Difference between revisions of "Team:Paris Bettencourt/Project/Enzyme"

Throughout human history, people have taken inspiration from nature to find solutions for everyday problems. A famous example is velcro. Velcro was inspired by the way the structure of plant burrs allowed them to attach to the furry bodies of animals.
As technology has progressed, we not only take inspiration from nature, but also control and enhance biology for our specific purposes.
We adopted this mindset when we approached this part of our project: the development of anthocyanin degrading enzymes that can be attached to a specific type of fabric.
First, we searched for enzyme candidates that could be suitable within this scope. We selected enzymes from literature based on the following three criteria:
1- the enzyme has the known ability to specifically degrade Anthocyanins, or, more broadly, polyphenols, which are the class of compounds to which red wine pigments (Anthocyanins) belong.
2 - the enzyme has been successfully expressed in E. coli
3 - the enzyme activity can be tested using spectrography

Polyphenol degrading enzymes

Our literature search resulted in six enzymes that met these criteria.
The first enzyme of interest was the natural anthocyanin degrading Polyphenol Oxidase (PPO) from Camellia sinensis (PPO) (Jaiswal 2009). In order to optimally express this enzyme, we removed the naturally occurring signal peptide region of PPO and annotated the resulting enzyme as PPO2. Wu and coworkers reported that this signal peptide hampered the heterologous expression in E. coli (Wu et al. 2010).
We also selected the Beta-galactosidase from Vitis bellula, BG1. This enzyme was a strong candidate for our assays since it was shows to degrade anthocyanins, but unfortunately we did not manage to express it correctly in E. coli.
Bacillus pumilus’ BpuI laccase caught our attention, as laccases have been shown to degrade anthocyanins (Rodriguez Couto 2006). Notably, its corresponding gene was available as BioBrick BBa_K863000 in the iGEM Part Registry.
Next, we wanted to test two Catechol-dioxygenases: XylE from Pseudonomas putida and CatA from Acinetobacter pittii. We hypothesized that these enzymes would be strong candidates for removal of red-wine stains because catechol shares important structural similarities with anthocyanin (Cerdan 1995, Kobayasi 1995 and Lin 2015).
Finally, we decided to work with Lactobacillus plantarum’s tannin acyl hydrolase TanLpI. Although the red wine colour is associated with anthocyanins, tannins are important colour stabilizers. Tannins molecules form covalent bonds with anthocyanins to provide molecular stability of the anthocyanins over time. By targeting the tannin molecules, we wanted to destabilize the anthocyanins and thereby indirectly remove the wine stain (Curiel 2009).

Figure 1: Targets of our enzymes within the malvidin molecule This figure shows the place in which each of the enzymes attack the malvidin molecule. The different coloured lightbolts represent the different enzymes.

Design of our expression system

We constructed recombinant E. coli strains carrying the selected candidate enzymes using the following strategy. First, all the genes corresponding to the enzymes described above were codon optimized for expression in E. coli and equipped with an histidine tag at their C-terminus to facilitate protein purification. We adopted the pCOLA plasmid from the pDUET system to overexpress each enzyme (Novagen). In order to make the pCOLA plasmid compatible with the Golden Gate assembly method, we deleted all BsaI restriction sites and replaced the original multiple cloning site downstream of the T7 promoter for a LacZ cassette. Two BsaI restriction sites were added to allow the insertion of any gene of interest between the T7 promoter and T7 terminator while removing LacZ.
This way, we had a simple and standardised workflow for the construction of our expression plasmids. Golden gate reactions were transformed to E. coli DH5-alpha cells and plated on xGal/IPTG plates to allow for blue/white screening. The verified constructs were transformed to E. coli BL21(DE3) to overexpress the proteins in the presence of IPTG.

Fabric Binding Domain fusion proteins

An important part of our project was to specify the localization of the potential stain removing enzymes. We hypothesized that by increasing the time of contact between our enzymes and the substrate, we could more efficiently remove stains. In nature, peptide-based binding domains play an important role in the localization of an enzyme. Using the same principle, we fused proteins to the peptides we found to have a binding affinity for cotton, wool, silk and linen. (More information on the Fabric Binding Domains.) These peptides were thoroughly tested to see whether they were able to bind to specific type of fabrics.
To quantify the fabric binding affinity in the context of a pure protein, we assayed their behavior when fused to GFP. For the near future, we would like to test a larger number of enzymatic detergents fused to a variety of fabric binding domains. This way, we have paved the way to expand the way stains can be treated with biological solutions.

Results

Outline of the project

Figure 2: Outline of the enzyme project The enzyme sub-project has two main parts: the first one is the testing of a set of polyphenol degrading enzymes on their ability to degrade Anthocyanins, and the second one is the directing of that ability for stain removal purposes by fusing the enzymes to the Fabric Binding Domains (FBD) found by our team.
First, the different genes are codon optimised and cloned in E. coli. The heterologously expressed enzymes are then tested with their natural substrate to check whether they are functional of not.
Once their functionality is confirmed, we move on to the fusion of the enzymes with the FBD. To test whether it is feasible to fuse proteins with the FBD without disrupting their activity, we first create GFP-FBD fusion proteins for each of the FBDs and test their fluorescence. We also test the affinity of the resulting fusion proteins on different fabrics. This allows us to confirm whether it is feasible to construct fusion proteins in which both the enzyme and the FBD's activities are functional.
Having confirmed that fusion proteins are able to maintain both their enzymatic activity and their binding activity, we create fusion proteins with our polyphenol degrading enzymes and the FBDs. We test the activity of the resulting fusion proteins on the natural substrate of each of the original enzymes.
The last step is to test the original enzymes and the GFP-fusion versions on their ability to degrade Anthocyanins. We envision doing this in two ways: by testing them on Malvidin directly and by testing them on fabrics stained with wine.

Confirming expression of our proteins

After obtaining E. coli BL21(DE3) strains able to express our polyphenol degrading enzymes, we decided to focus our work on four of them, BpuI, CatA, XylE and BG1. We decided to postpone the experiments with PPO2, since due to its origin in plants, its expression in E. coli was the most demanding out of the set of proteins. We also decided to postpone the experiments on TanLpI due to the fact that it forms inclusion bodies, which ones again demands a complicated setup to make functional.
The first step was to check whether we were correctly expressing our proteins of choice. Figure 3 shows an SDS-PAGE analysis of the cell extracts performed with the cells in which expression of the proteins was induced. The cell extract protocol can be seen in the methods part of this project.

Figure 3: SDS-PAGE gels for expression of BpuI, CatA and XylE. Sample preparation: E.coli BL21(DE3) cells expressing the proteins were induced for 5 hours with 0.5mM IPTG. After 5 hours, the OD(600nm) was measured, the cells centrifuged and the pellet was resuspended in Laemmili sample buffer to a final OD of 10. The cells were cooked at 95ºC for 10 min, and 10uL of the resulting solution were loaded on the gel. Ladder used: Kaleidoskope. Staining: The gel was washed 3x with miliQ water to remove the SDS and staining was performed using BioRad Comassie for 30min. De-staining was performed by leaving the gel in miliQ water for 1 hour with gentle shaking.
As expected, no overexpression was observed in BL21(DE3), whether it was induced or not induced. BpuI was correctly overexpressed, with the observed size being the expected, around 59kDa. CatA overexpression was very mild, but the correct sized band was observed, of around 34kDa. XylE was correctly overexpressed, with the correct band being observed, of around 36kDa.

Overexpression of three of the four enzymes was achieved. BG1, although very promising, was not correctly expressed, and therefore we continued our experiments with BpuI, Cat and XylE alone.

Testing our enzymes in their natural substrate

The next natural step was to test whether the overexpressed proteins were functional. We performed cell extracts of BpuI, CatA and XylE enzymes and measured their activity using the appropriate substrates.

Figure 4: Confirmation of enzymatic activity of heterologously expressed BpuI, CatA and XylE. Results obtained using the cell extracts prepared as explained on the methods section of this page. Control corresponds at BL21(DE3) background, using cell extracts of BL21(DE3) cells that do not express our proteins. In all cases, values measured correspond to reaction product.
A. BpuI activity was measured in Citrate Phosphate Buffer at pH 4, with 0.4mM of ABTS being used as substrate, as recommended in the literature. Measurements were taken after 90min, timepoint at which the substrate present in the reaction medium was finished.
B. CatA activity was measured in Sodium Phosphate 50mM at pH 7, with 30mM of Catechol as substrate, as recommended in the literature. Measurements were taken after 35 min, timepoint at which all the substrate had been consumed.
C. XylE activity was measured in Potassium Phosphate 100mM at pH 7.5, with 30mM of Catechol as substrate, as recommended in the literature. Measurements were taken after 12 min, timepoint after which all the substrate had been consumed.

In all cases, the proteins were demonstrated to be functional. All three cell extracts carrying them were able to degrade their natural substrates more than the control.

Testing the effect of fusing FBD with proteins

The following step was to construct our fusion proteins, fusing Fabric Binding Domains to our best performing enzymes.
First, we needed to understand whether it was feasible to build a functional FBD-enzyme fusion protein. We needed to understand if the joining together of the two parts would have a negative effect on the binding capacity or in the enzyme itself.
An easy way of testing the feasibility of these constructs is to test the FBDs with GFP first. We constructed fusion proteins with the several FBD on the N'-terminal, since it was the terminal the binding domains were recommended to be placed at.

Figure 5: Analysis of FBD-GFP fusion proteins. With this image we show that the GFP activity was not affected by the fusion between the FBDs and the GFP. CBD stands for the +control, cellulose binding domain from the Biobrick registry BBa_K1321357.
Cell extract from strains expressing our fusion proteins was incubated overnight with the fabrics. Afterwards two washes of the fabrics were performed. Water, PBS, BSA 5%, Ethanol 70% and Catechol correspond to the washing solutions we used to remove the non-bound GFP, as well as to test the binding strength of the different peptides. The data displayed corresponds to the values after the final wash, normalised using the values from the first wash.
The intensity of the colour corresponds to the GFP signal measured at that point.

In figure 5 we can see that some FBDs show high affinity to certain fabrics, and some are even specific. In the case of FBD2, it shows high affinity and specificity for both cotton and linen. This FBD binds poorly to silk and wool, which is interesting given that both cotton and linen are plant-fiber-based, and silk and wool come from animal sources.
FBD5 shows affinity to all four fabrics, and has no specificity for any of them.
Interestingly, most FBDs have affinity for silk, which is very useful in the development of our product given that through our human practises we were able to get to know that silk is one of the most demanding fabrics when it comes to stain removal.
Different washing solutions also influenced differently the several FBDs. For example, FB10 performs very well when washed with water in silk and wool, but performs very badly when washed with the same solution in cotton and linen.

Figure 6: Confirmation of enzymatic activity of the fusion proteins CatA-FBD1, CatA-FBD10, XylE-FBD1 and XylE-FBD10. Results obtained using the cell extracts prepared as explained on the methods section of this page. Control corresponds at BL21(DE3) background, using cell extracts of BL21(DE3) cells that do not express our proteins. In all cases, values measured correspond to reaction product.
In green. CatA fusion proteins' activity was measured in Sodium Phosphate 50mM at pH 7, with 30mM of Catechol as substrate, as recommended in the literature. Measurements were taken after 35 min, timepoint at which all the substrate had been consumed by the native protein.
In red. XylE fusion proteins' activity was measured in Potassium Phosphate 100mM at pH 7.5, with 30mM of Catechol as substrate, as recommended in the literature. Measurements were taken after 12 min, timepoint after which all the substrate had been consumed by the native protein.

Figure 6 can have two interpretations: either the binding of the FBDs affects the activity of the proteins, or it lowers their expression. Both interpretations would explained the decrease in the activity that we observed. Nonetheless, the enzymes were still active.

Testing our enzymes on their ability to degrade Anthocyanins

The next step is to test our enzymes fused with the FBDs on their ability to degrade anthocyanins. A similar assay to the one performed with Catechol will be performed using Malvidin or a solution of Anthocyanins.

Testing our enzymes on their ability to degrade wine stains

Following the assay with the Anthocyanins, a similar assay to that performed with the fabric assays on the FBD-CBD fusion proteins will be performed, only this time it will be carried out with fabrics stained with wine, on the assay developed by the assay team and analysed with the software designed by the same team as well.

Methods

Codon optimisation of plant and bacterial genes of interest

Because every organism has different codon preferences, Genes of enzyme candidates were codon optimized for expression of in E. coli using using the IDT Codon Optimisation tool.

Golden Gate and Gibson Assembly for Biobrick Construction

We used Golden Gater and Gibson Assembly as cloning methods to construct our plasmid In our Golden Gate reactions, we used NEB’s Golden Gate Assembly mix,. 40 fmol of backbone and 40 fmol of insert following the instructions provided by NEB. The mixture was incubated with the following cycle: 37ºC for 5 min, (37ºC for 2 min, 16ºC for 5min)x50, 37ºC for 5min, 50ºC for 10min and 80ºC for 10min.
When we performed Gibson Assembly reactions, 200ng of DNA (backbone and insert with a ratio of 1:2) were mixed in a 15uL total reaction, using NEB’s 2X Hifi DNA Assembly MasterMix. The mixure was incubated 1h at 37ºC.

Induced protein expression by the use of expression systems

We used E. coli BL21(DE3) as our expression strain. Cells transformed with our constructs were incubated at appropriate temperature until they reached an OD(600nm) of 0.4-0.6. They were then induced with 0.5mM IPTG for 5 hours.

Preparation of cell extract for protein activity testing

We performed our cell extract preparation using ThermoScientific’s B-PER Bacterial Protein Extraction Reagent with Enzymes.

His-tag purification of proteins

We performed the purification of our proteins using ThermoScientific’s His-Pur Ni-NTA Resin and all buffers as indicated by the manufacturer.

SDS-PAGE analysis

We analysed our proteins using pre-made, 10% SDS-PAGE gels from BioRad. As a ladder we used Kaledoskope from BioRad.

Assaying the binding strength of the Fabric Binding Domains.

We assayed the strength or the Fabric Binding Domains using the assay developed by our team, carrying cotton, wool, silk and linen untreated, unbleached fabrics. Cell extracts of BL21(DE3) strains carrying GFP-FBD fusion proteins were performed and 1/10 dilutions were made. The diluted cell extracts were incubated in the 96-well-plates-based assay with the fabrics overnight. Several washes of the fabrics were performed using water, PBS, BSA 5%, Ethanol 70% and Catechol 0.03M. The several measurements were performed in a Infinite M200-Pro Tecan plate reader.

Assaying the enhanced stain removal activity of fusion proteins

We assayed the stain removal activity of our fusion proteins in our assay, with the different fabrics stained with wine. The different measurements wera carried out in the Infinite M200-Pro Tecan plate reader.
BpuI activity was measured in Citrate Phosphate Buffer at pH 4, with 0.4mM of ABTS being used as substrate, as recommended in the literature. Measurements were taken at 420nm.
CatA activity was measured in Sodium Phosphate 50mM at pH 7, with 30mM of Catechol as substrate, as recommended in the literature. Measurements were taken at 260nm.
XylE activity was measured in Potassium Phosphate 100mM at pH 7.5, with 30mM of Catechol as substrate, as recommended in the literature. Measurements were taken at 475nm.

Attributions

This project was done by Alicia Calvo-Villamañán and Sebastián Sosa-Carrillo. BpuI data was obtained by Mislav Acman. A special thanks to our advisor Nadine Bongaerts for all the insight on Golden Gate and for the GFP sequences. A special thanks to our advisor Jake Wintermute for all the help with the figures and the pDUET system. Thank you to Afonso Bravo for the help with the his-tag purification. Thank you to IDT for the synthesis of our genes, and thanks to NEB for the enzymatic kits.

References

• Barbagallo, R.N., Palmeri, R. Fabiano, S.; Rapisarda, P.; ,Spagna, G (2007). Characteristic of beta-glucosidase from Sicilian blood oranges in relation to anthocyanin degradation.Enzymatic Microbial Technology. 41, 570-575
• Cerdan P, Rekik M and Harayama S (1995). Substrate specificity differences between two catechol 2,3-dioxygenases encoded by the TOL and NAH plasmids from Pseudomonas putida. European Journal of Biochemistry. 229, 113-118
• Curiel JA, Rodriguez H, Acebron I, Mancheño JM, de las Rivas B and Muñoz R (2009). Production and physicochemical properties of recombinant Lactobacillus plantarum tannase. Journal of Agricultural Food Chemistry. 57(14), 6224-6230
• Jaiswal V, DerManderosian A, and Porter JR. (2009). Anthocyanins and polyphenol oxidase from dried arils of pomegranate (Punica granatum L.). Food chemistry. 118(1), 11-16
• Kobayashi T, Ishida T, Horiike K, Takara Y, Numao N, Nakazawa A, Nakazawa T and Nozaki M (1995). Overexpression of Pseudomonas putida Catechol 2,3-Dioxygenase with High Specific Activity by Genetically Engineered Escherichia coli. Journal of Biochemistry. 117, 614-622
• Lin J and Milase RN (2015) Purification and characterization of catechol 1,2-dioxygenase from Acinetobacter sp. Y64 Strain and Escherichia coli transformants. Protein Journal. 34(6), 421-433
• Rodriguez Couto S, Toca Herrera JL, (2006). Industrial and biotechnological applications of laccases: a review. Biotechnology Advances. 24/5), 500-513
• Wu YL, Pan LP, Yu SL and Li HH (2010). Cloning, microbial expression and structure-activity relationship of polyphenol oxidases from Camellia sinensis. Journal of Biotechnology. 145(1), 66-72