Team:Hamburg/Results

Results

E. coli lab

To obtain the desired proteins, we successfully performed PCRs and constructed the plasmids via Golden Gate Assemblies (GGAs). As we wanted to investigate whether the efficiency of the Cytochrome P450 enzymes (CYP) increases when fused to certain Cytochrome P450 reductases (CPR) due to an improved electron transport between both enzymes, our aim was to construct plasmids for expressing the fusion proteins. (compare Fusionproteins Parts BBa_K3846200-220) Plasmids for the expression of both proteins separately served as controls (e.g. BBa_K3846368). For this purpose, we tried to fuse the CYP-enzymes and reductases in different combinations. The success of the assemblies was verified by sequencing.

Successfully assembled constructs:

OleT-BM3R-Fusion
OleT-BM3R-Control
OleT-CPR- Control
CYP1A1-BM3R-Control
CYP1A1-Ferredoxin-Fusion
CYP71AV1-BM3R-Fusion
CYP71AV1-BM3R-Control
CYP71AV1-CPR-Fusion
CYP71AV1-CPR-Control

Unsuccessfully assembled constructs:

OleT-CPR-Fusion
OleT-Ferredoxin-Fusion
CYP1A1-BM3R-Fusion
CYP1A1-CPR-Fusion
CYP1A1-CPR-Control
CYP71AV1-BM3R-Fusion
CYP71AV1-BM3R-Control
CYP71AV1-CPR-Fusion
CYP71AV1-CPR-Control
CYP71AV1-ATR2-Fusion

As shown above, numerous constructs could not be assembled successfully, although the according primers were redesigned several times.
Furthermore, the list of assembled constructs displays that there is only one combination of CYP-enzyme and reductase that was successfully assembled as fusion protein as well as a control containing both enzymes separately. Constructs that were only available as fusion protein or control could not be used to obtain significant data to confirm or refute our hypothesis. Reason for this is the requirement of a direct comparison of experimental results obtained with the fusion-protein on the one hand and results obtained with both enzymes separately on the other hand would be mandatory.
Nevertheless, our plan was to test the activity of CYP1A1 in combination with BM3R (unfused) to see if our experimental setup is working, as we hoped that we are going to have the corresponding fusion-protein soon afterwards. We wanted to use CYP1A1 for catalyzing the reaction of 7-ethoxycoumarin to 7-hydroxycoumarin. To monitor the reaction, we first investigated the emission of both substances at different wavelengths after excitation at 325 nm to make sure that we can detect the product in fluorescence measurements.

Figure 1: Fluorescence measurement of 7-ethoxycoumarin and 7-hydroxycoumarin after excitation at 𝜆 = 325 nm.

The fluorescence measurement revealed that both substances do have different emission maxima which do not not overlap (Figure 1). While 7-ethoxycoumarin reaches the emission maximum at a wavelength of 385 nm, the maximum for 7-hydroxycoumarin is 455 nm. Next, we used this information to investigate the enzyme activity. We used E. coli BL21 expressing CYP1A1 and BM3R separately to first measure the emission at 452 nm (emission of 7-hydroxycoumarin) of the cells without adding any substrate and then measured the emission again after adding 7-ethoxycoumarin. The difference in emission then should give the information whether 7-ethoxycoumarin is converted into 7-hydroxycoumarin or not. Additionally, we carried out the same experiment with wild type E. coli BL21 as a negative control.

Figure 2: Fluorescence measurement at 452 nm of E. coli BL21 with (left) and without (right) CYP1A1 and BM3R before (red) and after (blue) adding 7-ethoxycoumarin. The experiment was carried out with four different concentrations of the substrate resulting in different intensities of the emission.

The performed experiment showed that both - the cells carrying the plasmid for expressing CYP1A1 and BM3R and the wild type cells - converted 7-ethoxycoumarin into 7-hydroxycoumarin (Figure 2). The intensity in both approaches was nearly identical, indicating that the additional expression of CYP1A1 and BM3R did not influence the turnover of 7-ethoxycoumarin into 7-hydroxycoumarin. We were not able to carry out the experiment with the corresponding fusion-protein as the construct could not be assembled successfully. Moreover, we wanted to use the combination of the CYP-enzyme OleT and the reductase BM3R for in vitro experiments as the construct for the fusion-protein (GGA 1.2) and for the control (GGA 3.1) could be assembled successfully. All proteins were designed to contain a 6x-His-tag at the N-terminus to allow purification via an affinity chromatography using a Ni-NTA resin. To confirm a successful protein expression and purification, samples of the different purification steps during affinity chromatography were collected and analysed via SDS-PAGE. As no clear bands were visible in those gels (which was indicated to be possible by an expert of CYP-expression (Prof. Dr. Vlada B. Urlacher) beforehand), we tried to detect the proteins of interest in a dot blot experiment using Anti-His-Antibodies.

Figure 3: Dot blot of purified GGA 1.2 (OleT-BM3R fusion-protein) and GGA 3.1 (OleT and BM3R separately). Pink dots indicate the presence of a His-tag.

The dot blot confirmed that the proteins were expressed successfully and were present in the sample obtained after affinity chromatography (Figure 3). Next, we wanted to test the activity of the purified proteins by monitoring the conversion of myristic acid to 1-tridecen via mass spectrometry (EI). The obtained spectrum did not indicate the presence of the desired product. The most likely reason for this observation is that we were not able to properly dissolve the used carboxylic acid in anything but highly concentrated acetic acid and enzymes tend to dislike solid bulks of substrates and solutions of low pH. Furthermore, we wanted to use CYP71AV1 to produce artemisinin. For this purpose, we used additional plasmids (Keasling group) containing the Mevalonate pathway. Those were necessary to produce precursors (amorphadiene) needed for artemisinin production catalysed by CYP71AV1. To check whether the needed precursors were produced, we transformed E. coli BL21 with the Keasling-plasmids and induced the expression of the proteins involved in the Mevalonate pathway. Afterwards, we used dodecan to extract possible contained amorphadiene and investigated the sample via mass spectrometry. After realising that no substance of interest was present in the spectrum, we soon discovered the reason why. We lacked a plasmid responsible for the expression of amorphadiene synthase which is needed to convert FPP to amorphadiene. Due to time limitations, we were not able to correct this mistake and therefore did not obtain any results using this construct.
FPP supplied by the mevalonate pathway can furthermore be used by α-humulene synthase to produce α-humulene. Long story short, this didn’t work out either.

Cyano lab

The cultivation of cyanobacteria poses many obstacles, mainly due to slow growth. It is very important to keep wild type cultures of cyanobacteria axenic, as many bacterial and fungal contaminants have the potential to outgrow cyanobacteria. Therefore, the cultures were examined for contamination before three-parental-mating. This was done by plating out cultures on LB-agar plates and visually checking the cultures with a transmitted-light microscope. As pictured below, no contamination was observed, all colonies appeared uniformly coccal and greenish (Figure 4).

Figure 4: Wild type Synechocystis PCC 6803, transmission-light microscopy.

Conjugation was executed using three-parental-mating with the helper strain E. coli HB101 with plasmid pRK2012.2 and E. coli DH5a with PAM5411, coding for GFP as a three-parental-mating method control. Even though an unexpected tissue-localized-like fluorescence was observed in the unicellular organism Synechocystis PCC 6803, a fluorescence in colony-appearing structures was also seen, indicating a successful test-conjugation (Figure 5). After further investigation with confocal microscopy, it became obvious that no conjugation of the cyanobacteria occured, but one of the E. coli strains seemingly took up the construct. This was concluded as successful conjugates should show both red chlorophyll autofluorescence and GFP fluorescence, resulting in a yellow signal in the overlay picture. No such combined fluorescence was detected, rendering the conjugation protocol as not functional (Figure 6). Due to a very slow growth rate, no optimisation could be implemented.

Figure 5: Three-parental-mating samples, red fluorescence binoculum.

Figure 6: Three-parental-mating samples, confocal microscopy.

Software tool LEA

To predict linker sequences we successfully built the software tool LEA (Linker Extraction from Alignments). We were able to show that the results are reasonable linker sequences that could be used for fusion protein design. Nevertheless, we were not able to carry out experiments proving an enhanced protein activity by modeled linker sequences due to time limitation. More detailed information can be found on the software tool page.