2.1. Sequence availability
To find the Atezolizumab antibody 3D structure, the Protein data bank (PDB) was investigated. No record found for Atezolizumab antibody 3D structure on Protein data bank. further, registered patents Were searched for registered sequences of the antibody. PD-L1 and GrB sequences were achieved from UniProtKB.
2.2.Atezolizumab 3D structure prediction
Rosetta Antibody Protocol , PIGS )A web server for the automatic modeling of immunoglobulin variable domains based on the canonical structure method) and SAbPred (An antibody modelling tool and prediction software from the Oxford Protein Informatics Group (OPIG)) employed To predict the 3D structure of the Atezolizumab sequence.
RamPage Ramachandran Plot Analysis available was used to determine the quality of the built models and selection the best structure. A Ramachandran plot is a way to visualize energetically allowed regions for backbone dihedral angles ψ against φ of amino acid residues in protein structure.
2.3. Linker sequence Addition
As regards, the light and heavy chains of antibody are considered separately without any linking amino acids in 3D structure prediction, for recombinant production of the antibody, it seemed necessary to add a linker to join two chains. The (Gly4Ser)n is the most prevalent linker employed in joining two chains of9 antibody.
Three repeats of the sequence of linker, (Gly4Ser)3, were added to join the heavy and light chains. modeler 9.21 software was applied to predict the antibody 3D structure with linker. The predicted antibody 3D structure in previous step was used as the template structure in modeler 9.21 software for homology modelling. To estimate the homology between the final antibody after linker addition and the predicted antibody structure without linker, Discovery Studio software and Dali Server were used to compute the RMSD and Z-Score between them respectively.
2.4. Immunotoxin design and modelling
GrB structure as the toxic domain of the immunotoxin obtained from protein data bank (PDB). The acquired structure was in its dimeric form and with ligand interactions, accordingly MOE software was employed to eliminate its ligands and one of the monomers. In this regard, an adaptor sequence was attached before the interpolation of the antibody and the GrB to increase the GrB molecule toxicity. In the following, to predict the immunotoxin final structure combined of the GrB and antibody structures modeler homology modeling software was employed. Homology modeling predict based on template structure. In this regard modeler aligns the immunotoxin sequence with the pre-existing structures. Modeller software was applied to attain the immunotoxin structure based on Antibody (containing linker) and preparing GrB as templates.
2.5. Loop and structure refinement
Structure refinement was applied on the immunotoxin structure to achieve a more native and more stable molecule. We used the loop model class in MODELLER to refine the conformation of the loops. The loop optimization method relies on a scoring function and optimization schedule adapted for loop modeling.
ModRefiner at is an algorithm for atomic-level, high-resolution protein structure refinement, which can start from either C-alpha trace, main-chain model or full-atomic model. Both side-chain and backbone atoms are completely flexible during structure refinement simulations, where conformational search is guided by a composite of physics- and knowledge-based force field.
One aim of structure refinement is to draw the initial starting models closer to their native state, in terms of hydrogen bonds, backbone topology and side-chain positioning. It also generates significant improvement in physical quality of local structures. The standalone program also supports ab initio full-atomic relaxation, where the refined model is not restrainted by the initial model or the reference model.
2.6. Model validation
Estimating the quality of protein structure models is a vital step in protein structure prediction. QMEAN Qualitative Model Energy Analysis employed for the quality estimation of im0munotoxin structure model. RamPage (Ramachandran Plot Analysis) available was used to validate the quality of the immunotoxin 3D model before and after the refinement.
One of the ways that can corroborate the 3D structure accuracy is secondary structure determination of the sequence .in order to predict the immunotoxin secondary structure following software and services were used: Jpred, PSIPRED and SOPMA
In the next step, the secondary structure of the modeled molecules was compared with the predicted secondary structure acquired from MOE software to indicate if they are in concordance with each other.
2.7. Immunotoxin characterization
Further characterization of the immunotoxin was carried out using following software and services: ProtParam at is a tool which allows the computation of various physical and chemical parameters. The computed parameters include the molecular weight, theoretical pI, amino acid composition, atomic composition, extinction coefficient, estimated half-life, instability index, aliphatic index and grand average of hydropathicity (GRAVY).
Periscop is a sequence-based predictor that estimates the expression level and yield of soluble protein in the periplasm of E. coli.
ccSol predicts protein solubility using physico-chemical properties.
Further research were done to investigate the recombinant protein antigenicity and toxicity.
VaxiJen is a tool for prediction of protective antigens and subunit vaccines. ToxinPred at is an in silico method, which is developed to predict and design toxic/non-toxic peptides. AlgPred at was used to determine the possible allergic feature of the recombinant molecule. To specify and characterize the existing epitopes within the immunotoxin sequence, various software was used. Propred and Propred-1 were utilized to specify T-cell epitopes 1 and 2, respectively.
In order to identify the B cell epitopes, Ellipro , Bepipred and SVMtrip were used.
2.8. Immunotoxin binding efficiency
The interaction possibility between the immunotoxin and its receptor can be investigated using its 3D structure for molecular docking study. First, PD-1 3D structure was obtained from PDB database and edited using MOE software. The obtained structure was in Complex with Atezolizumab-Fab form. Hence, MOE software was applied to remove extra domains and ligands. The possible interaction between PD-1 and immunotoxin was predicted by Zdock server ClusPro server can be used to explore the designed immunotoxin functionality. To calculate the Gibbs free energy of reaction between these two proteins PRODIGY server was applied.