Molecular

Models

scFv

Alkaline Phosphatase

Bioreceptor

ospA

Bioreceptor Bonding

scFv

The 3-24, a single-chain variable fragment (scFv) has been previously studied by Ghosh et al. In their research, they had determined the complementarity determining regions (CDRs) of the scFv but had not solved its structure. Our molecular model was inspired by their research with slight modifications to the anchor regions. When expressing the anti-opsA scFv, the modifications we made to the protein needed not to affect the folding and the structure of the scFv, and its binding to ospA. As there was no protein data bank file to use, our team used ABodyBuilder to model the 3D structure of the anti-opsA scFv by aligning the sequence with templates that closely match different regions of the protein. Our 3D structure produced CDR regions with very close alignment to what is predicted by the paper, with the closest alignment being predicted by the Kabat system. The Kabat method was chosen not because it closely aligns with the results of the paper, but because in the 3D structure generated, the predicted CDRs form a well-defined face, contrary to the default North model, where the CDRs extended into the internal regions of the protein, which is incorrect.

Table 1 - Complimentary determining regions (CDRs) as predicted by Ghosh and Huber and ABodyBuilder.

Note that the independently determined CDRs are in complete agreement with each other, except for an additional four C-terminal amino acids in the ABodyBuilder VH CDR3. The predictive model used by ABodyBuilder which most closely aligned with the results of Ghosh and Huber was the Kabat model, which is shown in the table above, which differs from the default method, North.

Ghosh and Huber (1)	ABodyBuilder
VH CDR1: DYYLH	VH CDR1: DYYLH
VH CDR2: RINPSSGATYSPQRFQG	VH CDR2: RINPSSGATYSPQRFQG
VH CDR3: LTTFNIW	VH CDR3: LTTFNIWGFDY
VL CDR1: RASQSISTYLN	VL CDR1: RASQSISTYLN
VL CDR2: TASSLQS	VL CDR2: TASSLQS
VL CDR3: QQSYSATFTF	VL CDR3: QQSYSATFTF

Alkaline Phosphatase

As a method of detection, a reporter that could be detected even when the target protein is present in very small quantities needed to be included. The team selected alkaline phosphatase as it can convert a variety of clear substrates to coloured substrates which can be used to detect binding. There were several renditions of this enzyme’s design, which can be seen in the design and engineering success sections. After finding a successful version, we performed alignments of the iGEM part BBa_K1216001 alkaline phosphatases as well as the standard versions of the E.coli alkaline phosphatase with the highly active mutant alkaline phosphatase LINKFROMWORD. Sequence alignments via Seaview (1) revealed many differences in all regions of the protein, with large overhanging sequences on the 5’ and 3’ end and specific mutations in the binding site. 3D alignments of the structures using PyMol highlighted structural differences in the active site of the enzyme.

Figure 2 - Coloured 3D rendering of x40 Alkaline Phosphatase.

A & B domains shown in red and green respectively with zinc and magnesium ions in grey and blue respectively.

Figure 3 - 3D alignment of x40 alkaline phosphatase with BBa_K1216001 alkaline phosphatases.

x40 and BBa_K1216001 variants shown in cyan and green respectively.

Bioreceptor

To accurately predict the binding ability of our bioreceptor and thus, the viability of our proof of concept, we generated two separate fusion proteins using Chimera and PyMol. The first of these fusion proteins was the combination of the 3-24 ScFv bound via C-terminal through a (GGGS)₃ linker region to a GFP. This fusion protein was generated as a base test to aid in determining if correct protein folding will occur within an E. Coli system, in addition to being used as a basic test of ospA binding. The green fluorescent protein(GFP) was selected as the first stage in the fusion protein design as it is a simple & small protein that would have limited impact on the folding and binding of the scFv.

Figure 4 - 3D rendering of 3-24 ScFv bound to GFP.

ScFv shown in green with VH and VL CDR regions in blue and pink respectively. GFP shown in cyan bound to scFv via orange glycine linker.

The second fusion protein we created was the 3-24 sScFv which was bound via a (GGGS)₃ C-terminal linker to a mutated alkaline phosphatase (phoA). We created and modeled this fusion protein as the phoA would allow us to detect binding to low concentrations of OspA. We modeled this protein second because the size of phoA is much larger when compared to GFP and as such, we predicted it may interfere with the binding or folding of our sScFv to a greater extent.

Figure 5 - 3D rendering of 3-24 ScFv bound to x40 Alkaline Phosphatase (PhoA).

ScFv shown in green with VH and VL CDR regions in blue and pink respectively. PhoA shown in red bound to scFv via orange glycine linker.

ospA

OspA was modeled in preparation for binding testing using Cluspro (2-5) in Antibody mode. A PDB file was available for this protein which we sourced via the PDB ID 1fj1.

Figure 6 - 3D rendering of ospA.

Bioreceptor Bonding

To test the binding affinity of our generated 3-24 ScFv to ospA, we ran super-computer docking simulations through Cluspro(2-5). Cluspro is a supercomputer based in Boston , and run by the Boston University and Stony Brook University, which is designed to test billions of docking confirmations between two proteins. During analysis, the conformations with lowest-energy structures underwent root-mean-square deviation and energy minimization. Based on these predicted docking confirmations, we concluded that our sScFv would theoretically bind to ospA as an individual protein. This was further supported by investigating the exact residues within the CDR regions that are bound to ospA, which are highlighted in Figure_ & _.

Figure 7 - Cluspro 3D rendering of OspA bound to 3-24 scFv.

OspA and scFv shown in cyan and green respectively. ScFv VH and VL CDR regions highlighted in blue and magenta respectively. Full structure of CDR residues involved in binding shown.

Table 2 - Cluspro 3-24 ScFv binding to OspA model scores.

Cluster	Members	Representative	Weighted Score
0	118	Center	-323.9
0	118	Lowest Energy	-330.9
1	81	Center	-306.9
1	81	Lowest Energy	-320.1
2	40	Center	-275.4
2	40	Lowest Energy	-306.0
3	40	Center	-273.2
3	40	Lowest Energy	-306.0
4	37	Center	-337.2
4	37	Lowest Energy	-337.2
5	28	Center	-268.3
5	28	Lowest Energy	-298.1

Figure 8A - ScFv VH CDR region binding with OspA.

OspA and scFv shown in cyan and green respectively. ScFv VH CDR region highlighted in blue with the full structures of residues involved in binding shown.

Figure 8B - ScFv VL CDR region binding to ospA.

ScFv VL CDR region highlighted in magenta.

Once we understood that our scFv would bind, we wanted to determine what impact the additionof a GFP and alkaline phosphatase would have on binding affinity. Therefore, we ran two more Cluspro docking analyses both with the previously created 3-24 + GFP and 3-24 + PhoA fusion proteins. The results of these tests were promising because they indicated that both the GFP and PhoA would have limited impact on binding to OspA as the binding location and residues participating in binding are highly similar.

Figure 9 - Cluspro 3D rendering of ospA bound to 3-24 scFv + GFP Fusion Protein.

OspA, scFv, and GFP shown in cyan, green, and red respectively. Orange glycine linker shown connecting GFP to scFv.

Table 3 - Cluspro 3-24 ScFv + GFP binding to OspA model scores.

Cluster	Members	Representative	Weighted Score
0	100	Center	-321.2
0	100	Lowest Energy	-334.4
1	82	Center	-285.5
1	82	Lowest Energy	-322.9
2	39	Center	-275.4
2	39	Lowest Energy	-305.4
3	38	Center	-272.1
3	38	Lowest Energy	-292.2
4	35	Center	-277.5
4	35	Lowest Energy	-302.8
5	34	Center	-271.7
5	34	Lowest Energy	-325.0

Figure 10 - Cluspro 3D rendering of ospA bound to 3-24 scFv + PhoA Fusion Protein.

OspA, scFv, and PhoA shown in cyan, green, and red respectively. Orange glycine linker shown connecting phoA to scFv.

Table 4 - Cluspro 3-24 ScFv + PhoA binding to OspA model scores.

Cluster	Members	Representative	Weighted Score
0	49	Center	-301.6
0	49	Lowest Energy	-336.2
1	47	Center	-313.7
1	47	Lowest Energy	-341.9
2	42	Center	-355.1
2	42	Lowest Energy	-364.5
3	37	Center	-340.8
3	37	Lowest Energy	-349.8
4	35	Center	-360.1
4	35	Lowest Energy	-361.0

References

1. Gouy, M., Guindon, S., and Gascuel, O. (2010) Sea view version 4: A multiplatform graphical user interface for sequence alignment and phylogenetic tree building. Mol. Biol. Evol. 27, 221–224

2. Kozakov, D., Hall, D. R., Xia, B., Porter, K. A., Padhorny, D., Yueh, C., Beglov, D., and Vajda, S. (2017) The ClusPro web server for protein-protein docking. Nat. Protoc. 12, 255–278

3. Vajda, S., Yueh, C., Beglov, D., Bohnuud, T., Mottarella, S. E., Xia, B., Hall, D. R., and Kozakov, D. (2017) New additions to the ClusPro server motivated by CAPRI. Proteins Struct. Funct. Bioinforma. 85, 435–444

4. Desta, I. T., Porter, K. A., Xia, B., Kozakov, D., and Vajda, S. (2020) Performance and Its Limits in Rigid Body Protein-Protein Docking. Structure. 28, 1071-1081.e3

5. Kozakov, D., Beglov, D., Bohnuud, T., Mottarella, S. E., Xia, B., Hall, D. R., and Vajda, S. (2013) How good is automated protein docking? Proteins Struct. Funct. Bioinforma. 81, 2159–2166

Models scFv Alkaline Phosphatase Bioreceptor ospA Bioreceptor Bonding References Go to Top

Team:Queens Canada/Molecular-Modelling