Effect of Electric Field on α-Synuclein Fibrils: Revealed by Molecular Dynamics Simulations

The self-association of amylogenic proteins to the fibril form is considered a pivotal factor in the pathogenesis of neurodegenerative diseases, including Parkinson’s disease (PD). PD causes unintended or uncontrollable movements in its common symptoms. α-synuclein is the major cause of PD development and thus has been the main target of numerous studies to suppress and sequester its expression or effectively degrade it. Nonetheless, to date, there are no efficient and proven ways to prevent pathological protein aggregation. Recent investigations proposed applying an external electric field to interrupt the fibrils. This method is a non-invasive approach that has a certain benefit over others. We performed molecular dynamics (MD) simulations by applying an electric field on highly toxic fibrils of α-synuclein to gain a molecular-level insight into fibril disruption mechanisms. The results revealed that the applied external electric field induces substantial changes in the conformation of the α-synuclein fibrils. Furthermore, we show the threshold value for electric field strength required to completely disrupt the α-synuclein fibrils by opening the hydrophobic core of the fibril. Thus, our findings might serve as a valuable foundation to better understand molecular-level mechanisms of the α-synuclein fibrils disaggregation process under an applied external electric field.


Introduction
Neurodegenerative diseases, a group of late-onset progressive nervous system diseases, posed a severe challenge before modern medicine. Parkinson's disease (PD) is a widespread neurological disorder that is pathologically characterized by progressive loss of dopaminergic neurons [1,2]. Aside from common motor disorders such as bradykinesia, tremor, rigidity, and postural instability [3], PD also severely affects the quality of life through complications such as cognitive impairment, mental health disorders, sleep α-synuclein monomer and dimer were investigated by means of atomistic discrete MD simulations [44]. The modeling results predicted the formation of partial helices around the N-terminus (residues . The different types of β-sheet conformation occurred in the range of residues 35-56 (N-terminal tail) and residues 61-95 (nonamyloid β-component region). In α-synuclein dimers, some disordered parts of the α-synuclein conformationally transformed into the β-sheet conformation. Other simulation studies also show the importance of dimerization in triggering the α-synuclein aggregation by conformational transformations into both intramolecular β-hairpin and β-sheet [45]. Moreover, the effect of specific conditions, e.g., pH and ions and charge alterations, were also studied by applying specific computational methods [46][47][48]. Thus, knowledge about the nature of interactions between certain regions of α-synuclein plays a critical role in preventing its aggregation [49]. Such precise information and biochemistry findings will help us develop a mechanistic understanding of protein aggregation diseases and ultimately triumph over such disorders.
In the present research, we use MD simulations to study electric field-induced changes in αsynuclein fibril conformation. In addition, we unravel the threshold value of the electric field for total disaggregation of α-synuclein fibrils.

Results and Discussion
We carried out MD simulations to investigate the static EF effect on the conformational changes of the α-synuclein fibril. The chosen mutated α-synuclein H50Q narrow fibril structure displays a tendency for faster aggregation kinetics and higher toxicity in comparison to the wild type α-synuclein structure [50]. Thus, the disruption of such fibrils is important in combat against amyloid-based diseases, including PD. Figure 1 shows the final snapshots of a replica 1 (out of four) of the 600 ns MD simulation of the α-synuclein pentamer structure. As is clear, there is almost no change in the α-synuclein pentamer in the absence of EF, i.e., its conformation is quite similar to experimental findings. The low values of EF, e.g., 0.05 and 0.10 V/nm, induced negligible change. However, the N-terminal end of the α-synuclein pentamer (i.e., residues between 36-46) unfolded and moved further from the main core of fibril, starting from 0.05 V/nm EF. Similar alterations in conformation were observed in the case of 0.15, 0.20 and 0.25 V/nm EF. Moreover, the secondary structure As is clear, there is almost no change in the α-synuclein pentamer in the absence of EF, i.e., its conformation is quite similar to experimental findings. The low values of EF, e.g., 0.05 and 0.10 V/nm, induced negligible change. However, the N-terminal end of the α-synuclein pentamer (i.e., residues between 36-46) unfolded and moved further from the main core of fibril, starting from 0.05 V/nm EF. Similar alterations in conformation were observed in the case of 0.15, 0.20 and 0.25 V/nm EF. Moreover, the secondary structure analysis shows a 17% reduction of the β-sheet conformation for 0.25 V/nm (see Table 1). Note that the β-sheet conformation plays one of the dominant roles in stabilizing fibril-like structures [51]. The further increase in the EF strength from 0.30 to 0.40 V/nm, caused even more impact on fibril conformation. Specifically, β-sheet conformation decreased almost four times and the major percentage of this conformation transformed to the coil conformation, which was doubled in 0.4 V/nm in comparison to the absence of EF (cf. β-sheet and coil conformation in Table 1). Moreover, the helical conformation also started emerging in higher intensity of EF (see Supplementary information Figure S1). Thus, the low intensity of EF such as 0.05-0.25 V/nm was sufficient to induce conformational changes but the core of the fibril remained. In contrast, the higher values of EF, e.g., 0.3-0.4 V/nm, caused more changes that resulted in the opening of fibril's core and mainly turned it to coil conformation (cf. Figures 1 and S1 and Table 1). Table 1. Secondary structure analysis of the α-synuclein pentamer in each model system. The various conformation occurrences (%) of protein's various secondary structure components. The calculated backbone root mean square deviation (RMSD) plot shows that 0.05-0.25 V/nm EF strength disturbed fibril structure and caused higher fluctuations compared to the absence of the EF (see Figure 2). However, these fluctuations lead only to the unfolding of the N-terminal end, i.e., residues from 36 to 46, and the hydrophobic core remained unchanged (see Figure 1). However, 0.30-0.40 V/nm EF induced more changes in the conformation of the α-synuclein structure, in that the occurrence of β-sheet conformation decreased significantly, and this conformation mainly was altered into coil conformations (see Table 1 and Figure S1). Finally, in the case of 0.40 V/nm EF, the α-synuclein fibril completely unfolded (see Figure 1) and the hydrophobic core of the α-synuclein fibril completely opened which was remained at lower EF strengths. According to the RMSD, the major change occurs within the initial 100 ns simulation, i.e., in high EF intensity such as 0.3, 0.35, and 0.4 V/nm (see Figure 1). The secondary structure map also shows insignificant changes in conformation in the rest of the simulation time (see Figure S1).

EF (V/nm) α-Helix
The RMSD values of backbone atoms increased almost ten times as the EF strength increased (see Figure 2). Likewise, the solvent accessible surface area (SASA) and radius of gyration (R G ) of fibril increased at higher EF strengths. The SASA and R g of fibrils reached the highest values and the conformation of each peptide became almost linear, similar to the primary structure of proteins. Furthermore, under this condition, the narrow and uniform shape of the α-synuclein fiber turned into a flat form which marked a full disaggregation point for the α-synuclein fiber. In addition, the root mean square fluctuations (RMSF) was calculated to understand the flexibility and dynamics of different regions of the peptide located in the middle of the pentamer chain C (see Figure S2). The aim of choosing the latter is associated with its stability and this chain is highly buried by neighboring chains. It is also evident that the presence of EF influenced the mobility of residues. As a result, fluctuations of chain C had considerably changed and showed greater values between 60 and 90 residues at 0.3 and 0.4 V/nm EF intensity (see Figure S2).  The RMSD values of backbone atoms increased almost ten times as the EF strength increased (see Figure 3). Likewise, the solvent accessible surface area (SASA) and radius of gyration (RG) of fibril increased at higher EF strengths. The SASA and Rg of fibrils reached the highest values and the conformation of each peptide became almost linear, similar to the primary structure of proteins. Furthermore, under this condition, the narrow and uniform shape of the α-synuclein fiber turned into a flat form which marked a full disaggregation point for the α-synuclein fiber. In addition, the root mean square fluctuations (RMSF) was calculated to understand the flexibility and dynamics of different regions of the peptide located in the middle of the pentamer chain C (see Figure S2). The aim of choosing the latter is associated with its stability and this chain is highly buried by neighboring chains. It is also evident that the presence of EF influenced the mobility of residues. As a result, fluctuations of chain C had considerably changed and showed greater values between 60 and 90 residues at 0.3 and 0.4 V/nm EF intensity (see Figure S2).
Interprotein interactions, specifically in fiber-like proteins, lateral hydrogen bonds between peptides, hydrophobic packing of residues, and salt bridges play a vital role in stabilizing and stimulating further elongation of fibrils [52][53][54]. The extensive number of hydrogen bonds between individual β-strands and long-range interactions drive cytotoxic fibril formation [55]. Our results show that the average number of the inter-and intrapeptide hydrogen bonds per chain gradually decreased under the influence of EF (see Table 2). This hydrogen bond loss affects α-synuclein fiber stability by lowering the strength of intrapeptide interactions. Thus, highly ordered β-sheet-rich fibrils are quite sensitive to EF.  Interprotein interactions, specifically in fiber-like proteins, lateral hydrogen bonds between peptides, hydrophobic packing of residues, and salt bridges play a vital role in stabilizing and stimulating further elongation of fibrils [52][53][54]. The extensive number of hydrogen bonds between individual β-strands and long-range interactions drive cytotoxic fibril formation [55]. Our results show that the average number of the inter-and intrapeptide hydrogen bonds per chain gradually decreased under the influence of EF (see Table 2). This hydrogen bond loss affects α-synuclein fiber stability by lowering the strength of intrapeptide interactions. Thus, highly ordered β-sheet-rich fibrils are quite sensitive to EF. It is worth mentioning that each type of protein maintains a certain value of a dipole moment due to the presence of charged side chains [56,57]. The value of a total dipole moment can serve as one of the indicators that show conformational changes on the protein, i.e., its folded or denaturated state. An increase in total dipole moments in comparison to the native state of the protein is associated with conformational transitions towards to the denaturation state. Therefore, we also calculated the total dipole moments of the pentamer for each case of applied EF and the average over the all replicas (see Figure 3). It is worth mentioning that each type of protein maintains a certain value of a dipole moment due to the presence of charged side chains [56,57]. The value of a total dipole moment can serve as one of the indicators that show conformational changes on the protein, i.e., its folded or denaturated state. An increase in total dipole moments in comparison to the native state of the protein is associated with conformational transitions towards to the denaturation state. Therefore, we also calculated the total dipole moments of the pentamer for each case of applied EF and the average over the all replicas (see Figure 4). As is clear from Figure 4, in the absence of EF the total dipole moment of α-synuclein fiber is ~1400 Debye. Evidently, the presence of EF induces a force that acts on charged side chains. Consequently, the total dipole moment increased continuously and reached the highest value, i.e., ~4800 Debye, at 0.4 V/nm. In other words, the total dipole moment of the α-synuclein fiber rose more than three times compared to that in the absence of EF. Furthermore, we observed the fast change of orientation of α-synuclein fibril at higher values of applied EF. This in turn led the alignment of total dipole moments of α-synuclein fibril along the EF direction in a short period of time during the simulation (see Figure S3). The contribution of salt bridges is also substantial in holding the conformation of fibers [58][59][60]. In the current conformation, Glu46 and Lys80 form inter-and intrapeptide salt bridges, which prevent the opening of the fibril's hydrophobic core at lower values of EF.

Methods and Materials
The graphical processing unit (GPU) version of the GROMACS program package was employed to perform all simulations [61]. The united atom GROMOS 45a3 force field As is clear from Figure 3, in the absence of EF the total dipole moment of α-synuclein fiber is~1400 Debye. Evidently, the presence of EF induces a force that acts on charged side chains. Consequently, the total dipole moment increased continuously and reached the highest value, i.e.,~4800 Debye, at 0.4 V/nm. In other words, the total dipole moment of the α-synuclein fiber rose more than three times compared to that in the absence of EF. Furthermore, we observed the fast change of orientation of α-synuclein fibril at higher values of applied EF. This in turn led the alignment of total dipole moments of α-synuclein fibril along the EF direction in a short period of time during the simulation (see Figure S3). The contribution of salt bridges is also substantial in holding the conformation of fibers [58][59][60]. In the current conformation, Glu46 and Lys80 form inter-and intrapeptide salt bridges, which prevent the opening of the fibril's hydrophobic core at lower values of EF.

Methods and Materials
The graphical processing unit (GPU) version of the GROMACS program package was employed to perform all simulations [61]. The united atom GROMOS 45a3 force field parameters were used to generate the necessary files to run the model system [62]. The 3D coordinate structure of α-synuclein H50Q narrow fibril was obtained from the web page of the Protein Data Bank (PDB ID: 6PEO) [50].
In order to build the simulation system, the α-synuclein structure was centred in the dodecahedron box, and the dimensions of the current box were chosen to be 1.1 nm from atoms of α-synuclein to the edges of the box (see Figure 4a). Further, the system was filled by a simple point charge water model with 0.1 M NaCl to create a similar physiological environment (see Figure 4a   The calculated backbone root mean square deviation (RMSD) plot shows that 0.05-0.25 V/nm EF strength disturbed fibril structure and caused higher fluctuations compared to the absence of the EF (see Figure 3). However, these fluctuations lead only to the unfolding of the N-terminal end, i.e., residues from 36 to 46, and the hydrophobic core remained unchanged (see Figure 1). However, 0.30-0.40 V/nm EF induced more changes in the conformation of the α-synuclein structure, in that the occurrence of β-sheet conformation decreased significantly, and this conformation mainly was altered into coil conformations (see Table 1 and Figure S1). Finally, in the case of 0.40 V/nm EF, the αsynuclein fibril completely unfolded (see Figure 1) and the hydrophobic core of the αsynuclein fibril completely opened which was remained at lower EF strengths. According to the RMSD, the major change occurs within the initial 100 ns simulation, i.e., in high EF intensity such as 0.3, 0.35, and 0.4 V/nm (see Figure 3). The secondary structure map also shows insignificant changes in conformation in the rest of the simulation time (see Figure  S1). Initially, the energy minimization was run to remove the excess potential in the model system. During this simulation, atoms found the appropriate positions corresponding to the nearest local minimum energy conformation for the given model system. Next, short 100 ps NVT (canonical ensemble) and 500 ps NPT (isobaric-isothermal ensemble) simulations were performed by applying the position-restrained potential. Subsequently, a 600 ns four replica production run was performed by releasing position restrained potential and randomizing velocities by applying 0.05, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, and 0.40 V/nm static electric fields along the X direction. The velocity-rescaling thermostat [64] and the Parrinello-Rahman barostat [65] were applied at 310 K and 1 atm, respectively. A 1 nm cutoff radius was used in these simulations. The entire trajectory dataset was used to calculate the root mean square deviation (RMSD) [66], and the last 200 ns of trajectory data was employed to calculate the radius of gyration, solvent accessible surface area (SASA) [67], number of hydrogen bonds per chain, residual root mean square fluctuations (RMSF), and secondary structure analysis of α-synuclein. The DSSP tool was used to determine detailed conformational changes in the α-synuclein structure. Pymol and visual molecular dynamics viewer (VMD) software were used to create images [68,69]. Note that, the data and snapshots in the main text was obtained from replica 1 simulation. The rest of the data which belong to other replicas are given in supplementary material (see Figures S1-S4 and  Table S1).

Conclusions
PD is a progressive movement disorder with other nonmotor symptoms. Because of our brain's plasticity, PD symptoms appear only after more than 50-60% of dopaminergic neurons within the substantia nigra are already dead [70]. Electrical DBS of specific areas shows good results for PD. The associated drawback is that it loses its effectiveness over time. The main idea of DBS is to stimulate the electric activity of neurons, but its effect on a molecular level is not totally clear. In our research, we perform MD simulations to investigate the static EF effect on the conformational changes of the α-synuclein fibril. We showed that the application of 0.30, 0.35, and 0.4 V/nm EF during 600 ns disorganized α-synuclein fibrils. Typical DBS parameter settings of voltage, pulse width, and frequency range are from 1 to 3.5 V, 60 to 210 ms, and 130 to 185 Hz, respectively [71].
We believe that classical settings of DBS might be enough to disorganize α-synuclein fibrils in brain cells; however, it should be tested in further experiments.
The formation of the α-synuclein inclusions occurs by a generic process of misfolding, by which an ordinarily soluble protein converts into fibrillar aggregates via a series of oligomeric intermediates and, ultimately, the insoluble fibrils are deposited in the brain. Soluble oligomeric species generated during the formation of fibrils are the most neurotoxic species linked with the development of PD [72][73][74]. The kinetic of α-synuclein fibril formation can often be dominated by secondary nucleation events, such as fibril fragmentation, adding further elements of complexity to the kinetic process [75]. Disorganization and spread prevention of amyloid fibrils are some of the main goals for scientists involved in PD research. Disorganization of α-synuclein fibrils, which we saw during MD simulation, might possibly lead to the formation of toxic oligomeric structures, which might further undergo second nucleation events or structures that will be utilized by the protein quality control system. Thus, we assume that further research, both in silico and in vitro, is needed to understand whether the disruption of α-synuclein fibrils by EF has a positive or negative impact.