It is known from previous research that some lignin degrading microbial enzymes to be so tough that they can even deal with more difficult polymers: plastics. Some previous iGEM teams have exploited this and worked on enzymatic degradation of various kinds of plastic. We would like to build on this work and extend the plastic degradation capabilities of the synthetic biology community by improving PUR degradation in particular as this has not been successfully achieved before. We have identified 5 PUR-esterase enzymes from the literature that are capable of catalyzing the below reaction.

Poly-3-hydroxybutyrate(P3HB) is a bioplastic which is naturally produced inside bacteria such as Ralstonia eutropha, where it accumulates as globules inside the cell. In its native bacteria it is produced as an energy store(1) but it is as a plastic that it interests researchers and industrialists.

We have made P3HB in E.coli, transferring three genes, naturally found in Ralstonia eutropha into E.coli MG1655. These encode the three enzymes necessary for P3HB production; polyhydroxyalkanoate synthase(phaC), 3-ketothiolase(phaA) and acetoacetyl coenzyme A reductase(phaB). These are encoded by the phaCAB operon. We have altered the expression of these three genes to maximise the production of P3HB as high yields are required for it to be economically viable.

E.coli is commonly used as a chassis in innovative iGEM project that aim to prove the concept and make the case for a novel function in a biologically engineered machine. In our case, we aim to degrade and synthesise plastic and the degradation part of our system needs to be extracellular. There are not only such biotechnological but also many other reasons for the extracellular targeting of a protein: It can be a product with a medical function, like the [http://www.2009.igem.org/Team:METU-Gene/EGF_Transportation EGF hormone in a previous iGEM project] or be important in bioremediation such as in Dundee Team’s toximop system and much more. There are many strategies for secretion. On the following page, you can read about these and find out why we chose the pelB tag for all our degradation enzymes and how we improved the ESTCS2 esterase by changing the phoA outer-membrane anchoring tag in the original Biobrick to pelB. Our story will help you to choose a secretion strategy best suited to your project and also give you guidance with the technical details of using it.

Alternative strategies for protein secretion in E.coli:

N terminal secretion-signal peptides are recognised by the sec pathway . The pathway transports to the periplasm and additional mechanism are usually necessary to further export the protein to the extracellular space. An approach for using this pathway is the addition of an N terminal signal peptide to the target protein which can obtain as high as 90% efficiency in secretion to the extracellular space. Such signals are pelB and phoA tags which are available as biobricks. PelB can get cleaved off by pelB peptidase in the periplasm but phoA will anchor your protein to this localisation.

Fusion partners are endogenous proteins that naturally get secreted in E.coli and can be fused to the target protein. Some such proteins were demonstrated to be a powerful carriers of medically relevant human proteins in E.coli. The yebF and ompF proteins exit the cell via the sec pathway and use additional mechanism for leaving the periplasm where they interact with outer-membrane porins for extracellular secretion. The osmY is available as a Biobrick and we have submitted ompF this year.

The porin proteins such as ompF and ompA, can be used as fusion partners too and anchor proteins to the outer surface of E.coli.

ABC transporters use ATP for transport of specific proteins across the bacterial membrane. The ABC transporter of Erwinia chrysanthemi can be expressed in E.coli and export proteins that contain the LARD1 domain as a C terminal fusion. The system is biobricked in two separate plasmids (transporter, LARD1) with different antibiotic resistance and it is important to use both at the same time.

The [http://parts.igem.org/Protein_domains/Localization page] in the registry that lists localisation tags] is incomplete and contains eukaryotic and prokaryotic localisation signals mixed together. We have therefore put together the tables below for you to help choose an E.coli secretion strategy that is suitable for your project. </p>

Extracellular Secretion in E.coli:

Outer Membrane Anchors and Periplasmic Expression in E.coli:

Our choice is to use the pelB secretion tag as it has been demonstrated to work in many cases with sometimes as high as 90% transport efficiency. The pelB has been used in iGEM project for many years and is part of 50+ constructs. The UC-Davis team last year used it to secrete LC-Cutinase, a PET plastic degrading enzyme ( BBa_K936013) successfully which is somewhat similar to our plastic degradation enzymes.

Assembly methods and toolkit:

An advantage of the pelB is that it is relatively short (only 66 BP) and we could get our gene synthesised with the tag. A problem with standard biobrick assembly is that the scar site contains a STOP codon and therefore [http://parts.igem.org/Protein_domains alternative strategies are needed]. One of the ways around this is to use Infusion/Gibson assembly which we also had a go with and you can find instructions under our protocols section. We have constructed a biobrick where LARD1 is after a constitutive promoter and RBS in order to facilitate quicker cloning with less assembly steps, [http://parts.igem.org/Part:BBa_K1149021 BBa_K1149021].

Unfortunately, the pelB tag did not prove to be the best strategy for secreting our degradation enzymes in the MG1655 strain. Therefore, we are changing the strategy to use osmY.

References

1. Economou A. Following the leader: bacterial protein export through the Sec pathway. Trends in microbiology 1999;7(8) 315-320.
2. Thanassi DG, Hultgren SJ. Multiple pathways allow protein secretion across the bacterial outer membrane. Current opinion in cell biology 2000;12(4) 420-430.
3. Sletta H, Tondervik A, Hakvag S, Aune TEV, Nedal A, Aune R, et al. The presence of N-terminal secretion signal sequences leads to strong stimulation of the total expression levels of three tested medically important proteins during high-cell-density cultivations of E.coli. Applied and Environmental Microbiology 2007;73(3) 906-912.
4. Qian Z, Xia X, Choi JH, Lee SY. Proteome-based identification of fusion partner for high-level extracellular production of recombinant proteins in E.coli. Biotechnology and bioengineering 2008;101(3) 587-601.
5. Prehna G, Zhang G, Gong X, Duszyk M, Okon M, McIntosh LP, et al. A Protein Export Pathway Involving Escherichia coil Porins. Structure 2012;20(7) 1154-1166.
6. Binet R, Letoffe S, Ghigo JM, Delepelaire P, Wandersman C. Protein secretion by Gram-negative bacterial ABC exporters - A review. Gene 1997;192(1) 7-11.
7. Chung CW, You J, Kim K, Moon Y, Kim H, Ahn JH. Export of recombinant proteins in E.coli using ABC transporter with an attached lipase ABC transporter recognition domain (LARD). Microbial Cell Factories 2009;8 11.
8. Francetic O, Pugsley AP. Towards the identification of type II secretion signals in a nonacylated variant of pullulanase from Klebsiella oxytoca. Journal of Bacteriology 2005;187(20) 7045-7055.
9. Fu LL, Xu ZR, Li WF, Shuai JB, Lu P, hu CXH. Protein secretion pathways in Bacillus subtilis: Implication for optimization of heterologous protein secretion. Biotechnology Advances 2007;25(1) 1-12.
10. Cho HY, Yukawa H, Inui M, Doi RH, Wong SL. Production of minicellulosomes from Clostridium cellulovorans in Bacillus subtilis WB800. Applied and Environmental Microbiology 2004;70(9) 5704-5707.
11. Zhang XZ, Cui ZL, Hong Q, Li SP. High-level expression and secretion of methyl parathion hydrolase in Bacillus subtilis WB800. Applied and Environmental Microbiology 2005;71(7) 4101-4103.

We chose the MG1655 E.coli strain as chassis because it constitutively expresses genes that make it resistant to the potentially toxic molecule. The [http://parts.igem.org/Part:BBa_K892010 AldA] and [http://parts.igem.org/Part:BBa_K892009 FucO] genes have an important role in decreasing the toxic effects of Ethylene-glycol by converting it into Glycolaldehyde which is a link to the cell`s central metabloism. We have received these genes from the registry and future work could be to express these in Bacillus or cellulose degrading organisms to make them tolerant as well.

We will first use the phaCAB biobrick [http://parts.igem.org/Part:BBa_K934001 BBa_K934001] to produce P3HB. Our aim is the industrial production of P3HB and this require high yields to be economically viable. The results from the metabolic model suggest that to do this the amount of phaB should be increased. To do this we designed constructs with stronger promoters to upregulate the expression of the operon.

The operon was original under the native promoter from Ralstonia eutropha. We designed two promoters to increase transcription. The first uses the constitutive promoter [http://parts.igem.org/Part:BBa_J23104 J23104]. The second was designed as a hybrid promoter incorporating BBa_J23104 and the original promoter due to recent results of high expression from a hybrid promoter[http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0056321].

Bacteria strains

We used E. coli K-12 strains MG1655, NEB 10 beta, NEB 5 alpha and TOP10 (similar to E. coli K-12 DH 10 beta that fall under Risk Group 1 for the experiments, but eventually only MG1655 strain will be used in the bioreactors. This strain rarely survives outside laboratories, although minor symptoms may show if infected. Meanwhile, the parts we designed are specialised for degrading PUR and they should not harm human health.</p>

Chemical reagents

Some of the reagents we used were slightly toxic to human. In order to reduce the risks of using toxic reagents, we always wear gloves and labcoats. For some toxic reagents, we performed in fume cupboard. Click on the reagents to see the MSDS.

1. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=W222305&brand=ALDRICH&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Faldrich%2Fw222305%3Flang%3Den Tributyrin]

2. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=458392&brand=ALDRICH&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Faldrich%2F458392%3Flang%3Den Poly(diethylene glycol adipate)].

3. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=19123&brand=SIGMA&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Fsigma%2F19123%3Flang%3Den Nile red].

4. [http://www.sigmaaldrich.com/MSDS/MSDS/PleaseWaitMSDSPage.do?language=&country=GB&brand=SIGMA&productNumber=A3256&PageToGoToURL=http://www.sigmaaldrich.com/catalog/product/sigma/a3256?lang=en&region=GB Arabinose].

5. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=N9876&brand=SIGMA&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Fsigma%2Fn9876%3Flang%3Den 4-Nitrophenyl butyrate].

6. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=N8130&brand=SIGMA&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Fsigma%2Fn8130%3Flang%3Den 4-Nitrophenyl acetate].

7. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=S8511&brand=SIGMA&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Fsigma%2Fs8511%3Flang%3Den N-Succinyl-Ala-Ala-Pro-Leu p-nitroanilide].

8. [http://www.sigmaaldrich.com/MSDS/MSDS/DisplayMSDSPage.do?country=GB&language=en&productNumber=L3126&brand=SIGMA&PageToGoToURL=http%3A%2F%2Fwww.sigmaaldrich.com%2Fcatalog%2Fproduct%2Fsigma%2Fl3126%3Flang%3Den Lipase from porcine pancreas].

Bioreactor safety

We have come up with a few strategies to prevent release of our bacteria into the environment. Firstly, the bioreactor containing bacteria will have a kill switch that kills bacteria if the bioreactor is opened when operating. We also lyse the cells at the end of reactions to harvest our product P3HB, ensuring that no living bacteria escape the bioreactors.

The model was built and simulated in Simbiology, a Matlab package designed for modelling biological systems. In our model we have a compartment (labelled "E.coli_1") that represents our engineered cell. The compartment contains the reactions and species that interact with the polymer synthesis pathway.

The pathways on the right half of the cell are genetic expressions of the 3 enzymes involved in the plastic synthesis pathway. The pathways on the left represent the actual P(3HB) production pathway.

Genetic regulations and expressions

BDH2 (3-hydroxybutyrate dehydrogenase)

Schematic of the gene expression of BDH2 as seen in the model diagram

PhaB (Acetoacetyl-CoA reductase)and PhaC (P(3HB) synthase)

Schematic of the gene expression of PhaB and PhaC as seen in the model diagram

• Average molecular weight (Mw) of a base pair = 660g/mol[http://www.geneinfinity.org/sp/sp_dnaprop.html][http://www.lifetechnologies.com/uk/en/home/references/ambion-tech-support/rna-tools-and-calculators/dna-and-rna-molecular-weights-and-conversions.html]
• Average mass of a base pair = 660g/mol x 1.66x10^-24 = 1.1x10^-21g
• Volume of an E.coli cell = 1µm³[http://kirschner.med.harvard.edu/files/bionumbers/fundamentalBioNumbersHandout.pdf] = 1x10^-15L
  • ∴Mass concentration =
  • ∴Molar concentration of 1 base pair in the volume of E.coli = = 1.66x10^-6 mM
• BioBrick assembly plasmid pSB1C3 is a high copy number plasmid (100-300 copies per cell)[http://parts.igem.org/Part:pSB1C3?title=Part:pSB1C3]
  • assume 200 copies per cell
• ∴ concentration of the gene per cell = N x 200 x 1.66x10^-6mM, where N = number of base pairs
  • ∴ concentration of the gene BDH2 (N = 768) in the volume of an E.coli cell is = 0.25mM
• Transcription rate in E.coli= 80bp/s[http://kirschner.med.harvard.edu/files/bionumbers/fundamentalBioNumbersHandout.pdf] = 80 x 1.66x10^-6mM/s = 80 x 1.66x10^-6 x 60mM/min = 7.97x10^-3mM/min
• ∴ Rate of mRNA_BDH2 production under the control of pBAD = 7.97x10^-3 ÷ 0.25 = 0.032/min

2.Protein production rate of BDH2, k₂

• Average molecular weight(Mw) of an amino acid(aa)= 110g/mol[http://www.genscript.com/conversion.html][http://www.promega.com/~/media/Files/Resources/Technical%20References/Amino%20Acid%20Abbreviations%20and%20Molecular%20Weights.pdf]
• Average mass of an amino acid = 110g/mol x 1.66x10^-24=1.83x10^-22g/L
  • ∴Mass concentration of one aa in the volume of an E.coli = = 1.83x10^-6g/L
  • ∴Molar concentration of one aa = = 1.66x10^-5mM
• Translation rate = 20aa/s = (20 x 1.66x10^-5 x 60)mM/min = 0.020mM/min
• BDH2 comprises of 256aa[http://www.uniprot.org/uniprot/Q2PEN2&format=html]
  • ∴concentration of BDH2's aa in the volume of an E.coli= 1.66x10^-5mM x 256 = 4.25x10^-3mM
• ∴ Rate of protein production = 0.020 ÷ 4.25x10^-3 = 4.7/min

3.Protein production rate for PhaB and PhaC, k₃

• Relative promoter strengths: J23104 = 1.3RPU, J23101 = 1.0RPU.
  • ∴ 104 is 1.3x stronger than 101.
• In absolute units: take GFP synthesis rate (molecules per min per cell) and approximate that as a generic protein synthesis rate for the promoter.
  • GFP synthesis rate of 101 = 2232 molecules per min per cell.
  • ∴ GFP synthesis rate of 104 = 1.3 x 2232 = 2902 molecules per min per cell
• Assume 1 molecule in an E.coli cell gives a concentration of 1nM.
  • ∴ GFP synthesis rate of 104 = 2902 x 1nM = 2.9x10^-6nM/min per cell
• Plasmid copy number assumed as 200 (as in derivation 1)
  • ∴ GFP synthesis rate of 104 in our E.coli = 200 x 2.9x10^-6 = 0.00058nM/min = 0.58mM/min

BDH2 (3-hydroxybutyrate dehydrogenase)

Reaction R1:

Reaction R1r:

Values, sources and assumptions

atoAD (Acetyl-CoA:acetoacetyl-CoA transferase (α and β subunits))

Reaction R2:

Reaction R2r:

Values, sources and assumptions

phaB (Acetoacetyl-CoA reductase)

Reaction R3:

Reaction R3r:

Values, sources and assumptions

phaC (P(3HB) synthase)

Reaction R4:

Values, sources and assumptions

atoB (acetyl-CoA acetyltransferase)

Reaction R5:

Reaction R5r:

Values, sources and assumptions

blablabla

P(3HB) production

Sensitivity analysis: species concentrations

What is sensitivity analysis and why do it?

In order to look at how we could increase the production of P(3HB) we decided to run a sensitivity analysis (also in Simbiology) to identify what species in the model P(3HB) is sensitive to, given the specific conditions with which we set the model. In other words, we calculated the time-dependent sensitivity of P(3HB) with respect to the initial conditions of other species (species formed along the synthesis pathway and enzymes involved).[http://www.mathworks.co.uk/help/simbio/ug/calculating-sensitivities.html#brbumlr|ref]

For a species x, whose concentration depends on time, it can be expressed as x(t). To calculate the sensitivity of x(t) with respect to another species y(t), the sensitivity with normalisation relative to the numerator x(t) can be calculated as:

Therefore, the underlying algorithm for this sensitivity analysis of P(3HB) is:

Scan with different levels of PhaB

The plot below is to show that increasing PhaB concentration within the engineered organism would have a positive effect on the concentration of the P(3HB) synthesised. As so many parameters within the model were taken from different sources, it would be difficult to give a simulation plot that is accurate quantitatively. It was decided to show the single cell simulation for this because the scans were carried out up to the doubling time of 20mins, hence avoided the issue of scaling-up which would have otherwise subjected the simulation to more errors. Therefore, this plot should be interpreted from a qualitative perspective and the trend should be observed.

Difference between promoter expressions after optimisation

In our Module 1: Resourceful Plastic, we uses glucose as the product from degradation to feed in the system where P3HB is produced. The synthetic pathway needs to be reconsidered if we are using glucose as a solo input to the system. Here, 3HB monomer is no longer the source so BDH2 should be removed from the pathway. Acetyl-coA becomes the key resource to produce P3HB. A metabolic model is needed to show the production of Acetyl-coA from glucose. Furthermore, our synthetic pathway involved several key metabolites from the pathway such as NADPH, NADP+, NADH and NAD+. The initial concentration of those metabolites can be determined from the metabolic model by assuming they are in their steady states. More features and results of the metabolic model can be found in our Module 2.

Initial concentrations of metabolites

The metabolic model involves two pathways, Glycolysis pathway and TCA cycle. Where glucose is used as the single source of carbon. The reason is that Glucose is directly taken by Glycolysis pathway and there are some interactions between Acetyl-coA and TCA cycle. The model is based on original metabolic model by Angela Dixon.[Predictive Mathematical Model for Polyhydroxybutyrate Synthesis in Escherichia coli] The metabolic model is built in MATLAB Simbiology extension as well. All reactions and pathways are based on ODEs and kinetic data. The model can dynamically predict the concentration of all metabolites involved by a defined duration. <p>The metabolic model for determining steady state concentrations is shown as below:

The initial concentrations of all metabolites and kinetic data are referenced in the paper [http://digitalcommons.usu.edu/engineering_datasets/1/]

Production of P3HB with glucose as the only source

Synthesis model scaled up and what that means for industrial implementation and MAPLE.

Here are some specific objectives for the model to achieve:

1. The model should contain the gene expression model of the degradation enzymes because the enzyme concentration determines the rate of plastic degradation. In our case for PUR degradation, we used [http://parts.igem.org/Part:BBa_K206000 pBAD strong promoter K206000] for most enzymes. We built the gene expression model based on inducible pBAD promoter, which gene expression rate can be regulated by inducer concentration.

2. The model should show the efficiency of the enzyme secretion to the culture from the cells. It's also important because the enzyme concentration in the culture depends on it. Here we used pelB secretion tag for most enzymes in order to achieve a high efficiency.

3. The model basically predict how long will take to degrade a known concentration of soluble polyurethane. It is assumed that the enzyme in our assays has the same kinetic properties as the enzyme used in the literature. The model can suggest a suitable concentration of the plastic to use in order to get good results from the assays.

4. It is known that ethylene glycol is toxic to E.coli. However, it has no clear effect on the growth of our MG1655 strain when the concentration of ethylene glycol is below 200mM. Therefore, the model should suggest a safe range of PUR concentration to avoid a high concentration (>200mM) of ethylene glycol produced.

</div>