Index
Programme at a glance........................ Pages 2-3
Full Programme.................................... Pages 5-8
Speaker Abstracts
Monday, September 16..................... Pages 10-14
Tuesday, September 17.................... Pages 16-24
Wednesday, September 18............... Pages 26-34
Thursday, September 19................... Pages 36-40
Friday, September 20........................ Pages 42-46
Poster Abstracts.................................. Pages 48-82
List of Participants............................... Pages 83-88
1
MPA2013 Conference at a glance
MONDAY September 16
10:00-17:00 Thermodynamics in Metabolism Workshop
Speakers: Johann Rower, Andreas Hoppe, Avi Flamholz,
Oliver Ebenhöh, Daniele De Martino
18:00
Wine Reception
TUESDAY September 17
08:50
Welcome – Oliver Ebenhöh
09:00-14:40 Session I: Analysis Techniques for Metabolic Networks
Speakers: Steffen Klamt, Vitaly Selivanov, Isabel Rocha,
Arne Müller, Caroline Baroukh, Huili Yuan
14:40-18:00 Session II: Design and Evolution of Metabolic Systems
Speakers: Herbert Sauro, Steven Court, Kazuhiro Takemoto
WEDNESDAY September 18
09:00-12:50 Session III: Experimental Studies of Metabolic Systems
Speakers: Ralf Takors, Dany Beste, Jeroen Jeneson,
George Ratcliffe, Katja Tummler
14:00-18:00 Session IV: Biotechnological Applications
Speakers: Ralf Steuer, Brian Rudd/David Fell, Alison Smith,
Oliver Hädicke
2
THURSDAY September 19
09:00-10:20 Session IV: Biotechnological Applications continued
Speakers: Guillaume Cogne, Claude Aflalo
10:20-12:50 Session V: Interactions between Organisms
Speakers: Bas Teusink, Souvik Kusari, Ross Carlson
Afternoon free
19:30
Conference Dinner
FRIDAY September 20
09:00-12:50 Session VI: Modelling Photosynthetic Organisms
Chair: Ross Carlson
Speakers: Zoran Nikoloski, Dusán Lažar, Sudip Kundu,
Nadine Töpfer, Sophie Colombie
12:50-13:00 Closure – Mark Poolman
3
4
MPA2013 Conference full programme
MONDAY September 16
10:00-17:00 Thermodynamics in Metabolism Workshop
10:00-10:30 Oliver Ebenhöh - Introduction
10:30-11:30 Andreas Hoppe - Thermodynamic constraints improve flux
balance optimization
11:30-12:00 Coffee Break
12:00-13:00 Johann Rower - Separating thermodynamic and kinetic
contributions to enzyme control and regulation
13:00-14:00 Lunch
14:00-15:00 Avi Flamholz - Glycolytic strategy as a tradeoff between
energy yield and protein cost.
15:00-15:30 Oliver Ebenhöh - Statistical Thermodynamics meets polymer
biochemistry
15:30-16:00 Coffee Break
16:00-17:00 Daniele De Martino - Physical constraints and counting
problems in metabolic networks: exploiting duality.
18:00 Wine Reception
TUESDAY September 17
08:50 Welcome – Oliver Ebenhöh
09:00-14:40 Session I: Analysis Techniques for Metabolic Networks
Chair: Oliver Ebenhöh
09:00-09:30 Steffen Klamt - Enumeration of smallest intervention
strategies in genome-scale metabolic networks
09:40-10:10 Vitaly Selivanov - An analysis of compartmental structure of
central metabolic pathways based on 13C tracer distribution in metabolites.
5
10:20-10:50 Isabel Rocha - Improving in silico predictions in Metabolic
Engineering problems
11:00-11:30 Coffee Break
11:30-12:00 Arne Müller - Modules in Metabolic Networks
12:10-12:40 Caroline Baroukh - A new framework for metabolic
modelling under non-balanced growth. Application to photoautotrophic
microalgae in day/night cycle.
12:50-14:00 Lunch
14:00-14:30 Huili Yuan - Use of genome-scale metabolic models for plant
metabolism
14:40-18:00 Session II: Design and Evolution of Metabolic Systems
Chair: Mark Poolman
14:40-15:10 Herbert Sauro - Design of Self-Optimizing Metabolic
Networks
15:20-16:40 Coffee and Posters
16:40-17:10 Steven Court - Are metabolic pathways optimal?
17:20-17:50 Kazuhiro Takemoto - On the interaction between
environments and metabolic network structure: in the light of modularity
18:00 Close
WEDNESDAY September 18
09:00-12:50 Session III: Experimental Studies of Metabolic Systems
Chair: Stefan Schuster
09:00-09:30 Ralf Takors - Studies on the Coupling of Energy
Management and Cellular Metabolism
09:40-10:10 Dany Beste - Carbon fixation and a mixed diet for the
intracellular tuberculosis bacillus
10:20-10:50 Jeroen Jeneson - Experimental observations on glycolysis in
muscle: revisiting physiological versus biochemical regulation in the study
of metabolic pathway topology and function.
11:00-11:30 Coffee Break
11:30-12:00 George Ratcliffe - Analysing primary carbon metabolism in
heterotrophic plant cells with steady-state metabolic flux analysis
6
12:10-12:40 Katja Tummler - Dynamic mathematical modeling of time-
resolved quantitative metabolomics data explores the metabolic flexibility
of mycobacteria
12:50-14:00 Lunch
14:00-18:00 Session IV: Biotechnological Applications
Chair: Sudip Kundu
14:00-14:30 Ralf Steuer - Computational models of cyanobacterial
metabolism: Systemic properties of phototrophic growth
14:40-15:10 Brian Rudd/David Fell - Exploiting Metabolic Diversity for
Industrial Biotechnology
15:20-16:40 Coffee and Posters
16:40-17:10 Alison Smith - Exploiting metabolism for algal biotechnology
17:20-17:50 Oliver Hädicke - The AGE1.CR.pIX cell line as example
system for analysis of avian central metabolism and statistical evaluation of
inter-experimental variance
18:00 Close
THURSDAY September 19
09:00-10:20 Session IV: Biotechnological Applications continued
Chair: David Fell
09:00-09:30 Guillaume Cogne - Energetic analysis and modeling of
metabolic functioning of an eukaryotic microalgae under growth
conditions in photobioreactor
09:40-10:10 Claude Aflalo - Kinetic models for microalgal photosynthesis
coupled to cell motion in a photobioreactor.
10:20-12:50 Session V: Interactions between Organisms
Chair: David Fell
10:20-10:50 Bas Teusink - Optimal states of specific rate carry flux
through an elementary flux mode
11:00-11:30 Coffee Break
11:30-12:00 Souvik Kusari - Crosstalk between endophytic
microorganisms leads to biosynthesis of bioactive natural products
7
12:10-12:40 Ross Carlson - In silico and in vitro analysis of interacting
microbial consortia
12:50-14:00 Take away lunch
Afternoon free
19:30 Conference Dinner
FRIDAY September 20
09:00-12:50 Session VI: Modelling Photosynthetic Organisms
Chair: Ross Carlson
09:00-09:30 Zoran Nikoloski - Integration of metabolomics data in large-
scale models of photosynthetic organisms
09:40-10:10 Dusán La W
žar I- T
MoH
delD
lin R
g of A
va W
riabl N
e chlorophyll fluorescence
originating in photosystem I
10:20-10:50 Sudip Kundu - Systems level responses of rice metabolism
under varying conditions
11:00-11:30 Coffee Break
11:30-12:00 Nadine Töpfer - Integration of genome-scale modeling and
transcript profiling reveals metabolic pathways underlying light and
temperature acclimation in arabidopsis
12:10-12:40 Sophie Colombie - The tomato fruit development with a
constraint-based model
12:50-13:00 Closure – Mark Poolman
13:00-14:00 Take away lunch and Departures
8
SPEAKER ABSTRACTS
(In order of presentation)
Monday, September 16
9
Thermodynamic constraints improve flux balance optimization
Andreas Hoppe
Institute for Biochemistry- Charité Berlin, 10117 Berlin, Germany,
andreas.hoppe@charite.de
Flux balance analysis (FBA) has been successfully applied in a wide range of
applications such as growth prediction, strain optimization, and antimicrobial
target prediction. Its main strength is the low computational effort: a complete
flux prediction requires the solution of a LP (linear program), MILP (mixed-
integer), or QP (quadratic program) which is fast even for genome-scale
networks, and allows large simulation series in manageable time.
It became apparent that predictivity of FBA depends on additional side
constraints. One of these constraints is based on the second law of
thermodynamics as a directional constraint on the basis of Gibb’s free energies.
Its application is relatively simple as no special characterization of the enzymes is
necessary. Formation energies (the only necessary input) are available for a wide
range of metabolites,more can be estimated with prediction methods.
Using a smart mathematical reformulation, the FBA with the constraint can still be
implemented as a MILP, thus, the FBA remains to possess the computational
ease. As the thermodynamic constraint restricts flux directions, it is most
effective if the direction of fluxes are largely unknown, e.g. in less studied
organisms and cell types or under abnormal conditions. Here, we present the
validity of the constraint on experiments on E. coli in Zurich and Keio, and show
recent applications.
Literature
[1] Hoppe A, Hoffmann S, and Holzhütter HG. Including metabolite concentrations into
flux-balance analysis: Thermodynamic realizability as a constraint on flux distributions in
metabolic networks BMC Syst. Biol., (2007) 1(1): 23.
[2] Hoffmann S, Minimale Flussmoden als theoretisches Konzept für die funktionelle
Analyse und modulare Beschreibung zellulärer Stoffwechselnetzwerke, Dissertation (2012).
10
Separating thermodynamic and kinetic contributions to enzyme
control and regulation
Johann M. Rohwer1 and Jan-Hendrik S. Hofmeyr1,2
1Triple-J Group for Molecular Systems Biology, Department of Biochemistry, and
2Centre for Studies in Complexity, Stellenbosch University, Private Bag X1,
Stellenbosch ZA-7602
In this presentation I will describe concepts and tools for assessing and
quantifying the regulation of the rate of an enzyme-catalysed reaction, focussing
on the distinction between kinetic and thermodynamic contributions to the
control of reaction rate. To achieve this, I will show how a generic reversible rate
equation can be recast in two ways. The first makes the distance from equilibrium
explicit, thereby allowing a quantitative distinction between kinetic and
thermodynamic control of reaction rate, as well as between near-equilibrium and
far-from-equilibrium reactions. Specifically, the analysis shows that the traditional
criterion put forward by Rolleston in 1972 to separate non-equilibrium from
equilibrium reactions on the grounds of the disequilibrium ratio being less than
0.05 or larger than 0.2, is flawed. When this ratio equals 0.2, a reaction is still very
far from equilibrium.
Recasting the rate equation in the second form separates mass action from rate
capacity and quantifies the degree to which intrinsic mass action contributes to
reaction rate, and how regulation of an enzyme-catalysed reaction either
enhances or counteracts this mass-action behaviour. The contribution of enzyme
binding to regulation will be analysed in detail for a number of enzyme-kinetic
rate laws, separating it from the contribution of rate capacity and mass action.
For cooperative reactions, the contribution of enzyme binding can be further
partitioned into a single-subunit term (describing the contribution of a single
subunit on its own) and a subunit-interaction term (quantifying cooperative
interaction). Importantly, the above partitioning directly applies to the expressions
for the elasticity coefficients of the rate equation as well, allowing the individual
contributions of mass action, enzyme binding and cooperative interactions to the
regulation of enzyme activity to be quantified.
11
Glycolytic strategy as a tradeoff between energy yield and
protein cost
A Flamholz 2, E Noor 1, A Bar-Even 1, W Liebermeister 1, R Milo 1.
1Department of Plant Sciences, The Weizmann Institute of Science, Rehovot
76100, Israel.
2Department of Molecular and Cellular Biology at the University of California,
Berkeley.
Not all sugar-consuming organisms use the canonical Embden–Meyerhoff–
Parnass (including several alternative glycolytic pathways, the most common of
which is the Entner–Doudoroff (ED) pathway. The prevalence of the ED pathway is
puzzling as it produces only one ATP per glucose—half as much as the EMP
pathway. In this talk I argue that the diversity of prokaryotic glucose metabolism
may reflect a thermodynamically-driven tradeoff between a pathway’s energy
(ATP) yield and the amount of enzymatic protein required to catalyze pathway
flux. We will discuss the connection between pathway thermodynamics, kinetics
and protein levels and examine how this interrelationship affects glycolysis.
Specifically, we will see that the ED pathway is expected to require several-fold
less enzymatic protein to achieve the same glucose conversion rate as the EMP
pathway. Finally, I will show that this theoretical result is consonant with the
genomic record, which reveals that that bacteria and archaea employ different
glycolytic pathways depending on their energy supply. EMP) glycolytic pathway.
Prokaryotic glucose metabolism is particularly diverse,
12
Statistical Thermodynamics meets polymer biochemistry
Oliver Ebenhöh1,2
1) Institute for Complex Systems and Mathematical Biology, University of
Aberdeen, Meston Building, Aberdeen AB24 3UE
2) GFZ German Research Centre for Geosciences, Section 4.5, Telegrafenberg,
14473 Potsdam
The vast diversity of carbohydrates is generated by carbohydrate-active enzymes
(CAZymes) accepting many different substrates and catalyzing numerous
reactions. This promiscuity is in stark contrast to most enzymes active in central
metabolism which are highly specific and catalyze exactly one or a very small
number of reactions. The multitude of accepted substrates and catalyzed
reactions makes CAZymes hard to characterize in classical enzymological terms.
Equilibrium constants and Michaelis constants, which were developed for highly
specific enzymes, have no straight forward analogon for CAZymes.
We show how statistical thermodynamics can be employed to concisely describe
and explain the action of CAZymes, where the mixing entropy of the reactants
emerges as an important state variable of metabolic systems. We can thus
correctly predict equilibrium distributions and explain their dependence on the
initial conditions, which has not been possible with previous approaches.
Experimentally verified stochastic simulations confirm the validity of our approach
outside equilibrium.
Our proposed interpretation of polydisperse pools as statistical ensembles
facilitates a new perspective to understand many enzymatic processes. With a
mathematical model of the turnover of the soluble heteroglycan pool, we illustrate
how entropy gradients are exploited constructively in vivo to establish a robust
buffering and integrating metabolic function. The latter result motivates a novel
interpretation of metabolism as an interplay of energy- and entropy-driven
processes and hints at a novel evolutionary design principle.
Literature:
[1] Kartal, O.; Mahlow, S.; Skupin, A. & Ebenhöh, O. Carbohydrate-active enzymes
exemplify entropic principles in metabolism. Mol Syst Biol 2011, 7, 542
[2] Ebenhöh, O.; Skupin, A.; Kartal, Ö.; Mahlow, S. & Steup, M. Thermodynamic
Characterisation of Carbohydrate-active Enzymes. Proceedings of the “5th International
ESCEC Symposium on Experimental Standard Conditions of Enzyme Characterizations”,
2013, p.129-147
13
Physical constraints and counting problems in metabolic networks:
exploiting duality
Daniele De Martino
Center for Life Nano Science@Sapienza, Istituto Italiano di Tecnologia, Viale
Regina Elena 291, 00161,Roma.
The implementation of thermodynamics and the search for conservation laws are
two key steps for the physical modeling of metabolic networks. The solution of
these problems relies on finding structural motifs in the network corresponding to
closed cycles of reactions (infeasible loops) and closed cycles of metabolites
(conserved pools). From a mathematical viewpoint these two problems pertain to
the same issue: we have to find the integer basis of the solution space of a linear
system, an hard computational task for genome-scale models. Nevertheless, the
dual theorems of the alternative can provide important insights. On one hand,
they connect the infeasibility of closed loops to the definition of the chemical
potentials and thus connect the search for loops to the calculation of
concentrations and free energies. On the other hand, they connect the search for
conservation laws to the calculation of production profiles (Von Neumann
problem). The dual problems (calculation of free energies and production profiles)
are easier since they consist in solving a system of linear inequalities and efficient
relaxation algorithms can be used. However the two problems pertain to different
instances of the same issue: the implementation of thermodynamics is easier
than the search for conserved pools. Some insights will be given from the
analysis of the genome-scale metabolic network model of E. Coli iAF1260, where
we provide the complete list of infeasible loops and conserved pools.
Literature
D. De Martino, M. Figliuzzi, A.De Martino, and E. Marinari, PLoS Comput. Biol. 8, (2012).
D. De Martino, Phys. Rev. E 87, (2013).
14
Tuesday, September 17
15
Enumeration of Smallest Intervention Strategies in Genome-
Scale Metabolic Networks
Steffen Klamt1 and Axel von Kamp1
1Max Planck Institute for Dynamics of Complex Technical Systems,
Sandtorstraße 1, 39106 Magdeburg, Germany
One ultimate goal of metabolic network modeling is the rational redesign of
biochemical reaction networks to optimize the production of certain compounds
by micro-organisms. Although a number of techniques based on FBA have been
proposed for this purpose, a method for systematic enumeration of the most
efficient intervention strategies in large-scale networks is still missing. Minimal
Cut Sets (MCSs), in principle, provide such an exhaustive approach, however,
their disadvantage is the combinatorial explosion in larger networks and the
requirement to compute the elementary modes (EMs) in a preprocessing step
which itself is impractical in genome-scale networks.
Here we present a new framework by which thousands of the smallest (with
respect to the fewest number of required interventions) inclusion-minimal
intervention strategies can be computed in large-scale metabolic network
models. For this we combine two approaches, namely (i) the mapping of MCSs to
EMs in a dual network, and (ii) a modified algorithm by which shortest EMs can
be effectively determined in genome-scale networks. In this way, we can identify
the smallest MCSs by calculating the shortest EMs in the dual network. Using
several realistic application examples, we demonstrate that our approach is able
to list thousands of the most efficient MCSs in genome-scale networks for a great
variety of possible intervention problems. For example, for the first time we were
able to enumerate all synthetic lethals in E.coli for combinations of up to 5
reactions. We also employed the new algorithm for computing metabolic
engineering strategies that lead to coupled biomass and ethanol synthesis in
E.coli. The results show that a large ethanol yield of 1.8 (molecules ethanol per
molecules glucose) can be guaranteed by just three knockouts, independently of
the assumption of optimal growth of the mutant.
To summarize, the strength of the presented approach lies in the possibility to
quickly (compared to other approaches) calculate the smallest intervention
strategies with neither network size nor the number of required interventions
posing major challenges.
16
An analysis of compartmental structure of central metabolic
pathways based on 13C tracer distribution in metabolites
Vitaly Selivanov, Marta Cascante
Universitat de Barcelona, Department of Biochemistry and Molecular Biology,
08028, Barcelona, Spain.
Tracing metabolites through the network of reactions in the cell by 13C labeling is
a classical method used for the evaluation of metabolic fluxes in the metabolic
pathways. However, the existence of multiple compartmentalized pools of the
same metabolite complicates such an evaluation. The complications are related
with the fact that the same compound, located in different subcellular spaces,
likely possesses compartment-specific 13C signatures, whereas the
measurements average out the compartment-specificity. We analyzed
compartmental structure of TCA cycle metabolites based on 13C isotopic isomers
distributions measured in intermediates of TCA cycle metabolites. This analysis
was performed using a kinetic model adopted for the simulations of time course
of metabolite labeling. The dynamic approach is justified by the fact that neither
labeling, nor total metabolite concentrations (even those of intracellular
intermediates of TCA cycle) do not come to a steady state during 24 h of
incubation. Methodical aspects of the adaptation of kinetic model to the
simulation of dinamics of isotopic isomer distribution are considered in detail:
i) extension of the set of metabolic fluxes to account for all the isotope
exchanges, and ii) transformation of reactions between substances into reaction
between isotopomers. To reveal the compartmental structure of TCA cycle
metabolites several schemes were considered, and the scheme consistent with
the experimental data, as it was found using model discrimination analysis,
indicated the particular compartmentation of TCA cycle metabolites.
This work was supported by the European Commission Seventh Framework Programme
FP7 (COSMOS project grant agreement n° 312941; Etherpaths KBBE-grant n°222639;
Synergy-COPD project grant agreement n° 270086 ); the Spanish Government and the
European Union FEDER funds (SAF2011-25726).
17
Improving in silico predictions in Metabolic Engineering problems
Isabel Rocha
IBB - Institute for Biotechnology and Bioengineering, Centre of Biological
Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga –
PORTUGAL
The field of Metabolic Engineering (ME) has gained a major importance, since it
allows the design of improved microorganisms for industrial applications, starting
with wild-type strains that usually have low production capabilities in terms of the
target compounds. The ultimate aim of ME is to identify genetic manipulations in
silico leading to improved microbial strains, that can be implemented using novel
molecular biology techniques. This task, however, is a complex one, requiring the
existence of reliable metabolic models for strain simulation and robust
optimization algorithms for target identification.
Strain simulation is usually performed by using Linear or Quadratic Programing
methods that assume a steady state over the intracellular metabolites. However,
most of the available genome-scale models do not allow to make good
predictions on flux distributions, ultimately leading to ineffective ME strategies.
Another important aspect associated with model predictions is the influence of
the biomass equation added to the model. Since most simulation tools require
directly or indirectly the computation of maximal biomass formation, this
composition has a great impact in the predictive power of these models.
Moreover, biomass composition is intrinsically related with essentiality
predictions.
In this talk, a detailed analysis will be presented on the present prediction power
of genome-scale metabolic models and simulation tools and on the impact of
having accurate experimental measurements for model validation. Moreover,
improved simulation and optimization tools based on different optimization
formulations and on the concept of control effective fluxes will be introduced that
allow to perform metabolic engineering tasks in a more reliable fashion.
Relevant Literature:
Gonçalves, E. et al. J Comp Biol, 19, 102, 2012.
Machado, D et al. Bioinformatics 28, i515, 2012.
Rocha I. et al. BMC Systems Biology, 4(45), 1-12, 2010.
Soons, Z. et al. PLoS One, 8(4), e61648, 2013.3
18
Modules in Metabolic Networks
Arne C. Müller1;2;3, Brett G. Olivier5;6;8, Leen Stougie5;7, Frank J. Bruggeman9,
Alexander Bockmayr1,4
1 Department of Mathematics and Computer Science, Freie Universität Berlin,
Arnimallee 6, 14195 Berlin, Germany; 2 International Max Planck Research School
for Computational Biology and Scientific Computing (IMPRS-CBSC), Max Planck
Institute for Molecular Genetics, Ihnestr 63-73, D-14195 Berlin, Germany; 3 Berlin
Mathematical School (BMS), Berlin, Germany; 4 DFG-Research Center Matheon,
Berlin, Germany; 5 Life Science, Centre for Mathematics and Computer Science
(CWI), Science Park 123, 1098 XG Amsterdam, The Netherlands; 6 Molecular Cell
Physiology, VU University, De Boelelaan 1087, 1081 HV, Amsterdam, The
Netherlands; 7 Operations Research, VU University, De Boelelaan 1085, 1081 HV,
Amsterdam, The Netherlands; 8 Netherlands Institute for Systems Biology,
Amsterdam, The Netherlands; 9 Systems Bioinformatics, VU University, De
Boelelaan 1087, 1081 HV, Amsterdam, The Netherlands
Flux balance analysis (FBA) is one of the most often applied methods on genome-
scale metabolic networks. Although FBA uniquely determines the optimal yield,
the pathway that achieves this is usually not unique. The analysis of the optimal-
yield flux space has been an open challenge. Flux variability analysis is only
capturing some properties of the flux space, while elementary mode analysis is
intractable due to the enormous number of elementary modes. However, it has
been found that the space of optimal-yield fluxes decomposes into modules [1].
These decompositions allow a much easier but still comprehensive analysis of
the optimal-yield flux space.
We will present methods on how each module can be analyzed by itself. Together
with visualization methods for the interplay of the modules, this gives an intuitive
understanding of the optimal-yield flux space of genome-scale metabolic
networks.
Using the mathematical definition of module introduced by [2], we are now able
to compute the decomposition into modules in a few seconds for genome-scale
networks. Hence, we expect the new method to replace flux variability analysis in
the pipelines for metabolic networks.
Literature
[1] Steven M. Kelk, Brett G. Olivier, Leen Stougie, and Frank J. Bruggeman. Optimal flux
spaces of genomescale stoichiometric models are determined by a few subnetworks.
Scientific Reports, 2:580, 2012.
[2] Arne Müller and Alexander Bockmayr. Flux modules in metabolic networks. Journal of
Mathematical Biology, 2013. submitted, preprint: urn:nbn:de:0296-matheon-12084.
19
A new framework for metabolic modelling under non-balanced
growth. Application to photoautotrophic microalgae in day/night
cycle.
Caroline Baroukh*,**, Rafael Muñoz-Tamayo**, Jean-Philippe Steyer*, Olivier
Bernard**,***
*Laboratoire des biotechnologies de l’environnement, INRA UR050, avenue des
étangs, 11100 Narbonne, FRANCE; (e-mail: {caroline.baroukh ; jean-philippe.steyer}@
supagro.inra.fr). **INRIA-BIOCORE, 2004 route des lucioles, 06250 Sophia-Antipolis,
FRANCE; (e-mail:{rafael.munoz_tamayo ; olivier.bernard}@inria.fr); ***LOV-UPMC-
CNRS, UMR 7093, Station Zoologique, B.P. 28, 06234 Villefranche-sur-mer, France
Metabolic modelling is a powerful tool to understand, predict and improve
bioprocesses, particularly when they imply intracellular molecules of interest.
Unfortunately, the use of metabolic models for time varying metabolic fluxes inducing
accumulation of intracellular compounds is hampered by the lack of experimental data
required to define and calibrate the kinetic reaction rates of the metabolic pathway. For
this reason, metabolic models are often used under the balanced growth hypothesis,
that is internal metabolites are assumed not to accumulate inside the microorganisms.
However, when analysing the photoautotrophic metabolism of microalgae, the
balanced growth assumption appears to be unreasonable. Indeed, because of their
synchronization of their circadian cycle on the daily light, microalgae are forced to
store energy and carbon during the day so as to continue their growth and cell
maintenance during the night. Yet, understanding microalgae carbon storage
metabolism is necessary to improve bioprocesses for the production of third-
generation biofuels. .
In this work, we propose a new dynamic modelling framework that handles the non-
balanced growth condition and hence accumulation of intracellular metabolites. The
first stage of the approach consists in splitting the metabolic network into sub-
networks, which are assumed to satisfy balanced growth condition. The left
metabolites interconnecting the sub-networks are allowed to behave dynamically.
Then, thanks to Elementary Flux Mode analysis, each sub-network is reduced to
macroscopic reactions, for which simple kinetics are assumed. The whole process
allows to obtain an Ordinary Differential Equation system predicting substrate
consumption, biomass production, products excretion and accumulation of some
internal metabolites.
Our approach was applied to the accumulation of lipids and carbohydrates of the
microalgae Tisochrysis lutea under day/night cycles. The metabolic network was split
into five sub-networks, which yielded seven macroscopic reactions. Metabolites
allowed to accumulate were glyceraldehyde-3-phosphate, glucose-6-phosphate,
carbohydrates, lipids and phosphoenolpyruvate. The resulting model described
accurately experimental data obtained in day/night conditions; it efficiently predicts the
accumulation and consumption of triacylglycerol and carbohydrates.
20
Use of genome-scale metabolic models for plant metabolism
H.L.Yuan1, B. Wijnen1, G.F. Zhou1,2, P.A.J. Hilbers1, and N.A.W. van Riel1
1 Eindhoven University of Technology, Eindhoven, Netherlands,
2 Philips Research Asia-Shanghai, Shanghai, China.
Genome-scale metabolic models (GSMM) have been extensively applied for a
wide variety of organisms. However, for plants genome-scale modeling is still in
its infancy. It has been successfully developed for Arabidopsis in the last few
years. In this study we appraise the existing GSMM for Arabidopsis by computing
biomass accumulation using flux balance analysis (FBA) and attempt to outline
potential applications for horticultural crops in the future. The predicted growth
rates for the Poolman et al. (2009) and Dal’Molin et al. (2010) model were quite
similar, despite different carbon sources were provided (glucose and sucrose,100
mmol/g*h DW). The exchange fluxes for respiration in nonphotosynthetic cells
confirmed that plants prefer NH3 in the process of nitrogen assimilation. When
carbohydrates were used as substrate for respiration the Respiratory quotient
(RQ) of Poolman’s model approached 1, which is consistent with experimental
data. For the Dal’Molin’s model the RQ is less than 1. H2S is utilized in the
Dal’Molin model as sulfur source, whereas sulfate (SO 2-
4 ) is consumed in the
Poolman’s model. The Poolman’s model seems to reflect biological reality more
closely than the Dal’Molin model. This could be due to the subcellular
compartmentation included in the Dal’Molin model, adding subcellular
compartmentation of reactions was done manually. In conclusion, to apply
constraints-based reconstruction and analysis (COBRA) methods, including FBA,
to plant metabolism requires careful analysis and possibly curation of existing
GSMM’s.
21
Design of Self-Optimizing Metabolic Networks
Herbert M Sauro
University of Washington, Department of Bioengineering, Seattle, WA, 98195-
5061, USA.
Biological organisms acclimatize to varying environmental conditions via active
self-regulation of internal gene regulatory networks, metabolic networks, and
protein signaling networks. In this talk I will discuss some of the work we have
done on designing self-optimizing metabolic networks. That is networks that
continuously adjust enzyme levels to maximize flux even though the environment
is changing. We carried out in silico evolution experiments to develop such
networks using a two-step approach: (1) We optimized metabolic networks given
a constrained amount of available enzyme. Network optimization was performed
by iteratively redistributing enzyme until flux through the network was maximized.
Optimization was performed for a range of boundary metabolite conditions to
develop a profile of optimal enzyme distributions as a function of environmental
conditions. (2) We generated and evolved randomized gene regulatory networks
to modulate the enzymes of a target metabolic pathway. The objective of the
gene regulatory networks was to produce the optimal distribution of metabolic
network enzymes given specific boundary metabolite conditions of the target
network. Competitive evolutionary algorithms were applied to optimize the
specific structures and kinetic parameters of the gene regulatory networks.
22
Are metabolic pathways optimal?
Steven J. Court, Bartlomiej Waclaw and Rosalind J. Allen
School of Physics and Astronomy, University of Edinburgh, UK
The core metabolic pathways that are central to the function of living organisms
are remarkably conserved across all domains of life. Is this because they are
optimal solutions to an evolutionary problem, or simply because they happened
to evolve by chance, very early in the history of life on Earth? We have developed
a computational algorithm that allows us to search for all possible alternatives to
existing core metabolic pathways and test their feasibility. Our approach involves
creating a chemical network containing all compounds and reactions that are
feasible from a biochemical point of view, including many which are not observed
in contemporary organisms. Applying this to the trunk pathway of glycolysis and
gluconeogenesis, we find evidence that although a large number of alternative
pathways exist, the extant versions of these pathways produce optimal metabolic
fluxes under physiologically relevant intracellular conditions.
Literature
Glycolysis and gluconeogenesis provide maximal biochemical flux. Steven J. Court,
Bartlomiej Waclaw and Rosalind J. Allen, (in preparation)
23
On the interaction between environments and metabolic network
structure: in the light of modularity
Kazuhiro Takemoto
Department of Bioscience and Bioinformatics, Kyushu Institute of Technology,
Iizuka, Kawazu 680-4, Fukuoka 820-8502, Japan
The hypothesis that variability in natural habitats promotes modular organization
is widely accepted for metabolic networks, and it is an important concept in
evolutionary biology and ecology. However, since many factors influence the
structure of the metabolic network, this hypothesis might be limited and there
may be other explanations for the change in the network modularity. Therefore,
we focus on archaea because they show variability in growth conditions such as
growth temperature, but not in habitats because of their specialized growth
conditions. We first show the absence of a relationship between network
modularity and habitat variability in archaea. Rather, growth conditions, optimal
growth temperature, in particular, affect the network modularity. Moreover, we
demonstrate, quantitatively, that metabolic network modularity can arise from
simple growth processes, independent of the change in the evolutionary goal, in
both bacteria and archaea using a qualitative, novel evolving network model
without tuning parameters. These results have begun to cast doubt on the impact
of habitat variability on the modularity. Therefore, we re-evaluated this hypothesis
using current metabolic information. We were unable to conclude that an increase
in modularity was the result of habitat variability. Although horizontal gene
transfer was also considered because it may contribute for survival in a variety of
environments, closely related to habitat variability, and is positively correlated
with network modularity, such a positive correlation was not concluded in the
latest version of metabolic networks. Furthermore, we demonstrated that the
previously observed increase in network modularity due to habitat variability and
horizontal gene transfer was probably due to a lack of available data on
metabolic reactions. Instead, we determined that modularity in metabolic
networks is dependent on species growth conditions. These findings encourage
a reconsideration of the widely accepted hypothesis, and they enhance our
understanding of adaptive and evolutionary mechanisms in metabolic networks.
24
Wednesday, September 18
25
Studies on the Coupling of Energy Management and Cellular
Metabolism
Tobias Vallon, Tom Schuhmacher and Ralf Takors
Institute of Biochemical Engineering, University of Stuttgart, Allmandring 31,
70569 Stuttgart, Germany; takors@ibvt.uni-stuttgart.de
The adenlyate energy charge (EC) is a well-known and well-accepted measure to
qualify the cellular activity status. Values above 0.8 are typically regarded as
distinctive for well-equilibrated catabolic and anabolic activities. Values below
0.8, refer to sensitive cellular viability and/or activity.
Studies on n-butanol consumption with P. putida KT2440 reveal that EC values
show the anticipated high-levels if n-butanol is used as the sole carbon source.
However, in combination with glucose, EC levels fall down to 0.4, still allowing
stable growth. 13C flux analysis unravelled intracellular flux patterns at this
condition.
Phosphate-limited production processes are often applied to limit growth in fed-
batch processes. Studying E. coli- based L-tryptophan production at these
conditions, falling EC values were observed. Flux balance studies were
successfully applied to predict cellular behaviour further allowing hypothesis why
EC values dropped to low levels during the production period of the said amino
acid.
26
Carbon fixation and a mixed diet for the intracellular tuberculosis
bacillus
Dany J V Beste1, Katharina Nöh2, Sebastian Niedenführ2, Tom A Mendum1,
Nathaniel D Hawkins3, Jane L Ward3, Michael H Beale3, Wolfgang Wiechert2 and
Johnjoe McFadden1
1Faculty of Health and Medical Sciences, Department of Microbial and Cellular
Sciences, University of Surrey, Guildford, GU2 7XH, UK.
2Forschungszentrum Jülich GmbH, IBG-1: Biotechnology & JARA-HPC, Jülich,
Germany
3National Centre for Plant and Microbial Metabolomics, Rothamsted Research,
Harpenden, Herts, AL5 2JQ, UK.
Intracellular carbon metabolism has emerged as an attractive drug target yet the
carbon sources of intracellularly replicating pathogens such as the tuberculosis
bacillus Mycobacterium tuberculosis, which causes long-term infections in one
third of the world’s population, remain mostly unknown. Here, we use a novel
systems-based approach, 13C-flux spectral analysis (FSA) to measure for the
first time the metabolic interaction between M. tuberculosis and its macrophage
host cell. We demonstrate that amino acids, C1, C2 and C3 substrates are
contributing to the intracellular nutrition of M. tuberculosis and that the C1 source
was derived from fixed CO2. The finding that intracellular M. tuberculosis has
access to diverse carbon sources within a macrophage has significant
implications for the development of therapeutics targeting intracellular bacteria
and drug-screening protocols.
27
Experimental observations on glycolysis in muscle: revisiting
physiological versus biochemical regulation in the study of metabolic
pathway topology and function.
JAL Jeneson1, JPJ Schmitz2, NAW van Riel3.
1 Laboratory for Liver, Digestive and Metabolic Disease, Beatrix Children’s
Hospital/University Medical Center Groningen, the Netherlands
2 Systems Bioinformatics Group, Department of Molecular Cell Biology, Division
of Earth and Life Sciences, Vrije Universiteit, Amsterdam, the Netherlands
3 Computational Biology Group, Department of Biomedical Technology,
Eindhoven University of Technology, the Netherlands
It has been known for decades that phosphofructokinase (PFK), a key enzyme in
the glycolytic pathway, can reversibly bind to cytoskeletal structures in red blood
cells and fish muscle (1,2). This binding and unbinding to actin has significant
effects on PFK activity (and thereby the glycolytic pathway as a whole) and is
triggered by metabolic cues arising from functional activity (e.g., contraction in
the case of muscle) (1). These findings raised questions on proper experimental
models and conditions to study glycolysis in order to elucidate and understand
its physiology as well as the cell functions it supports in living cells and tissues of
organisms. For example, the finding that pH changes steeply affected actin-
binding activity of PFK isolated from fish but not rabbit muscle led the authors to
question whether a necessary factor was perhaps missing from the muscle
extracts used in the assay (1). Here, we present real-time non-invasive
experimental observations of hexose-mono-phosphate (HMP) metabolite
concentration dynamics in human muscle during exercise and subsequent
recovery that revisit this old, standing problem. Specifically, we show that a
computational model that captured the accumulated information on pathway
topology and biochemical regulatory mechanisms for muscle glycolysis failed to
predict the experimental HMP dynamics measured non-invasively in living muscle
in situ. The aim of this contribution is to spark debate on the need for and
solutions to obtaining and incorporating information on ‘physiological’ regulation
of metabolic pathway topology and function in the MPA community.
Literature
1. T Higashi, CS Richards, K Uyeda (1979). J Biol Chem 254: 9542-50.
2. SJ Roberts, M Lowery, GN Somero (1988). J Exp Biol 137: 13-27.
28
Analysing Primary Carbon Metabolism in heterotrophic plant cells
with steady-state metabolic flux analysis
R.G. Ratcliffe
University of Oxford, Department of Plant Sciences, South Parks Road, OXFORD
OX1 3RB, UK
Steady-state metabolic flux analysis (MFA) strategies for defining plant metabolic
phenotypes will be described with reference to heterotrophic metabolism in an
Arabidopsis cell suspension culture. These studies have revealed physiologically
important features of heterotrophic metabolism, such as the robustness of the
network, the balance between cytosolic and plastidic metabolism, and the
response of the flux distribution to nutritional and other stresses. For example,
analysis of flux maps obtained from cultures grown on different sources of
nitrogen showed: (i) that flux through the oxidative pentose phosphate pathway
varied independently of the reductant demand for biosynthesis; (ii) that non-
plastidic processes made a significant and variable contribution to the provision
of reducing power for the plastid; and (iii) that the inclusion of ammonium in the
growth medium increased cell maintenance costs. These conclusions highlight
the complexity of the metabolic response to the change in nitrogen nutrition, as
well as providing support for the futile cycling model of ammonium toxicity. High
quality experimental flux maps can also be used to evaluate the predictive
accuracy of constraints-based flux balance analysis (FBA) and a recent analysis
has shown that accounting for transport and maintenance costs improved the
accuracy of the predicted fluxes in heterotrophic Arabidopsis cells substantially.
Moreover the results of this analysis suggest that an increase in cell maintenance
costs might be viewed as a metabolic signature of stress conditions. Finally a
major current challenge is to find ways of generating cell-specific flux information
in differentiated tissues, and recent progress in the implementation of an
experimental strategy based on the use of cell-specific reporter proteins will be
described.
29
Dynamic mathematical modeling of time-resolved quantitative
metabolomics data explores the metabolic flexibility of mycobacteria
Katja Tummler1, Michael Zimmermann2, Olga Schubert2, Clemens Kühn1, Ruedi
Aebersold2, Uwe Sauer2, Edda Klipp1
1Theoretical Biophysics, Humboldt Universität zu Berlin, Invalidenstraße 42, D-
10115 Berlin
2Institute of Molecular Systems Biology, ETH Zürich, Wolfgang-Pauli-Straße 16,
CH-8093 Zürich
Tuberculosis remains one of the most dangerous infectious diseases and claims
over one million deaths each year, especially in developing countries. Metabolic
flexibility is thought to be a major survival strategy of the causative agent
Mycobacterium tuberculosis (Mtb) within its human host. Mtb effectively utilizes
the wide range of limited carbon sources available within the macrophage.
Thorough understanding of the dynamic properties of the central carbon
metabolism (CCM) will point to fragile points of this highly specialized system and
help to improve therapeutic strategies.
To systematically explore the capabilities of Mtb’s CCM, a dynamic ordinary
differential equation model of the central reaction system was constructed. The
model topology, biomass composition and a set of experimentally derived kinetic
parameters were implemented based on literature. To estimate the unknown
parameters, we experimentally traced the dynamic changes of intracellular
metabolites following a switch in the major carbon source. Experimental planning
was based on model simulations and included pairwise switches between 7
different carbon sources, each entering the network in distinct regions. Enzyme
concentrations were measured where possible and incorporated into the model.
The resulting high dimensional parameter estimation problem is approached with
a likelihood based method.
The calibrated model aims to predict adaptations of Mtb in experimentally
inaccessible situations, e.g. during growth in macrophages. It allows the direct
translation of observed changes in enzyme concentrations, which can be
measured from mixed samples of bacteria and macrophages, to metabolic flux.
The interplay between transcriptional and allosteric regulation can be
characterized. Our data suggests this to be critical for propionyl detoxification
and it is likely to be relevant in other pathways. Simulations of our model will
allow for efficient experimental planning, considering e.g. the effectivity of drug
treatment strategies under various nutritional conditions.
This work was supported by the European Union FP7 program (SysteMTb 241587).
30
Computational models of cyanobacterial metabolism: Systemic
properties of phototrophic growth
Ralf Steuer
Humboldt University Berlin, Department of Biology, Institute for Theoretical
Biology (ITB), Berlin, Germany
Cyanobacteria are the only known prokaryotes capable of oxygenic
photosynthesis. Due to their capability to utilize atmospheric CO2 for the
synthesis of organic carbon compounds, the manifold biotechnological
applications of cyanobacteria are at the forefront of current global challenges.
The domestication of phototrophic microorganisms, such as cyanobacteria,
therefore remains one of the grand challenges of the 21st century.
One step towards such a domestication of cyanobacteria are integrated
experimental and computational approaches to understand the functional
properties of phototrophic growth. The focus of this contribution is to describe
the construction of computational models of cyanobacteria and the analysis of
such models using kinetic and constraint-based methods. Specifically, we follow
a modular approach to integrate the diverse cellular processes, such as the
photosynthetic light reactions, cellular metabolism, the circadian clock, and
carbon sequestration, into a coherent minimal cell model of phototrophic growth.
The minimal cell model is supplemented with a model of carbon fluxes in a
laboratory scale photobioreactor that allows to determine all CO2 related
exchange rates of a cyanobacterial culture. Taken together, our approach allows
to elucidate the internal mechanisms of carbon fixation, and will be instrumental
in future applications of cyanobacterial biotechnology.
[1] H. Knoop, M. Gründel, Y. Zilliges, R. Lehmann, S. Hoffmann, W. Lockau, R. Steuer
(2013) Flux Balance Analysis of Cyanobacterial Metabolism: The metabolic network of
Synechocystis sp. PCC 6803. PLoS Computational Biology
[2] R. Steuer, H. Knoop, and R. Machne (2012) Modelling cyanobacteria: from metabolism
to integrative models of phototrophic growth. J Exp Bot. 63(6):2259-74.
[3] C. Beck, H. Knoop, I. M Axmann and R. Steuer (2012) The diversity of cyanobacterial
metabolism: genome analysis of multiple phototrophic microorganisms. BMC Genomics.
2;13:56.
[4] H. Knoop, Y. Zilliges, W. Lockau, and R. Steuer (2010) The metabolic network of
Synechocystis sp. PCC 6803: Systemic properties of autotrophic growth. Plant Physiol.
154: 410-422.
31
Exploiting Metabolic Diversity for Industrial Biotechnology
Brian Rudd1 & David A Fell2
1BioSyntha Technology Limited, Welwyn Garden City, AL7 3AX, UK; 2Dept
Biological & Medical Sciences, Oxford Brookes University, Oxford OX3 0BP, UK.
Concerns about the depletion of non-renewable resources and the environmental
impact of the chemical industry have spurred efforts to manufacture fuels and
industrial chemicals using biological systems fed with renewable resources or
waste materials. Rather than the organisms used in traditional biotechnology,
this requires organisms that can grow at higher temperatures and use carbon
sources other than simple sugars. Hence there is a need to search out
organisms with unusual metabolic networks that can be reconfigured to produce
new products. For example, the acetogens include organisms that can ferment
the CO2 in flue gases or syngas (CO2, CO and H2) to grow, producing acetate as
a by-product. They use the Woods-Ljungdahl pathway, but the energy yield of
this pathway is small, which places limitations on the scope for diverting acetate
into larger end-product molecules. Few of these organisms have been
extensively studied, so building genome-scale metabolic models from their
genome sequences could offer a means to understand these organisms and
exploit their metabolic capabilities.
A model of the Woods-Ljungdahl pathway will be described that incorporates an
unusual energy-coupling mechanism found in some acetogens and that
illustrates the marginal ATP yields if acetyl CoA is diverted into other products.
32
Exploiting Metabolism for Algal Biotechnology
Alison G Smith, Katherine E. Helliwell, Trinh (Ginnie) Nguyen and Mark A. Scaife
Dept of Plant Sciences, University of Cambridge, Downing Street, Cambridge
CB2 3EA, UK
There is enormous potential to use microalgae as feedstocks for everything from
recombinant proteins, to biofuels, to high value chemicals, but to implement this
technology in a sustainable and economic manner, it will be necessary to
optimize many parameters, and metabolic engineering strategies will be
essential. However, in comparison with the well-developed molecular biology
approaches available for manipulation of bacteria, yeast, and even land plants,
those for algae are limited, even for the well-studied Chlamydomonas reinhardtii.
Knowledge of endogenous metabolic pathways and their regulation are likely to
provide a rich source of information and tools that can help in establishing algal
biotechnology. Our interest is in vitamin metabolism, and we have discovered two
vitamin-inducible genetic elements in Chlamydomonas: the METE promoter,
which is repressed by vitamin B
1
12 , and riboswitches in the THI4 & THIC genes,
which undergo alternative splicing in the presence of thiamine pyrophosphate2.
Here I will describe how studies of the molecular action of the riboswitches to
develop them as tools, is in turn leading to insights into regulation of thiamine
biosynthetic pathway.
Literature
[1] Croft MT, Lawrence AD, Raux-Deery E, Warren MJ, Smith AG. Algae acquire vitamin
B12 through a symbiotic relationship with bacteria. Nature (2005) 438: 90-93
[2] Croft MT, Moulin M, Webb ME, Smith AG Thiamine biosynthesis in algae is regulated by
riboswitches. Proc Natl Acad. Sci USA (2007) 104: 20770-20775
33
The AGE1.CR.pIX cell line as example system for analysis of avian
central metabolism and statistical evaluation of inter-experimental
variance
Oliver Hädicke1, Verena Lohr1,2, Yvonne Genzel1, Udo Reichl1,3, Steffen Klamt1
1 Max Planck Institute for Dynamics of Complex Technical Systems,
Sandtorstrasse 1, 39106 Magdeburg, Germany
2 ProBioGen AG, Goethestrasse 54, 13806 Berlin, Germany
3 Chair of Bioprocess Engineering, Otto von Guericke University, Universitätsplatz
2, 39106 Magdeburg, Germany
In human vaccine manufacturing some pathogens such as Modified Vaccinia
Virus Ankara, measles, mumps virus as well as influenza viruses are still
produced on primary material derived from embryonated chicken eggs.
Processes depending on primary cell culture, however, are difficult to adapt to
modern vaccine production. The avian designer cell line AGE1.CR.pIX, derived
from muscovy duck is therefore of special interest as it is can be cultivated in
chemically-defined media. To better understand vaccine production processes,
we built a stoichiometric model of the central metabolism for avian cells and
applied flux variability and metabolic flux analysis. Obtained flux distributions are
compared to distributions observed for mammalian producer cell lines (e.g. CHO,
MDCK). A distinguishing feature of the avian cell line is that glutaminolysis plays
only a minor role in energy generation and production of precursors, resulting in
low ammonia concentrations.
In cell culture process development, key parameters like metabolic rates are
monitored and analyzed to demonstrate specific advantages of one experimental
setup compared to another. In such comparisons, false-positive results are
crucial pitfalls as they can lead to time-consuming and costly changes of an
established biotechnological process. We demonstrate how well-established
statistical tools can be employed to analyze systematically different sources of
variations in metabolic rate determinations and to assess, in an unbiased way,
their implications on the significance of the observed differences. As a case
study, we evaluate differing growth characteristics and metabolic rates of the
AGE1.CR.pIX cell line cultivated in a stirred tank reactor and in a wave bioreactor.
Although large differences in metabolic rates and cell growth were expected (due
to different aeration, agitation, pH-control, etc.) and partially observed (up to 79
%), our results show that the inter-experimental variance is a major contributor to
the overall variance of metabolic rates that potentially leads to false-positive
results.
34
Thursday, September 19
35
Energetic analysis and modeling of metabolic functioning of an
eukaryotic microalgae under growth conditions in photobioreactor
G. Cogne1, A. Martzolff1, A. Bockmayr2, C.-G. Dussap3, J. Legrand1
1 GEPEA – UMR CNRS 6144, Université de Nantes, Bât. CRTT, 37 bd de
l’Université,
BP 406, F-44602 Saint-Nazaire Cedex, France (e-mail: guillaume.cogne@univ-
nantes.fr)
2 DFG Research Center Matheon, Freie Universität Berlin, Arnimallee 3,
D-14195 Berlin, Germany
3 Clermont Université, Université Blaise Pascal, EA 3866 – Laboratoire de Génie
Chimique et Biochimique, BP 10448, F-63000 Clermont-Ferrand Cedex, France
A constraint-based model for autotrophic growth of Chlamydomonas reinhardtii
was developed. Special efforts were made to gather and integrate the set of
known valuable constraints that can be expressed in biological systems such as
microalgae. The model explicitly includes thermodynamic and energetic
constraints on the functioning metabolism. A mixed integer linear programming
method was used to determine the optimal flux distributions with regard to this
set of constraints [1]. Constraint-based modeling methods that facilitate
predictions of reactant concentrations, reaction potentials, and enzyme activities
are introduced to identify putative regulatory and control sites in biological
networks by computing the minimal control scheme necessary to switch between
metabolic modes. The developed model is used to apprehend bioenergetic
adaptations in cells growing with reduced kinetic performance due to the
existence of an increasing dark zone inside photobioreactors.
Literature
[1] Cogne G., Rügen M., Bockmayr A., Titica M., Dussap C.-G., Cornet J.-F., Legrand J. A
model-based method for investigating bioenergetic processes in autotrophically growing
eukaryotic microalgae: application to the green alga Chlamydomonas reinhardtii.
Biotechnol. Prog., 2011, article in press.
36
Kinetic models for microalgal photosynthesis coupled to cell motion
in a photobioreactor
Claude Aflalo
Microalgal Biotechnology Laboratory, Blaustein Institutes for Desert Research,
Ben Gurion University of the Negev, Sede Boqer, Israel
New modeling approaches for algal growth are needed to describe dense
illuminated microalgal cultures, in which the light intensity varies in space. In
addition to the photosynthetic reactions, the motion of cells in the variable light
field must be accounted for.
A cyclic deterministic three-state model serves as the standard formalism for
photosynthesis, in which a (generic) resting reaction center is first activated, and
then irreversibly inhibited by photon absorption. The activated reaction center
may return to the resting state, after either unproductive energy dissipation, or
productive energy transduction; the inhibited reaction center may also recover to
the resting state after a separate repair process.
This model was tested on two cases: (i) static illuminated suspension at steady
state, (ii) dynamic exposure to variable light in a mixed suspension, simulated by
random cell motion. While the former represents a well-defined reference state
(approximated at low cell density), the latter relates better to real life in a
photobioreactor (substantial density).
At moderate to high irradiance/low cell density, the static approximation holds
and the model is able to fairly account for experimentally observed behavior.
However as cell density is increased, it fails to reproduce the experimentally
observed reduction in photosynthetic rate. This lack of fit cannot be amended by
allowing for cell motion. Furthermore, critical dynamic aspects of microalgal
growth such as its dependence on mixing are not supported by the model.
Only a more elaborate model involving the sequential absorption of four photons
to yield an activated reaction center, coupled to simulated stochastic motion of
particles can satisfactorily mimic the experimental results for moderate to high
density cultures.
The results of simulations using both models will be discussed in terms of their
implications on microalgal growth in commercially realistic photobioreactors.
37
Optimal states of specific rate carry flux through an elementary flux mode
Meike T.Wortel1,2,3,4, Han Peters6, Josephus Hulshof5, Bas Teusink1,2,3,4 & Frank J.
Bruggeman1,2,3,4
1Systems Bioinformatics, IBIVU, VU University, 1081 HV, Amsterdam, The
Netherlands
2Kluyver Centre for Genomics of Industrial Fermentation/NCSB, 2600 GA, Delft,
The Netherlands
3Netherlands Institute for Systems Biology (NISB), 1098 XH, Amsterdam, The
Netherlands
4Amsterdam Institute for Molecules, Medicines and Systems (AIMMS), VU
University, 1081 HV, Amsterdam, The Netherlands
5Department of Mathematics, VU University, De Boelelaan 1081a, 1081 HV
Amsterdam, The Netherlands
6Korteweg de Vries Institute for Mathematics, University of Amsterdam, Science
Park 904, Amsterdam, The Netherlands
In biotechnology, specific production rates, including specific growth rate, are
important determinants for success. However, at genome-scale, only
stoichiometric approaches are currently available, which by necessity can only
predict yields. In silico optimisation of production rates at genome-scale are at
the moment severely compromised by (i) the lack of kinetic parameters and (ii) in
silico methods to optimise large systems of non-linear equations. We present a
theoretical breakthrough that largely solves the latter computational challenge,
opening up the possibility for genome-scale exploration of kinetic models.
We proof that optimal states of specific rate always carry flux through one
specific elementary flux mode, irrespective of the kinetic peculiarities of the
enzymes involved. This simplifies the calculation of optimal states tremendously.
Strikingly, elementary flux modes are minimal pathways through a metabolic
network that depend on stoichiometry only, and thus we provide a new link
between the topology of the metabolic network and the optimal dynamic states it
can attain.
38
Crosstalk between endophytic microorganisms leads to biosynthesis
of bioactive natural products
Souvik Kusari1,†, Gail M. Preston2, and Michael Spiteller1
1Institute of Environmental Research (INFU), Department of Chemistry and Chemical
Biology, Chair of Environmental Chemistry and Analytical Chemistry, TU Dortmund,
Otto-Hahn-Str. 6, D-44221 Dortmund, Germany; 2Department of Plant Sciences,
University of Oxford, South Parks Road, Oxford OX1 3RB, United Kingdom
†Present address: Department of Plant Sciences, University of Oxford, South Parks
Road, Oxford OX1 3RB, United Kingdom. Email: souvik.kusari@plants.ox.ac.uk
Endophytic microorganisms constitute a remarkably multifarious group of
organisms ubiquitous in plants, and maintain a dynamic association with their hosts
for at least a part of their life cycle[1]. Their enormous biological diversity coupled
with their capability to biosynthesize bioactive secondary metabolites has provided
the impetus for a number of investigations on endophytes. The potential of
competent endophytes capable of biosynthesizing bioactive target and non-target
metabolites has undoubtedly been recognized. However, it is disappointing that
there is still no known breakthrough in the commercial production of these bioactive
secondary metabolites using endophytes[2]. Thus, it is important to scrutinize the
diverse crosstalk that endophytes have with coexisting endophytes, host plants,
and associated specific and non-specific pathogens[3]. The chemistry behind these
associations involved in the endophytic production of functionally important
compounds, including those mimetic to associated host plants, should be
elucidated to apprehend the ‘triggers’ of such interaction[4]. It is important to
elucidate the biosynthetic pathways in endophytes correlating to associated
organisms on a case-by-case basis to understand how the biogenetic gene clusters
are regulated and their expression is affected in planta and ex planta. The above
mentioned issues will be addressed in detail with recent examples from our group.
[1]Chem. Biol. (2012) 19, 792-798. [2]Nat. Prod. Rep. (2011) 28, 1203-1207. [3]Curr. Opin.
Microbiol. (2011) 14, 31-38. [4]Phytochemistry (2013) 91, 81-87.
39
in silico and in vitro analysis of interacting microbial consortia
Ross P. Carlson
Department of Chemical and Biological Engineering, Center for Biofilm
Engineering, Thermal Biology Institute, Montana State University, Bozeman, MT
59717
Microbes exist rarely as monocultures in nature; they exist primarily in consortia.
Consortial populations often evolve to interact via the exchange of extracellular
metabolites. Understanding the basis of these mass and energy exchanges is
foundational to understanding the operation of microbial communities that drive
many of the Earth’s biogeochemical cycles and to controlling engineered,
synthetic consortia being used for bioprocessing. This presentation explores both
in silico methods and in vitro laboratory systems for analyzing consortial
interactions.
Three in silico stoichiometric modeling methodologies were developed for
studying mass and energy flows in microbial communities. Each approach has
distinct advantages and disadvantages suitable for analyzing systems with
different degrees of complexity and different a priori knowledge. These
methodologies were tested and compared using extensive data from the
phototrophic, thermophilic mat communities at Octopus Spring in Yellowstone
National Park. The in silico models were used to explore fundamental microbial
ecology questions including the prediction and explanation for measured relative
abundances of different system bacteria. The three approaches represent a
flexible toolbox which can be adapted to study other microbial systems with a
variety of electron donors and acceptors on scales ranging from individual cells in
monoculture to ecosystems represented by metagenomic data.
The modeling efforts informed the design of in vitro engineered consortia. The
simplest consortium was comprised of two engineered, interacting E. coli strains.
The binary consortium was designed to mimic a ubiquitous, naturally occurring
ecological template of primary productivity supported by secondary
consumption. The engineered consortium utilized strategic division of labor to
optimize simultaneously multiple tasks. Consortial interactions resulted in the
emergent property of enhanced biomass productivity. Interestingly, the consortial
interactions also produced biofilms with self-assembling, laminated
microstructures. This consortium establishes an engineering paradigm which can
be adapted to existing bioprocesses to improve productivity based on a robust
ecological theme.
40
Friday, September 20
41
Integration of Metabolomics Data in Large-Scale Models of
Photosynthetic Organisms
Sabrina Kleessen, Nadine Töpfer, Anne Arnold, Robert Heise, Max Sajitz-
Hermstein, Zoran Nikoloski
Systems Biology and Mathematical Modeling Group, Max Planck Institute of
Molecular Plant Physiology, Potsdam-Golm, 14476, Germany
Recent modeling attempts of metabolism of photosynthetic organisms have
witnessed a shift from mathematical descriptions and computations with small
pathways towards investigations of genome-scale compartmentalized
tissue/organ-specific models. Following the surge of constraint-based methods
for analysis of large-scale metabolic models of prokaryotes, this modeling
framework has been adapted to gain better understanding of metabolism in
photosynthetic organisms. Potential drawbacks of the constraint-based modeling
framework include the requirement of condition-dependent objective function
and the existence of multiple optimal (steady-state) flux distributions compatible
with the imposed constraints. Here we present a class of constraint-based
methods suitable for investigating steady-state as well as non-stationary flux
distributions which allow integration of (time-resolved) metabolomics data. We
argue that integration of high-throughput data, especially about the relative or
absolute levels of metabolites, can considerably reduce the space of feasible flux
distributions that optimize a condition-specific objective function and are
compliant with the observed physiology. We illustrate the advantage of using
metabolomics data, in addition to transcriptomics and proteomics data, to
discriminate hypotheses about (in)active pathways and flux distributions based
on large-scale models of three photosynthetic organisms: Haematococcus
pluvialis under nitrogen starvation, Chlamydomonas reinhardtii with rapamycin
treatment under photoautotrophic and mixotrophic growth conditions, as well as
photorespiratory- and TCA-cycle mutants of Arabidopsis thaliana.
42
Modelling of Varaible Chlorophyll Fluorescence Originating in
Photosystem I
Dušan Lazár
Department of Biophysics, Centre of the Region Haná for Biotechnological and
Agricultural Research, Palacký University, Šlechtitel 586/11, 783 71 Olomouc–
Holice, Czech Republic
Photosystem I (PSI) is generally assumed not to emit variable chlorophyll (Chl)
fluorescence during light-induced Chl fluorescence rise (FLR), which occurs in a
time window up to 1 second under high intensity of excitation light. Therefore, the
measured FLR and its changes caused by any treatment are usually interpreted
by changes only in photosystem II (PSII) fluorescence. But examples can be
found in the literature indicating that PSI can emit variable Chl fluorescence at
least under certain conditions. As it is impossible to determine the PSI variable
Chl fluorescence in vivo solely based on experiments, a way to explore a possible
existence of PSI variable Chl fluorescence is to construct a mathematical model
of reactions occurring inside and around PSI and to simulate a hypothetical FLR.
Based on our present knowledge about the function of PSI, a detailed model
describing reac W
tions o I
cc T
urrin H
g insi D
de an RAWN
d around PSI was constructed and used
for the simulation of FLR originating exclusively in PSI. These simulations show
that PSI, in principle, can emit variable Chl fluorescence. Several in silico
experiments are performed showing the effect of particular reactions on the FLR.
The theoretical PSI variable Chl fluorescence is also compared with theoretical
variable fluorescence originating in PSII simulated on the basis of an improved
model of PSII showing that variable fluorescence originating in PSI can be as high
as 8 – 17 % of overall maximal fluorescence signal originating in both
photosystems. The overall FLR obtained as a sum of the simulated FLRs
originating in PSI and PSII shows a peak which is similar to an H-peak measured
with certain type of samples. We suggest that new experiments be planned to
prove the new concept of variable PSI fluorescence.
Supported by the Centre of the region Haná for biotechnological and agricultural research
(Grant No. ED0007/01/01).
43
Systems level responses of rice metabolism under varying conditions
Sudip Kundu1, Mark G Poolman2, Rahul Shaw1, and David A. Fell2
1Department of Biophysics, Molecular Biology, and Bioinformatics, Calcutta
University, Kolkata 700009, India
2Department of Biology and Medical Science, Oxford Brookes University,
Headington, Oxford OX3 OBP, United Kingdom
Rice (Oryza sativa) makes up nearly 20% of the total caloric intake of the world
human population. To assist the efforts of rice biotechnologists’ to design more
efficient rice cultivars, our aim is to understand the rice metabolism.
Here, we shall present a partially compartmentalized genome scale metabolic
model of rice, primarily built based on the RiceCyc database. The flux balance
analysis shows that the model is capable of producing biomass precursors (in
experimentally reported proportions) from its inorganic nutrients with the help of
light energy. We further simulate this model at varying light intensities and
observe the redox shuttles between the chloroplast, cytosol and mitochondrion
at low light levels. The results also indicate that the photorespiration can act to
dissipate excess energy at high light levels. We further observe a variation in the
mitochondrial metabolism at different light intensities.
We would also like to present the in-silico reaction deletion study of this
metabolic model. Results indicate that while deleting non-essential reaction there
are alternative possible metabolic pathways to produce the biomass in the fixed
proportion; however, sometimes the photon demands vary significantly. It also
demonstrates the interaction between different compartments.
44
Integration of Genome-Scale Modeling and Transcript Profiling
Reveals Metabolic Pathways Underlying Light and Temperature
Acclimation in Arabidopsis
Nadine Töpfer1, Camila Caldana2, Sergio Grimbs3, Lothar Willmitzer4, Alisdair R.
Fernie5, Zoran Nikoloski1
1Systems Biology and Mathematical Modeling Group, Max Planck Institute of
Molecular Plant Physiology, 14476 Potsdam-Golm, Germany
2Brazilian Bioethanol Science and Technology Laboratory (CNPEM/ABTLuS),
Campinas, Brazil
3Institute of Biochemistry and Biology, University of Potsdam, 14476 Potsdam-
Golm, Germany
4Genes and Small Molecules Group, Max Planck Institute of Molecular Plant
Physiology, 14476 Potsdam-Golm, Germany
5Central Metabolism Group, Max Planck Institute of Molecular Plant Physiology,
14476 Potsdam-Golm, Germany
Understanding metabolic acclimation of plants to challenging environmental
conditions is essential for dissecting the role of metabolic pathways in growth
and survival. As stresses involve simultaneous physiological alterations across all
levels of cellular organization, a comprehensive characterization of the role of
metabolic pathways in acclimation necessitates integration of genome-scale
models with high-throughput data. Here, we present an integrative optimization-
based approach, which, by coupling a plant metabolic network model and
transcriptomics data, can predict the metabolic pathways affected in a single,
carefully controlled experiment. Moreover, we propose three optimization-based
indices that characterize different aspects of metabolic pathway behavior in the
context of the entire metabolic network. We demonstrate that the proposed
approach and indices facilitate quantitative comparisons and characterization of
the plant metabolic response under eight different light and/or temperature
conditions. The predictions of the metabolic functions involved in metabolic
acclimation of Arabidopsis thaliana to the changing conditions are in line with
experimental evidence and result in a hypothesis about the role of homocysteine-
to-cysteine interconversion and asparagine biosynthesis. The approach can also
be used to reveal the role of particular metabolic pathways in other scenarios,
while taking into consideration the entirety of characterized plant metabolism.
45
The tomato fruit development with a constraint-based model
Sophie Colombié1, Christine Nazaret2, Bertrand Beauvoit1, Marion Solé1, Martine
Dieuaide-Noubhani1, Benoit Biais1, Jean-Pierre Mazat3 and Yves Gibon1
1INRA, Univ. Bordeaux, UMR1332 Fruit Biology and Pathology unit, F-33883
Villenave d’Ornon, France
2IMB, Université of Bordeaux 2, 3ter place de la Victoire. F-33076 Bordeaux
Cedex France
3University of Bordeaux 2, 146 rue Léo Saignat, F-33076, Bordeaux France
Commercial fruit production is under significant pressure from environmental
stresses but also by changes in the consumer’s demand for taste and nutritional
value. One of the key goals of fruit biology is therefore to understand the factors
that influence the levels of metabolites in cells and tissues, ultimately with a view
to manipulate these levels of improvement of fruit traits. To this aim, building a
metabolic model of tomato fruit is very helpful.We intend to model changes in
metabolic fluxes that occur in the tomato fruit during the three main phases
(division, expansion and ripening) of its development. We developed a metabolic
model encompassing central metabolism (about 70 reactions and 40 metabolites
including glycolysis, pentose phosphate pathway, TCA cycle, sugars cycle and
biomass synthesis…) and assumed a metabolic steady state (no change in
concentrations of internal metabolites) at each stage of fruit development.
A large number of metabolite concentrations have been determined at each
phase of fruit development at INRA Bordeaux (sugars, organic acids, amino
acids), and have been completed with data for significant biomass constituents
(protein content, cell wall, lipids, nucleic acids…). The best fit of these ‘external’
metabolite concentrations has been searched in order to precisely determine the
rates towards these metabolites, rates used as bounds of the corresponding
fluxes in the network. Some other fluxes of the metabolic network were
constrained with the maximal capacity of the enzymes that have also been
determined experimentally at each stage. Then, the constraint-based modelling
was applied to solve the metabolic system applying an objective function, which
is to minimize the sum of fluxes, leading to a minimum distribution of fluxes. We
looked up how fluxes change during fruit development and analysed our results
in term of carbon, nitrogen and energy metabolism. We also compared our results
with the behaviours of kinetic or ecophysiological models.
46
Poster ABSTRACTS
47
Poster number 01
On the interaction between environments and metabolic network
structure: in the light of modularity
Kazuhiro Takemoto
Department of Bioscience and Bioinformatics, Kyushu Institute of Technology,
Iizuka, Kawazu 680-4, Fukuoka 820-8502, Japan
The hypothesis that variability in natural habitats promotes modular organization
is widely accepted for metabolic networks, and it is an important concept in
evolutionary biology and ecology. However, since many factors influence the
structure of the metabolic network, this hypothesis might be limited and there
may be other explanations for the change in the network modularity. Therefore,
we focus on archaea because they show variability in growth conditions such as
growth temperature, but not in habitats because of their specialized growth
conditions. We first show the absence of a relationship between network
modularity and habitat variability in archaea. Rather, growth conditions, optimal
growth temperature, in particular, affect the network modularity. Moreover, we
demonstrate, quantitatively, that metabolic network modularity can arise from
simple growth processes, independent of the change in the evolutionary goal, in
both bacteria and archaea using a qualitative, novel evolving network model
without tuning parameters. These results have begun to cast doubt on the impact
of habitat variability on the modularity. Therefore, we re-evaluated this hypothesis
using current metabolic information. We were unable to conclude that an increase
in modularity was the result of habitat variability. Although horizontal gene
transfer was also considered because it may contribute for survival in a variety of
environments, closely related to habitat variability, and is positively correlated
with network modularity, such a positive correlation was not concluded in the
latest version of metabolic networks. Furthermore, we demonstrated that the
previously observed increase in network modularity due to habitat variability and
horizontal gene transfer was probably due to a lack of available data on
metabolic reactions. Instead, we determined that modularity in metabolic
networks is dependent on species growth conditions. These findings encourage
a reconsideration of the widely accepted hypothesis, and they enhance our
understanding of adaptive and evolutionary mechanisms in metabolic networks.
48
Poster number 02
An analysis of multistationary and oscillatory behavior of
mitochondrial respiratory chain interacting with central energy
metabolism
Vitaly Selivanov, Marta Cascante
Universitat de Barcelona, Department of Biochemistry and Molecular Biology,
08028, Barcelona, Spain.
Mitochondrial respiratory chain (RC) performs an important step of transformation
of chemical energy that is necessary for ATP synthesis, and also generates
reactive oxygen species (ROS) that play signaling/destructive role on the cell. The
RC is integrated in the central energy metabolism oxidizing NADH produced in
glycolysis and TCA cycle, and performing a step of TCA cycle oxidizing
succinate.
A detailed mathematical model of the RC revealed the multistationarity of the RC
operation as a characteristics inherent in the mechanism of electron transport.
The existence of more than one steady states at the same conditions (values of
parameters) practically means that the RC can operate in various modes.
Operating in normal mode it effectively transports electrons to reduce oxygen to
H2O, and translocating protons from the matrix to the intermembrane space.
Operating in this mode the RC produces small amount of ROS. Also it can
operate in a different mode, significantly reducing oxygen to superoxide radical
thus producing ROS, and less efficiently translocating protons. Transient hypoxia
can be a factor switching the RC from normal to ROS producing mode. This
finding allows explaining the intrinsic mechanisms of such phenomena as the
increased ROS production observed in reoxygenation after hypoxia. The
respiratory chain operation is sensitive to the state of whole central energy
metabolism, and, vice versa, it produces signals affecting the whole cell
metabolism through the common effectors, as NADH and ROS. Moreover,
interaction of the RC with central energy metabolism can result in more complex
characteristics, such as oscillatory behavior of whole energy metabolism.
This work was supported by the European Commission Seventh Framework Programme
FP7 (COSMOS project grant agreement n° 312941; Etherpaths KBBE-grant n°222639;
Synergy-COPD project grant agreement n° 270086 ); the Spanish Government and the
European Union FEDER funds (SAF2011-25726).
49
Poster number 03
From metabolic interventions to combinational drug therapy: fast and
efficient calculation of minimal hitting sets
Christian Jungreuthmayer (1,2), and Jürgen Zanghellini (1,2)
1 Austrian Centre of Industrial Biotechnology
2 Department of Biotechnology, University of Natural Resources and Life
Sciences, Vienna, Austria
The calculation of minimal hitting sets (also called minimal cut sets) is a frequent
problem in combinatorial mathematics with many applications in computer
science, machine learning but also in the live sciences. In fact, diverse problems
such as calculating minimal intervention strategies in metabolic engineering,
predicting minimal combinational drug therapies in cancer research, or finding
optimal multi-genome alignments are hitting set problems. Here, we present our
open source software ``MCScalculator’’, which allows for a fast calculation of
hitting sets. MCScalculator provides two different solver for predicting hitting
sets. One is based on a classical algorithm (Berge algorithm) while the other is
based on integer programing. However, our implementation uses advanced pre-
processing techniques and bit-pattern trees for efficient and fast calculations. We
show that ``MCScalculator’’ is able to solve problems containing tens of
thousands of variables (nodes) or billions of lines (edges), and outperforms
currently existing software.
50
Poster number 04
What is the “best” deletion strategy to construct a metabolic network
of minimal functionality?
David Ruckerbauer (1,2), Christian Jungreuthmayer (1,2), and Jürgen Zanghellini (1,2)
1 Austrian Centre of Industrial Biotechnology
2 Department of Biotechnology, University of Natural Resources and Life
Sciences, Vienna, Austria
Microbial cell factories are an increasingly important source of biotechnologically
produced chemicals. Typically however, microorganisms need to be genetically
modified in order to produce the product of interest as efficiently as possible.
Several tools for the prediction on rational strain designs are available. In
particular, Hädicke and Klamt [1] recently introduced the concept of constrained
minimal cut sets (cMCS), which allows predicting minimal genetic interventions.
The method is based on the analysis of the complete set of elementary flux
modes. In our contribution, we show that the problem of finding cMCS can be
(re-)formulated as in integer program [2]. Not only is the integer program formally
equivalent to the original concept of cMCS, it also adds extra flexibility to the
design process as it allows integrating regulation in the design process as well.
Moreover (and in contrast to Hädicke and Klamt’s implementation), given an
objective criterion (e.g. most efficient production of a product of interest) our
integer program will always return the best possible strain design for any (desired)
number of genetic interventions without a preceding analysis, which first
identifies desirable and undesirable network states. This allows identifying an
optimal deletion sequence for constructing metabolic networks of minimal
functionality. That is, at any step in the deletion strategy, the mutant will show
optimally shaped metabolic capabilities. We will illustrate our approach using a
classical example, that is the design of an minimal E. coli cell for the most
efficient production of ethanol [3].
Literature
[1] O Hädicke and S Klamt, Metab Eng. 2011 Mar;13(2):204-13. PMID: 21147248
[2] C Jungreuthmayer and J Zanghellini, BMC Syst Biol. 2012; 6: 103. PMCID:
PMC3560272
[3] CT Trinh, P Unrean, and F Srienc, Appl Environ Microbiol. 2008; 74(12): 3634–3643.
PMCID: PMC2446564
51
Poster number 05
Energetic analysis and modeling of metabolic functioning of an
eukaryotic microalgae under growth conditions in photobioreactor
G. Cogne1, A. Martzolff1, A. Bockmayr2, C.-G. Dussap3, J. Legrand1
1 GEPEA – UMR CNRS 6144, Université de Nantes, Bât. CRTT, 37 bd de
l’Université, BP 406, F-44602 Saint-Nazaire Cedex, France (e-mail:
guillaume.cogne@univ-nantes.fr)
2 DFG Research Center Matheon, Freie Universität Berlin, Arnimallee 3, D-14195
Berlin, Germany
3 Clermont Université, Université Blaise Pascal, EA 3866 – Laboratoire de Génie
Chimique et Biochimique, BP 10448, F-63000 Clermont-Ferrand Cedex, France
A constraint-based model for autotrophic growth of Chlamydomonas reinhardtii
was developed. Special efforts were made to gather and integrate the set of
known valuable constraints that can be expressed in biological systems such as
microalgae. The model explicitly includes thermodynamic and energetic
constraints on the functioning metabolism. A mixed integer linear programming
method was used to determine the optimal flux distributions with regard to this
set of constraints [1]. Constraint-based modeling methods that facilitate
predictions of reactant concentrations, reaction potentials, and enzyme activities
are introduced to identify putative regulatory and control sites in biological
networks by computing the minimal control scheme necessary to switch between
metabolic modes. The developed model is used to apprehend bioenergetic
adaptations in cells growing with reduced kinetic performance due to the
existence of an increasing dark zone inside photobioreactors.
Literature
[1] Cogne G., Rügen M., Bockmayr A., Titica M., Dussap C.-G., Cornet J.-F., Legrand J. A
model-based method for investigating bioenergetic processes in autotrophically growing
eukaryotic microalgae: application to the green alga Chlamydomonas reinhardtii.
Biotechnol. Prog., 2011, article in press.
52
Poster number 06
modelBorgifier, a Tool for Semi-automated Assimilation of Metabolic
Reconstructions
John T. Sauls and Joerg M. Buescher
BRAIN Aktiengesellschaft, Microbial Production Technologies Unit, Quantitative
Biology and Sequencing Platform, Darmstädter Str. 34-36, 64673 Zwingenberg,
Germany
Phone: +49 6251 9331 64, jrb@brain-biotech.de, www.brain-biotech.de
Historically, genome-scale metabolic reconstructions come in a great variety of
file formats and use different naming schemes for reactions and metabolites.
While SBML is on good track to become the standard file format, no generally
accepted naming scheme has emerged, yet. However, a unified naming scheme
greatly simplifies sharing of reaction and metabolite-associated information
across reconstructions, comparing and combining reconstructions, and
integrating reconstructions with measured data.
We developed modelBorgifier to facilitate the process of assimilating
reconstructions. In an automated first step, a pairwise comparison of all reactions
and metabolites in 40 parameters such as ID, name, and stoichiometry is
computed. This scoring is the basis for a GUI-guided user oversight of the semi-
automated reaction- and metabolite matching. Machine learning is iteratively
employed to enhance the automated scoring based on user input. ModelBorgifier
is implemented in Matlab and integrates seamlessly with the popular
openCOBRA toolbox. We demonstrate the advantage of assimilated models with
examples of industrial applications of stoichiometric modeling.
53
Poster number 07
Modules in Metabolic Networks
Arne C. Müller1;2;3, Brett G. Olivier5;6;8, Leen Stougie5;7, Frank J. Bruggeman9,
Alexander Bockmayr1,4
1 Department of Mathematics and Computer Science, Freie Universität Berlin,
Arnimallee 6, 14195 Berlin, Germany; 2 International Max Planck Research School
for Computational Biology and Scienti_c Computing (IMPRS-CBSC), Max Planck
Institute for Molecular Genetics, Ihnestr 63-73, D-14195 Berlin, Germany; 3 Berlin
Mathematical School (BMS), Berlin, Germany; 4 DFG-Research Center Matheon,
Berlin, Germany; 5 Life Science, Centre for Mathematics and Computer Science
(CWI), Science Park 123, 1098 XG Amsterdam, The Netherlands; 6 Molecular Cell
Physiology, VU University, De Boelelaan 1087, 1081 HV, Amsterdam, The
Netherlands; 7 Operations Research, VU University, De Boelelaan 1085, 1081 HV,
Amsterdam, The Netherlands; 8 Netherlands Institute for Systems Biology,
Amsterdam, The Netherlands; 9 Systems Bioinformatics, VU University, De
Boelelaan 1087, 1081 HV, Amsterdam, The Netherlands
Flux balance analysis (FBA) is one of the most often applied methods on genome-
scale metabolic networks. Although FBA uniquely determines the optimal yield,
the pathway that achieves this is usually not unique. The analysis of the optimal-
yield flux space has been an open challenge. Flux variability analysis is only
capturing some properties of the flux space, while elementary mode analysis is
intractable due to the enormous number of elementary modes. However, it has
been found that the space of optimal-yield fluxes decomposes into modules [1].
These decompositions allow a much easier but still comprehensive analysis of
the optimal-yield flux space.
We will present methods on how each module can be analyzed by itself. Together
with visualization methods for the interplay of the modules, this gives an intuitive
understanding of the optimal-yield flux space of genome-scale metabolic
networks.
Using the mathematical definition of module introduced by [2], we are now able
to compute the decomposition into modules in a few seconds for genome-scale
networks. Hence, we expect the new method to replace flux variability analysis in
the pipelines for metabolic networks.
Literature
[1] Steven M. Kelk, Brett G. Olivier, Leen Stougie, and Frank J. Bruggeman. Optimal flux
spaces of genomescale stoichiometric models are determined by a few subnetworks.
Scientific Reports, 2:580, 2012.
[2] Arne Müller and Alexander Bockmayr. Flux modules in metabolic networks. Journal of
Mathematical Biology, 2013. submitted, preprint: urn:nbn:de:0296-matheon-12084.
54
Poster number 08
Metabolic Databases and the Benefits of Rule-Based Annotation
Stephan Richter1,2, Ingo Fetzer3, Martin Thullner2, Florian Centler2 and Peter
Dittrich1
¹Friedrich-Schiller-Universität, Ernst-Abbe-Platz 2, 07743 Jena, Germany
²Helmholtz-Centre for Environmental Research - UFZ, Permoserstr. 15, 04318
Leipzig, Germany
³Stockholm Resilience Centre, Stockholm University, Kräfriket 2B, 11419
Stockholm, Sweden
Current research in biology and biochemistry utilizes huge amounts of qualitative,
quantitative and descriptive data. Such data are derived from experiments and
model simulations, and in turn are is used to create models, parameterize
experiments and enable simulations and predictions, whose predictions in turn
can gain further insights. Therefore such data has to be complied, stored, and
made available for use by the scientific community in an organized manner.
For biological, chemical, and metabolic data several online databases featuring
various kinds of data have emerged in the last decades and are serving this
purpose.
Today, many researchers in bioinformatics, systems biology and related scientific
fields benefit from those databases and rely on their content. As we pass the step
from small application-centered network models to whole cell investigations and
further to models covering multi-species assemblages, there is an increasing
need to systematically take care of the reliability of the utilized data:
For smaller models, scientists were able to spot and manually curate the limited
number of utilized datasets. With the growing size of models, the number of
employed datasets increases and a manual curation of integrated knowledge
becomes infeasible. Moreover, the formatting capabilities provided for storage in
online databases becomes a limiting factor for the facility of assembled models.
Models can only be general to the extend allowed by the form of the given data
they are based on.
In this talk, we would like to present an automated, qualitative and quantitative
analysis of common shortcomings of metabolic databases. We will further
discuss how these shortcomings will hinder model assembly and how they may
be resolved.
Most examples are based on the prominent KEGG database, but are transferable
to similar online databases.
55
Poster number 09
Kinetic models for Microalgal photosynthesis coupled to cell motion
in a photobioreactor
Claude Aflalo
Microalgal Biotechnology Laboratory, Blaustein Institutes for Desert Research,
Ben Gurion University of the Negev, Sede Boqer, Israel
New modeling approaches for algal growth are needed to describe dense
illuminated microalgal cultures, in which the light intensity varies in space. In
addition to the photosynthetic reactions, the motion of cells in the variable light
field must be accounted for.
A cyclic deterministic three-state model serves as the standard formalism for
photosynthesis, in which a (generic) resting reaction center is first activated, and
then irreversibly inhibited by photon absorption. The activated reaction center
may return to the resting state, after either unproductive energy dissipation, or
productive energy transduction; the inhibited reaction center may also recover to
the resting state after a separate repair process.
This model was tested on two cases: (i) static illuminated suspension at steady
state, (ii) dynamic exposure to variable light in a mixed suspension, simulated by
random cell motion. While the former represents a well-defined reference state
(approximated at low cell density), the latter relates better to real life in a
photobioreactor (substantial density).
At moderate to high irradiance/low cell density, the static approximation holds
and the model is able to fairly account for experimentally observed behavior.
However as cell density is increased, it fails to reproduce the experimentally
observed reduction in photosynthetic rate. This lack of fit cannot be amended by
allowing for cell motion. Furthermore, critical dynamic aspects of microalgal
growth such as its dependence on mixing are not supported by the model.
Only a more elaborate model involving the sequential absorption of four photons
to yield an activated reaction center, coupled to simulated stochastic motion of
particles can satisfactorily mimic the experimental results for moderate to high
density cultures.
The results of simulations using both models will be discussed in terms of their
implications on microalgal growth in commercially realistic photobioreactors.
56
Poster number 10
Are metabolic pathways optimal?
Steven J. Court, Bartlomiej Waclaw and Rosalind J. Allen
School of Physics and Astronomy, University of Edinburgh, UK
The core metabolic pathways that are central to the function of living organisms
are remarkably conserved across all domains of life. Is this because they are
optimal solutions to an evolutionary problem, or simply because they happened
to evolve by chance, very early in the history of life on Earth? We have developed
a computational algorithm that allows us to search for all possible alternatives to
existing core metabolic pathways and test their feasibility. Our approach involves
creating a chemical network containing all compounds and reactions that are
feasible from a biochemical point of view, including many which are not observed
in contemporary organisms. Applying this to the trunk pathway of glycolysis and
gluconeogenesis, we find evidence that although a large number of alternative
pathways exist, the extant versions of these pathways produce optimal metabolic
fluxes under physiologically relevant intracellular conditions.
Glycolysis and gluconeogenesis provide maximal biochemical flux. Steven J. Court,
Bartlomiej Waclaw and Rosalind J. Allen, (in preparation)
57
Poster number 11
Integration of Genome-Scale Modeling and Transcript Profiling
Reveals Metabolic Pathways Underlying Light and Temperature
Acclimation in Arabidopsis
Nadine Töpfer1, Camila Caldana2, Sergio Grimbs3, Lothar Willmitzer4, Alisdair R.
Fernie5, Zoran Nikoloski1
1Systems Biology and Mathematical Modeling Group, Max Planck Institute of
Molecular Plant Physiology, 14476 Potsdam-Golm, Germany
2Brazilian Bioethanol Science and Technology Laboratory (CNPEM/ABTLuS),
Campinas, Brazil
3Institute of Biochemistry and Biology, University of Potsdam, 14476 Potsdam-
Golm, Germany
4Genes and Small Molecules Group, Max Planck Institute of Molecular Plant
Physiology, 14476 Potsdam-Golm, Germany
5Central Metabolism Group, Max Planck Institute of Molecular Plant Physiology,
14476 Potsdam-Golm, Germany
Understanding metabolic acclimation of plants to challenging environmental
conditions is essential for dissecting the role of metabolic pathways in growth
and survival. As stresses involve simultaneous physiological alterations across all
levels of cellular organization, a comprehensive characterization of the role of
metabolic pathways in acclimation necessitates integration of genome-scale
models with high-throughput data. Here, we present an integrative optimization-
based approach, which, by coupling a plant metabolic network model and
transcriptomics data, can predict the metabolic pathways affected in a single,
carefully controlled experiment. Moreover, we propose three optimization-based
indices that characterize different aspects of metabolic pathway behavior in the
context of the entire metabolic network. We demonstrate that the proposed
approach and indices facilitate quantitative comparisons and characterization of
the plant metabolic response under eight different light and/or temperature
conditions. The predictions of the metabolic functions involved in metabolic
acclimation of Arabidopsis thaliana to the changing conditions are in line with
experimental evidence and result in a hypothesis about the role of homocysteine-
to-cysteine interconversion and asparagine biosynthesis. The approach can also
be used to reveal the role of particular metabolic pathways in other scenarios,
while taking into consideration the entirety of characterized plant metabolism.
58
Poster number 12
Propagation of Bounds on Gibbs Free Energy of Formation
Arne C. Mueller1;2;3
1 Department of Mathematics and Computer Science, Freie Universit at Berlin ,
Arnimallee 6, 14195 Berlin, Germany
2 International Max Planck Research School for Computational Biology and
Scienti_c Computing (IMPRS-CBSC), Max Planck Institute for Molecular
Genetics, Ihnestr 63-73, D-14195 Berlin, Germany
3 Berlin Mathematical School (BMS), Berlin, Germany
Reaction irreversibilities are important in constraint-based methods like Flux
balance analysis (FBA). Irreversibilities exclude thermodynamically infeasible
pathways from the space of feasible solutions. However, reaction irreversibilities
are usually the result of manual curation. This can be a tedious and error-prone
process. Hence, we are interested in automated methods for irreversibility
detection. If the Gibbs free energy of formation of every metabolite would be
known, then reaction directions can be easily computed using the second law of
thermodynamics. However, the Gibbs free energy of formation is usually not
known precisely (e.g. because of unknown metabolite concentration, errors in the
estimation of the equilibrium constants). We investigate systems methods to
derive non-heuristic reaction irreveribilities where the irreversibility cannot be
directly derived from the thermodynamic information given. In contrast to
heuristic methods [1] the computed irreversibilities will not block all
thermodynamically infeasible pathways, but the method guarantees that there will
exists no thermodynamically feasible pathway that violates one of the computed
irreversibilities. We analyze our method on E. coli iAF1260 where all internal
reactions are initially assumed to be reversible.
Literature
[1] Anne Kuemmel, Sven Panke, and Matthias Heinemann. Systematic assignment of
thermodynamic constraints in metabolic network models. BMC Bioinformatics, 7:512,
2006.
59
Poster number 13
Kinetic Modelling of the Metabolism and Compartmentation of
Sugars During Tomato Fruit Development
Bertrand Beauvoit1,2, Sophie Colombié1, Christine Nazaret2, Marie-Hélène
Andrieu1, Martine Dieuaide-Noubhani1,2, Benoit Biais1, Jean-Pierre Mazat2 and
Yves Gibon1
1INRA, UMR1332 Biologie du Fruit et Pathologie, F-33140 Villenave d’Ornon,
France
2University of Bordeaux 2, 146 rue Léo Saignat, F-33076, France
In a previous work, changes in the metabolite and enzyme composition of tomato
fruit were analysed as a function of development, thus allowing the description of
several stage-specific enzymes clusters. Here we intend to analyse the control of
sucrose metabolism during fruit growth by integrating these experimental data
into a kinetic model.
The core of the model is based on the sugarcane model of Uys et al. (2007) with
significant modifications. Fourteen enzyme-catalysed reactions were
parameterized assuming reversible and irreversible Michaelis-Menten rate laws
and combining about 80 fixed parameters taken from the literature and about 20
measured parameters. Moreover, the model takes into account the time-course
evolution of the vacuole volume, of the enzyme Vmax and of the relative growth
rate. Finally, the values of the sucrose uptake and of the Vmax of vacuolar
carriers have been adjusted to fit the measured sucrose, glucose and fructose
contents.
After the resolutions of the ODEs system, for each stage of fruit development, the
system reaches a steady-state where the sucrose uptake balances the glycolysis,
the starch and cell wall synthesis and the growth-dependent sugar accumulation.
Output of the model is a set of 12 concentrations and 24 fluxes at each stage.
The first interesting result is that most of the enzymes are working at less than
5% of their maximal capacity regardless of the developmental stage. Also, the
behaviour of the model has been analysed under conditions where the
mechanism, the specificity and the affinity of the vacuolar carriers were
modulated. In addition, the effect of a progressive shift of the carbon unloading
from the symplasmic to the apoplasmic pathway has been investigated. Overall,
our calculations pinpoint the crucial role of the vacuolar sucrose- and hexose-
proton antiporters in the control of the sugar accumulation into the vacuole.
60
Poster number 14
Metabolic phenotype profiling of Dickeya species, a significant
pathogen of potato
Souvik Kusari1,†, Sarah L. McCraw2, Leighton Pritchard3, Gail M. Preston2,
Nicholas J. Kruger2, and R. George Ratcliffe2
1Institute of Environmental Research (INFU), Department of Chemistry and
Chemical Biology, Chair of Environmental Chemistry and Analytical Chemistry, TU
Dortmund, Otto-Hahn-Str. 6, D-44221 Dortmund, Germany; 2Department of Plant
Sciences, University of Oxford, South Parks Road, Oxford OX1 3RB, United
Kingdom; 3Information and Computational Sciences, James Hutton Institute,
Dundee DD2 5DA, Scotland, United Kingdom
†Present address: Department of Plant Sciences, University of Oxford, South
Parks Road, Oxford OX1 3RB, United Kingdom. Email:
souvik.kusari@plants.ox.ac.uk
Successful colonization of plant tissues by pathogenic microorganisms is a
constant threat to world agriculture. The incidence and development of plant
disease is dependent on the uptake and assimilation of nutrients by pathogens
from their host plants, and the ability of pathogens to synthesize and secrete
virulence factors that disable plant defenses[1]. Both pathogen growth and
virulence gene expression are strongly affected by host metabolites and strain-
specific alterations in pathogen metabolism[2,3]. Thus, understanding the
metabolic phenotypes[4,5] of pathogens during growth inside plant tissues could
lead to innovative strategies to prevent plant disease. Dickeya spp. (D. solani, D.
dadanti, and D. dianthicola) are plant pathogenic bacteria that causes tuber soft
rots, wilt and blackleg, and D. solani has recently emerged as a major concern for
potato growers in Europe, Israel and the UK[6]. This highly aggressive pathogen,
formerly belonging to the genus Erwinia, is a member of the Enterobacteriaceae,
which includes the model organism Escherichia coli. We are developing
metabolic flux maps for the Dickeya species and are further attempting to
demonstrate the viability of cell-specific labeling approaches for performing
stable-isotope metabolic flux analysis (13C-MFA) in Dickeya during infection. The
results of our ongoing studies on probing the metabolic phenotypes of Dickeya
spp. will be presented.
Literature
[1]Chem. Biol. (2012) 19, 792-798. [2]Mol. Plant-Microbe. Interact. (2008) 21, 269-282. [3]Curr.
Opin. Microbiol. (2011) 14, 31-38. [4]Plant J. (2006) 45, 490-511. [5]J. Exp. Bot. (2012) 63,
2243-2246. [6]Plant Pathol. (2011) 60, 385–399.
61
Poster number 15
Multiscale metabolic modeling: dynamic flux balance analysis on a
whole plant scale
Eva Grafahrend-Belau1, Astrid Junker1, André Eschenröder3, Johannes Müller3,
Falk Schreiber1,2,5 and Björn Junker1,4
1 Leibniz Institute of Plant Genetics and Crop Plant Research Gatersleben (IPK),
Corrensstrasse 3, D-06466 Gatersleben, Germany
2 Institute of Computer Science, Martin Luther University Halle-Wittenberg, Von-
Seckendorff-Platz 1, D-06120 Halle, Germany
3 Institute of Agricultural and Nutritional Sciences, Martin Luther University Halle-
Wittenberg, Betty-Heimann-Strasse 5, D-06120 Halle, Germany
4 Institute of Phamacy, Martin Luther University Halle-Wittenberg, Hoher Weg 8,
D-06120 Halle, Germany
5 Clayton School of Information Technology, Monash University, Victoria 3800,
Australia
Plant metabolism is characterized by a unique complexity on the cellular, tissue
and organ level. On a whole plant scale changing source- and sink-relations
accompanying plant development add another level of complexity to metabolism.
With the aim of getting a spatio-temporal resolution of source-sink interactions in
crop plant metabolism, a multiscale metabolic modeling (MMM) approach was
applied which integrates static organ-specific metabolic models with a whole
plant dynamic model. Allowing for a dynamic flux balance analysis (dFBA) on a
whole plant scale, the MMM approach was used to decipher the metabolic
behavior of source and sink organs during the generative phase of the barley
plant.
62
Poster number 16
Stoichiometric modeling approach for organisms interaction
Kambiz Baghalian (a), Eva Grafahrend-Belau (a), Falk Schreiber (a,b)
(a) Leibniz Institute of Plant Genetics and Crop Plant Research(IPK),
Corrensstrasse 3, 06466 Gatersleben, Germany
(b) MartinLuther University Halle-Wittenberg, Institute of Computer Science, Von-
Seckendorff-Platz 1, 06120, Halle, Germany
Constraint-based modelling is a promising approach for analysing the structure,
dynamics, and behaviour of complex metabolic networks. Stoichiometric
modelling is one of the most widely used methods to estimate the flux
distribution in cellular systems during steady-state situations. This approach has
been successfully applied to models on the scale of cells, tissues, and very
recently multi-organs. So far this approach has been focused on metabolic
modelling within single organisms. However, there are several important
biological cases that occur based on microorganism-organism interactions and
are of vital importance for human life. Better understanding of these interactions
will be very helpful in areas such as human health (pathogen-human interactions),
plant production (plant-microbe interactions), environmental issues (global
biogeochemical cycling of energy and- carbon), and so on. The major challenge
in these studies is the complexity of the interaction network that occurs between
the partner organisms. The aim of this project is to determine whether
stoichiometric modelling approaches can efficiently help us to understand how
metabolism is impacted during the interaction of two organisms. This includes
also questions on how to reduce the size and complexity of metabolic interaction
networks while preserving their predictive power. The stoichiometric modelling of
the interaction between simple and complex organisms will provide a better
understanding of the relevant metabolic processes and build the base for
applying engineering or optimizing strategies in a way that is more compatible
with human goals and requirements. As a highly relevant test case for such an
integrated modelling approach, we will investigate the symbiotic nitrogen fixation
(SNF) between legume and bacteria.
63
Poster number 17
Damage analysis of the global catabolic core network in Salmonella
Typhimurium
Hassan Hartman, Mark G. Poolman, David A. Fell
Cell Systems Modelling Group, Department of Medical and Biological Sciences,
Faculty of Health and Life Science, Oxford Brookes University
Because of widespread antibiotic-resistance among bacteria, there is a need for
new antimicrobials. This is particularly urgent for Gram-negative pathogens.
Salmonella Typhimurium is a commonly used model organism for this group of
pathogens. During infection S. Typhimurium is able to use a broad range of
nutrients for growth and energy, which in combination with a robust metabolic
network makes identification of metabolic enzymes suitable as targets for drug
development difficult. Here we identify sets of reactions that when removed from
the network will impair its ability to generate energy, regardless of nutrient
availability.
It has previously been shown that catabolic core models can be extracted from
genome-scale models (GSMs), by varying the rate of ATP hydrolysis using linear
programming to identify reaction fluxes that co-varies with ATP hydrolysis. In
these analyses a single carbon source has been assumed. Here the analysis was
repeated for several carbon sources and the superset of responsive reactions
analysed. All reaction-sets in the core model of size one to three that when
removed from the network abolished energy generation were identified. The
carbon sources used in this analysis had all been verified to support growth. In
order to use as many growth supporting carbon sources as possible, this
selection was based upon data obtained using Phenotype Microarray technology,
which allows simultaneous characterisation of a microorganisms ability to grow
on hundreds of nutrient sources.
The results suggest that a total of 500 reactions are needed to generate energy
on all carbon sources. This could be reduced to 150 reactions by removing those
that were unique for a given carbon source. Damage analysis suggested 63
reactions sets, involving two or three reactions, that caused significant damage to
the GSM. The majority of the lethal sets involved reactions associated with
glycolysis, Entner-Doudoroff and the pentose phosphate pathways.
64
Poster number 18
Use of genome-scale metabolic models for plant metabolism
Yuan, H.L.1, Wijnen B.1, Zhou, G.F.1,2, Hilbers, P.A.J.1, and N.A.W. van Riel1
1. Eindhoven University of Technology, Eindhoven, Netherlands,
2. Philips Research Asia-Shanghai, Shanghai, China.
Genome-scale metabolic models (GSMM) have been extensively applied for a
wide variety of organisms. However, for plants genome-scale modeling is still in
its infancy. It has been successfully developed for Arabidopsis in the last few
years. In this study we appraise the existing GSMM for Arabidopsis by computing
biomass accumulation using flux balance analysis (FBA) and attempt to outline
potential applications for horticultural crops in the future. The predicted growth
rates for the Poolman et al. (2009) and Dal’Molin et al. (2010) model were quite
similar, despite different carbon sources were provided (glucose and sucrose,100
mmol/g*h DW). The exchange fluxes for respiration in nonphotosynthetic cells
confirmed that plants prefer NH3 in the process of nitrogen assimilation. When
carbohydrates were used as substrate for respiration the Respiratory quotient
(RQ) of Poolman’s model approached 1, which is consistent with experimental
data. For the Dal’Molin’s model the RQ is less than 1. H2S is utilized in the
Dal’Molin model as sulfur source, whereas sulfate (SO 2-
4 ) is consumed in the
Poolman’s model. The Poolman’s model seems to reflect biological reality more
closely than the Dal’Molin model. This could be due to the subcellular
compartmentation included in the Dal’Molin model, adding subcellular
compartmentation of reactions was done manually. In conclusion, to apply
constraints-based reconstruction and analysis (COBRA) methods, including FBA,
to plant metabolism requires careful analysis and possibly curation of existing
GSMM’s.
65
Poster number 19
Heterotrophic plant cell network analysis: comparison between EFMs
and MCSs methods
Nguyen Vu Ngoc Tung, Beurton-Aimar Marie and Colombie Sophie.
LaBRI, Laboratoire Bordelais de Recherche en Informatique, UMR5800, Univ
Bordeaux, France
Faculty of Science and Technology, Hoa Sen University, Vietnam.
INRA Bordeaux, UMR1332, Fruit Biology and Pathology, France.
Plant metabolic networks are known to be complex with isoenzymes, redundant
pathways, and metabolite exchanges between subcellular compartments. For
that, a topological analysis can be helpful to better understand the metabolic
properties of the considered network.
Different tools have emerged last ten years to analyze such metabolic networks
at the cell levels. Searching feasible pathways has one of the first ones, using
different methods like Elementary Flux Modes (EFMs) or Extreme Pathways
computations (Schuster et al. 1999 and Papin et al. 2002) . More recently a dual
approach called minimal cut sets (MCSs), has been proposed (Klamt, 2006) to
identify minimal set of reactions able to prevent the execution of an objective
reaction. EFMs computation is well-known to produce a large dataset,
nevertheless, MCSs are supposed to be less numerous than EFMs. Then it could
be easiest to analyze MCS than EFMs. We have performed our analysis with a
plant cell network containing 78 reactions (with 33 reversible ones) and 55
metabolites (Beurton et al 2011) by comparing both, EFMs and MCSs results. We
have obtained 114, 614 EFMs and 93,009 MCSs. Even this difference seems to
be not really significative, one can note that the average length of EFMs (37.7) is
substantially different from the average size of MCSs (8.5) Then, we have
focused on dataset with output metabolites selected for their relevance in plant
metabolism such as glucose, starch…. For instance the number of feasible
pathways involved in the production of glucose is 19,608, whereas 5,500 MCSs
of reactions have been identified as able to stop the glucose accumulation. In this
case, interestingly, the import of sucrose is systematically present in the MCSs
selected for glucose accumulation but not always present in the MCSs selected
for starch accumulation. From a physiological point of view, it means that prevent
glucose accumulation is possible only if sugar imported by the cell is blocked.
66
Poster number 20
Metabolic modelling for optimizing hydrocarbon yield in E. coli
Zia Fatma1, Hassan Hartman2, Mark. G. Poolman2, David. A. Fell2, Syed Shams
Yazdani1
1-Synthetic Biology and Biofuels Group, International center of genetic
engineering and biotechnology, New-Delhi, India
2-Cell Systems Modelling Group, Department of Medical and Biological
Sciences, Oxford Brookes University2, UK
Alternative transportation fuels derived from biomass (biofuels) are in demand
because of availability of declining fossil fuels, and increasing energy demand
due to population growth, industrialization and prosperity. Hydrocarbon alkanes
are considered important next-generation biofuels owing to their chemical
similarity to contemporary diesel and jet fuels, which makes them compatible
with existing advanced internal combustion engines. However, low yield of
alkanes in biotechnological production is a major obstacle for commercialization.
The aim of this work was to identify genes suitable for deletion or overexpression
that would force flux towards alkane biosynthesis. This is done by using
established methods from structural modelling.
The analysis was done using a structural model of E. coli which is based on
model published in Trinh et al (2008). Model contains 76 reactions (including
transporter) and 82 metabolites. To identify genes suitable for overexpression,
variations in alkane production were simulated using linear programming (LP).
Reactions that co-varied with the imposed alkane production were considered
candidates for overexpression. This yielded set reactions, of which majorities
were associated with Glycolysis and Pentose phosphate pathway (PPP).
Since forcing increased flux through certain reactions will force the flux towards
different carbon sink apart from alkane, candidates for deletion could be
identified based on their association with carbon sink. A total of 7 deletions were
identified this way. Here, we have shown that by imposing increased flux through
1 reaction and removing 7 reactions from the network, and assuming only flux
minimization as the objective, we obtain an optimal solution that has an alkane
yield which is 97% of the theoretical maximum.
67
Poster number 21
Dynamic mathematical modeling of time-resolved quantitative
metabolomics data explores the metabolic flexibility of mycobacteria
Katja Tummler1, Michael Zimmermann2, Olga Schubert2, Clemens Kühn1, Ruedi
Aebersold2, Uwe Sauer2, Edda Klipp1
1Theoretical Biophysics, Humboldt Universität zu Berlin, Invalidenstraße 42, D-
10115 Berlin
2Institute of Molecular Systems Biology, ETH Zürich, Wolfgang-Pauli-Straße 16,
CH-8093 Zürich
Tuberculosis remains one of the most dangerous infectious diseases and claims
over one million deaths each year, especially in developing countries. Metabolic
flexibility is thought to be a major survival strategy of the causative agent
Mycobacterium tuberculosis (Mtb) within its human host. Mtb effectively utilizes
the wide range of limited carbon sources available within the macrophage.
Thorough understanding of the dynamic properties of the central carbon
metabolism (CCM) will point to fragile points of this highly specialized system and
help to improve therapeutic strategies.
To systematically explore the capabilities of Mtb’s CCM, a dynamic ordinary
differential equation model of the central reaction system was constructed. The
model topology, biomass composition and a set of experimentally derived kinetic
parameters were implemented based on literature. To estimate the unknown
parameters, we experimentally traced the dynamic changes of intracellular
metabolites following a switch in the major carbon source. Experimental planning
was based on model simulations and included pairwise switches between 7
different carbon sources, each entering the network in distinct regions. Enzyme
concentrations were measured where possible and incorporated into the model.
The resulting high dimensional parameter estimation problem is approached with
a likelihood based method.
The calibrated model aims to predict adaptations of Mtb in experimentally
inaccessible situations, e.g. during growth in macrophages. It allows the direct
translation of observed changes in enzyme concentrations, which can be
measured from mixed samples of bacteria and macrophages, to metabolic flux.
The interplay between transcriptional and allosteric regulation can be
characterized. Our data suggests this to be critical for propionyl detoxification
and it is likely to be relevant in other pathways. Simulations of our model will
allow for efficient experimental planning, considering e.g. the effectivity of drug
treatment strategies under various nutritional conditions.
This work was supported by the European Union FP7 program (SysteMTb 241587).
68
Poster number 22
Accounting for maintenance costs in flux balance analysis improves
the prediction of plant cell metabolic phenotypes under stress
conditions
C.Y. Maurice Cheung1, Thomas C. R. Williams2, Mark G. Poolman3, David. A.
Fell3, R. George Ratcliffe1 and Lee J. Sweetlove1.
1 Department of Plant Sciences, University of Oxford, South Parks Road, Oxford,
OX1 3RB, UK
2 Universidade de Brasília, Departamento de Botânica, Brasília, 70910-000
3 School of Life Sciences, Oxford Brookes University, Headington, Oxford OX3
0BP, UK
Flux-balance models of metabolism generally utilise synthesis of biomass as the
main determinant of intracellular fluxes. However, the biomass constraint alone is
not sufficient to predict realistic fluxes in central heterotrophic metabolism of
plant cells because of the major demand on the energy budget from transport
costs and cell maintenance. This major limitation can be addressed by
incorporating transport steps into the metabolic model and by implementing a
procedure that uses Pareto optimality analysis to explore the trade-off between
ATP and NADPH production for maintenance. This leads to a method for
predicting cell maintenance costs on the basis of the measured flux ratio
between the oxidative steps of the OPPP and glycolysis. We show that
accounting for transport and maintenance costs substantially improves the
accuracy of fluxes predicted from a flux-balance model of heterotrophic
Arabidopsis cells in culture, irrespective of the objective function used in the
analysis. Moreover when the new method was applied to cells under control,
elevated temperature and hyperosomotic conditions, only elevated temperature
led to a substantial increase in cell maintenance costs. It is concluded that the
hypersomotic conditions tested did not impose a metabolic stress in so far as the
metabolic network is not forced to devote more resources to cell maintenance.
69
Poster number 23
Virtual Mitochondrion
Jean-Pierre Mazat1, Christine Nazaret2, Stéphane Ransac1, Margit Heiske1,3
1IBGC CNRS UMR 5095 & Universite Bordeaux 2 1, rue Camille Saint-Saëns
33077 BORDEAUX-cedex, France ; 2 Institut de Mathématiques de Bordeaux,
ENSTBB-Institut Polytechnique de Bordeaux 3, Place de la Victoire 33076
Bordeaux Cedex ; 3 Institut für Biologie, Theoretische Biophysik, Humboldt-
Universität zu Berlin, Invalidenstr. 42, 10115 Berlin, Germany
Virtual Mitochondrion is a project of a multilevel modelling of mitochondrial
bioenergy metabolism. It involves:
- A molecular/atomic level with stochastic nanodynamic modelling (Gillespie) of
electrons and protons transfers in respiratory the chain complexes and super
complexes of respiratory chain A stochastic approach is particularly well adapted
to describe the time course of the redox reactions that occur inside the respiratory
chain complexes, because they involve only a few number of electrons (1-10). It
allowed us to predict a natural bifurcation of electrons in complex III (proof of Q-
cycle hypothesis of Mitchell) to clarify the antimycin inhibition constraints and to
simulate the ROS production in complex I and III. The stochastic approach also
permits to jump to the upper level of enzyme kinetics.
- A mitochondrial level with the global modelling of the respiratory chain using
simple but thermodynamical correct kinetics equations developed for the
respiratory chain complexes (Michaelis-Menten like equations with the
introduction of the proton gradient). The aim is to understand how local changes
(pathological mutations for instance, drug effect, competition between respiratory
substrates) in respiratory complexes influence the global behaviour of the
oxidative phosphorylation. (In collaboration with Edda Klipp, Berlin).
- A cell level with the description of simple(s) model(s) of central energy
metabolism easy to manipulate and to understand. The aim is to coherently
integrate various types of data, metabolomics, fluxomics, transcriptomics etc. or
to evidence inconsistencies revealing new mechanisms.
- At the cell level we would like to model the dynamics of mitochondrial network
(fusion/fission) and its dependency upon the energy metabolism.
Conclusion
We would like to emphasize the connection between lower and upper levels: how
the functioning at a given level explains (or does not explain) the functioning at
the upper (more integrated) level? Thus, in this work, the purpose of a model is
not only to fit the experimental results accurately but rather to evidence
inconsistencies that will lead to unveil mechanisms or properties which were
hitherto not taken into account or even unknown.
70
Literature
- Stéphane Ransac, Margit Heiske and Jean-Pierre Mazat : From in silico to in spectro
Kinetics of Respiratory Complex I. Biochim. Biophys. Acta 2012, 1817: 1958-1969.
- Jean-Pierre Mazat, Jonathan Fromentin, Margit Heiske, Christine Nazaret, Stéphane
Ransac : Virtual Mitochondrion: Towards An Integrated Model Of Oxidative
Phosphorylation Complexes and Beyond. Biochem. Soc. Trans. (2010) 38, 1215-1219.
- Stéphane Ransac, Jean-Pierre Mazat : How antimycin inhibits the bc1 complex? A part-
time twin. Biochim Biophys Acta. 2010 1797: 1849-1857.
- Stéphane Ransac, Clément Arnarez and Jean-Pierre Mazat : The flitting of electrons in
complex I: A stochastic approach. Biochim Biophys Acta. 2010 . ; 1797: 641-648.
- F. Medja, S. Allouche, P. Frachon, C. Jardel, M. Malgat, B. Mousson de Camaret, A.
Slama, J. Lunardi, J.P. Mazat, A. Lombès (2009) Development and implementation of
standardized respiratory chain spectrophotometric assays for clinical diagnosis
Mitochondrion 9 (2009) 331–339.
- Christine Nazaret, Margit Heiske, Kevin Thurley, Jean-Pierre Mazat : Mitochondrial
energetic metabolism : A simplified model of TCA cycle with ATP production. J Theor Biol.
2009 Jun 7;258(3):455-64.
- Ransac, S., Parisey, N. and Mazat J.-P.: The loneliness of electrons in the bc1 complex.
Biochim Biophys Acta. 2008 Jul-Aug;1777(7-8):1053-9.
Keywords: metabolism modelling, stochastic nano-dynamics, complex I, complex III (bc1
complex), mitochondria, oxidative phosphorylation.
71
Poster number 24
Development of a carbon neutral biofuel process
Luz P. Gomez de Cadiñanos, Pedro Peña, Friedrich Srienc.
Department of Chemical Engineering and Material Science, University of
Minnesota, 421 Washington Avenue SE, Minneapolis, MN 55455. Biotechnology
Institute, University of Minnesota, 1479 Gortner Avenue, St. Paul, MN 55108.
Ralstonia eutropha H16 is a facultative chemolithoautotrophic bacterium that has
served as a model organism for polyhydroxyalkanoates (PHAs) metabolism but
its metabolic network also enables the synthesis of useful biofuels and chiral
chemicals. This versatile microorganism is able to grow using CO2 as sole
carbon source when H2 and O2 are present to provide the energy for the fixation
of CO2. Therefore we may control the carbon flux in metabolic engineered R.
eutropha strains to obtain higher efficiency microbes (HEMs) for the conversion of
CO2 into ethanol and isobutanol. A metabolic model for R. eutropha has been
developed and Elementary Modes Analysis (EMA) has been used to identify the
minimum number of enzymes to be knocked out for developing an optimised
strain for the production of the desired compounds. Preliminary data suggest that
8 to 12 gene knockout (KO) mutations would reduce the number of possible
pathways to the desired production ones. The model has been validated by
knocking out the ldh gene which codifies for the main lactate dehydrogenase
activity in R. eutropha, leading to a non-lethal mutant. This gene is also one of the
predicted KOs to reduce the number of possible pathways to obtain HEMs for
the desired compounds. In all biotechnology processes that are based on
heterotrophic growth, the carbon yield is substantially reduced because a third
part of the carbon present in sugars is released as CO2. Using a co-culture
system, monitored by mass spectrometry, in which Sacharomyces cerevisiae
yeasts produce the CO2 that the developed R. eutropha can fix and convert into
additional ethanol or isobutanol, we can achieve a full conversion of the sugar
carbon into the desired product.
72
Poster number 25
Unraveling the Systems Biology of Grain Fillling in Rice
Marlen Müller, Navreet Bhullar, Wilhelm Gruissem
Group of Plant Biotechnology, Institute of Agricultural Sciences, ETH Zürich
Rice (Oryza sativa L.), together with maize and wheat, constitutes an important
cereal worldwide but popular rice mega-varieties lack sufficient key
micronutrients, vitamins and a balanced amino acid composition that are
essential for a healthy diet. The major bottleneck for improving the nutritional
quality of popular rice varieties through conventional breeding or gene technology
is our lack of an integrated understanding of the biochemical and molecular
processes that occur during rice grain filling.
We investigate what limits the import of micronutrients into the endosperm
through metabolic model reconstruction of rice coupled with integration of tissue-
and time- dependent OMICS data. In the first part, molecular expression profiling
on specific tissues of the rice grain as well as the flag leaves is being performed,
specifically during the grain filling period. Laser-microdissection is used to obtain
sections of different tissues comprising the grain. The dissected samples will be
subjected to RNA-Sequencing and the subsequent data analysis is expected to
provide a basis for comprehensive understanding of the molecular processes
operating in the specific tissues of the rice grain and at different developmental
stages. Thereafter, a constraint-based genome-scale metabolic model of the
developing rice seed based on a combination of mathematical modeling (Flux
Balance Analysis) with biochemical, genetic and genomic knowledge will be
constructed. For this, a model transfer of the genome-scale metabolic model of
Arabidopsis thaliana developed in our group to the rice organism will be
performed. By integrating the OMICS data, tissue- and time- specific metabolic
changes will be observed.
This project will contribute relevant knowledge, required to further improve the
nutritional value of rice grain and yield, therefore, contributing to global
improvement of human nutrition and health.
73
Poster number 26
Metabolism of cell cycle
Elahe Radmaneshfar1, Matteo Barberis2, Mark Petronczki3 and Markus Ralser4
1 Institute of complex and mathematical biology, University of Aberdeen, UK
2 Swammerdam Institute for Life Sciences, University of Amsterdam, Netherland
3 Cell Division and Aneuploidy Laboratory, Cancer Research UK, London
Research Institute
4 Molecular Biology of Metabolism Laboratory, Department of Biochemistry,
University of Cambridge, UK
Building a new cell is an extraordinary task that requires a myriad of metabolic
and biosynthetic reactions throughout the cell cycle. The cell cycle requires the
faithful duplication and partitioning of the genome. The cell cycle of eukaryotes
consists of four phases (G1, S, G2 and M), during which the cell substantially
increases in size, replicates its genome and finally divides. The metabolism
network provides and controls the nutrients and energy required for specific
phases and activities throughout the cell cycle.
The molecular basis of cell cycle regulation and the control of cellular metabolism
have been studied in isolation. However, for obvious reasons these processes
require coordination. For instance, the synthesis of DNA requires high levels of
nucleotides, while during cell division, an extreme level of lipids is required for
building novel membranes. Misregulation of either metabolism or cell cycle
control can have fatal consequences on the physiology of an organism.
In this talk I will present the progress on systematically quantification the
metabolism at distinct cellular events and cell cycle stage by using metabolic
pathway analysis.
74
Poster number 27
Towards the virtual chloroplast
Gilles Curien, Christophe Bruley, Florence Combes, Myriam Ferro, Marianne
Tardiff, Yves Vandenbrouck, Giovanni Finazzi and Norbert Rolland
From CNRS, UMR 5168, CEA, DSV, iRTSV, Laboratoire de Physiologie Cellulaire
et Végétale, 17 rue des Martyrs, F-38054 Grenoble, France-INSERM, Laboratoire
d’Etude de Dynamique des Protéomes, U880, Université Grenoble 1, F-38054
Grenoble, France.
The chloroplast is a great example of a complex and integrated metabolic
network that produces a high number of metabolites of industrial interest (sugars,
vitamins, lipids and pigments). One way to improve our knowledge of such a
“metabolic factory” and how it can be successfully modulated by synthetic
biology is to build automatic metabolic pathways reconstructions. Unfortunately,
current knowledge of the plastidial metabolism is still dispersed in the scientific
literature. Existing protein and metabolite databases do not allow quantitative
estimation of metabolic fluxes and do not take into account the suborganellar
localization of each reaction or the metabolite trafficking between cell
compartments.
We therefore decided to create a virtual chloroplast, by manually integrating all
the qualitative and quantitative data currently available. It will contain a user-
friendly interface allowing rapid visualization and virtual modulation of metabolic
fluxes, for research or teaching purposes.
We started by building a series of metabolic maps of the Arabidopsis thaliana
chloroplast by using the software Cell Designer. At present, this preliminary
database contains 30 maps for a total of 400 reactions and metabolites. It allows
easy access to the subplastidial localization of each reaction, EC numbers, gene
and metabolite identifiers, link to bibliographic sources and semi-quantitative
data on protein abundance obtained from the AT_CHLORO database.
These maps are already extremely useful, since they allow identifying the
metabolic steps that still need to be characterized at the gene level and provide a
better understanding of the cross-talk between different metabolisms.
75
Poster number 28
Cellular heterogeneity confounds steady-state metabolic flux
analysis in complex systems in the absence of cell-specific
labelling data
Nicholas J. Kruger, R. George Ratcliffe, Merja T. Rossi, Chaiyakorn
Srisakvarakul
Department of Plant Sciences, University of Oxford, Oxford, OX1 3RB, UK
Steady-state 13C-metabolic flux analysis (13C-MFA) is a well-established method
for quantifying flux through the reactions of the central network of carbon
metabolism in micro-organisms. The approach is based on deducing the flux
values that best describe the measureable redistribution of label within the
intermediates and products generated by a network supplied with a non-
uniformly labelled [13C]substrate. Application of this approach to multicellular
eukaryotic systems is more challenging, in part because of the presence of
heterogenous tissues. These typically contain multiple cell types that are likely to
differ in the amount and composition of biosynthetic products and in the fluxes
through the central metabolic network needed to support these differing
biosynthetic activities. To date, most analyses of such systems have ignored
such heterogeneity and have sought to obtain flux maps that report the
aggregate metabolic activity of the tissue. Here we test the assumption implicit in
this approach; namely, that the flux map generated from the average labelling
patterns of individual metabolites in the tissue represents the aggregate activity
of the metabolic networks of the component cell types of the tissue. In silico
analyses of model networks using 13CFLUX2 establish that measurements of the
average isotopomer abundance of specific metabolites do not provide a reliable
description of the aggregate flux through a combination of networks, except in
the simplest of systems containing a single variable flux. This is a consequence
of the non-linear relationship between fluxes and the isotopomer composition of
metabolite pools encountered in typical metabolic networks. We conclude that
accurate flux determination in heterogenous metabolic systems is likely to require
additional cell-type specific information.
76
Poster number 29
Resources for plant metabolic network analyses: Gramene, VitisCyc,
and FragariaCyc
Sushma Naithani1, Justin Elser1, Rajani Raja1, Palitha Dharmawardhana1, Justin
Preece1, Vindhya Amarasinghe1, Guanming Wu2, Peter D’Eustachio3, Marcela
Monaco4, Dylan Beorchia1, Kindra Amoss1, David Croft5, Robin Haw2, Lincoln
Stein2, Doreen Ware4,6 and Pankaj Jaiswal1
1Oregon State University, Dept. of Botany and Plant Pathology, Corvallis, OR,
97331, 2Ontario Institute for Cancer Research, Toronto, M5G0A3, Canada, 3NYU
School of Medicine, Dept. of Biochemistry & Molecular Pharmacology, New York,
NY, 10016, 4Cold Spring Harbor Laboratory,New York, NY, 11724, 5European
Bioinformatics Institute, Cambridge, CB10 1SD, United Kingdom, 6USDA ARS
NAA, Robert W. Holley Center for Agriculture and Health, Ithaca, NY, 14853
Pathway databases – when integrated with the wide array of available plant
genomic data sets – can provide a framework for understanding the metabolism
of a plant species and how it is affected by a variety of abiotic and biotic
stresses. Gramene (http://www.gramene.org/pathway) hosts ten plant pathways
databases: the Gramene-developed RiceCyc (Oryza sativa japonica),
SorghumCyc (Sorghum bicolor), MaizeCyc (Zea mays) and BrachyCyc
(Brachypodium distachyon), and the mirrored databases AraCyc (Arabidopsis
thaliana), MedicCyc (Medicago truncatula), PoplarCyc (Populus trichocarpa),
CoffeaCyc (Coffea canephora), LycoCyc (Solanum lycopersicum) and PotatoCyc
(Solanum tuberosum). These BioCyc-based databases facilitate searches for
gene products, enzymes, reactions, metabolites, and pathways; cross-species
comparisons; and upload and analyses of expression datasets using the
Omicsviewer tool.
Naithani Laboratory, in collaboration with the Jaiswal Lab at Oregon State
University, has additionally developed metabolic Cyc pathway databases for two
crops: grapevine (VitisCyc:
http://palea.cgrb.oregonstate.edu:1555/jaiswallab/vitiscyc.shtml) and strawberry
(FragariaCyc:
http://palea.cgrb.oregonstate.edu:1555/jaiswallab/fragariacyc.shtml). We expect
VitisCyc and FragariaCyc to become a resource for evidence based, well
annotated secondary metabolic pathways in fleshy fruits, and to facilitate
understanding the regulation of non-climacteric fruit ripening and associated
metabolic processes.
Gramene is developing Plant Reactome (http://plantreactome.oicr.on.ca) in
collaboration with its international partners and the human Reactome project. The
Plant Reactome platform extends the BioCyc-based pathway databases in
77
several useful ways: 1) it facilitates graphical display of the pathway networks in
the context of their subcellular locations, 2) it has additional analytical capability
including pathway overrepresentation analysis, inter-species comparison and
overlaying of expression and gene-gene interaction data, 3) it links pathways,
reactions and gene entries to external resources like UniProt, PIR, ChEBI,
PubChem, PubMed and the Gene Ontology (GO), and 4) it allows curation,
annotation and connection between metabolic and regulatory pathways.
Currently, Plant Reactome features rice, and Arabidopsis. We expect Plant
Reactome to become a primary investigative resource for plant pathways,
network analysis and community curation.
78
Poster number 30
Investigating kinetic parameters that couple metabolic and gene
expression networks: Mapping epistasis
Zachary A. King1, Joshua A. Lerman1, Edward J. O’Brien1, Bernhard Ø. Palsson1
1. Department of Bioengineering, University of California, San Diego, 9500
Gilman Drive MC0412, La Jolla, 92093-0412, CA, USA
The metabolic networks of model organisms can be analyzed using constraint-
based modeling, and this approach has been useful for predicting metabolic flux
states, gene knockout phenotypes and essentiality, and genetic designs for
bioprocessing strains. Constraint-based models of metabolism have a number of
drawbacks (e.g. they do not express the enzyme-cost of metabolic pathways)
that can be addressed by expanding the model scope to include other cellular
networks. We report on the use of an expanded constraint-based model that
incorporates metabolism (M) with the gene expression network (E) in Escherichia
coli. The reconstructed gene expression network includes all the machinery for
gene transcription and translation. The gene-protein relationship in the “ME”
model is now explicitly represented in the conversion of genes to mRNA and
mRNA to enzymes, which catalyze metabolic reactions. In order to compute flux
states with the ME model, one must consider the effective catalytic efficiency of
metabolic enzymes in order to determine how much enzyme is necessary to
catalyze the simulated flux through a reaction. These effective catalytic
efficiencies are model parameters that represent condition-specific enzyme
activity, and they affect predictions of metabolic phenotypes. Thus, by sampling
the space of possible effective catalytic efficiencies, we can make predictions of
how enzyme kinetics will impact global metabolic phenotypes. We demonstrate
epistatic effects between the effective catalytic efficiencies of enzymes in the ME
model for E. coli—effects cannot be recapitulated by analyzing just the metabolic
network. We then categorize metabolic enzyme interactions according to the
character of the epistasis. The classification of epistatic interactions is important
for understanding the evolutionary history of E. coli, because epistasis has been
shown to impact evolutionary trajectories, and also for selecting targets for
metabolic and enzyme engineering.
79
Poster number 31
Substrate use and robustness in human retinoic acid metabolic
pathways
Jennifer R. Chase
Department of Biology, Northwest Nazarene University, Nampa, Idaho, USA
Retinoic acids participate in cellular regulation of differentiation to apoptosis in a
variety of human tissues. Retinoic acids (atRA, 9cRA, 4oRA) are made in target
tissues from protein-bound retinoids in the blood. Their cellular levels are a
complex function of nutritional intake, storage and catabolism of esters.
Dysregulation of these pathways has been implicated in alcohol-related diseases
(e.g, cancer, fetal terratogenesis). While most studies have focused on
understanding the role of single participating proteins, it is now recognized that
only when the whole system of enzymes, binding proteins, and compartments
are taken into account will we be able to assess when and where atRA
biosynthesis occurs. Therefore, two different models of human retinoic acid
metabolism have been constructed, based on KEGG hsa00830 (38 reactions) and
HumanCyc pathways (24 reactions, including 3 with isoenzymes). KEGG includes
retinoid catabolic pathways, but 2 of the 3 retinoid sources had to be added. For
the HumanCyc model, retinol and atRA pathways were combined and
compartmentation (with transport steps) was added, but no additional retinoid
sources were added. NAD(P)H turnover was also added to each model. The
systems were evaluated using CellNetAnalzyer to identify EFMs and robustness,
which is the reciprocal of reaction participation. The KEGG system had over 5000
EFMs but the “Body” model had only 10, if one ignores isoenzymes. There were
significant differences in robustness depending on the retinoid source and model.
The robustness of both systems for (retin)aldehyde dehydrogenases was very
low. The KEGG model also had very low robustness for NAD(H) turnover. These
results suggest that these relatively less-studied enzymes and their interactions
with the system may be significant sites of dysregulation by ethanol in the
development of disease.
80
Poster number 32
Toward Multiscale Metabolic Network Analysis of an Anaerobic
Microbial Community
Kristopher Hunt, (kristopher.hunt@biofilm.montana.edu) Zach Adam, Tisza Bell,
Laura Camilleri, James Connolly, Egan Lohman, Alex Michaud, Heidi Smith, Reed
Taffs, Michelle Tigges, James Folsom, Ross Carlson (rossc@coe.montana.edu),
Matthew Fields, Christine Foreman, Robin Gerlach, William Inskeep,
Montana State University, Bozeman MT 59717
Microbial communities represent a growing interest to many disciplines including
ecology, microbiology and engineering. Understanding microbial interactions
within these communities provides fundamental insight into properties such as
nutrient and energy cycling and provides a rational basis for the optimization of
technologies such as anaerobic biogas synthesis. One method for probing
community behaviors, including microbial composition and product yields, is
through the construction of stoichiometric metabolic models and their
investigation using elementary flux mode analysis. Unfortunately, this metabolic
modeling method has limits regarding the complexity of systems it can examine;
the method is generally limited to simplified, sub-genome scale analyses. Here,
an automated, demand-based dissection of the metabolic networks into smaller
more manageable sub-networks is shown to make elementary flux mode analysis
feasible at increased levels of system complexity. This research expands the
application of elementary flux mode analysis allowing for examination of more
complex models such as genome-scale eukaryotic systems and anaerobic
microbial communities.
81
Poster number 33
Optimal Species Structure in Biofilms
Isaac Klapper
Department of Mathematics, Temple University
Biofilms are not well-mixed structures with the consequence that diffusive
transport, a mechanism not part of metabolic models, becomes important. A
natural question one might ask is, given models of individual organisms within a
biofilm, how might one expect those organisms to be distributed in that biofilm,
or at least what might be an optimal distribution? This poster presents some
preliminary answers in the context of a one dimensional biofilm model.
82
List of Attendees
ADHIKARI Kailash, Oxford Brookes University, Oxford, UK,
k.adhikari@brookes.ac.uk
AFLALO Claude, Institutes for Desert Research, Ben Gurion University of the
Negev, Israel, aflaloc@bgu.ac.il
ANDRADE Ricardo, INRIA / LBBE, Villeurbanne, France, ricardo.de-andrade-
abrantes@inria.fr
BAGHALIAN Kambiz, Leibniz Institute of Plant Genetics and Crop Plant Research
(IPK), Gatersleben, Germany, baghalian@ipk-gatersleben.de
BAROUKH Caroline, INRA LBE, Narbonne, France,
caroline.baroukh@supagro.inra.fr
BAUDET Christian; INRIA, Grenoble Rhône-Alpes, France,
christian.baudet@inria.fr
BEAUVOIT Bertrand, INRA-BFP, Villenave d’Ornon, France,
bertrand.beauvoit@bordeaux.inra.fr
BESTE Dany, Faculty of Health and Medical Sciences, University of Surrey,
Guildford, UK, d.beste@surrey.ac.uk
BEURTON-AIMAR Marie, LaBRI, Laboratoire Bordelais de Recherche en
Informatique, University of Bordeaux, France, marie.beurton@labri.fr
BUESCHER Joerg, BRAIN AG, Zwingenberg, Germany, public@brain-biotech.de
CARITI Federica, University of Geneva, Geneva, Switzerland,
federica.cariti@gmail.com
CARLSON Ross, Dept. Chemical and Biological Engineering, Montana State
University, Bozeman, USA, rossc@erc.montana.edu
CHASE Jennifer, Dept of Biology, Northwest Nazarene University, Nampa, Idaho,
USA, jrchase@nnu.edu
CHEUNG Maurice, Dept of Plant Sciences, University of Oxford, Oxford, UK,
maurice.cheung@plants.ox.ac.uk
COGNE Guillaume, University of Nantes, Saint-Nazaire, France,
guillaume.cogne@univ-nantes.fr
83
COIMBRA KLEIN Cecilia, INRIA / LBBE Universite Lyon 1, Villeurbanne, France,
cecilia.coimbra-klein@inria.fr
COLOMBIE Sophie, INRA, Villenave d’Ornon, France, scolombi@bordeaux.inra.fr
COURT Steven, School of Physics and Astronomy, University of Edinburgh, UK,
S.J.Court@sms.ed.ac.uk
CURIEN Gilles, CEA, Grenoble, France, gilles.curien@cea.fr
DE MARTINO Daniele, Center for Life Nano Science@Sapienza, Istituto Italiano di
Tecnologia, Rome, Italy, demartid@roma1.infn.it
DIKAIOS Ioannis, University of Verona, Verona, Italy, ioannisdikaios@hotmail.com
EBENHÖH Oliver, University of Aberdeen, Aberdeen, UK, ebenhoeh@abdn.ac.uk
FATMA Zia, ICGEB, New Delhi, India, ziafatma1@gmail.com
FELL David, Oxford Brookes University, Oxford, UK, dfell@brookes.ac.uk
FLAMHOLZ Avi, University of California- Berkeley, Berkeley, USA,
flamholz@gmail.com
FLORI Serena, CEA, Grenoble, France, serenaflo@hotmail.it
GHIN Leonardo, University of Verona, Verona, Italy, leonardo.ghin@gmail.com
GRAFAHREND-BELAU Eva, Leibniz Institute of Plant Genetics and Crop Plant
Research (IPK), Gatersleben, Germany, grafahr@ipk-gatersleben.de
GRIGORE Alexandra, Free University of Berlin, Berlin, Germany,
alexandra.grigore@fu-berlin.de
GROSSMANN Peter, Friedrich-Schiller-University Jena, Jena, Germany,
m6grpe2@uni-jena.de
HÄDICKE Oliver, Max Planck Institute for Dynamics of Complex Technical
Systems, Magdeburg, Germany, haedicke@mpi-magdeburg.mpg.de
HARTMAN Hassan, Oxford Brookes University, Oxford, UK, hassan.hartman-
2010@brookes.ac.uk
HEILAND Ines, University of Tromsø, Tromsø, Norway, ines.heiland@uit.no
HOPPE Andreas, Institute for Biochemistry- Charité Berlin, Berlin, Germany,
andreas.hoppe@charite.de
84
HUNT Kristopher, Montana State University, Bozeman, USA,
kristopher.hunt@biofilm.montana.edu
JENESON Jeroen, Beatrix Children’s Hospital/University Medical Center
Groningen, The Netherlands, j.a.l.jeneson@umcg.nl
JUNGREUTHMAYER Christian, ACIB GmbH, Graz, Austria,
christian.jungreuthmayer@acib.at
KAHN Daniel, INRA / Lyon University / CNRS, Villeurbanne, France,
Daniel.Kahn@univ-lyon1.fr
KEIBLER Mark, Massachusetts Institute of Technology, Cambridge,
Massachusetts, USA, mkeibler@mit.edu
KING Zachary, Dept of Bioengineering, University of California, San Diego, La
Jolla, USA, zaking17@gmail.com
KLAMT Steffen, Max Planck Institute for Dynamics of Complex Technical
Systems, Magdeburg, Germany, klamt@mpi-magdeburg.mpg.de
KOCH Ina, Goethe-University Frankfurt (Main), Frankfurt (Main), Germany,
ina.koch@bioinformatik.uni-frankfurt.de
KRUGER Nicholas, Dept of Plant Science, University of Oxford, Oxford, UK,
nick.kruger@plants.ox.ac.uk
KUDAHL Ulrich Johan, University of Cambridge, Cambridge, UK,
ulrich.kudahl@gmail.com
KUNDU Sudip, Department of Biophysics, Molecular Biology, and Bioinformatics,
Calcutta University, Kolkata, India, skbmbg@caluniv.ac.in
KUSARI Souvik, Institute of Environmental Research (INFU), TU Dortmund,
Dortmund, Germany; souvik.kusari@gmail.com
LISOWSKA Beata, University of Bath, Bath, UK, B.K.Lisowska@bath.ac.uk
MALATINSZKY David, University of Turku, Turku, Finland, davmal@utu.fi
MATUSZYNSKA Anna, University of Aberdeen, Aberdeen, UK,
anna.b.matuszynska@gmail.com
MAZAT Jean-Pierre, IBGC CNRS UMR 5095 & Universite Bordeaux 2, Bordeaux,
France, jpm@u-bordeaux2.fr
85
MOEJES Fiona, Daithi O’Murchu Marine Research Station, Bantry, Ireland,
fionamoejes@hotmail.com
MÜLLER Arne, Free University of Berlin, Berlin, Germany, arne.mueller@fu-
berlin.de
MÜLLER Marlen, Group of Plant Biotechnology, Institute of Agricultural Sciences,
ETH Zürich, marlmuel@ethz.ch
NAITHANI Sushma, Dept. of Botany and Plant Pathology, Oregon State
University, Corvallis, USA, naithans@science.oregonstate.edu
NAZARET Christine, IMB Bordeaux 2 University, Bordeaux, France,
christine.nazaret@u-bordeaux2.fr
NIKOLOSKI Zoran, Max Planck Institute of Molecular Plant Physiology, Potsdam,
Germany, nikoloski@mpimp-golm.mpg.de
PEÑA Pedro, University of Minnesota, Minneapolis, USA, ppena@umn.edu
PEREZ Luz, University of Minnesota, Minneapolis, USA, lperez@umn.edu
POOLMAN Mark, Oxford Brookes University, Oxford, UK,
mgpoolman@brookes.ac.uk
RADMANESHFAR Elahe, University of Aberdeen, Aberdeen, UK,
e.radmaneshfar@abdn.ac.uk
RATCLIFFE George, Department of Plant Sciences, University of Oxford, Oxford,
UK, george.ratcliffe@plants.ox.ac.uk
RICHTER Stephan, Friedrich-Schiller-Universität Jena, Jena, Germany,
stephan.richter@uni-jena.de
ROCHA Isabel, Institute for Biotechnology and Bioengineering, University of
Minho, Braga, Portugal, irocha@deb.uminho.pt
ROHWER Johann, Dept. of Biochemistry- Stellenbosch University, Matieland,
South Africa, jr@sun.ac.za
RUDD Brian, BioSyntha Technology Ltd, Welwyn Garden City, UK,
Brian.Rudd@biosyntha.com
SAURO Herbert, Department of Bioengineering- University of Washington,
Seattle, USA, hsauro@u.washington.edu
86
SCHUSTER Stefan, Friedrich-Schiller-University Jena, Dept of Bioinformatics,
Jena, Germany, stefanschu@gmail.com
SRIENC Friedrich, University of Minnesota, Minneapolis, USA, srienc@umn.edu
SELIVANOV Vitaly, Department of Biochemistry and Molecular Biology,
Universitat de Barcelona, Barcelona, Spain, seliv55@gmail.com
SINGH Dipali, Oxford Brookes University, Oxford, UK, dipali.d4@gmail.com
SMITH Alison, Dept of Plant Science, University of Cambridge, Cambridge, UK,
as25@cam.ac.uk
STELLA Giulio, Cellectis, Paris, France, giulioste@gmail.com
STEUER Ralf, Humboldt University Berlin, Berlin, Germany, ralf.steuer@hu-
berlin.de
TADDEI Lucilla, UMR 7238 Génomique des Microorganismes, Paris, France,
lucilla.taddei@upmc.fr
TAKEMOTO Kazuhiro, Department of Bioscience and Bioinformatics, Kyushu
Institute of Technology, Iizuka, Kawazu 680-4, Fukuoka 820-8502, Japan,
kz.takemoto@gmail.com
TAKORS Ralf, Institute of Biochemical Engineering, University of Stuttgart,
Stuttgart, Germany; takors@ibvt.uni-stuttgart.de
TEUSINK Bas, VU University Amsterdam, Amsterdam, The Netherlands,
b.teusink@vu.nl
TOEPFER Nadine, Max Planck Institute of Molecular Plant Physiology, Potsdam,
Germany, toepfer@mpimp-golm.mpg.de
TUMMLER Katja, Humboldt University Berlin, Berlin, Germany,
katja.tummler@hu-berlin.de
URBAIN Brieuc, University of Nantes, Saint-Nazaire, France,
brieuc.urbain@gmail.com
VAN RIEL Natal, Eindhoven University of Technology, Eindhoven, The
Netherlands, n.a.w.v.riel@tue.nl
VILLANOVA Valeria, Fermentalg, Libourne, France, villanova.valeria@gmail.com
WANNAGAT Martin, INRIA / LBBE Universite Lyon 1, Villeurbanne, France,
martin.wannagat@inria.fr
87
YUAN Huili, Eindhoven University of Technology, Eindhoven, The Netherlands,
H.Yuan@tue.nl
ZABKE Claudia, University of Aberdeen, Aberdeen, UK, c.zabke@abdn.ac.uk
ZANELLA Martina, ETH Zürich, Zürich, Switzerland, martina.zanella@biol.ethz.ch
ZANGHELLINI Juergen, ACIB GmbH, Graz, Austria, juergen.zanghellini@acib.at
ZHAO Zheng, DSM Biotechnology Center, Delft, The Netherlands,
zheng.zhao@dsm.com
88
NOTES
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108