Part:BBa_K3589105
Mutant laccase from Botrytis aclada for Escherichia coli
This basic part contains the coding sequence of the mutant form L499F of the laccase from the ascomycete Botrytis aclada (hereafter referred to as BaLac). This part is codon-optimized for Escherichia coli. Combined a promoter and a terminator, this basic part mediates the oxidation of a wide variety of substrates including phenolic compounds and aromatic amines. This part can be cloned into standard E coli. expression vectors like the commercially available pGEX-6P-1 vector. The mutant laccase L499F has a high redox potential and shows activity near neutral pH (Scheiblbrandner et. al).
Summary of the Results from Team Kaiserslautern 2020 for part BBa_K3589105
- Expression could be achieved in different E. coli strains when cloned in the pGEX-6P-1 expression vector
- Protein showed activity in ABTS-assay
- Protein could oxidize Diclofenac
Design of the constructs
For the recombinant expression of the laccase genes from Botritis aclada(baLac) the E. coli vector pGEX-6P-1 was used (Fig. 1). This expression vector is used to construct a translation fusion protein of Glutathione S-transferase (GST) and our laccase. The expression is regulated by a tac promotor. This promotor combines the strong expression rate from the tryptophan promotor and can be induced with IPTG like the lac operon. To make sure the promotor is inhibited if there is no induction with IPTG the vector also includes the genetic code for the lac-inhibitor that can bind the lac-operon. Because GST has a high affinity for glutathione, the fusion of the laccases with GST allows the purification by affinity chromatography using glutathione agarose. In addition, a protease cleavage site is incorporated between the GST and our fusion protein (BaLac-GST). This allows the separation of GST and the laccase using PreScission Protease. BaLac has a size of 89.3 kDa with GST and a size of 61.6 kDa without GST. For selection of plasmid containing cells the expression vector carries an ampicillin resistance gene.
In order to be able to do further enzyme assays with the laccases, the vector pGEX-6P-1-balac has to be transformed in the E. coli expression strain BL21(DE3). Additionally, we transformed them in E. coli DH5α for isolation the vectors. The strain has an endA1 mutation. This leads to the inactivity of an intracellular endonuclease, which degrades plasmid DNA. Therefore, plasmid DNA isolation is more efficient.
Fig. 1: pGEX-6P-1_baLac . It includes a tac promotor that can be regulated with the lac-inhibitor (expressed with lacI) which can be inactivated with IPTG. The protein is fused to GST, with a PreScission protease cleavage site in between. The vector has an ampicillin resistance gene, which is used for selection.
Growth test
First, we wanted to find out the best conditions for the plasmid containing E. coli BL21(DE3) cells to produce BaLac. We did a test expression where the cells grow at different temperatures (37°C, 30 °C and 17°C; Fig. 2). After every hour we took a sample for an SDS-PAGE (Fig. 3) and measured the optical density (OD) at 600 nm. We designed the experiment based on Kittl et al., 2012 where they used copper sulfate in the media for the protein production. So, we tested the growth at all the temperatures, one time with CuSO4 and one time without, to be sure that the copper sulfate doesn’t influence the growth (Fig. 2).
To see if our protein is soluble or insoluble, we lysed the cells and separated the pellet and the soluble fraction with SDS-PAGE (Fig.3).
Fig. 2: Growth curve from E. coli BL21(DE3) pGEX-6P-1_baLac producing cells at different temperatures. The expression was done over 19 hours, the cells were induced at x=0. The growth of E. coli BL21(DE3) pGEX-6P-1_baLac is shown. It was tested at 37°C, 30°C and 17°C and all temperatures both with [w] and without CuSO4.
Fig. 3: SDS-PAGE of the test expression with different temperatures. Samples of E.coli BL21(DE3) pGEX-6P-1_baLac were taken before induction and after inducing with IPTG after every hour for every temperature. The LB medium contains CuSO4.The cells were disrupted by sonication and insoluble and soluble fraction were separated. The red boxes show the produced and BaLac. Marker: New England BioLabs ® Blue Protein Standard Broad Range. The protein is at the level of a relative molecular mass of the marker between 75 and 100 kDa (BaLac, size 89.3 kDa). The positive control is a GST fusion protein with a size of 26 kDa.
Furthermore, we wanted to test in which medium the cells grow best. We compared LB-Medium with 2YT-Medium in two different temperatures (37°C, 30°C; Fig. 4)
Fig. 4: SDS-PAGE and western blot of the test expression with LB medium and 2YT medium. (A) shows BaLac producing cells growing in LB medium, (B) shows BaLac producing cells growing in 2YT medium. Each medium contains CuSO4. After induction, the cells grow at 37 °C and 30 °C for 3 h and 5 h. The cells were disrupted by sonification and insoluble and soluble fraction were separated. The red boxes show the produced BaLac. The fusion protein (BaLac and GST) was detected by anti-GST-antibodies (first antibody) and anti-Goat alkaline phosphatase conjugated antibodies (second antibody). Marker: New England BioLabs ® Blue Protein Standard Broad Range.
As it is shown in Figure 3 and 4 the protein seems to be insoluble. So, we took the conditions where we had the biggest amount of soluble protein. Because of that we decided to produce BaLac at 30°C for 5 hours (Fig. 3) in 2YT-Medium (Fig. 4). Because laccases are enzymes with copper-centers and the cells growth isn’t inhibited by the copper sulfate (Fig. 2), we decided to use CuSO4 in our production medium every time.
With the knowledge about the best conditions we started the production of our protein in the transformed E. coli BL21(DE3) strain. The expression was induced with IPTG. After 5 hours we harvested the cells. To purify the protein, we lysed the cells and worked on with the soluble proteins in the cytoplasma. The first step of purification is to separate the translation fusion protein from other soluble proteins using affinity chromatography (Fig. 5a, 6a). After dialysis with PreScission Protease a second affinity chromatography removes the GST from the solution (Fig. 5b, 6b). Then our purified protein was ready to be tested for activity. For BaLac we got around 0.15 mg protein per gram cell wet weigth.
Fig. 5: Purification of BaLac with affinity chromatography. The cells grew at 30°C and were harvested after 5 h then lysed with sonification and centrifuged to receive the soluble fraction. The BaLac in the soluble fraction was purified with glutathione-agarose affinity chromatography. The SDS-PAGE shows the samples taken after every step. The fusion protein (BaLac and GST) was detected by anti-GST-antibodies (first antibody) and anti-Goat alkaline phosphatase conjugated antibodies (second antibody). (A) shows the first step of the purification before dialysis with PreScission Protease. The cell lysate was applied to the column. The fusion protein with GST-tag was able to bind to the column, the remaining lysate passed the column (flow through). This was followed by a washing step with washing buffer. Finally, the fusion protein was eluted with elution buffer containing glutathione (eluates 1-6). This was followed by dialysis with the PreScission protease to separate the laccase from the GST-tag. (B) shows the second step after dialysis. The GST binds to the glutathione agarose due to its affinity. The laccase flows through the column (D1-3). After a washing step with PBS (D4), the GST is eluted using an elution buffer containing glutathione (elution). The produced BaLac is shown in the red boxes. Marker: New England BioLabs ® Blue Protein Standard Broad Range.
Fig. 6: SDS-PAGE and western blot of the test expression with different E. coli strains. The following strains are shown: (A) E.coli Rosetta gami (B) E.coli BL21 Codon Plus RIL (C) E.coli DH5α (D) E.coli BL21(DE3) (E) E.coli Origami (F) E.coli AD494(DE3). After induction the cells are grown at 17°C for 3 h and for 19 h. The samples taken were lysed with sonication. The fusion protein (BaLac and GST) was detected by anti-GST-antibodies (first antibody) and anti-Goat alkaline phosphatase conjugated antibodies (second antibody). Marker: Promega GmbH © Broad Range Protein Molecular Markers.
On the SDS-PAGE showed in Figure 7 we could see, that compared to E. coli BL21(DE3) we have much more soluble protein in the strains E. coli AD494(DE3) and E. coli Rosetta gami. We started a new production and purification with these two strains (Fig. 8) and got a higher protein yield in the end (E.coli Rosetta gami pGEX-6P-1_baLac: 0,21 mg/g cell wet weight; E. coli AD494 (DE3) pGEX-6P-1_baLac: 0,18 mg/g cell wet weight). With this we started a new ABTS assay.
Fig. 7: SDS-PAGE of the BaLac production in E. coli AD494(DE3) and E.coli Rosetta gami. The cells grew at 17 °C and were harvested the next day, then lysed with ultrasonic and centrifuged to receive the soluble fraction. The BaLac in the soluble fraction is purified with glutathione-agarose affinity chromatography. The SDS-PAGE shows the samples taken after every step. The fusion protein (BaLac and GST) was detected by anti-GST-antibodies (first antibody) and anti-Goat alkaline phosphatase conjugated antibodies (second antibody). (A) shows E. coli AD494(DE3), (B) shows E. coli Rosetta gami. Marker: Promega GmbH © Broad Range Protein Molecular Markers.
E.coli BaLac strain BL21:
BaLac produced in E. coli BL21 (DE3), henceforth BL21, was the primary sample we ran assays with, due to it being produced as the initial construct. The assay was performed with the T. versicolor positive control at 40 μM and a negative control without enzyme, with all samples receiving 250 μM ABTS (Fig. 8). The assay was run identically to the positive controls, 4 hours at 30˚C. All following assays were performed this way.
There was activity shown, however because initially concentration of the enzyme (30 μM for the sample documented in Fig. 8) was determined using spectroscopy, an error prone method to determine the protein concentration, it was uncertain if this was the actual concentration due to imprecision in the method. In later assays a Bradford assay analysis was performed to confirm concentration more accurately. ABTS activity appears very similar to the T. versicolor positive control at a similar concentration run in pH 4 buffer, suggesting it could be, however concentration produced by this strain was never confirmed with the Bradford due to time constraints. Due to this and low concentration issues, only one positive result was achieved in ABTS assays, though later analysis suggests it had activity higher or similar to T. versicolor (see HPLC results below LINK).
Fig. 8: BL21 ABTS assay performed in pH 4 Phosphate-Citrate buffer. 1: Positive (+) control is 40 μM T. versicolor 2: BaLac BL21 sample containing 30 μM. 3: Negative (-) control contains no enzyme. Each well in row A contains 250 μM ABTS, while all B rows have ABTS substituted with buffer.
Fig. 9: BL21 ABTS assay analysis. a) Raw data: Positive (+) control is 40 μM T. versicolor. Negative (-) control contains no enzyme. BaLac (30 μM determined by spectra) shows similar activity to positive control. b) Normalized data (Raw absorbance/Initial absorbance) demonstrates a change from initial to visualize the reaction more clearly, showing the BaLac and T. versicolor sample reacting at similar rates over the course of 4 hours.
E.coli BaLac strain AD494:
In later assays, it became apparent that the concentration provided for the assays was a large issue, as yield was low and much of the protein was insoluble. As a result, we decided to try two new strains of chassis. The first was E. coli AD494, referred to as AD494, and while the yield and concentration provided for the assay after purification was slightly better than previous reactions, now being verified with the Bradford assay (26.8 μM per around 500 μL obtained), the reaction appeared less dramatic.
Fig. 10: AD494 ABTS assay analysis. a) Raw data: Positive (+) control is 40 μM T. versicolor. Negative (-) control contains no enzyme. BaLac (26.8 μM determined by Bradford assay) shows some activity, however much lower compared to positive control, which is about 1.5 times more concentrated. b) Normalized data (Raw absorbance/Initial absorbance) demonstrates a change from initial to visualize the reaction more clearly, showing the BaLac sample reacting more linearly over the course of 4 hours when zoomed in and compared to the stable negative control without enzyme.
However, the sample did react, as shown in Fig. 9b and in well A3 in Fig. 7, demonstrating another successful construct. It is important to note that due to time constraints because of COVID19, we were unable to perform additional replications or an HPLC with these samples, so more study into this construct would be necessary, however our positive results through assay and photographic evidence are indicative of a valid construct and a promising direction.
E.coli BaLac Strain Rosetta gami:
The final enzyme was produced in E. coli Rosetta gami, referred to as Rosetta gami. This construct had a yield of 27.5 μM for around 200 μL of purified protein, preventing any additional replications due to time constraints and logistics. This variant did appear to demonstrate some oxidation, however it was on par with what was seen in AD494 in reaction strength, and did not produce a blue coloration in the well in Fig. 10 (though this could have faded as there was a delay in photographing the wells after the experiment). Again, the results are promising, however more replications are needed to confirm and perfect the construct.
Fig. 11: Rosetta gami ABTS assay performed in pH 4 Phosphate-Citrate buffer. Positive (+) control is 40 μM T. versicolor with wells A1-A3 containing 250 μM ABTS. Negative (-) control contains no enzyme, while wells C1 and B3 had ABTS replaced with buffer. There was not enough enzyme to have an ABTS negative control for the BarLac sample.
Fig. 11: Rosetta ABTS assay analysis. a) Raw data: Positive (+) control is 40 μM T. versicolor . Negative (-) control contains no enzyme. BaLac (27.5 μM determined by Bradford assay) shows some activity, however much lower compared to positive control, which is about 1.5 times more concentrated. b) Normalized data (Raw absorbance/Initial absorbance) demonstrates the change from initial to visualize the reaction more clearly, showing the BaLac sample reacting more linearly over the course of 4 hours when zoomed in and compared to the stable negative control.
Sources of Error:
One source of error was likely the length of time the concentration and plating steps took, which possibly denatured some of the proteins before they could react in the assay. We also experienced very low yield, which made technical replicates very difficult to run. Due to the nature of the assay, some runs were performed delayed when the software froze on the TECAN plate reader, causing the initial values, which were used to determine the normalized reaction curves, to be elevated.
Elution of protein did occur for some of the produced proteins after time passed, especially in warmer temperatures. Occasionally BaLac did demonstrate this tendency, which obviously would lead to a skewed result when reading absorbance.
Bubbles and foam due to denatured proteins was also another very large issue, especially when resuspending or simply adding sample to the plate, and so after several runs with BL21, we realized it was important to inject the ABTS into the well to begin the reaction rather than inject the enzyme into the well as we had previously. Any resulting bubbles could shift the reads on the plate, or leave enzyme in the epi when pipetting, so we tried to centrifuge these tubes very briefly to reduce the bubbles without causing a pellet to form and need to resuspend. This caution also may have led to undermixing of the wells to prevent bubbles forming, causing the reader to potentially miss the very low concentrated enzyme if it wasn’t reacting close enough to the sensor.
HPLC:
Since dicolfenac and other common micropollutants dissolved in the water do not change color as soon as they are oxidized by a laccase, it is necessary to separate and determine the products by means of spectroscopic detection using the HPLC column. We use HPLC to detect such products to verify the HPLC method.
Fig. 12: Oxidation of diclofenac by laccase. Literature proposes that the radical cation (b) of Diclofenac (a), which is generated by a laccase-mediated oxidation reaction, reacts with water to form the para-hydroxy substituted intermediate (c). This can undergo further oxidation to form the para-benzoquinone imine derivate (d).1 Formation of 4‘-Hydroxydiclofenac, where the hydroxylation takes place in para-position to the nitrogen atom on the chlorinated benzene ring has also been described. 8 Note that the structure of product (d) shown above was established for the laccase from Trametes versicolor. 1
HPLC BaLac BL21:
After testing the T. versicolor samples, produced BaLac was incubated with diclofenac in an identical method as seen in the positive control. Due to time constraints, no replications of the BaLac samples were able to be produced to run on the HPLC. Concentration of BaLac was accurately determined by Bradford assay to only be 3.5 μM within 200 μL, enough for 4 samples. We predicted the reaction would take much longer to occur due to the limited concentration and as a result, we decided to incubate over the course of an hour to give the enzymes the most optimal conditions.
Surprisingly, the reaction appeared to progress at nearly the same rate or slightly slower than the positive control, displaying an area of about 900 mAU*s under the peak compared to the nearly 600 mAU*S initial T. versicolor positive control. The reaction progressed at a similar rate, though due to limited sample and time constraints, replications and shorter time intervals were impossible to perform.
Fig. 13: Diclofenac elution peak during reaction with BaLac. Each sample held 3.5 μM BarLac BL21 laccase and 250 μM diclofenac and reactions were halted at assigned time with heat treatment and filtration. Diclofenac retention time was 13.1 minutes and aligned with the literature value of around 14.1 minutes. 7 No obvious product peaks are visible and the area underneath the initial noise between 2 and 4 minutes does not change. Area underneath the diclofenac peaks appear to decrease at a linear rate until T10 at which time they are consistently below 100 mAU*s. BaLac t30 does not have a peak at 13.1, but instead at 13.4 with an area that corresponds with t60.
With such a low concentration to begin with, this sample reacting in such a clear way is suggestive that the produced laccase is not only effective, but possibly even more effective at oxidizing diclofenac than seen in the ABTS assay. It is possible that due to the errors with obtaining a protein concentration through raw absorbance calculations that the previously measured sample that was recorded at 30 μM was actually much lower, which would align with the E. coli team’s protein purification results tracing the loss of protein through each stage of the process. If this were the case, because this sample was the first to be accurately measured by Bradford assay, it is possible the yield was always so low and simply not recorded as such, suggesting the protein was much more productive than previously imagined. It is possible that the protein shows higher affinity to diclofenac than ABTS, which could also explain the results. This would need to be further explored by future work.
Sources of Error:
It is important to mention due to the quality of the column that was used, measurement was difficult and required a lot of troubleshooting. Some controls showed no results, which required rerunning samples and making new stocks or reactions of the control samples that we were able to remake.
The phosphoric acid and methanol buffer gradient seemed to display a large amount of noise in our well used column (kindly donated by the Chemistry Department), which made spotting peaks in the beginning of the readouts very difficult. The machine was also a manual loading older model, meaning that accuracy in inserting samples and the duration of the injection was key to aid in precise measurements, and while accuracy was intended and caution was taken in every step, because this was the first use of the machine by the Assay team members, errors could have occurred in technique.
Additional sources of error may stem from the preparation of the samples. The injection needle was a reusable glass needle which needed to be cleaned thoroughly between uses to avoid cross contamination. Samples were all mixed individually rather than in a single epi and then split between several for incubation to ensure equal distribution of both diclofenac and enzyme. This was initially done because of the importance of stopping the reaction for t0 immediately, however with such low concentrations as those provided in the produced sample, thorough mixing is essential and should be regarded in future assays to be sure all samples contain the reaction equally.
References
(1) Hahn, V. Enhanced Laccase-Mediated Transformation of Diclofenac and Flufenamic Acid in the Presence of Bisphenol A and Testing of an Enzymatic Membrane Reactor. 2018, 11.
Sequence and Features
- 10INCOMPATIBLE WITH RFC[10]Illegal EcoRI site found at 775
Illegal PstI site found at 463
Illegal PstI site found at 550 - 12INCOMPATIBLE WITH RFC[12]Illegal EcoRI site found at 775
Illegal PstI site found at 463
Illegal PstI site found at 550 - 21INCOMPATIBLE WITH RFC[21]Illegal EcoRI site found at 775
Illegal BamHI site found at 920 - 23INCOMPATIBLE WITH RFC[23]Illegal EcoRI site found at 775
Illegal PstI site found at 463
Illegal PstI site found at 550 - 25INCOMPATIBLE WITH RFC[25]Illegal EcoRI site found at 775
Illegal PstI site found at 463
Illegal PstI site found at 550
Illegal NgoMIV site found at 612
Illegal AgeI site found at 136
Illegal AgeI site found at 976
Illegal AgeI site found at 1059
Illegal AgeI site found at 1315 - 1000COMPATIBLE WITH RFC[1000]
| None |