Engineering Success

The first and most time-consuming challenge for our team was the overall project planning. By reviewing the literature, we adapted a hypothetical enzymatic pathway for the degradation of phthalates, utilizing two enzymes from two different bacterial species. We selected two esterases and four decarboxylases that could potentially catalyze the chemical reactions we identified.

Simultaneously, we considered various methods for enzyme immobilization, such as binding to cellulose or encapsulation in alginate beads. We also contemplated the use of genetically modified bacterial cells for pollutant degradation.

We created a graphical design of the enzymatic pathway, connecting potentially useful enzymes from various publications (Fig. 3A). Taking inspiration from the Brno team (https://2020.igem.org/Team:Brno_Czech_Republic) and literature , we designed a protein scaffold in the form of a cellulosome (Fig. 3B).

Figure 3. Graphical representation of our ideas

a) The diagram represents the enzymatic pathway we utilized for the degradation of phthalates. In our project, we selected two esterases (PnbA, EstJ6) derived from Bacillus subtilis strain BJQ0005 and soil metagenomic libraries, and four decarboxylases (Dcx1, Dcx2, Dcx3, Dcx4) from Pantoea dispersa BJQ0007.

b) Structure of the cellulosome. which binds to cellulose fibers via CBD (cellulose binding domain) on one side, and on the other side binds enzymes that are in fusion with dockerins via cohesins.

We conducted numerous consultations of our ideas with specialists from various fields, including Professor Obuchowski, Professor Ołdziej and our advisors: Professor Czajkowski and Professor Węgrzyn.

Together, we chose the cell-free path and a system that utilizes the cellulosome. We learned that during the design, we would have to consider aspects such as protein purification methods (e.g., adding a His-Tag) and a way to separate the dockerin from the enzymes (e.g., adding a sequence recognized by the TEV protease) to enable testing of native enzymes. Detailed project planning helped us streamline further processes, and having a general plan allowed us to proceed with its implementation.

Once the details were established, we could proceed with designing sequences in the form of biobricks. Our aim was to maximize the modularity of our device so that we could utilize and exchange its elements, such as promoters or vectors, using the iGEM DNA Distribution Kit database. Given that the bacteria we will be working with are E. coli, we decided to prepare composite parts using the "E. coli Protein Expression Toolkit" (Table 1, Fig. 4).

Table 2. Parts from iGEM Distribution kit plate we used in our project.

Our name system	Registry ID part	Prefix and suffix	Length
Promoter	BBa_J435350	AB	103 bp
Vector pUC	BBa_J435330	AF	1769 bp
RBS	BBa_J435385	BC	84 bp
linkerDT4	BBa_J435309	DT4	29 bp
linkerDT4tev	BBa_J435395	DT4	35 bp
linkerEF	BBa_J435361	EF	65 bp
linkerD4T5	BBa_J435349	T4T5	38 bp
linkerT5Ehis	BBa_J435369	T5E	27 bp
linkerDF	BBa_J435393	DF	371 bp
pJUMP 23-1A	BBa_J428347	AF	3113 bp
pJUMP 26-1A	BBa_J428350	AF	2323 bp
pJUMP 28-1A	BBa_J428353	AF	2520 bp
pJUMP 29-1A	BBa_J428341	AF	2976 bp
GFP	BBa_J435328	CD	722 bp

Figure 4. . Assembly schematic of the E. coli expression toolkit. Parts in the E. coli expression toolkit (orange boxes) are flanked by Golden-Gate compatible fusion sites (grey boxes). Graphic based on: https://stanford.freegenes.org/products/expression-tookit#description

We optimized the designed sequences in the OPTIMIZER program (http://genomes.urv.es/OPTIMIZER/) to ensure efficient expression in E. coli cells. We added appropriate suffixes and prefixes to the basic parts and then designed composite parts with elements from the "E. coli Protein Expression Toolkit" (Figure 5A and 5B). Before ordering synthetic sequences for basic parts, we conducted trial ligations in silico using the SnapGene program.

Figure 3. Designed components for synthesis. a) Composite parts design of ScfL-His device (scaffoldin). b) Composite parts design of enzymes in fusion with dockerin (esterases: PnbA-DocScaB device, EstJ6-DocScaB device; decarboxylases: Dcx1-DocXylY device, Dcx2-DocXylY device, Dcx3-DocXylY device, Dcx4-DocXylY device).

The diagram shows the elements of the 'E. coli Protein Expression Toolkit' (orange), the corresponding prefixes and suffixes (blue) and our designed basic parts (green).

We attempted to ligate the synthesized sequences with elements from the "E. coli Protein Expression Toolkit" using the Golden Gate method, but they did not yield the expected results. After attempting the ligation of our enzymes in fusion with the dockerin, we obtained results from restriction analysis that were inconsistent with our expectations. In the case of the scaffoldin, we did not observe protein production in the transformed bacteria. More details in engineering cycles: Obtaining the protein scaffold – scaffoldin, Obtaining the enzymes.

We determined that we need an alternative ligation method to Golden Gate, one that increases the chances of success and reduces the issues with matching the appropriate quantity of individual parts in the reaction mixture. We speculate that less complex reaction mixtures should lead to more efficient ligations.

To streamline our work, we decided to combine the elements of our system through artificial DNA synthesis. Due to financial constraints, we could not order all the elements, so we created three constructs containing two esterases (PnbA-DocScaB device, EstJ6-DocScaB device) and one decarboxylase (Dcx1-DocXylY device) in fusion with dockerins.

We performed ligations of the ordered sequences with plasmids: vector pUC and pJUMP 23-1A, then transformed E. coli cells with these constructs.

We conducted restriction analyses and PCR (polymerase chain reaction) assays, which confirmed successful ligation and transformation. Furthermore, we were able to overproduce all three enzymes in E. coli cells. More details in engineering cycles: Obtaining the protein scaffold – scaffoldin, Obtaining the enzymes.

We successfully designed basic parts and then composite parts. The transformed bacteria carry plasmids containing sequences PnbA-DocScaB device and Dcx1-DocXylY device.

In the first stage, we planned to transform E. coli cells with selected elements from the iGEM distribution plate kit. We chose the E. coli DH5α strain, which is well-regarded for plasmid overproduction.

We carried out the first transformation with the elements: promoter, vector pUC and RBS.

After the transformation, we plated the cells on selective media containing kanamycin and ampicillin. We successfully obtained individual colonies containing the vector pUC (Fig. 6).

Figure 6. Cells plated on selective medium containing kanamycin and ampicillin after transformation with vector pUC part.

We successfully obtained cells containing the plasmid with vector pUC part. We learned that we should improve the method to increase the transformation efficiency.

We modified the basic transformation protocol by extending the incubation times of the samples and introducing a heat shock step.

We carried out another transformation, following the modified protocol, with the following elements: promoter and RBS.

We plated the cells after transformation on selective media containing ampicillin. Unfortunately, not a single bacterial colony grew.

We concluded that the method we chose, or the bacterial strain may not be suitable for these plasmids in this case.

We decided to test additional transformation methods, including electroporation, and used NEB® 10-beta Competent E. coli cells, which were transformed according to the producer's recommendations.

As planned, we carried out further transformations with promoter and RBS parts.

Once again, we plated the cells after transformation on selective media containing ampicillin. Unfortunately, as before, not a single bacterial colony grew.

We learned new transformation methods, but none of them yielded the expected results.

During consultations with our advisor, we realized that we did not centrifuge the iGEM distribution kit plates before sampling the material. It was likely that the DNA was stuck to the walls, which resulted in a lower concentration in the obtained suspension than we had expected.

We repeated the transformation according to the producer’s instructions, using NEB10-beta competent cells. Before extracting genetic material, we centrifuged the iGEM distribution kit plate.

We plated the cells after transformation on selective media containing ampicillin. This time, we observed the presence of bacteria on all plates. We isolated the plasmids from them, and subsequent restriction analyses confirmed the presence of the desired parts (Fig. 7).

Figure 7. Selected results of two restriction analyses. Plasmids from the iGEM distribution kit plate were transformed into bacteria, and the isolated plasmids were digested with the BsaI enzyme. The images show DNA separation in an agarose gel

a) In the middle well, the band represents the isolated vector pUC. In the well on the right, the restriction enzyme-cut product is visible, with two bands corresponding to the speculated product lengths of the analysis (2120bp and 1773bp).

b) In the middle well is a plasmid containing the DT4 linker element. Three bands in this well correspond to plasmid conformations: supercoiled, circular, and linear. After restriction enzyme digestion (well on the right), only one plasmid conformation is visible - linear. The product of the restriction analysis has a length of 29 bp, so it is not visible in the gel.

Plate centrifugation increased the amount of the product in the reaction mixture and allowed us to carry out successful transformations. This enabled us to prepare a cell bank containing plasmids with the parts necessary for the subsequent stages.

Having synthesized the basic parts (see Cycle 1) and the plasmids containing the rest of the elements needed to construct the composite parts (see Cycle 2), we could proceed to the first ligation. We decided to place our elements on the vector pUC, which carries genes that determine resistance to two antibiotics: kanamycin and ampicillin. During ligation, the ampicillin resistance determinant element is cut out, so the sensitivity of the bacteria to this antibiotic is a selection factor for correct ligation and transformation.

We performed Golden Gate ligation and then transformed NEB® 10-beta Competent E. coli cells with the resulting mixture.

We seeded the transformed cells on selection medium containing kanamycin. We then made replicate colonies on ampicillin medium for negative selection (Fig. 8a). We selected bacterial colonies that did not grow on ampicillin medium, isolated their plasmids and performed restriction analysis (Fig. 8b).

Figure 8. First BioBrick constructs based ligation testing.

a) Bacterial colony replicates at the top cultured on kanamycin medium (positive selection), at the bottom cultured on ampicillin medium (negative selection). b) Gel photo after agarose gel electrophoresis of samples subjected to restriction analysis. On the left, simulated analysis in SnapGene software. After treating our devices with the restriction enzyme SapI, we did not observe cutting of the plasmid into two parts. The control samples are post-ligation mixtures.

Restriction analysis showed that the bacteria that grew after transformation do not have the desired plasmids. We need a new method for ligation of the parts.

We decided to ligate elements of our expression system through DNA synthesis (see Cycle 1 iteration 3). We cloned the synthesized constructs: PnbA-DocScaB device, EstJ6-DocScaB device, Dcx1-DocXylY device into the commercial pBAD24 plasmid using the blunt-end ligation method.

After ligation with the pBAD24 vector, we transformed E. coli Rosetta 2(DE3) cells.

We seeded the transformed cells onto the selection medium. We obtained post-transformation colonies potentially containing the constructs PnbA-DocScaB device and Dcx1-DocXylY device. We isolated plasmids from them and subjected them to restriction analysis, which did not yield results as expected (Fig. 9).

Figure 9. Restriction analysis after ligation to the pBAD24 vector. The constructs PnbA-DocScaB device and Dcx1-DocXylY device were treated with EcoRI and EcoRV enzymes. As within the sequence of the constructs there is one site recognized by the EcoRI enzyme, we expected to observe one band on the gel. EcoRV enzyme was used to perform ligation, so after ligation there should be no sites recognized by the enzyme within the sequence.

The method using blunt ends is inefficient for chosen plasmids.

We decided to try GoldenGate ligation to the pUC vector once again, this time having the devices assembled by DNA synthesis (see Cycle 1 iteration 3).

Having the elements ligated using DNA synthesis, we performed their ligation with the vector pUC using the GoldenGate method. Then we carried out the transformation using NEB10-beta competent cells.

We seeded the transformed cells onto the selection medium containing ampicillin. We obtained transformed colonies potentiated with the PnbA-DocScaB device, EstJ6-DocScaB device and Dcx1-DocXylY device constructs. We isolated plasmids from them and subjected them to restriction analysis, which this time gave us expected results (Figure 10).

Figure 10. Gel obtained by agarose electrophoresis showing the result of restriction analysis. After ligation, the constructs lose sequences recognized by the BsaI enzyme (no cut - one band), while they gain sequences recognized by SapI (cut - two bands).

Restriction analysis confirmed that the cells carry a plasmid containing our constructs. We also learned that the pUC vector is high-copy and is not suitable for overproduction of all proteins.

We re-examined the iGEM distribution kit plate, where we found plasmids belonging to the JUMP collection. We identified 4 potential vectors that we could use: pJUMP 23-1A, pJUMP 26-1A, pJUMP 28-1A, pJUMP 29-1A. Plasmids, according to the description, were characterized by different copy number , but we decided to confirm it.

We performed bacterial transformation with all four selected vectors.

We seeded the transformed bacteria onto a selection medium containing kanamycin. We transferred the resulting colonies to liquid medium and cultured them overnight. Under UV light, we analyzed the fluorescence intensity of our cultures (Figure 11), which correlated with plasmid copy number.

Figure 11. Photo of cultures containing cells transformed with the following plasmids: pJUMP23-1A, pJUMP 26-1A, pJUMP 28-1A and pJUMP 29-1A. Photo was taken while the tubes were illuminated with UV light.

For further studies, we selected plasmid pJUMP23-1A, which had the lowest copy number , as confirmed by the lowest fluorescence intensity among the plasmids tested.

The next step in our project was to ligate the constructs PnbA-DocScaB device, EstJ6-DocScaB device and Dcx1-DocXylY device with the vector pJUMP23-1A.

We ligated our constructs with the pJUMP23-1A vector using the GoldenGate method and then transformed E. coli BL21 cells with the resulting ligation mixture.

We seeded the transformed bacteria onto a selection medium containing kanamycin. None of the bacterial colonies that grew fluoresced under UV light, which preliminarily suggested that ligation was successful and the correct part of the plasmid was excised. We then isolated the plasmids and performed restriction analysis (Fig 12a). We also performed PCR reactions analysis to confirm the presence of all elements of the construct on the isolated plasmid containing the PnbA-DocScaB device. (Fig 12b).

Figure 12. a) Gel obtained by agarose electrophoresis showing the result of restriction analysis with enzymes: BsaI and XbaI. After ligation, the constructs lose sequences recognized by the BsaI enzyme (no cut - one band). All our devices have two sequences recognized by XbaI (cut - two bands). b) Agarose gel with PCR products using primers complementary to the sequences of all biobricks used to create the PnbA-DocScaB device construct. On the left, visible simulation performed in SnapGene software. The control sample was the post-reaction mixture samples without DNA template.

Restriction analysis showed that we were able to obtain plasmids containing 2 of the 3 enzymes: pJUMP23-1A-Dcx1-DocXylY and pJUMP23-1A-PnbA-DocScaB. In addition, PCR analysis showed that the pJUMP23-1A-PnbA-DocScaB construct contains all the necessary elements.

At this stage, we could move on to try to overproduce our enzymes.

We carried out protein overproduction by adding IPTG (Isopropyl β-d-1-thiogalactopyranoside) to cell cultures containing Dcx1-DocXylY and PnbA-DocScaB elements on pJUMP23-1A plasmid.

After induction, we collected culture samples and resuspended them in Laemmli buffer (a sample buffer for denaturing and loading of protein samples). We subjected the samples to polyacrylamide gel electrophoresis in the presence of SDS (SDS-PAGE) to quickly check whether protein overproduction had occurred (Fig. 13).

Figure 13. Gel obtained by SDS-PAGE (sodium dodecyl sulphate-polyacrylamide gel electrophoresis). Arrows indicate proteins that were significantly increased. The predicted mass of dockerin-esterase (PnbA-DocScaB) is 64.6 kDa, while that of dockerin-decarboxylase (Dcx1-DocXylY) is 28.6 kDa.

We managed to overproduce proteins: esterase and decarboxylase.

The final step in this cycle was the purification of the resulting protein - PnbA-DocScaB. Our proteins had a His-Tag, so we chose Immobilized Metal Affinity Chromatography (IMAC) for purification.

We lysed the cells using sonication and then separated the proteins on a Ni-NTA resin column. After washing the bed with buffer, we recovered the proteins bound to the bed by washing with an imidazole gradient.

We visualized the fractions obtained by washing and protein recovery using SDS-PAGE (Fig. 14).

Figure 14. Gel obtained by SDS-PAGE after purification of proteins. We can see the sample applied to the column (supernatant load), the fraction not bound to the bed (flow through), selected wash fractions (11,12,14) and selected fractions extracted from imidazole gradient protein recovery (5, 10, 12, 15, 18, 21, 24, 27).

We were able to obtain a purified enzyme (esterase) in fusion with dockerin: PnbA-DocScaB. We were able to proceed to enzyme assays, which are shown in Cycle 5.

We first wanted to obtain bacteria carrying a plasmid containing scaffoldin: ScfL. For this purpose, we decided to ligate selected elements from iGEM distribution kit plate with our protein, thus obtaining ScfL-His device.

We first performed the ligation of scaffoldin (ScfL) with iGEM distribution kit plate elements such as a promoter, RBS, terminator, linkers, and vector pUC. We performed the ligation according to the Golden Gate protocol.

We then cultured the bacteria onto a positive selection medium containing kanamycin, and then prepared replicates on a negative selection medium containing ampicillin (after successful ligation, the bacteria should be sensitive to this antibiotic). We isolated plasmids from selected colonies and performed restriction analysis.

Figure 15. Gel obtained by agarose electrophoresis showing the result of restriction analysis. After ligation, the constructs gain sequences recognized by SapI (cut - two bands). The control sample was a ligation mixture. On the left, visible simulation performed in SnapGene software.

Restriction analysis showed that bacteria that grew after transformation lacked the desired plasmid.

We decided to repeat the ligation using the GoldenGate method but making a few modifications to the protocol. We also decided to use a different vector, this time choosing pJUMP23-1A.

We modified the previously used GoldenGate ligation protocol by increasing the amount of ligase and the volume of the reaction mixture, as well as increasing the duration of the ligation cycles. The ligation mixture was then used to transform E. coli BL21 cells.

Transformed cells were cultured onto a selection medium containing kanamycin. We selected colonies that did not fluoresce under UV light, isolated plasmids from them and performed restriction analysis (Fig. 16).

Figure 16. Gel obtained by agarose electrophoresis showing the result of restriction analysis. After ligation, the constructs lose sequences recognized by the BsaI enzyme (we predict if no cut - one band), while they gain sequences recognized by SapI (we predict if cut - two bands).

We were unable to confirm by restriction analysis that the bacteria carry a plasmid containing ScfL-His device.

We decided to repeat the ligation for the third time, making further modifications to the GoldenGate ligation protocol. We also planned different testing methods this time because we wanted to make sure that in the case of scaffoldin, the results of the restriction analysis were not false negatives.

We modified the GoldenGate ligation protocol used by increasing the number of ligation cycles (from 30 to 60). We also increased the amount of the restriction enzyme and ligase. We then transformed E. coli NEB 10beta cells with the ligation mixture.

We modified the GoldenGate ligation protocol used by increasing the number of ligation cycles (from 30 to 60). We also increased the amount of the restriction enzyme and ligase. We then transformed E. coli NEB 10beta cells with the ligation mixture.

Figure 17. ScfL-His device ligation correctness test. A) Gel obtained by SDS-PAGE after induction of scaffoldin overproduction. We can observe thickened bands in control samples corresponding to PnbA-DocScaB (64.6 kDa), Dcx1-DocXylY (28.6 kDa) and GFP [fluorescent colony] (26.9 kDa). For research samples, we did not observe significant differences at the height corresponding to scaffoldin mass (77,8 kDa). b) Gel obtained by agarose electrophoresis of the PCR reaction products. On the left, visible simulation performed in SnapGene software. Five of the seven analyzed reaction products are in a different location than predicted.

The results of our tests were inconclusive. In the case of SDS-PAGE, perhaps protein production was too low to detect with our method. In the PCR analysis, only two products were in the correct location, so perhaps ligation did not occur properly. In the case of scaffoldin, we ran out of time to complete the research.

In order to measure the concentrations of substrates and products of enzymatic reactions [i.e., dibutyl phthalate (DBP), monobutyl phthalate (MBP), butyl benzoate (BB), phthalic acid (PA), benzoic acid (BA)] by high-performance liquid chromatography (HPLC), we had to plot a calibration curve.

We prepared serial dilutions of the analyzed HPLC standards in a 20% solution of acetonitrile in phosphate-buffered saline (PBS). The samples were then loaded onto a chromatography column and separated for 1 hour.

We plotted calibration curves based on the chromatography results (Fig. 18).

Figure 18. Calibration curves of substrates and products of enzymatic reactions i.e., dibutyl phthalate (DBP), monobutyl phthalate (MBP), butyl benzoate (BB), phthalic acid (PA), benzoic acid (BA).

alibration curves plotted for PA, MBP and BA on the first chromatograph have the coefficient of determination (R2) above 0,99 giving a reasonably good fit. Calibration curves plotted for DBP and BB, on the other hand, have quite low R2, therefore values calculated using them would not be very reliable.

After obtaining the purified PnbA-DocScaB and standard curves, we were able to move on to testing phthalate degradation activity.

The analyses of enzymatic activity were done in 20% ACN in PBS to increase substrate solubility. The reaction was carried out for 30 minutes at 37°C. We tried dibutyl phthalate (DBP) and butyl benzoate (BB) as substrates.

We used high-performance liquid chromatography to conduct the analyses (Fig. 19).

Figure 19. Diagrams prepared based on the results of HPLC analysis. Control samples did not contain esterase. a) Mean dibutyl phthalate concentration in samples – substrate of the reaction. b) Mean monobutyl phthalate concentration in samples – product of the reaction. c) Mean butyl benzoate concentration in samples – substrate of the reaction. d) Mean benzoic acid concentration in samples – product of the reaction.

For the degradation reaction of the substrates (dibutyl phthalate and butyl benzoate) in the test sample, we can observe a significant increase in the concentration of the degradation reaction product (monobutyl phthalate and benzoic acid, respectively).

Unfortunately, the substrate concentration changed significantly after 30 minutes in both the control and test sample. These unexpected changes in substrate concentration may be related to the change in temperature during incubation, since DBP and BB did not form a homogeneous solution in 20% ACN in PBS, so an increase in temperature may have altered the substrate distribution between phases.

In conclusion, although this experiment did not provide data on reaction kinetics and enzyme activity, it still confirms that PnbA-DocScaB catalyzes the hydrolysis of ester bonds in MBP and BB at room temperature.