolecular synthesis
[51,52] and signaling [53,54]), the suite of available computational analysis tools extends. While many of the
COBRA methods have been developed for analysis of
a single organism metabolism, they may be computationally too inefficient for use with larger multi-species
models.
A grand challenge of systems biology is the integrative
analysis of heterogeneous, large-scale omics data sets [55].
The high resolution, detailed representation captured in
COBRA models permits mapping and integrated analysis
of such data by providing a context for content. Moreover,
the COBRA host–microbe interaction models can also be
employed as a starting point for kinetic modeling, thus,
permitting to address questions regarding time varying
interactions.
Computational models of host–microbe interactions may
profitably be used to improve our understanding of complex biological systems and will thus assist in unraveling
mechanisms underlying physiological and pathophysiological states.
Acknowledgements
This work was supported by a Marie Curie International Reintegration
Grant awarded to IT (no. 249261) within the 7th European Community
Framework Program. The authors thank Ms. M. Galhardo and Ms. S.
Magnúsdóttir for critical reading of the manuscript as well as Dr. P. Wilmes
for valuable discussions.
www.sciencedirect.com
Papers of particular interest, published within the period of review,
have been highlighted as:
of special interest
of outstanding interest
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/
j.copbio.2012.10.001.
1.
Palsson B: Systems Biology: Properties of Reconstructed
Networks. Cambridge University Press; 2006.
2.
Kussmann M, Van Bladeren PJ: The extended nutrigenomics understanding the interplay between the genomes of food, gut
microbes, and human host. Front Genet 2011, 2:21.
3.
Guarner F, Malagelada JR: Gut flora in health and disease.
Lancet 2003, 361:512-519.
4.
Barrila J, Radtke AL, Crabbe A, Sarker SF, Herbst-Kralovetz MM,
Ott CM, Nickerson CA: Organotypic 3D cell culture models:
using the rotating wall vessel to study host–pathogen
interactions. Nat Rev Microbiol 2010, 8:791-801.
5.
Payne AN, Zihler A, Chassard C, Lacroix C: Advances and
perspectives in in vitro human gut fermentation modeling.
Trends Biotechnol 2012, 30:17-25.
6.
Tsilingiri K, Barbosa T, Penna G, Caprioli F, Sonzogni A, Viale G,
Rescigno M: Probiotic and postbiotic activity in health and
disease: comparison on a novel polarised ex-vivo organ
culture model. Gut 2012, 61:1007-1015.
7.
Claus SP, Ellero SL, Berger B, Krause L, Bruttin A, Molina J,
Paris A, Want EJ, de Waziers I, Cloarec O et al.: Colonizationinduced host-gut microbial metabolic interaction. MBio 2011,
2 e00271-00210.
In this study, the body fluid and tissue metabolite profile of germfree
(sterile) and conventional mice is measured through NMR spectroscophy.
Metabolites associated with gut colonization after are identified in urine,
plasma, liver, kidney, and colon after exposing germfree mice to a typical
microbial environment.
8.
Mahowald MA, Rey FE, Seedorf H, Turnbaugh PJ, Fulton RS,
Wollam A, Shah N, Wang C, Magrini V, Wilson RK et al.:
Characterizing a model human gut microbiota composed of
members of its two dominant bacterial phyla. Proc Natl Acad
Sci USA 2009, 106:5859-5864.
In this study, germfree mice are selectively colonized with Bacteroides
thetaiotaomicron and Eubacterium rectale, prominent Bacteroidetes and
Firmicutes representatives, respectively. This gnotobiotic mouse model
shows both microbes’ niche specification and adaptation to each other’s
presence.
9.
Jumpertz R, Le DS, Turnbaugh PJ, Trinidad C, Bogardus C,
Gordon JI, Krakoff J: Energy-balance studies reveal
associations between gut microbes, caloric load, and nutrient
absorption in humans. Am J Clin Nutr 2011, 94:58-65.
10. Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, Manichanh C,
Nielsen T, Pons N, Levenez F, Yamada T et al.: A human gut
microbial gene catalogue established by metagenomic
sequencing. Nature 2010, 464:59-65.
11. Yatsunenko T, Rey FE, Manary MJ, Trehan I, DominguezBello MG, Contreras M, Magris M, Hidalgo G, Baldassano RN,
Anokhin AP et al.: Human gut microbiome viewed across age
and geography. Nature 2012, 486:222-227.
12. Kovatcheva-Datchary P, Egert M, Maathuis A, RajilicStojanovic M, de Graaf AA, Smidt H, de Vos WM, Venema K:
Linking phylogenetic identities of bacteria to starch
fermentation in an in vitro model of the large intestine by RNAbased stable isotope probing. Environ Microbiol 2009,
11:914-926.
13. Thiele I, Palsson BO: A protocol for generating a high-quality
genome-scale metabolic reconstruction. Nat Protoc 2010,
5:93-121.
Current Opinion in Biotechnology 2012, 24:1–9
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
8 Analytical Biotechnology
14. Thorleifsson SG, Thiele I: rBioNet: a COBRA toolbox extension
for reconstructing high-quality biochemical networks.
Bioinformatics 2011, 27:2009-2010.
31. Haraldsdottir HS, Thiele I, Fleming RM: Quantitative assignment
of reaction directionality in a multicompartmental human
metabolic reconstruction. Biophys J 2012, 102:1703-1711.
15. Henry CS, DeJongh M, Best AA, Frybarger PM, Linsay B,
Stevens RL: High-throughput generation, optimization and
analysis of genome-scale metabolic models. Nat Biotechnol
2010, 28:977-982.
The Model SEED pipeline is introduced. Model SEED is a web-based
resource that generates automated draft metabolic reconstructions from
the genome sequence. The draft reconstructions are ready for analysis
but may be further curated and validated manually against available
experimental data.
32. Reed JL, Vo TD, Schilling CH, Palsson BO: An expanded
genome-scale model of Escherichia coli K-12 (iJR904 GSM/
GPR). Genome Biol 2003, 4:R54.
16. Feist AM, Herrgard MJ, Thiele I, Reed JL, Palsson BO:
Reconstruction of biochemical networks in microorganisms.
Nat Rev Microbiol 2009, 7:129-143.
17. Orth JD, Thiele I, Palsson BO: What is flux balance analysis? Nat
Biotechnol 2010, 28:245-248.
18. Feist AM, Palsson BO: The growing scope of applications of
genome-scale metabolic reconstructions using Escherichia
coli. Nat Biotechnol 2008, 26:659-667.
19. Oberhardt MA, Palsson BO, Papin JA: Applications of genomescale metabolic reconstructions. Mol Syst Biol 2009, 5:320.
20. Lewis NE, Nagarajan H, Palsson BO: Constraining the metabolic
genotype-phenotype relationship using a phylogeny of in
silico methods. Nat Rev Microbiol 2012, 10:291-305.
21. Gianchandani EP, Oberhardt MA, Burgard AP, Maranas CD,
Papin JA: Predicting biological system objectives de novo from
internal state measurements. BMC Bioinform 2008, 9:43.
22. Schuetz R, Kuepfer L, Sauer U: Systematic evaluation of
objective functions for predicting intracellular fluxes in
Escherichia coli. Mol Syst Biol 2007, 3:1-15.
23. Nagrath D, Avila-Elchiver M, Berthiaume F, Tilles AW, Messac A,
Yarmush ML: Integrated energy and flux balance based
multiobjective framework for large-scale metabolic networks.
Ann Biomed Eng 2007, 35:863-885.
33. Stolyar S, Van Dien S, Hillesland KL, Pinel N, Lie TJ, Leigh JA,
Stahl DA: Metabolic modeling of a mutualistic microbial
community. Mol Syst Biol 2007, 3:92.
34. Klitgord N, Segre D: Environments that induce synthetic
microbial ecosystems. PLoS Comput Biol 2010, 6:e1001002.
This work presents a framework for modeling microbial species–species
interactions. Published reconstructions are joined through an in silico
environment and allowed to exchange nutrients. Previously unknown
neutral, commensal, and mutualistic relationships could be predicted.
35. Freilich S, Zarecki R, Eilam O, Segal ES, Henry CS, Kupiec M,
Gophna U, Sharan R, Ruppin E: Competitive and cooperative
metabolic interactions in bacterial communities. Nat Commun
2011, 2:589.
In this work, interactions between 118 metabolic reconstructions
retrieved from the Model SEED pipeline are systematically explored.
The cooperative and competitive potential of 6903 bacterial pairs is
predicted, and bacterial interactions are further explored using 2801
environmental samples.
36. Zhuang K, Izallalen M, Mouser P, Richter H, Risso C,
Mahadevan R, Lovley DR: Genome-scale dynamic modeling of
the competition between Rhodoferax and Geobacter in anoxic
subsurface environments. ISME J 2011, 5:305-316.
37. Zomorrodi AR, Maranas CD: OptCom: a multi-level optimization
framework for the metabolic modeling and analysis of
microbial communities. PLoS Comput Biol 2012, 8:e1002363.
38. Greenblum S, Turnbaugh PJ, Borenstein E: Metagenomic
systems biology of the human gut microbiome reveals
topological shifts associated with obesity and inflammatory
bowel disease. Proc Natl Acad Sci USA 2012, 109:594-599.
24. Schellenberger J, Palsson BO: Use of randomized sampling for
analysis of metabolic networks. J Biol Chem 2009, 284:54575461.
39. Abubucker S, Segata N, Goll J, Schubert AM, Izard J, Cantarel BL,
Rodriguez-Mueller B, Zucker J, Thiagarajan M, Henrissat B et al.:
Metabolic reconstruction for metagenomic data and its
application to the human microbiome. PLoS Comput Biol 2012,
8:e1002358.
25. Price ND, Reed JL, Palsson BO: Genome-scale models of
microbial cells: evaluating the consequences of constraints.
Nat Rev Microbiol 2004, 2:886-897.
40. Borenstein E: Computational systems biology and in silico
modeling of the human microbiome. Brief Bioinform 2012 http://
dx.doi.org/10.1093/bib/bbs022.
26. Schellenberger J, Que R, Fleming RM, Thiele I, Orth JD, Feist AM,
Zielinski DC, Bordbar A, Lewis NE, Rahmanian S et al.:
Quantitative prediction of cellular metabolism with constraintbased models: the COBRA Toolbox v2.0. Nat Protoc 2011,
6:1290-1307.
The current version of the COBRA (constraint-based reconstruction and
analysis) Toolbox, 2.0, is presented. The COBRA Toolbox is Matlabbased and contains methods for in silico metabolic modeling, such as flux
balance analysis, random sampling, 13C flux analysis, mapping of highthroughput data, and strain design for metabolic engineering.
41. Sigurdsson G, Fleming RM, Heinken A, Thiele I: A systems
biology approach to drug targets in Pseudomonas aeruginosa
biofilm. PLoS ONE 2012, 7:e34337.
27. Duarte NC, Becker SA, Jamshidi N, Thiele I, Mo ML, Vo TD,
Srivas R, Palsson BO: Global reconstruction of the human
metabolic network based on genomic and bibliomic data. Proc
Natl Acad Sci USA 2007, 104:1777-1782.
This work introduces the first genome-scale human reconstruction,
Recon1, which represents the metabolic functions found in a generic
human cell. Recon1 has been applied successfully for the prediction of
potential drug targets and the construction of human tissue-specific
models.
42. Oberhardt MA, Goldberg JB, Hogardt M, Papin JA: Metabolic
network analysis of Pseudomonas aeruginosa during chronic
cystic fibrosis lung infection. J Bacteriol 2010, 192:5534-5548.
43. Bordbar A, Lewis NE, Schellenberger J, Palsson BO, Jamshidi N:
Insight into human alveolar macrophage and M. tuberculosis
interactions via metabolic reconstructions. Mol Syst Biol 2010,
6:422.
The first constraint-based host–pathogen model is presented. A tissuespecific reconstruction of a human alveolar macrophage based on
Recon1 is joined with a metabolic reconstruction of Mycobacterium
tuberculosis. Potential drug targets and metabolic changes during different infectious states are analyzed.
44. Huthmacher C, Hoppe A, Bulik S, Holzhutter HG: Antimalarial drug
targets in Plasmodium falciparum predicted by stage-specific
metabolic network analysis. BMC Syst Biol 2010, 4:120.
28. Bordbar A, Palsson BO: Using the reconstructed genome-scale
human metabolic network to study physiology and pathology.
J Intern Med 2012, 271:131-141.
45. Heinken A, Sahoo S, Fleming RMT, Thiele I: Systems-level
characterization of a host-microbe metabolic symbiosis in the
mammalian gut. Gut Microbes. 2013, 4(1):1-13.
29. Sigurdsson MI, Jamshidi N, Steingrimsson E, Thiele I, Palsson BO:
A detailed genome-wide reconstruction of mouse metabolism
based on human Recon 1. BMC Syst Biol 2010, 4:140.
46. Thiele I, Fleming RM, Bordbar A, Schellenberger J, Palsson BO:
Functional characterization of alternate optimal solutions of
Escherichia coli’s transcriptional and translational machinery.
Biophys J 2010, 98:2072-2081.
30. Selvarasu S, Karimi IA, Ghim GH, Lee DY: Genome-scale
modeling and in silico analysis of mouse cell metabolic
network. Mol Biosyst 2010, 6:152-161.
Current Opinion in Biotechnology 2012, 24:1–9
47. Chavali AK, D’Auria KM, Hewlett EL, Pearson RD, Papin JA: A
metabolic network approach for the identification and
www.sciencedirect.com
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001
COBIOT-1104; NO. OF PAGES 9
The role of microbes in human health Thiele, Heinken and Fleming 9
prioritization of antimicrobial drug targets. Trends Microbiol
2012, 20(3):113-123.
48. Resendis-Antonio O, Hernandez M, Salazar E, Contreras S,
Batallar GM, Mora Y, Encarnacion S: Systems biology of
bacterial nitrogen fixation: high-throughput technology and its
integrative description with constraint-based modeling. BMC
Syst Biol 2011, 5:120.
49. Thomas GH, Zucker J, Macdonald SJ, Sorokin A, Goryanin I,
Douglas AE: A fragile metabolic network adapted for
cooperation in the symbiotic bacterium Buchnera aphidicola.
BMC Syst Biol 2009, 3:24.
50. Belda E, Silva FJ, Pereto J, Moya A: Metabolic networks of
Sodalis glossinidius: a systems biology approach to reductive
evolution. PLoS ONE 2012, 7:e30652.
51. Thiele I, Jamshidi N, Fleming RM, Palsson BO: Genome-scale
reconstruction of Escherichia coli’s transcriptional and
www.sciencedirect.com
translational machinery: a knowledge base, its mathematical
formulation, and its functional characterization. PLoS Comput
Biol 2009, 5:e1000312.
52. Thiele I, Fleming RMT, Que R, Bordbar A, Diep D, Palsson BO:
Multiscale modeling of metabolism and macromolecular
synthesis in E. coli and its application to the evolution of codon
usage. PLoS ONE 2012, 7(9):e45635.
53. Lee JM, Gianchandani EP, Eddy JA, Papin JA: Dynamic analysis
of integrated signaling, metabolic, and regulatory networks.
PLoS Comput Biol 2008, 4:e1000086.
54. Covert MW, Xiao N, Chen TJ, Karr JR: Integrating metabolic,
transcriptional regulatory and signal transduction models in
Escherichia coli. Bioinformatics 2008, 24:2044-2050.
55. Joyce AR, Palsson BO: The model organism as a system:
integrating ‘omics’ data sets. Nat Rev Mol Cell Biol 2006,
7:198-210.
Current Opinion in Biotechnology 2012, 24:1–9
Please cite this article in press as: Thiele I, et al.: A systems biology approach to studying the role of microbes in human health, Curr Opin Biotechnol (2012), http://dx.doi.org/10.1016/
j.copbio.2012.10.001