Overview
Working with proteins in our project, we first needed to design the genes coding for them. After that, we needed to find out the right conditions for the following steps. These steps include the transformation, expression, purification and immobilization. These had to be done before we could start with the functional testing of our proteins. The general difficulty in all steps was that no data was given for our newly designed proteins. So we needed to try out different conditions for every single step to find out what worked best for our proteins.
Gene Design
Our goal was to design a protein, which we can control regarding ligand binding and release. Since we already knew the binding properties of 2ABH [1], we chose this one and tried to modify it. Because this protein is a small monomer, we agreed to add a photosensitizer to it that would destroy the protein after being extended to blue light. As the photosensitizer, we chose SOPP3.
Testing a reversible approach, we chose VVD as the switchable protein. For this photoreceptor protein, we needed a phosphate binding protein that consists of two strands. Therefore, we chose CMI as the phosphate binding part.
To minimize the influence of the photosensitive proteins on the phosphate binding ability, we used a linker polypeptide [2] to connect the two proteins.
After that, we optimized the genes to be expressed in Escherichia coli BL21 (DE3) gold.
We used the pET-28a(+) vector to utilize the kanamycin resistance and obtain a polyhistidine tag at the end of the protein. The polyhistidine tag consists of 6 histidine residues and was added at the N-terminal ending of the proteins because the active site of 2ABH is near the C-terminal ending. We used the restriction sites NdeI and XhoI to insert each fusion protein, as shown in figure 1. pET-28a(+) has also a lac operon (lac I) to control the expression. It can be induced by either lactose or IPTG.
Figure 1: Gene map of the designed VVD-CMI fusion protein in pET-28a(+). Blue: self-designed fusion protein; purple: lac operon; light green: kanamycin resistance (made by Benchling)
To save time, we ordered the gene constructs for SOPP3-2ABH and VVD-CMI already inserted in the vector. The single 2ABH was used as a negative control for our experiment. To do this, we ordered the single gene and inserted it into the vector ourselves, using Gibson Assembly.
Gibson Assembly
Since our 2ABH gene was still in isolated form, we had to insert it into the vector we were going to use for transformation, namely pET-28a(+). To accomplish that, we performed a Gibson Assembly.
For the Gibson Assembly, the target gene and the vector needed overlapping gene sequences. These overlapping gene sequences can be achieved through PCR. Therefore, primers containing an end identical to the vector segment and an end annealing to the insert sequence were used for the amplification of the insert (primers P1 and P3). The same procedure was used to amplify and create the overlapping sequences for the vector (primer P2 and P4). All primers are listed in table 1.
Table 1: Sequences of PCR primers for DNA amplification
| Primer Name | Sequence (5'- 3') |
|---|
| P1 | GTTTAATTTAAGAAGGAGATATACCATGGAAGCGAGCCTGACGGCG |
| P2 | CATGGTATATCTCCTTCTTAAAGTTAAAC |
| P3 | TCAGTGGTGGTGGTGGTGGTGCTCGAGATAGAGAGGTTTACCAG |
| P4 | TCGAGCACCACCACCACCACCACTG |
Afterward, we checked the success of the vector amplification with an agarose gel electrophoresis in 1% agarose gel. As shown in figure 2 by the band at 5 kb, pET-28a(+) could be amplified. The concentration was determined to be 57.9 ng/µL.
Figure 2: 1 (left) Agarose gel electrophoresis of the 2ABH gene (A: DNA marker; B: PCR sample 2ABH); 2. (right) Agarose gel electrophoresis of the pET-28a(+) - vector (A: DNA marker; B: PCR sample pET-28a(+))
We were able to amplify the 2ABH gene as the agarose gel electrophoresis showed a distinct band at 1 kb, where the gene was expected. The concentration of the amplified 2ABH gene was measured to be 96 ng/µL.
Now that we had enough DNA, we were able to start with the actual Gibson Assembly. To do this, we pipetted the master mix together on ice and added the DNA. After incubating at 50 °C for 20 minutes, we added competent E. coli cells and performed a transformation. We then plated the cells on an LB medium containing kanamycin and incubated them overnight at 37 °C.
The next day, we picked nine clones and performed a colony PCR to check for cells that contained the pET-28(+) vector carrying the 2ABH insert. For the PCR, we used primers P1 and P3 again to ensure that only the insert was amplified and visible on the following agarose gel electrophoresis. The agarose gel electrophoresis (figure 3) revealed a clone carrying the desired vector construct in lane D, which showed a band at 1 kb.
Figure 3: Colony PCR to find colonies carrying the pET-28a(+) construct with 2ABH. A: NEB QuickLoad 1kb extend DNA ladder; B-J: Colony samples of transformed cells
This clone was used to inoculate a preculture at 37 °C overnight. The plasmid DNA was then purified from the preculture using the Machery-Nagel NucleoSpin protocol, resulting in a concentration of 101 ng/µL, and sent to Microsynth for sequencing. This revealed three mutations at wobble bases within our 2ABH gene that did not impact the amino acid sequence. This concluded the Gibson Assembly and proved the successful insertion of 2ABH in pET-28a(+), which could then be used for transformation.
Transformation
Transferring our genes into the E. coli BL21 (DE3) gold cells, we performed transformations. Because E. coli cells are not naturally competent, we made them chemically competent and used a heat shock afterward to insert the DNA.
Subsequently, we plated the cultures to grow single colonies. For this purpose, we used agar containing kanamycin. Because the vector we used for our genes contains a gene for kanamycin resistance, only cells carrying the plasmid could grow. As you can see in the results and in figure 4, we managed to grow single colonies. This indicates that these cells took up the plasmid and could be further used for production.
Figure 4: LB agar plate with transformed E. coli BL21 (DE3) grown over night
Expression
Because we designed new proteins, there was no data available for the best expression conditions, so we tested different media and temperatures. We used data describing the expression of individual parts of the fusion proteins as an orientation. To check if the expression was successful or not, we conducted SDS-PAGEs.
Since we knew that 2ABH can be expressed in E. coli at 37 °C [1], we started with that temperature for 2ABH, SOPP3-2ABH, and VVD-CMI. The medium we used was TB medium, as it was recommended by our supervisors as a good standard expression medium. To obtain the highest possible yield, we inoculated the main expression culture with a preculture. This ensured that the cells were already fit and entered the exponential growth phase more quickly. After an OD of about 1 was reached, the main culture was induced with IPTG. This led to the activation of the lac operon and subsequently the start of the expression. The expression itself ran for 20 to 24 hours.
2ABH showed good results at 37 °C (see Results). Therefore, we decided to use 37 °C as our standard expression temperature for 2ABH in the following experiments.
As you can see in figure 5, the expression for SOPP3-2ABH (50 kDa) and VVD-CMI (30 kDa) at 37 °C did not work very well, so we tested an expression temperature of 30 °C.
Figure 5: Expression control of SOPP3-2ABH and VVD-CMI. A: Size marker; B: SOPP3-2ABH supernatant; C: SOPP3-2ABH pellet; D: VVD-CMI supernatant; E: VVD-CMI pellet
In figure 6 (part I), you can see the SDS-PAGE showing the samples of SOPP3-2ABH that were expressed at 30 °C. There are big spots slightly under the 55 kDa mark in lane D and G. This is the area where we expected our protein to be. This suggests that our expression of SOPP3-2ABH was successful.
Figure 6: Expression of SOPP3-2ABH and VVD-CMI. Part I: SOPP3-2ABH, A: Size marker; B: Flow fraction; C: Wash fraction; D: Eluate; E: Flow fraction; F: Wash fraction; G: Eluate. Part II: VVD-CMI, H: Flow fraction LB medium; I: Wash fraction LB medium; J: Eluate LB medium; K: Flow fraction TB medium; L: Wash fraction TB medium; M: Eluate TB medium; N: Expression control autoinduction medium
Even though CMI is also from E. coli, the fusion protein VVD-CMI could not be expressed as easily as expected. The expressions at 37 °C and 30 °C (see Results) were not successful. First, we continued to express in TB medium and just lowered the expression temperature to 20 °C, but this still did not work. After that, we consulted with experts and decided to try out different media and lower the temperature to 18 °C. Because the metabolism of the cells is very slow at 18 °C, we gave them 3 days to express the protein. We compared LB medium, TB medium, and autoinduction medium. VVD-CMI is a homodimer that is present as two single amino acid chains in the absence of light. Each chain has a size of 30 kDa, so we expected the protein to be found between the 35 kDa and the 25 kDa size mark. As you can see in figure 6 part II, the expression of VVD-CMI worked when using the autoinduction medium.
Purification
To purify our proteins, we used the $His_{6}-Tag$ which was already coded on the pET-28a(+) vector and fused it to our protein by inserting the target gene in the suitable position on the vector. For the purification, we used Ni-IDA columns from Macherey-Nagel. The purification posed the same problem as before, as there were no exact data available for how to purify the fusion proteins. We used a standard protocol from the company and consulted with our supervisors. We decided to try out different imidazole concentrations to find out which concentration worked best for elution. We first tested it with SOPP3-2ABH. As you can see in figure 7, at a concentration of 500 mM imidazole proteins were still being eluted. Because no other proteins were eluted, we decided to use the highest concentration of imidazole we tested for elution. Since the highest imidazole concentration worked for the purification of SOPP3-2ABH, we used the same concentration for the other protein purifications as well.
Figure 7: Purification of SOPP3-2ABH using different imidazole concentrations for eluation. A: Size marker, B: Flow fraction, C: Wash fraction, D: 5 mM imidazole, E: 25 mM imidazole, F: 50 mM imidazole, G: 100 mM imidazole, H: 200 mM imidazole, I: 300 mM imidazole, J: 400 mM imidazole, K: 500 mM imidazole
After optimizing the expression conditions for VVD-CMI, we followed the same protocol for purification. However, after purification, we detected low protein concentrations using the BCA Assay. In the SDS-PAGE, the purified samples of VVD-CMI also did not show the expected results. At the expected level for VVD-CMI, no clear bands were detectable (see Results). Due to the advanced time, we were not able to adjust the purification conditions for VVD-CMI.
Immobilization
To test if our protein can bind phosphate and to test the impact of blue light exposure, we had to take samples of the supernatant at different time points. To hold back our protein during the sampling, we immobilized it. For the immobilization, we used EziG beads from EnginZyme AB. Those beads are made out of pored glass coated with an organic polymer and chelated Fe(III) ions [4]. The coating can bind the polyhistidine rest we already used for purification. Additionally, there are three different variants of beads which differ in their properties, so that they can be used for proteins with different needs. To control how much protein we bound to the beads, we used the BCA Assay or the Bradford Assay. We took samples before and after the binding and compared them. This was done for the three different variants of the EziG beads (amber, coral and opal), to see what type was best with our protein SOPP3-2ABH. We expected a decrease in the protein concentration in the supernatant. As shown in figure 8 our expectations were fulfilled and we could observe a decrease in the protein concentration in the supernatant. For the exact values, have a look at our results.
Figure 8: Protein concentration of the supernatant before and after immobilization of SOPP3-2ABH for different EziG beads
As a negative control for SOPP3-2ABH, we used the single protein 2ABH. To be able to compare the results, we treated 2ABH the same way. That means that we used the same protocols for purification and immobilization.
In the further course of the project, the binding capacity, and release of phosphate of 2ABH and SOPP3-2ABH were tested. Please have a look at the proof of concept for the results.
