1. VL Netzwerke, WS 2007/08 Edda Klipp 1 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Networks in Metabolism and Signaling Edda Klipp Humboldt University Berlin Lecture 7 / WS 2007/08 Flux Balance Analysis

2. VL Netzwerke, WS 2007/08 Edda Klipp 2 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Reminder: Stoichiometry Stoichiometric matrix Vector of metabolite concentrations Vector of reaction rates Parameter vector Systems equations in matrix form In steady state: K represents the basis vector for all possible steady state fluxes.

3. VL Netzwerke, WS 2007/08 Edda Klipp 3 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Elementary Modes/Extreme Pathways

4. VL Netzwerke, WS 2007/08 Edda Klipp 4 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Flux Balance Analysis in Metabolic Networks

5. VL Netzwerke, WS 2007/08 Edda Klipp 5 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics InSilico Network Reconstruction

6. VL Netzwerke, WS 2007/08 Edda Klipp 6 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics

7. VL Netzwerke, WS 2007/08 Edda Klipp 7 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics

8. VL Netzwerke, WS 2007/08 Edda Klipp 8 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Re-Formulation Additional constraints: Capacity:  j ≤ v j ≤  j

9. VL Netzwerke, WS 2007/08 Edda Klipp 9 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Additional constraints Capacity and Thermodynamic Constraints:  j ≤ v j ≤  j

10. VL Netzwerke, WS 2007/08 Edda Klipp 10 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Phenotypic Phase Plane Each phase is denoted by Pn x,y, where “n” denotes the number of the demarcated phase, and “x,y” denotes the two uptake rates on the axis of the PhPP. The PhPP can also be generated for a mutant genotype; represented as P gene n x,y. One demarcation line in the PhPP is defined as the line of optimality (LO). This line represents the optimal relation between respective metabolic fluxes. The LO is identified by varying the x-axis flux and calculating the optimal y-axis flux with the objective function defined as the growth flux

11. VL Netzwerke, WS 2007/08 Edda Klipp 11 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Additional Information An E. coli (ECK12) system definition is available on the website of Palsson’s group. website of Palsson’s group Free LP package (GLPK, simplex) is freely available on the Internet.GLPK LP solvers are available in Matlab ® optimization toolbox.Matlab ® optimization toolbox FBA course at Palsson’s group website http://gcrg.ucsd.edu/classes/be203.htm http://gcrg.ucsd.edu/classes/be203.htm

12. VL Netzwerke, WS 2007/08 Edda Klipp 12 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Optimization of the system with different objective functions (Z). Case I gives a single optimal point, whereas case II gives multiple optimal points lying along an edge.

13. VL Netzwerke, WS 2007/08 Edda Klipp 13 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Predicting the E.coli Optimal Growth Ibarra et al. Escherichia coli k-12 undergoes adaptive evolution to achiev in silico predicted optimal growth. Nature 2002. Daniel Segre, Dennis Vitkup, and George M. Church. Analysis of optimality in natural and perturbed metabolic networks. PNAS, vol. 99, 2002. Edwards et al. Characterizing the metabolic phenotype. A phenotype phase plan. Biotechnology and bioengineering. 2002 Kenethh et al. Advances in flux balance analysis. Current Opinion in Biotechnology. 2003. Schillling et. Al Combining pathway analysis with flux balance analysis for the comprehensive study of metabolic systems. Biotechnology and bioengineering, 2001.

14. VL Netzwerke, WS 2007/08 Edda Klipp 14 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Last lecture - a short reminder… What is the biological interpretation of any point in the flux cone ?

15. VL Netzwerke, WS 2007/08 Edda Klipp 15 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics (I) Narrowing the Steady State Flux Cone The steady state flux cone contains infinite flux distributions! Only a small portion of them is physiologically feasible. More constraints on the external fluxes. These depend on factors as: –Organism –Environment and accessibility substrates –maximum rates of diffusion mediated transport –Etc…

16. VL Netzwerke, WS 2007/08 Edda Klipp 16 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics (II) Calculating Optimal Flux Distribution The constrained flux cone in E.coli contains ~10^6 (Schilling 2001) How can we identify a “biologically meaningful” flux? Assumption … the metabolic network is optimized with respect to a certain objective function Z. Z will be a linear function. Later, we will deal with how exactly to choose Z

17. VL Netzwerke, WS 2007/08 Edda Klipp 17 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Minimize/Maximize S.T + inequality constraints What we want to do is find the vector v in the flux cone which maximizes Z. This optimization problem is a classical linear programming (LP) problem that can be solved using the simplex algorithm. W. Wiechert. Journal of biotechnology(2002) (II) Calculating Optimal Flux Distribution …this can be can formulated as an optimization problem:

18. VL Netzwerke, WS 2007/08 Edda Klipp 18 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics (III) How to Choose the Objective Function Z We want to choose a Z that is biologically meaningful. Reasonable options could be: 1.Z: Cellular growth (maximization) 2.Z: Particular metabolite engineering (maximization) 3.Z: Energy consumption (minimization) We want a v that: (A) Resides in side the cone. (B) maximizes Z=B+D+2E. Example: cellular growth is correlated with the production of B,D and 2E.

19. VL Netzwerke, WS 2007/08 Edda Klipp 19 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics 1. “It has been shown that under rich growth conditions (i.e. no lack of phosphate and nitrogen), E. Coli grows in a stoichiometrically optimal manner.” (Schilling 2001, Edwards 1994) We shall use Z which reflects: Cellular Growth (III) How to Choose the Objective Function Z 2. “It is reasonable to hypothesize that unicellular organisms have evolved toward maximal growth performance.” (Segre, 2002.)

20. VL Netzwerke, WS 2007/08 Edda Klipp 20 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics (IV) Phenotype phase planes - PPP Predicting Cellular Growth X axis – Succinate uptake rate Observations: Schilling 2001 Y axis – Oxygene uptake rate Z axis - Growth rate (maximal value of the objective function as function of succinate and oxygen uptake) Metabolic network is unable to utilize succinate as sole carbon source in anaerobic conditinos. Region 1: oxygen excess – this region is wasteful – (less carbon is available for biomass production since it is oxidized to eliminate the excess oxygen.) Line of optimality Succinate Oxygene Growth rate

21. VL Netzwerke, WS 2007/08 Edda Klipp 21 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics (IV) Phenotype phase planes - PPP Predicting Cellular Growth X axis – Succinate uptake rate Observations: Schilling 2001 Y axis – Oxygene uptake rate Z axis - Growth rate (maximal value of the objective function as function of succinate and oxygen uptake) Region 2 – limitation on both oxygen and succinate Region 3- the uptake of additional succinate has a negative effect. Cellular resources are required to eliminate excessive succinate. Succinate Oxygene Growth rate

22. VL Netzwerke, WS 2007/08 Edda Klipp 22 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics (IV) Phenotype phase planes - PPP Predicting Cellular Growth The EPs can be projected onto the plane. EPs are used to explain the different regions from a pathway perspective PPPs were also constucted for Malate/oxygen and Glucose/oxygen

23. VL Netzwerke, WS 2007/08 Edda Klipp 23 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Robustness Flux dependencies. Centre: phenotypic phase plane (PhPP). It shows the maximum biomass production that is achievable at every possible combination of O2 and succinate uptake rates. The line of optimality corresponds to the conditions that are necessary for maximal biomass yield (g DW cell mmol–1 carbon source, where DW is dry weight). Robustness analysis of the two uptake rates is shown in the two side panels. Left: effect on growth rate of varying O2 uptake at a fixed succinate uptake rate. Right: the effect on biomass generation of varying the succinate uptake rate at a fixed O2 uptake rate.

24. VL Netzwerke, WS 2007/08 Edda Klipp 24 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Model vs. biological experiments

25. VL Netzwerke, WS 2007/08 Edda Klipp 25 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Does E. coli behave according to optimal behavior predictions? E. coli was grown with malate as sole carbon source. A range of substrate concentrations and temperatures were used in order to vary the malate uptake rate (MUR). Oxygen uptake rate (OUR) and growth rate were measured.

26. VL Netzwerke, WS 2007/08 Edda Klipp 26 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Does E. coli behave according to optimal behavior predictions? Malate/oxygen PPP Ibarra et al., Nature 2002 1- The experimentally determined growth rate were on the line of optimality of the PPP !

27. VL Netzwerke, WS 2007/08 Edda Klipp 27 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Does E. coli behave according to optimal behavior predictions? Malate/oxygen PPP Ibarra et al., Nature 2002 Is the optimal performance on malate stable over prolonged periods of time? Evolution of E. coli on malate was studied for 500 generations in a single condition… 2- An adaptive evolution was observed with an increase of 19%in growth rate! 3- Same adaptive evolution was observed for succinate and Malate!

28. VL Netzwerke, WS 2007/08 Edda Klipp 28 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Does E. coli behave according to optimal behavior predictions? Why does this adaptive evolution occur? In other words why is the starting point at the bottom of the hill?

29. VL Netzwerke, WS 2007/08 Edda Klipp 29 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Does E. coli behave according to optimal behavior predictions? Same experiments were made using glycerol as sole carbon source Day 0 – Sub optimal growth Day 1-40 – evolution toward optimal growth Day 40 –optimal growth Day 60 –optimal growth (no change) Why?

30. VL Netzwerke, WS 2007/08 Edda Klipp 30 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Exploitation of Constraints A growing toolbox for constraint-based analysis. The two steps that are used to form a solution space — reconstruction and the imposition of governing constraints — are illustrated in the centre of the figure. As indicated, several methods are being developed at various laboratories to analyze the solution space. Ci, concentration of compound i; EP, extreme pathway; vj, flux through reaction j; vnet, net flux through loop. Price, 2004

31. VL Netzwerke, WS 2007/08 Edda Klipp 31 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Considering instances where FBA predictions are inaccurate… MOMA What happens to the metabolism in the case of a mutation/genetically engineered bacteria? What happens in terms of the flux cone? 0 0 0=Nv N

32. VL Netzwerke, WS 2007/08 Edda Klipp 32 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Considering instances where FBA predictions are inaccurate… FBA – assumes optimality of growth for wild type This assumption is not necessarily correct some instances… Knockout to pyruvate kinase

33. VL Netzwerke, WS 2007/08 Edda Klipp 33 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Considering instances where FBA predictions are inaccurate… MOMA Is there any other objective function Z that can capture the biological essence of these mutations? Perhaps another model… MOMA - minimization of metabolic adjustments [Segre, Vituk and Church 2002]

34. VL Netzwerke, WS 2007/08 Edda Klipp 34 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics MOMA Relaxes the assumption of maximal optimal growth. Wild-type growth Mutant optimal growth Mutant growth actual FBA MOMA a mutant is likely to display a suboptimal flux distribution between wild-type optimum and mutant optimum. Uses the same steady state flux cone as FBA.

35. VL Netzwerke, WS 2007/08 Edda Klipp 35 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics How does MOMA work? Assumption – “Initially, the mutant remains as close as possible to the wild- type optimum in terms of flux values.” Mutant optimal growth Mutant growth actual Wild-type growth FBA MOMA

36. VL Netzwerke, WS 2007/08 Edda Klipp 36 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics How does MOMA work? In other words: MOMA searches for the flux distribution in the “mutant flux space” which is closest to the optimal flux distribution in the “wild-type flux space”. Mutant growth actual Optimal growth wild type

37. VL Netzwerke, WS 2007/08 Edda Klipp 37 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics How does MOMA work? Formally: V wt – the wild-type optimal growth vector. V m – a vector in mutant flux space. Find V m which minimizes the Euclidian distance to V wt : This can be stated as a QP problem. That is, minimize Under a set of linear constraints.

38. VL Netzwerke, WS 2007/08 Edda Klipp 38 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Comparing MOMA and FBA on mutant strains

39. VL Netzwerke, WS 2007/08 Edda Klipp 39 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Testing robustness… “To be or not to be…”

40. VL Netzwerke, WS 2007/08 Edda Klipp 40 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Conclusions selection pressure results in optimal performance through evolutionary process. This optimal performance can be predicted using in-silico modeling. Unicellular evolution can be thought in terms of an iterative optimization procedure whose objective function maximizes the organisms ability to survive and proliferate. If given enough time (iterations) a local maxima is struck….

41. VL Netzwerke, WS 2007/08 Edda Klipp 41 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Bibliography [1] Daniel Segre, Dennis Vitkup, and George M. Church. Analysis of optimality in natural and perturbed metabolic networks. PNAS, vol. 99, 2002. [2] C. H. Schilling, D. Letscher and Bernhard Palsson. Theory for the Systemic Definition of Metabolic Pathways and their use in Interpreting Metabolic Function from a Pathway-Oriented Perspective. J. theor. Biol. (2000) [3] Schillling et. Al Combining pathway analysis with flux balance analysis for the comprehensive study of metabolic systems. Biotechnology and bioengineering, 2001. [4] Edwards et al. 2002. Characterizing the metabolic phenotype” A phenotype phase plan. Biotechnology and bioengineering [5] Kenethh et al. Advances in flux balance analysis. Current Opinion in Biotechnology. [6] Ibarra et al. Escherichia coli k-12 undergoes adaptive evolution to achiev in silico predicted optimal growth. Nature 2002. [7] W. Wiechert. Modeling and simulation: tools for metabolic engineering. Journal of biotechnology(2002) [8] Cornish-Bowden. From genome to cellular phenotype- a role for meatbolic flux analysis? Nature biotechnology, vol 18, 2000. [9] Schuster et al. Detection of elelmtary flux modes in biochemical networks: a promising tool for pathway analysis and metabolic engineering. TIBTECH 1999 [10] J. Papin, Nathan D Price, B. Palsson. Extreme pathway lengths and reaction participation in genome scale metabolic networks. Genome research, 2002. [11] Stelling eta l. Metabolic netwrok structure determines key aspects of functionality and regulation. Nature 2002. [12] A general definition of metabolic pathways useful for systematic organization and analysis of complex metabolic networks.

42. VL Netzwerke, WS 2007/08 Edda Klipp 42 Max Planck Institute Molecular Genetics Humboldt University Berlin Theoretical Biophysics Growth Phenotypes of insilico deletion strains The biomass yields are normalized with respect to the results for the full metabolic genotype. The α and β value for the constraints on the external fluxes for glucose and oxygen uptake are defined as follows (units- mmole g-1 hr-1): Phase 1 - vglc = 10, voxy = 23; LO - vgic = 10, voxy = 20.3; Phase 2 - vglc = 10, voxy = 17; Phase 3 - vglc = 10, voxy = 12; Phase 4 - vglc = 10, voxy = 8; Phase 5 - vglc = 10, voxy = 3; Phase 6 - vglc = 10, voxy = 0. Maximal biomass yields on glucose for all possible single gene deletions in central intermediary metabolism. The environmental variables (uptake rate/external metabolic fluxes) are set to correspond to a point within each of the phases of the wild-type PhPP (figure inset). The maximal yields were calculated using flux-balance analysis with the objective of maximizing the growth flux.