2. Molecular Dynamics Simulations
- Figure 6: MD simulations performed with CHARMM force-field and parameter files using VMD and NAMD. All simulations were performed for 1000ns, which simulates human body conditions for 50000 frames.
A Mechanistic Study of Antibody Binding to Alpha-Synuclein for the Treatment of Parkinson’s Disease
Malcolm C. Harrison, Pin-Kuang Lai
Department of Chemical Engineering and Materials Science, Stevens Institute of Technology, Hoboken, NJ, USA
- Parkinson’s Disease (PD) is the second most common neurodegenerative disorder
- No cure has been found for the disease, which affects 2-3% of the population above 65.
- Degeneration of dopaminergic neurons caused by the inclusion of cytotoxic Lewy bodies made up of misfolded alpha-synuclein (α-syn) aggregates.
- Roche and Prothena are developing a monoclonal antibody (mAb) called Prasinezumab (PRX002), which targets the C-terminus of α-syn.
- Computational methods of protein modeling, preparation, analyzation, and docking were performed on the mAb and α-syn (PDB ID: 2KKW).
- Molecular dynamics (MD) simulations were performed with VMD and NAMD to simulate binding in silico with intercellular conditions.
- MDAnalysis used to calculate hydrogen bond frequency and location
- DeepSCM used to test PRX002’s developability for subcutaneous injection
This work was funded by National Science Foundation, REU/RET Site Grant #2050921.
We acknowledge the computing resource from ACCESS (BIO220114 and CHM210013)
- H-bonds form at between PRX002 and the target epitope of a-syn, primarily at HCDR2 and LCDR1
- RMSD analysis shows that once bound, the CDR regions do not detach or deviate during simulation
- The SCM score for PRX002 is 695.50, predicting it will have low viscosity in high concentrations for subcutaneous injection, and the TAP analysis showed low aggregation potential and similarity to clinical-stage therapeutics
- More analysis on the binding free energy and affinity need to be performed to
Conclusions and Future Work
>chain_H
EVQLVESGGGLVQPGGSLRLSCAASGFTFSNYGMSWVRQAPGKGLEWVASISSGGGSTYYPDNVKGRFTISRDDAKNSLYLQMNSLRAEDTAVYYCARGGAGIDYWGQGTLVTVSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSCDKTHT
>chain_L
DIQMTQSPSSLSASVGDRVTITCKSIQTLLYSSNQKNYLAWFQQKPGKAPKLLIYWASIRKSGVPSRFSGSGSGTDFTLTISSLQPEDLATYYCQQYYSYPLTFGGGTKLEIKRTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC
- Figures 7a-c: Graphs of RMSD of the three heavy-chain CDR (HCDR) and light chain CDR (LCDR) regions of PRX002
- Figures 8a-c: Bonds that form most often throughout the simulation, with hydrogen bonds displayed in yellow.
1. Docking, Minimization, and Solvation
- Figure 5: The minimized complex was then added to a water box, and ions were added to create an overall neutral charge
- Figure 4: The PRX002 Fab structure and α-syn underwent protein-protein docking using PIPER, and energy minimization was performed using NAMD
- Figure 1: PD symptoms and pathological hallmarks
- Figures 2-3: Sequence for PRX002 was obtained from the Kyoto Encyclopedia of Genes and Genomes, and the structure of α-syn from the Protein Data Bank. BioLuminate was used to create the fragment antigen binding (Fab) region of PRX002
3. Hydrogen Bonding, RMSD, and Developability
- MDAnalysis, a Python library was used to calculate the amount and frequency of hydrogen bonds (h-bonds) during the simulation
- VMD’s Root-mean-square deviation (RMSD) analysis module calculated the RMSD of the complementarity determining regions (CDR) every frame in the simulation using the initial frame as a reference point
- DeepSCM and Therapeutic Antibody Profiler (TAP) were used to calculate the theoretical spatial charge map (SCM) score to predict viscosity in high concentrations and the aggregation potential as well as similarity to other clinical-stage therapeutics respectively