Team:Aboa/Design
Design
In this page we describe how we designed our antibody fragments and what tests they went through when we tested if antibody modifications have any effect on antibody functionality. If you want to know more about the project, please read our Project Description .
Figure 1. Project timeline
Antibody fragment modification
We, like any other iGEM team, had very little time to spend on the project. The iGEM competition takes less than year and it does include lots of things to do, so we needed to keep our project as simple as possible. Usually antibodies are produced with the same Esherichia coli strains and their sequences need to be optimized and modified so as not to be toxic to the producing organism (1) .
However, our idea included expanded genetic code and using an E. coli strain, 321.deltaA, that was pretty unknown to us (2) . We did not want to take any risks, so we decided to give up our idea of producing capture antibodies to Troponin (you can read why we first thought about troponin from our Human Practices ). Instead, we decided that, realistically, we should complete a proof-of-concept study about site specific immobilization of antibody fragments on DBCO beads. In the end we decided to use the anti-digoxigen Fab in our research as it is a widely used codon harmonized antibody for different E. coli -strains (1). As a joker card, we designed and tried to produce anti-NS1 fabs for non-structural protein of Dengue Virus, but that didn't go too well.
Then we had to figure out how we should modify the antibody so it would attach to the DBCO bead and would not lose its ability to bind Digoxigenin.
We used known 3d models of humanized Fabs from public databases, like National Center for Biotechnology Information (NCBI). We searched for 3D models of amino acids that had a solvent accessibility value greater than 0.3 and had a large side chain. Then we replaced an amino acid codon with a TAG -codon that is coding p-azidophenylalanine intead of stopping the translation. We also added 6x histidine tail on the C-terminal end of the heavy chain.
In the end we ordered and constructed the following gene variants from IDT:
Anti-digoxigenin:
- Sequence with no TAG-codons = Digox-Fab0
- Digox-Fab0 with Histag-TAGTAA = Digox-Fab460
- Light chain 191th Lysine changed to TAG-codon = Digox-Fab191.
- Light chain with added 216th TAG-codon = Digox-Fab216.
- Light chain 108th lysine changed to TAG-codon = Digox-Fab108.
- Heavy chain 263rd cysteine changed to TAG-codon = Digox-Fab263.
- Heavy chain with added TAG-codon to 265th codon = Digox-Fab265.
- Heavy chain with added TAG-codon to 271th codon = Digox-Fab271.
Anti-DengueNS1:
- Sequence with no TAG-codons = Dengue Fab0
- Light chain 152th aspargine changed to TAG-codon = Dengue-Fab152
- Heavy chain 221th serine changed to TAG-codon = Dengue-Fab221
- Heavy chain 196th threonine changed to TAG-codon = Dengue-Fab196
We use the pLK04RCH plasmid from University of Turku for production of our Fabs. The main function of this vector is to confer the required properties to any Fab protein it produces. First of all, it contains a SfiI restriction site inside a PelB signal sequence. This signal sequence is used to relocate the produced protein to the periplasmic space. Expression can be induced with IPTG, as the vector contains a Lac promoter upstream of the PelB sequence. It also contains an ampicillin resistance gene for antibiotic selection. This vector backbone has been added to iGEM registry as a BBa_K2941004
Figure 2. pLK04RCH plasmid used for Fab production
Production and validation
Protein production and purification are done as they usually are. You can see our protocols from here . The next step is to validate if our designed Fabs have incorporated p-azidophenylalanine into the protein.
Figure 3. A plasmid containing an orthogonal tRNA/tRNA-synthetase pair that is able to incorporate p-azido-l-phenylalanine in the place of a TAG-stop codon (pEVOL-pAzF) and a plasmid containing our TAG modified Fab fragment (pLk04RCH) are inserted into E.coli 321.deltaA.exp. When induced correctly, these plasmids will start to produce the modified protein and secrete it into the periplasmic space of the cell to fold and form disulphide bridges correctly. The protein of interest is relatively easy to purify with the recombinant 6x-His-tag, which is located at the pLK04RCH plasmid.
To find out whether the Fabs we produced had p-azidophenylalanine incorporated in them or not, we labeled them with DBCO-Cy5.5. We have to take into account possible unspecific binding, so Fab0 samples - that have no p-azidophenylalanine incorporated into the protein - serve as a negative control. After this the labeled proteins are analyzed with SDS-PAGE. If there is p- azidophenylalanine incorporated in the Fabs, they will be labeled with Cy5.5 and can then be visualized with an appropriate label reader. The signal in Fab0 should be a lot lower, if there is any signal at all.
Figure 4. Strain promoted Azide Alkyne Cycloaddition (SPAAC). This reaction is a refined version of the original copper catalyzed Huisgen cycloaddition that no longer needs a catalytic copper(+1). The end result is a stable covalent bond that is created with high specificity between DBCO and azide groups.
Immobilizing and testing capture surface
The last step of our project is to immobilize produced capture antibody Fab fragments onto a test surface and test the surface capacity and make sure that the Fab is still functional and binds the antigen as intended.
In this case we immobilize antibodies onto DBCO functionalized beads. DBCO-azide reaction is rapid and spontaneous and it does not need any specific buffer to happen, so immobilization is quite a straight forward process (3) .
We also need to immobilize Fab0s' onto the same kind of beads to have something whit which to compare our site-specific immobilizations. Fab0s' do not contain azide-group naturally, so we have to use NHS-Azide -reagent that reacts with the antibody's amine groups like conventional biotin (4). Then they are allowed to react with DBCO beads, but this time they will orient randomly, just like they do in current tests.
Figure 5. DBCO coated magnetic bead
The final step in our project is to compare the beads that have the capture antibody that’s been immobilized using site-specific with click-chemistry to the beads that have the capture antibody that’s been immobilized randomly with nhs-chemistry. This will also reveal if there is any change in the antibody function caused by non-canonical amino acid in the protein.
We saturate the antibody coated beads with Cy5 labeled digoxigenin. We analyze the signal each bead gives with flow-cytometry and compare the site-specifically coated beads' and the randomly coated beads' signal intensity and variation. The higher the signal intensity, the more digoxigenin there is in the bead and less variation means a more homogenous capture surface.
References
1. Kulmala, A., Huovinen, T. & Lamminmäki, U. (2017) Effect of DNA sequence of Fab fragment on yield characteristics and cell growth of E. coli. Sci rep. doi: 10.1038/s41598-017-03957-6.
2. pEVOL-pAzF plasmid from Addgene. [html document visited 10/12/2019]
3. (2028) DBCO magnetic beads. Jena bioscience. PDF document.
4. Hermanson, G.T. (1996) Bioconjugate Techniques, Acad. Press, San Diego.