Metabolic pathway alteration, regulation and control (3)

Xi Wang

01/29/2013

Spring 2013BsysE 595 Biosystems Engineering for Fuels and Chemicals

2

Important Concepts

1. Genome-scale model

The model is based on Genome data (sequenced organisms, KEGG database, annotations of genes, etc.), and therefore the size of the model is genome-scale.

Example:

Yeast iMM904: 905 genes (1577 reactions, 1228 metabolites)

3

2. What's the content of this model?

• All the metabolic information reported (KEGG etc.) (Metabolomics)

Example:

• E.coli iAF1260: 1261 genes (2382 reactions, 1668 metabolites)

• Yeast iMM904: 905 genes (1577 reactions, 1228 metabolites)

4

Read the model of iMM904

5

Model in Excel

6

3. What can we do with this model?

[1] Optimize the particular reaction rate (flux)

[2] Predict all the flux information of each reactions under certain conditions

(Schellenberger J, et al, 2011)

7

4. Linear Programming

Example:

Find numbers X1 and X2 that maximize the sum Z= X1 + X2, subject to the constraints:

X1 + 2X2 ≤ 4

4X1 + 2X2 ≤ 12

-X1 + X2 ≤ 1

X1≥0, X2≥0

Z=X1+X2

8

5. Objective function of FBA

Max Z = c1v1 + c2v2 + c3v3 + … + cn vn

S. t. S v=0 (steady-state equation)

ai<vi<bi (lb, ub)

(Orth JD et al., 2010; Kauffman KJ et al., 2003)

9

Examples in running FBA by COBRA toolbox

10

Example 1 Growth rate optimization

• Task Use FBA to predict flux distribution for optimal growth by yeast model iMM904 under different carbon source and aerobic conditions.

(Becker SA et al., 2007)

Growth rate (h-1)

Mets uptake rate upper limit (10 mmol gDW-1 h-1)

Aerobic Anaerobic

D-Glucose

D-Xylose

Acetate

Glycerol

11

Code

D-Glucose:initCobraToolboxmodel=readCbModel('iMM904');model1=changeRxnBounds(model,'EX_glc(e)',-10,'l')solution1=optimizeCbModel(model1)model2=changeRxnBounds(model1,'EX_o2(e)',0,'l')solution2=optimizeCbModel(model2)

D-Xylosemodel1=changeRxnBounds(model,'EX_glc(e)’,0,'l')model2=changeRxnBounds(model1,'EX_xyl_D(e)',-10,'l')

Acetate:model2=changeRxnBounds(model1,'EX_ac(e)',-10,'l')

Glycerol:model2=changeRxnBounds(model1,'EX_glyc(e)',-10,'l')

12

Growth rate (h-1)

Mets uptake rate upper limit (10 mmol gDW-1 h-1)

Aerobic Anaerobic

D-Glucose 0.2879 0

D-Xylose 0.1493 0

Acetate 0.0256 0

Glycerol 0.0861 0

13

Example 2 Robustness analysis

Task Study the sensitivity of the predicted growth rate to changing the flux of reactions, including O2 and glucose exchange (‘EX_o2(e)’, ‘EX_glc (e)’), as well as ethanol secretion (‘EX_etoh(e)’). Perform the analysis and generate plots that show how optimal growth rates vary as a function of flux.

O2 exchange:model=readCbModel('iMM904');robustnessAnalysis(model,'EX_o2(e)',20)

Glucose exchange: robustnessAnalysis(model,'EX_glc(e)',20)

Ethanol secretion: robustnessAnalysis(model,’EX_etoh(e)',20)

14

Anticipated results

15

Example 3 Dynamic FBA

Task Simulate aerobic batch growth in glucose media conditions using Yeast iMM904 model. The maximum uptake rates of glucose and oxygen are changed to 10 and 2 mmol gDW-1 h-1, respectively, and the initial glucose concentration is set to 10 mM. The initial biomass concentration is set to 0.035 gDW-1 L-1. The time step is 30 min and the maximum number of steps is set to 20 in order to allow observing the full diauxic shift.

model = changeRxnBounds(model, {'EX_glc(e)','EX_o2(e)'}, [-10 -2], 'l');substrateRxns = {'EX_glc(e)'};initConcentrations = 10; initBiomass =.035;timeStep = .50; nSteps = 20;plotRxns = {'EX_glc(e)','EX_ac(e)','EX_etoh(e)'}dynamicFBA(model,substrateRxns,initConcentrations, initBiomass, timeStep,nSteps,plotRxns)

(Becker SA et al., 2007)

16

The expected result

17

Practice: Single reaction knockout study

Task Perform the single reaction knockout study by using the E. coli iAF1260 model. Delete the reaction of ‘TPI’ and simulate the optimized growth rate after deletion

>> model=readCbModel('iAF1260')

>> solution=optimizeCbModel(model)

>> model1=removeRxns(model,'TPI')

>> solution1=optimizeCbModel(model1)

18

The expected result

model =

rxns: {2382x1 cell}

mets: {1668x1 cell}

S: [1668x2382 double]

rev: [2382x1 double]

lb: [2382x1 double]

ub: [2382x1 double]

c: [2382x1 double]

model1=

S: [1668x2381 double]

rxns: {2381x1 cell}

lb: [2381x1 double]

ub: [2381x1 double]

rev: [2381x1 double]

c: [2381x1 double]

Biomass flux (h-1)

WT 0.7367

KO 0.7098

Table P.1 The expected result

19

Outline (Topics)

1. Introduction of Cellular Metabolism (Lec 5)

2. Constraint-based Metabolic Network Reconstruction (Lec 5)

3. Flux Balance Analysis of Metabolic Network (Lec 6)

4. COBRA Toolbox (Lec 6)

5. Regulation of Metabolic Pathway (Lec 7)

6. Examples of Metabolic Pathway Modification (Metabolic

Engineering) (Lec 7)

20

Examples of Metabolic Pathway Modification (Metabolic Engineering)

Applications of metabolic engineering:

• Improving the production of chemicals already produced by the host organism,

• Extending the range of substrate for growth and product formation• Adding new catabolic activities for degradation of toxic chemicals• Producing chemicals new to the host organism were produced • Creating situations that contributed to a drastic modification of the

overall cellular properties

21

Examples of Metabolic Pathway Modification (Metabolic Engineering)

1. Improve ethanol production by blocking major by-product.

(Jeppsson M et al., 2002)

22

Prediction by COBRA (iMM904)

Code：model=readCbModel('iMM904')solution=optimizeCbModel(model)id=findRxnIDs(model,'EX_etoh(e)')solution.x(id)

model1=removeRxns(model,'ALDD2y')solution1=optimizeCbModel(model1)id=findRxnIDs(model1,'EX_etoh(e)')solution1.x(id)

Biomass flux (h-1) Ethanol (mmol g DW-1 h-1)

WT 0.2879 15.8155

KO 0.2875 15.8202

The expected result

23

2. Enable the utilization of xylose for growth

R. jostii XYLAB; ● R. jostii XYLA; R. jostii/pNV18.

(Xiong X et al, 2012; Jackson S et al, 2002)

24

Reference• Becker SA, et al. Quantitative prediction of cellular metabolism with constraint-based

models: the COBRA Toolbox. Nature Protocols, 2007, 2(3): 727–738• Kauffman KJ, Prakash P, Edwards JS. Advances in flux balance analysis. Current

Opinion in Biotechnology. 2003, 14: 491–496.• Jackson S, Nicolson SW. Xylose as a nectar sugar: from biochemistry to ecology.

Comparative Biochemistry and Physiology Part B. 2002, 131: 613-620• Jeppsson M, et al. Reduced Oxidative Pentose Phosphate Pathway Flux in

Recombinant Xylose-Utilizing Saccharomyces cerevisiae Strains Improves the Ethanol Yield from Xylose. Applied and Environmental Microbiology. 2002, 68(4): 1604-1609.

• Orth JD, Thiele I, Palsson B. What is flux balance analysis. Nature Biotechnology. 2010, 28(3): 245–248.

• Schellenberger J, et al. Quantitative prediction of cellular metabolism with constraint-based models: the COBRA Toolbox v2.0. Nature Protocols, 2011, 6(9):1290–2307.

• Xiong XC, Wang X, Chen SL. Engineering of a xylose metabolic pathway in Rhodococcus strains. Applied and Environmental Microbiology. 2012, 78(16): 5483-5491.

25

Thank you for your attention!

Questions?