Firstly, we obtained the amino acid sequences of two single chain antibodies through literatures, and used MOE (Molecular Operating Environment) [1] to measure the sequence, complementarity determining regions (CDRs), and framework regions (FRs) of monoclonal antibodies.
Next, the amino acid sequences of VH and VL of the monoclonal antibody were homologous compared by fragment database using MOE, to predict the NeuA and NeuB three-dimensional structure and extract the most accurate data from the best scores to estimate the modeling quality. We used Mol Probity [2] to carry out structural analysis and quality evaluation of the 3D structural model predicted by MOE, and performed conformational optimization to approximate the native state by Rossetta Relax program [3-6].
of NeuA (left) and NeuB (right).
The 3D structure of the Fel d 1 from protein data bank (PDB 1PUO) was used. We deleted the last two amino acids in the 1PUO pdb file to make its sequence consistent with that used by the experimental group.
Since the binding of the Fel d 1 protein to IgE in human blood is the mechanism that causes feline allergy, we hope that our antibody will be able to block IgE binding by covering the antigen's IgE binding epitopes. Therefore, the next step is to consult the literature and analyze the epitope of the antigen.
Three important IgE epitopes have been defined by van Milligen et al on the Fel d 1 allergen using 14-residue overlapping peptides spanning both chains, two in chain 1 (residues 117-130 and 138-151) and one in chain 2 (residues 15-28). These three epitopes are located on the neighboring helices H6-H7, H7-H8, and H1-H2, respectively, and connecting loops [7]. Amino acids of these three IgE binding peptides that are surface exposed are shown in green boxes.
Later, a novel IgE-binding epiotpe of Fel d 1 was characterized by Tasaniyananda et al, and it is located in chain 1 with residues 104-117 [8]. Residues that formed the epitope are in red boxes.
Then, the refined model of the scFv and the Fel d 1 structure were docked according to the Fast Fourier Transform (FFT)-basedd program, i.e., PIPER. The antibody mode available on the automated ClusPro 2.0 protein-protein docking server was used for the docking [7,9]. Ten clusters with highest model scores was chosen and statistical analysis of interaction sites was performed by MOE. The ChimeraX and Pymol software (The PyMOL Molecular Graphics System, Schrodinger, LLC) was used for visualizing the intermolecular interaction.
Y-axis represents the amino acid sequence of Fel d 1 and the X-axis represents the number of interactions.
The statistical results of rigid docking showed that NeuA protected two peptide regions in Fel d 1, corresponding to amino acids 93-123, 138-158; NeuB protected two peptide regions corresponding to amino acids 5-31 and 49-73, basically achieving complete coverage of antigen epitopes.
It was found that our antibody can not only achieve the coverage of the relevant epitope region, but also bind to the specific IgE binding amino acid residues exposed to the solvent. The results showed that NeuA covered the IgE conformational epitope on Fel d 1 chain 2, which consisted of F15, N19, L23 and L24 between H1 and H2. NeuB covers IgE conformational epitopes on Fel d1 chain 1, L104, T107, T109, E112, E115, and Q119 between H5 and H6. The combined use of the two antibodies covered 3 IgE binding regions, suggesting that NeuA/B had a high ability to block Fel d 1 binding of serum IgE in allergic patients, and could play a high blocking effect.
Peptides covered by NeuA and NeuB are shown in purple and blue letters respectively. Amino acids specifically bound by antibodies are shown in gray boxes, and sites that coincide with amino acids of the four surface-exposed IgE-binding peptides are shown in dark red and dark green boxes.
Finally, through the analysis of the results, we found that two ScFvs do not have steric hindrance and competition with each other because they bind to the two surfaces of the antigen, so we can combine the co-expressed NeuA and NeuB to achieve a large increase in blocking efficiency. Experiments have proved our conclusion.
[1]Chemical Computing Group ULC. Molecular Operating Environment (MOE), 2022.02; Chemical Computing Group, Inc.: Montreal, QC, Canada, 2023.
[2]Brenke R, Hall DR, Chuang GY, Comeau SR, Bohnuud T, Beglov D, Schueler-Furman O, Vajda S, Kozakov D. Application of asymmetric statistical potentials to antibody-protein docking. Bioinformatics. 2012 Oct 15;28(20):2608-14.
[3]Nivón LG, Moretti R, Baker D. A Pareto-optimal refinement method for protein design scaffolds. PLoS One. 2013;8(4):e59004.
[4]Conway P, Tyka MD, DiMaio F, Konerding DE, Baker D. Relaxation of backbone bond geometry improves protein energy landscape modeling. Protein Sci. 2014 Jan;23(1):47-55.
[5]Khatib F, Cooper S, Tyka MD, Xu K, Makedon I, Popovic Z, Baker D, Players F. Algorithm discovery by protein folding game players. Proc Natl Acad Sci U S A. 2011 Nov 22;108(47):18949-53.
[6]Tyka MD, Keedy DA, André I, Dimaio F, Song Y, Richardson DC, Richardson JS, Baker D. Alternate states of proteins revealed by detailed energy landscape mapping. J Mol Biol. 2011 Jan 14;405(2):607-18.
[7]van Milligen FJ, van 't Hof W, van den Berg M, Aalberse RC. IgE epitopes on the cat (Felis domesticus) major allergen Fel d I: a study with overlapping synthetic peptides. J Allergy Clin Immunol. 1994 Jan;93(1 Pt 1):34-43.
[8]Tasaniyananda N, Tungtrongchitr A, Seesuay W, Sakolvaree Y, Indrawattana N, Chaicumpa W, Sookrung N. A novel IgE-binding epitope of cat major allergen, Fel d 1. Biochem Biophys Res Commun. 2016 Feb 12;470(3):593-598.
[9]Kozakov D, Beglov D, Bohnuud T, Mottarella SE, Xia B, Hall DR, Vajda S. How good is automated protein docking? Proteins. 2013 Dec;81(12):2159-66.