In vivo validation of in silico predicted
metabolic engineering strategies for terpenoid production in yeast
Evamaria Gruchattka, Oliver Kayser
Technische Universität Dortmund, Technische Biochemie
Terpenoids
Objectives
Gene
transfer
·
Possess important medicinal and industrial properties
·
Some are rare and produced in low amounts in plants
Native host
A
Cell Wall
GLC
NADP[c]
G6P
F
A
D
6PG
G15L
B
ATP
ADP
NADP[c] NADPH[c]
CO2
NADPH[c]
RUP
BIOMASS
GLYC
F6P
FBP
GLY-3-P
XUP
R5P
GAP
S7P
E-4P
F6P
Uniporter
Antiporter
NAD[c]
Heterologous microbial production may help to
overcome supply problems and high purification costs
NADH[c]
GAP
DHAP
P
P
Cultivation
C
CO2[ext]
Mitochondrion
CO2
EtOH[m]
3PG
O2[ext]
NADH
NAD[m]
O2
NAD
NADH[m]
2PG
OAA[c]
NADH[m]
NADP[m]
NADPH[m
]
PYR[m]
ACE[m]
CO2
EtOH[ext]
EtOH[c]
NAD[m]NADH[m]
CO2
CO2
NADH[c]
PYR[m]
ACEADH[c]
NADP[c]
NAD[c]
CO2
NADH[m]
NAD[m]
NADPH[c]
ACE[ext]
NADH[c]
ACE[c]
OAA[m]
AcCoA[c]
PYR[c]
In silico
analysis
OAA[c]
AMP
EtOH[m]
AcCoA[m]
CO2
NADPH[m]
ACEADH[m]
NADP[m]
NADH[m]OAA[m]
ACE[m]
ICI
NAD[m]
MAL
NADPH[m]
CO2
AcCoA[c]
EtOH[c]
ACEADH[c]
CIT
NADP[m]
AMP
NAD
ACEADH[m]
NAD[m]
CO2
PEP
PYR[c]
NAD[c]
NADH
ACE[c]
NAD[m]
NADH[m]
CO2
AKG
AcetoAcCoA
NAD[m]
FUM
AcCoA[c]
HMG-CoA
FADH2
SUCC-CoA
FAD
NADH[m]
CO2
SUCC
2 NADP[c]
MEV
24 ADP + 20 NADH[m] + 10 O2
5-P-MEV
Necessity of optimization of yield and productivity in
yeast e.g. via metabolic engineering [1]
24 ADP + 20 FADH2 + 10 O2
24 ATP + 20 NAD[m]
Engineered yeast
Plant terpenoid
24 ATP + 20 FAD
5-diP-MEV
ATP
CO2
IPP [ext]
ATP + AMP
ADP
2 ADP
IPP
Prenylated proteins
Dolichol
Ubiquinone
0.6
0.4
0.2
0
0.8
0.6
0.4
0.2
0
control
tHMG
Fusion
control
tHMG
Theoretical maximum
0.53
0.53
0.53
0.53
0.56
0.68
140
35
120
30
100
80
60
yield
in mg/g glucose
0.8
sesquiterpenoid titer in mg/L
1
squalene content
in mg/g CDW
sesquiterpenoid yield
in mg/g glucose
C-source
Glucose
Galactose
Fructose
Xylose
Glycerol
Ethanol
Theory
Fusion of FPP synthase with patchoulol synthase
Results
Theory
Reduce flux
draining of FPP
Patchoulol
FPP
Results
Theoretical terpenoid yields in wild type yeast
in C-mol/C-mol
(based on elementary mode analysis)
Overexpression of tHMG1
1
40
20
0
glucose-phase ethanol-phase
Fusion
10
9
8
7
6
5
4
3
2
1
0
Ø Increased squalene content
(triterpenic intermediate to
ergosterol)
Ø Increased terpenoid yield
ØDifferent carbon sources have
different theoretical potential
ØGlycerol and especially ethanol are
more promising than sugars
[3]
80
25
20
15
10
5
0
Ø Increased sesquiterpenoid titer
[4]
75
70
65
60
55
50
45
40
glucose-phase ethanol-phase
glucose-phase ethanol-phase
glucose-phase
ergosterol
squalene
ethanol-phase
sesquiterpenoids
Ø Increased fraction of
sesquiterpenoids in total
terpenoids
[4]
Ø Increased (total)
terpenoid content
Step 3: Validation of in silico predicted metabolic engineering strategies: knockout in citric acid cycle
In silico method: constrained minimal cut sets - theory
·
·
Minimal set of structural interventions (gene knockouts)
Repressing a certain functionality (deletion task: low product yield)
Preserving a certain functionality (desired modes: high product yield)
biomass/glucose in C-mol/C-mol
Wild type
biomass/glucose in C-mol/C-mol
Theory
·
cMCS for enhanced terpenoid yield: feasible set of interventions
GLC
cMCS knockout mutant
Cytosol
1. Prevention of acetate secretion
2. Prevention of ethanol secretion or production
3. Partial disruption of citric acid cycle, e.g.:
- mitochondrial α-ketoglutarate dehydrogenase
- mitochondrial succinyl-CoA ligase
Mitochondrion
PYR
PYR
Citric
acid
cycle
AcCoA
Consequence for flux distribution
· Carbon cannot be oxidized completely to CO2
· Carbon flux is redirected towards AcCoA
MVA
Terpenoids
IPP/glucose In C-mol/C-mol
IPP/glucose in C-mol/C-mol
Ø Coupling of a specified minimum terpenoid yield to growth
[2]
[3]
Disruption of α-ketoglutarate dehydrogenase gene (KGD1) of α-ketoglutarate dehydrogenase complex in citric acid cycle
300 bp
Resistance
cassette
removed
Ø Hardly any effect on
terpenoid production
8
6
Ø No effect on physiological
parameters
4
2
0
Ø Cre-loxP mediated gene
disruption successful
Future prospects
wild type
kgd1∆
Ø Reason: citric acid cycle
activity not sufficient
7
wild type
6
5
4
25
kgd1∆
70
60
50
3 40
2 30
wild type
1 20
70
0 10
00
60
50
40
30
20
10
0
0
kgd1∆
50
100
50 in h 100
time
time in h
Ø Reduced growth
150
150
20
15
10
5
0
wild type
Ø Reduced terpenoid production
ØAcetate formation needs to be prevented (neg. effect on
growth
and
terpenoid
production)
0
50
100
150
ØIf acetate would be efficiently converted to terpenoids, titers in the
g/L
time in
h range would be possible
Acknowledgements and References
This work was partly funded by the Ministry of Innovation, Science and Research of the German Federal State of North Rhine-Westphalia
(NRW) and by TU Dortmund through a scholarship to EG from the CLIB-Graduate Cluster Industrial Biotechnology (CLIB2021).
We thank Steffen Klamt and Oliver Hädicke for their assistance, Prof. Jens Nielsen (Chalmers University of Technology) for supplying us with
the plasmid pSP-GM1, Verena Schütz for initial support in this study, Michael Felten and Nicole Jurisch for support in the lab.
[1] Chang, M.C.Y. and Keasling, J.D. (2006) Nat. Chem. Biol. 2: 674-681
[2] Hädicke, O. and Klamt, S. (2011) Metab. Eng. 13:204-213
[3] Gruchattka, E. et al. 2013, Microb Cell Fact, 12:84
[4] Gruchattka, E. et al. 2015, submitted
kgd1∆
70
kgd1∆
wild type
60
70
50
60
50
40
40
30
20
20
10
0
0
30
wild type
0
0
70
50
100
60
50
100
time in h
50
acetate concentration in mM
500 bp
Resistance
cassette
replaces
gene
sesquiterpenoid yield
in mg/g CDW
2000bp
1500bp
10
CDW in g/L
M WT ∆ ∆2
Growth on ethanol-phase
in mM
concentration
acetate
in mM
concentration
acetate
Growth on glucose-phase
acetate
in mMconcentration in mM
acetate concentration
Gene disruption
Results
Prerequisites and in vivo validation of in silico predicted strategies
IPP
Farnesol
Ergosterol
Heme A
Step 2: Choice of carbon source
Well described engineering strategies according to literature:
Eliminate bottleneck in
MVA pathway
In a follow-up, we identified considerations for practical realization and
validated selected strategies in vivo [4]
·
Metabolic
network
Step 1: Engineering the mevalonate pathway (MVA)
Acetyl-CoA
In a previous study, we compared E. coli and S. cerevisiae as heterologous
hosts and identified promising metabolic engineering strategies using
elementary mode analysis and constrained minimal cut sets [2,3]
·
NAD[c]
NADH[c]
1,3PG
2 NADPH[c]
·
E
GLC[ext]
GLYC[ext]
·
Development of a platform organism for the efficient supply of
biosynthetic precursors for the production of terpenoids using an in silico
stoichiometric metabolic network analysis
·
fraction C15-terpenoids in %
One of the largest classes of natural products
content in mg/g CDW
·
sesquiterpenoid yield
in mg/g CDW
Introduction
[email protected]; [email protected]
150
150
time in h
40
30
20
10
0
Ø Strong acetate formation
0
kgd1∆
[4]
50
100
time in h