systems metabolic engineering (2) - university of...

28
Systems M Engine Adriana B Consultant Metabolic eering Botes (PhD) Director & CSO

Upload: vuongtu

Post on 26-Mar-2018

218 views

Category:

Documents


1 download

TRANSCRIPT

Systems Metabolic Engineering

Adriana Botes (PhDConsultant

Systems Metabolic Engineering

Adriana Botes (PhD) Director & CSO

Contents

Introduction10 Systems strategies for developing industrial microbial strains

1. Project Design2. Selection of host strain3. Metabolic Pathway reconstruction4. Increasing tolerance to product4. Increasing tolerance to product5. Remove negative regulatory circuits limiting overproduction6. Rerouting fluxes to optimise cofactor & precursor availability7. Diagnose & optimise metabolic fluxes toward product8. Diagnose & optimise culture conditions9. System- wide manipulation of the metabolic network10. Scale-up fermentation and diagnosis

Conclusions & Perspectives

10 Systems strategies for developing industrial microbial strains

Metabolic Pathway reconstruction

Remove negative regulatory circuits limiting overproductionRerouting fluxes to optimise cofactor & precursor availabilityDiagnose & optimise metabolic fluxes toward productDiagnose & optimise culture conditions

wide manipulation of the metabolic networkup fermentation and diagnosis

Introduction

• Development of an industrial process to produc

• 50-300 person years of work

• Several $100M investment

• Despite revolutionary technological developments, only a few bio

• Researchers fail to consider a fully integrated industrial bioprocess when developing new microbial cell factories

• Companies lagged behind academia in adopting SOTA metabolic engineering techniques

Time/Cost is prohibitive for small companies – the few examples of technology translation to commercial processes • Time/Cost is prohibitive for small companies – the few examples of technology translation to commercial processes required early partnering with big players

• Need for academia/industry to collaborate more effectively & transfer knowledge more quickly

• 10 Stepwise strategies

• underpin successful development of industrial micr

• Implementation of Systems Metabolic Engineering:

• E. coli production strains for L-valine & L-threonine developed in 10 person years

• Feasible for small companies to develop production strains for target products

PronkNat.

uce bioproducts takes a great deal of time & effort:

Despite revolutionary technological developments, only a few bio-processes had been commercialised to date

fail to consider a fully integrated industrial bioprocess when developing new microbial cell factories

lagged behind academia in adopting SOTA metabolic engineering techniques

the few examples of technology translation to commercial processes the few examples of technology translation to commercial processes

collaborate more effectively & transfer knowledge more quickly1

icrobial strains through systems metabolic engineering (SME)

Implementation of Systems Metabolic Engineering:

threonine developed in 10 person years

Feasible for small companies to develop production strains for target products

Pronk, J.T. et al. How to set up collaborations between academia and industrial biotech companNat. Biotechnol. 33, 237–240 (2015).

Status Product MCFC Acetone C. acetobutylicumC citric acid Aspergillus nigerC lactic acid Issatchenkia orientalisC succinic acid E. coliC E. coliC S. cerevisiaeC B. succiniproducensC itaconic acid Aspergillus terreusC ‘06 1,3-propanediol E. coliD 1,3-butanediol E. coliC 1,4-butanediol E. coliD 2,3-butanediol C. autoethanogenum

Status of commercialization of microbial cell factories

D 2,3-butanediol C. autoethanogenumC PHA E. coliD Isoprene S. cerevisiae (E.coli)D Isobutene E. coliC L-Lysine & L-Arginine C. glutanicum? L-valine L-threonine E. coliC 1,5-pentanediamine C. glutanicumD adipic acid C. tropicalisD sebacic acid C. tropicalisD-C dodoecanedioic acid C. tropicalisC artemisinic acid S. cerevisiaeC-D squalene S. cerevisiaeC farnesene S. cerevisiaeD valencene S. cerevisiaeC vanillic acid S. cerevisiae

Feedstock CompanyCorn sugar Green Biologicssugar, molassesCorn sugars (dextrose) NatureWorkscorn sugar BioAmbersucrose MyriantStarch sugars Reverdiasugar, glycerol Succinitysugar, molasses Qingdao Kehaisugar DuPont Tate&Lyle Metabolic Ex sugar Genomatica & Versalissugar Genomatica & DuPont Tate&Lylesyngas LanzaTech

Status of commercialization of microbial cell factories

syngas LanzaTechsugar Metabolixsugar, cellulose Amyris, Braskem, MichelinGlucose, sucrose Global Bioenergiessugar SA Bioproductssugar KAISTsugar Cathay Industrial Biotechfatty acids (plant oil) Verdezynefatty acids (plant oil) Verdezynefatty acids (plant oil) VerdezyneSugar AmyrisSugar AmyrisSugar AmyrisSugar EvolvaSugar Evolva

Introduction

• Development of an industrial process to produc

• 50-300 person years of work

• Several $100M investment

• Despite revolutionary technological developments, only a few bio

• Researchers fail to consider a fully integrated industrial bioprocess when developing new microbial cell factories

• Companies lagged behind academia in adopting SOTA metabolic engineering techniques

Time/Cost is prohibitive for small companies – the few examples of technology translation to commercial processes • Time/Cost is prohibitive for small companies – the few examples of technology translation to commercial processes required early partnering with big players

• Need for academia/industry to collaborate more effectively & transfer knowledge more quickly

• 10 Stepwise strategies

• underpin successful development of industrial micr

• Implementation of Systems Metabolic Engineering:

• E. coli production strains for L-valine & L-threonine developed in 10 person years

• Feasible for small companies to develop production strains for target products

PronkNat.

uce bioproducts takes a great deal of time & effort:

Despite revolutionary technological developments, only a few bio-processes had been commercialised to date

fail to consider a fully integrated industrial bioprocess when developing new microbial cell factories

lagged behind academia in adopting SOTA metabolic engineering techniques

the few examples of technology translation to commercial processes the few examples of technology translation to commercial processes

Need for academia/industry to collaborate more effectively & transfer knowledge more quickly1

icrobial strains through systems metabolic engineering (SME)

Implementation of Systems Metabolic Engineering:

threonine developed in 10 person years

Feasible for small companies to develop production strains for target products

Pronk, J.T. et al. How to set up collaborations between academia and industrial biotech companNat. Biotechnol. 33, 237–240 (2015).

Systems Metabolic EngineeringSME integrates traditional metabolic engineering approaches with other fields

• Systems biology• -omics analysis and genome scale computational simulation

• Synthetic biology• Genetic engineering approaches, tools and pathway modules that allow fine control of gene expression levels and precise genome editingthat allow fine control of gene expression levels and precise genome editing

• Evolutionary engineering • Evolution of strains in the lab for enhanced product tolerance

While taking into account • Techno-economic factors• Tolerance to product and inhibitors in the feedstock• Genetic stability & strain robustness under actual fermentation conditions

Systems Metabolic EngineeringSME integrates traditional metabolic engineering

analysis and genome scale computational simulation

Genetic engineering approaches, tools and pathway modules that allow fine control of gene expression levels and precise

Design the cell factorto fit the techno-

economics and process

Do not design the process around the

strain-that allow fine control of gene expression levels and precise

Evolution of strains in the lab for enhanced product tolerance

While taking into account

Tolerance to product and inhibitors in the feedstockGenetic stability & strain robustness under actual

strain-

the techno-economicswill never work!

3 Bioprocess stages

Selection of microbial host

Construction of biosynthetic

Use of effective, low cost easily available C

� Performance

Construction of biosynthetic pathway

Improvement of tolerance against target product and inhibitors in

feedstock

Removal of negative regulations

Flux rerouting for cofactor and precursor optimisation

Optimisation of metabolic fluxes through pathway(s)

Systems level metabolic analysis

easily available Cchemically defined medium

Optimisation of culture conditions & feeding strategies

Performance of batch/fedbatch/semi-continuous cultures

Evaluation of production performance under scale

conditions

Scale-up of bioreactors

Iterative design and construction of strains

Use of effective, low cost easily available C-sources &

Use of effective, low cost easily available C-sources &

� Cost

easily available C-sources & chemically defined medium

Optimisation of culture conditions & feeding strategies

Performance of batch/fed-continuous cultures

Evaluation of production performance under scale-down

conditions

up of bioreactors

easily available C-sources & chemically defined medium

Minimisation of by-products

High [product]

Iterative design and construction of strains

Adapted from Lee & Kim, Nature Biotechnology, 2015, 33

10 Stepwise strategindustrial microbial strainsThe Public Sector aspires to a Circular Biochemicals and fuels from renewable nonconcerns about climate change & depletion of fossil resourcesconcerns about climate change & depletion of fossil resources

The Private Sector (bulk chemicals/fuels) aspires to increase profits and market share by reducing production costs using lower cost raw materials with stable supply, fewer unit operations, decrease CAPEX and OPEX

while protecting their existinvia controlled roll-out of disruptive

tegies for developing industrial microbial strains

aspires to a Circular Bio-economy to produce chemicals and fuels from renewable non-food biomass due to concerns about climate change & depletion of fossil resourcesconcerns about climate change & depletion of fossil resources

(bulk chemicals/fuels) aspires to increase profits and market share by reducing production costs using lower cost raw materials with stable supply, fewer unit operations, decrease CAPEX and OPEX

ting investments in production plants disruptive technologies or new products

1. Project DesignSelection of a BioSelection of a Bio

Fermentation Process

CostVolumeSupply chain

Selection of a Feedstock

Commodity, Speciality

Technical, economical, legal & regulatory factors

Preliminary techno

Fermentation Process

Operating ConditionsAerobic, Anaerobic, MicroaerobicpH, Temperature, Sterility Fed Batch/ContinuousType of reactor (shear)

Low value/Waste streams‘clean’: glycerol‘dirty’ organic acids (paper & pulp)‘intermediate’ fatty acids (tall oil, edible oil)C1 feedstocks

4; CH3OH; CO2/H2; CO/ CO2/H2

Cellulosic/starch sugarsLignocellulosic Sugars

Selection of a host strain

Selection of a Bio-ProductSelection of a Bio-Product

Fermentation Process

Downstream Process

Properties:Volatile, Soluble (aq), Insoluble (oil/ppt)Physical propertiesPurity required, ‘bad’ impurities

Technical, economical, legal & regulatory factors

Final Product

Preliminary techno-economic analysis

Fermentation Process

Operating ConditionsAerobic, Anaerobic, MicroaerobicpH, Temperature, Sterility Fed Batch/ContinuousType of reactor (shear) Titer (g/L), Yield (g/g) Productivity (g/L/h

Estimate performance metrics of different strains using genome scale metabolic simulati

Unit operationsEquipment & Siz

Selection of a host strain

Overview of the microbial cell factodesign process.

For successful commercial implementationfull picture should be consideredthroughout the design process.

A. Considerations relating to the choice orenewable feedstock, including the locatiothe production facility in close proximity available feedstocks, as indicated by blue(correct) and red (incorrect) concentric c

B. The metabolic engineeringprocess, start from selection of productioprocess, start from selection of productioorganisms and iterating through design-butest-learn cycles until the process requirements are met.

C and D. Critical parameters for the production process and downstream purifito final products respectively.

Taken from Gustavvson & Lee, Microbial Biotechnology9:610

2. Selection of host strain

Tractable to Genetic Manipulation Quality metabolic models

E. coliS. cerevisiae

Native producer of product or precursor • Amino acids: • Succinic acid: • Fatty acids• Antibiotics: Feedstock utilisation• C1 feedstocks: • Extensive genetic engineering to • C1 feedstocks:

• Organic acids, fatty acids

Product tolerance

• Need to develop toolkit for genetic manipulation• Extensive data generation required to refine

metabolic model

• Extensive genetic engineering to obtain desired traits

• Host may never produce product as efficiently as a non-conventional host

Tools to manipulate non-conventional hosts are becoming less of an issue

Systems biology advances

Synthetic biology tools CRIPR-Cas9

2. Selection of host strainInnate characteristics of host

Native producer of product or precursor Amino acids: Corynebacterium glutanicumSuccinic acid: Mannheimia succiniciproducensFatty acids: Y. lipolytica vs Rhodosporidium toruloidesAntibiotics: Streptomyces

Feedstock utilisationC1 feedstocks: C1 feedstocks:

• Clostridium autoethanogenum• Cupriavidus necator• Synechosystis sp.• Methylococcus capsulatus (Bath)

Organic acids, fatty acids• Cupriavidus necator, Candida tropicalis

Product tolerance• Pseudomonas putida

Need to develop toolkit for genetic manipulationExtensive data generation required to refine metabolic model

Host selection Needs analysis

Mature Molecular biology toolkit

Basic Toolkit, less established than E. coli

Genome scale model available

Basic model published, needs to be curated with data

Tolerance to target product

Requires evaluation

Degradation of Known degradation, analysisDegradation of Product/Precurosrs

Known degradation, requires KO strategy

Efflux of target product

Requires evaluation for active transported products

By-product formationRespiratory metabolism

negates need to co-produce by-products such as ethanol

Criteria defined for selecting a host organism suitable for the productionscale, and how Cupriavidus necator fit the criteria.� & � criteria are essential, � criteria are possible to address through

Host selection Needs analysis

Flexible and extensive Metabolic capabilities

Aerobic & anaearobic growth alpathways with oxygen requiring

oxygen sensitive enzymes

Phenotype characteristics that can be exploited

Stringent response under limitconditions allows continued uptof C & generates an NADPH poo

reduction

Feedstock flexibilityAutotrophic growth on CO2/H

Glycerol, TAG’s, organic acids, Faacids, aromatics

High growth rate on low cost defined

Carbon source + analysislow cost defined

media

Carbon source + mineral salts

ACDP classification of pathogenicity

Class 1

Robustness at industrial scale

Large scale production of PHBs to industrial scale

production of bulk chemicals from low cost feedstocks at indust

through strain evolution and genetic engineering, if identified early.

3. Metabolic Pathway reconstruction

Feedstock uptake

Central metabolism Product

pathway

Biomass, energy

Secreted Product

Product non-natural or inefficiently produced in

natural host

Gene Discovery: Enzymes required to complete pathway from bacterial, plant, fungal, mammalian origin?

Enzymes required to complete pathway kinetic properties?

Functional expression in selected host?

3. Metabolic Pathway reconstructionPathway Modeling Tools in SME:

• Computational tools for rational enzyme engineering

• Chemo-bioinformatic tools for pathway construction

• Constraint-based reconstruction and analysis (COBRA) of genome scale modelsanalysis (COBRA) of genome scale models

• 13C flux analysis

• Elementary mode analysis

Identify optimal metabolic pathways to drive fluxes from one metabolite to another

Enzymes required to complete pathway kinetic properties?

Enzyme selection & screening

Enzyme engineering

Semi-synthesis of

IPP ���� DMAPP

Glucose

G3P

Pyruvate

Acetyl-CoA

DXP Pathway

E. coli

DXP Pathway

E. coli

FPPFPPAmorphadiene

Synthase

Amorphadiene

SynthaseAcetyl-CoA

Mevalonate Pathway

S. cerevisiae

FPPFPP

AmorphadieneAmorphadiene

25 g/L E. coli25 g/L E. coli

40 g/L S. cerevisiae

SynthaseSynthase

Squalene Synthase

� Cu / Methionine

Squalene Synthase

� Cu / Methionine

ErgosterolErgosterol

synthesis of Artemesinin

Amorphadiene

Synthase

Amorphadiene

Synthase

Artemesinic acidArtemesinic acid

25 g/L S. cerevisiae

CYP71AV1, CPR1, CYB5

ADH1, ALDH1

CYP71AV1, CPR1, CYB5

ADH1, ALDH1

ArtemesininArtemesinin

Chemical

conversion

Chemical

conversion

S. cerevisiae

SynthaseSynthase

4. Increasing tolerance to productTest product toxicity and stability early on

� product tolerance once the strain under development produces close to

inhibitory [product]

Strain with � tolerance at high [product] does not necessarily correlate with �

productivity

Adaptive Evolution

• Serial subculturing with �[product] or product analogswith or w/o mutagen treatment

• Increase dilution rate during continuous culture

• Identification of cells with highest growth rate

4. Increasing tolerance to product

Rational Engineering

• Efflux pump for biofuel in E. coli

• Manipulation of ionic membrane gradients in S. cerevisiae for EtOH production

• Overexpression of L-valine exporter �titer by 40% in E. coli

Engineering efflux pumps is a powerful strategy to improve product tolerance

Bioprocess Design

• Couple in situ product removal with fermentation if no better ways of increasing product tolerance can be found

Competition assay efficiently identifies ef

©2011 by European Molecular Biology Organization

efflux pumps that provide biofuel tolerance.

Mary J Dunlop et al. Mol Syst Biol 2011;7:48

5. Remove negative regulatory circuits5. Remove negative regulatory circuitsTranscriptional regulation

• Replace native promoters

• KO transcription factors

Metabolic engineering of a C. glutanicumstrain overproducing L-Arg

(9 person years)(9 person years)

Negative feedback regulation in the AR1 strain was removed by inactivating two regulatory genes, argR and farR.

The resulting AR2 strain was able to produce 61.9 g/L of L-arginine by fed-batch culture compared to 34.2 g/L of the AR1 strain

5. Remove negative regulatory circuits5. Remove negative regulatory circuitsAllosteric regulation of enzymes

Feedback inhibition of enzyme by product or pathway intermediate

Metabolic engineering of a C. glutanicumstrain overproducing L-Lysine & PMD

The LysC gene encoding aspartatokinase was mutated to T311I to release feedback mutated to T311I to release feedback inhibition by L-Lysine & L-Threonine (Lys-1)

6. Rerouting fluxes to optimise cofactor & precursor availability

Co-factors & Precursors

• NADH, NADPH, ATP, CoA are involved in 100’s of reactions in the cell

• Acetyl-CoA, TCA cycle metabolites, amino acids

• Rerouting of metabolic fluxes is • Rerouting of metabolic fluxes is required to optimize the availability of cofactors and metabolic precursors of pathway

Optimisation of cofactors and precursors require systems-wide

approaches

Global mass, energy and redox balances must be considered

6. Rerouting fluxes to optimise cofactor & precursor

Manipulation of co-factors & precursors

• Remove competing pathways (gene KO)- time consuming, only applicable to non-essential gene

• Gene attenuation (gene knock-down) if gene is essential

• Combinatorial knockdown targets (synthetic • Combinatorial knockdown targets (synthetic biology tools & high-throughput screening of strains with combinations of downregulated gen

• Manipulation of co-factor specificity (swop NADPH-dependent enzymes and NADH dependenzymes)

Metabolic engineering of a C. glutanicustrain overproducing L-Lysine & PMD:

Co-factor & precursor availability

Flux rerouting to the pentose phosphate pathwa

Fluxes to L-lysine were reinforced by overexpressing ddh and removing competing pathways (knockout of pck and downregulation ohom).

Flux rerouting to the pentose phosphate pathwa(PPP) was conducted for NADPH generation by overexpressing the gluconeogenic gene fbp.

For PMD production, NCgl1469 encoding N-acetyltransferase and lysE encoding L-lysine exporter were both removed as they divert L-lysine away from 1,5-diaminopentane.

7. Diagnose & optimise meta8. Diagnose & optimise culture conditions

Diagnosis of the metabolic state

• Experiments must be performed under conditions as similar as possible to the final industrial fermentation conditions

• Fed-batch or continuous fermentation under scale-down conditions are essential for the ‘test’ element

• Identify bottlenecks and by-products for further metabolic engineering

etabolic fluxes toward product8. Diagnose & optimise culture conditions

Performance of intermediate strain under scale-down conditions facilitates evaluation & diagnosis of production

performanceperformance

Titer, Yield, Productivity

• Define new objectives for the next round of metabolic engineering

Metabolic engineering of a C. glutanicustrain overproducing L-Lysine & PMD:

Metabolic Flux Optimisation

Increase fluxes to L-Lys biosynthesis:

Overexpressing genes involved in the L-lysine biosynthetic pathway (dapB, lysA anthe mutated lysC),

mutating a pycA gene and downregulating the icd gene

For PMD production, amplification of its export(cg2893) led to further improvement in the production titer.

Amplification of PPP operon was amplified in theengineered C. glutamicum strain to enhance L-lysine production.

mutating a pycA gene and downregulating the icd gene

A complex medium based on molasses was used for the fed-batch culture ofC. glutamicum LYS-12 in order to evaluate its L-lysine production performance in an industrial setting.

Industrial glucose medium was used focultivating 1,5-diaminopentane-overproducing C. glutamicum DAP-16 strain.

Strain LYS 12

(L-Lys)

DAP-16

(PMD)

Titer (g/L) 120 88

Yield (g/g) 0.55 0.29

Productivity

(g/L/h)

4.0 2.2

9. System- wide manipulation of the metabolic network

Systems Biology Approaches

Cultivation profile-based system-wide analysis (‘fermentome’)

• -omics- based approaches

Final rounds of engineering required to construct the industrial strain

• -omics- based approaches

• In silico metabolic simulations

Current challenges

• Low transformation efficiency of host strains

• Screening methods for mutants overproducing a desired products

wide manipulation of the metabolic

Synthetic Biology Approaches

High-throughput genome scale engineering

• Multiplex automated genome engineering

• Trackable multiplex recombineering

Final rounds of engineering required to construct the industrial strain

• Trackable multiplex recombineering

• Synthetic small regulatory RNAs

• Auto-inducers for dynamic control of fluxes

Identify optimal combinations of genetic targets quickly

Isolation of mutants with desired phenotpes

10. Scale-up fermentation and diagnosis

• Aerobic fermentations are particularly affected by scale-up issues

• Mixing & aeration differences between lab and pilot scale

Fine

Pilot/Demo plant validation of industrial strain

lab and pilot scale

• Mass transfer rates of nutrients & oxygen

• Genetic instability (chromosomal manipulation)

up fermentation and diagnosis

Industrial Fermentation engineers

Fine-tune fermentation conditions using

• industrial grade feedstocks and medium components

Pilot/Demo plant validation of industrial strain

components

• Manipulate pH, temperature & oxygen transfer

• Contamination control (phage infection)

Conclusions & Perspectives

The process of bioengineering strains for commodity chemicals(process engineering and implementation)

Conclusions & Perspectives

als from initial concept (target molecule selection) to scale up

Victor Chubukov et al. npj Systems Biology and Applications (2016) 2, 16

Acknowledgements• Former colleagues

• Alex Conradie (�), • Changlin Chen & Ramdane Haddouche (�), • Unni Chokkathukalam & Satnam Surae (�)

• CPI • Frank Millar /Kris Wardrop• Robin Mitra & Steve Pearson

Further Reading• Sang Yup Lee & Hyun Uk Kim (2015). Systems strategies for developing industrial microbial strains. Nature Biotechnology, 33 (10):1061. doi:10.1038/nbt.3365

• Martin Gustavsson & Sang Yup Lee (2016). Prospects of microbial cell factories developed through systems metabolic engineering. Microbial Biotechnology 9 (5): 610. doi:10.1111/17517915.12385.

• Victor Chubukov et al. (2016). Synthetic and systems biology for microbial production of commodity chemicals. npj Systems Biology and Applications 2, 16009.

Kim (2015). Systems strategies for developing industrial microbial strains. Nature Biotechnology, 33 (10):1061. doi:10.1038/nbt.3365

& Sang Yup Lee (2016). Prospects of microbial cell factories developed through systems metabolic engineering. Microbial Biotechnology 9 (5): 610. doi:10.1111/1751-

et al. (2016). Synthetic and systems biology for microbial production of Systems Biology and Applications 2, 16009.