pH Effects Can Dominate Chemical Shift Perturbations in ¹H,¹⁵N-HSQC NMR Spectroscopy for Studies of Small Molecule/α-Synuclein Interactions

Abstract

¹H,¹⁵N-Heteronuclear Single Quantum Coherence (HSQC) NMR is a powerful technique that has been employed to characterize small-molecule interactions with intrinsically disordered monomeric α-Synuclein (aSyn). We report how solution pH can impact the interpretation of aSyn HSQC NMR spectra and demonstrate that small-molecule formulations (e.g., complexation with acidic salts) can lower sample pH and confound interpretation of drug binding and concomitant protein structural changes. Through stringent pH control, we confirm that several previously identified compounds (EGCG, Baicalin, and Dopamine (DOPA)) as well as a series of potent small-molecule inhibitors of aSyn pathology (Demeclocycline, Ro90–7501, and (±)-Bay K 8644) are capable of direct target engagement of aSyn. Previously, DOPA−aSyn interactions have been shown to elicit a dramatic chemical shift perturbation (CSP) localized to aSyn’s H50 at low DOPA concentrations then expanding to aSyn’s acidic C-terminal residues at increasing DOPA levels. Interestingly, this CSP profile mirrors our pH titration, where a small reduction in pH affects H50 CSP, and large pH changes induce robust C-terminal CSP. In contrast, under tightly controlled pH 5.0, DOPA induces significant CSPs observed at both ionizable and nonionizable residues. These results suggest that previous interpretations of DOPA−aSyn interactions were conflated with pH-induced CSP, highlighting the need for stringent pH control to minimize potential false-positive interpretations of ligand interactions in HSQC NMR experiments. Furthermore, DOPA’s preferential interaction with aSyn under acidic pH represents a novel understanding of DOPA−aSyn interactions that may provide insight into the potential gain of toxic function of aSyn misfolding in α-synucleinopathies.

Graphical Abstract

INTRODUCTION

α-synuclein (aSyn) is an intrinsically disordered protein (IDP) whose misfolding and aggregation into toxic oligomers and fibrils (e.g., Lewy bodies) are implicated in numerous α-synucleinopathies, including Parkinson’s disease (PD). Small-molecule inhibitors that arrest this IDP transformation are highly desired. However, IDPs are difficult drug targets due to a lack of stable tertiary or secondary structures and no welldefined binding sites to facilitate structure-based or rational drug discovery campaigns.^1–5 Characterizing a direct mechanism of action (MOA) of small-molecule aggregation inhibitors–e.g., direct binding to monomeric aSyn–has been extremely challenging.

NMR spectroscopy has been successfully employed in several drug discovery campaigns for structured protein targets but has found limited success in the investigation of IDPs.⁶ There have been several NMR studies that explore small-molecule binding to mature aSyn fibrils^7,8 as well as a few reportsusing ¹H,¹⁵N-HSQC NMR spectroscopy (HSQC)—where small molecules (including epigallocatechin gallate (EGCG), Baicalin, and dopamine (DOPA)) have been shown to engage monomeric or low-molecular-weight aSyn species.^9–11

We recently published a series of small molecules that reduce aSyn toxicity in multiple cell lines with high potency.¹² In a first attempt to identify a mechanism of action, we showed that these molecules (Demeclocycline (DEM), Ro90 7501 (RO), and (±)-Bay K 8644 (BAY)) directly inhibit aSyn fibrillization and alter the distribution of nonfibrillar aSyn oligomeric species in vitro. While we showed these direct effects on aSyn assemblies, a combination of single-molecule FRET and ¹⁹F-NMR studies suggested that there is no direct binding to monomer. However, those techniques rely on sitespecific labeling that can miss small-molecule-induced perturbations to local structure absent large conformational rearrangements of the protein.

In this study, we employ ¹H,¹⁵N-HSQC NMR and chemical shift perturbation (CSP) to examine direct target engagement for these hit compounds. CSP is a very sensitive measure of local protein structure. It is commonly employed to determine the direct binding of ligands, identify potential binding sites, and determine ligand affinity. The most common use of CSP is characterization of changes in ¹H,¹⁵N-HSQC NMR. For a detailed review on CSP, see ref 13. In preliminary experiments, it quickly became apparent that small-molecule formulations (specifically those complexed with acidic salts) can significantly influence the pH of the NMR sample, confounding interpretation of the HSQC spectrum and CSP profile. Such pH changes in vitro arise from the higher compound concentrations typically required for the 5:1 stoichiometric ratio of compound to protein at protein concentrations required for HSQC NMR. These compound-induced pH effects are likely irrelevant in the context of physiological aSyn in cells—where drug dosing is performed at low μM concentrations.

It has been known for some time that pH alters the conformational ensemble of monomeric aSyn and subsequently accelerates aSyn fibrillization.^14–23 This includes an important HSQC study by Eliezer et al. that evaluated aSyn HSQC spectra at pH 7.4 and 3.0 and demonstrated that even at low pH, both N- and C-terminal domains’ long-range interactions with the NAC core are conserved.²³ We confirmed these observations through a series of aSyn aggregation experiments and highlighted a set of transition pH values where significant increases in fibrillization occur. Next, to decouple compound binding from small-molecule-induced effects on solution pH, we undertook a detailed investigation of the pH effects on aSyn HSQC spectra, confirming the importance of stringent pH control when exploring small-molecule binders of monomeric aSyn.

RESULTS AND DISCUSSION

Changes in pH Increase Fibrillization Rate and Induce Large-Scale Perturbation of aSyn HSQC Spectra.

We examined pH-induced fibrillization and structural changes of aSyn using Thioflavin-T (ThioT) aggregation and ¹H,¹⁵N-HSQC NMR across a tightly controlled pH range from 7.4 to 4.5, just below aSyn’s isoelectric point (pI 4.7), expanding the resolution evaluated in previous studies.^14–23 aSyn is commonly divided into three domains, a positively charged (basic) N-terminus (residues 1−60), a mostly hydrophobic non-amyloid component (NAC) domain (residues 61−95), and a highly acidic C-terminus (residues 96−140) (Figure 1a). aSyn contains a total of 24 acidic residues (7 in the N-terminal domain, 2 in the NAC domain, and 15 in the C-terminal domain). Nested between the N-terminal and NAC domains is aSyn’s sole histidine (H50). Due to histidine’s model pK_a of 6.5, it may titrate near physiological pH and therefore be the cause of pH sensitivity in this range. On the other hand, acidic residues aspartic acid and glutamic acid have model pK_a’s near 4 and are therefore typically responsible for pH sensitivity in the acid range.²⁰

Figure 1.

pH effects on aSyn. (a) Sequence of aSyn. N-terminal segment (1−60) is shaded in light blue, NAC region (61−95) in light yellow, and C-terminal segment (96−140) in pink. Basic residues, lysine residues are colored blue, and aspartic and glutamic acids are colored red; (b) pH titration of aSyn at pH 4.5 (red), 5.0 (pink), 5.5 (orange), 6.0 (magenta), 6.5 (black), 7.0 (green), and 7.4 (blue). Residues adjacent to acidic amino acids are colored red; (c−f) pH effects on residues His50, Ser129, Val52, Tyr136, and Gly7. (g) Summary of ΔCSP due to pH changes. * indicates an unresolved resonance at all pH; # indicates an unresolved resonance at pH 4.5, 5; ** resonance 52/74 are indistinguishable.

We first explored the pH sensitivity of aSyn fibrillization using a monomer-only ThioT aggregation assay with an 8-point pH titration (7.4, 7, 6.5, 6, 5.5, 5, 4.5, and 4; ±0.1). Aggregation experiments were run in triplicate using 105 μM (1.5 mg/mL) recombinant aSyn under continuous shaking. Not surprisingly, we observed an increase in an aSyn fibrillization rate (quantified by t_1/2, defined as the time to reach half-maximum fluorescence intensity) that trended with reducing pH (see Supporting Information Figure 1). Interestingly, there were two significant changes in t_1/2 that occurred at specific pH transition points (pH 6.5 → 6.0 and pH 5.5 → 5.0). These coincide closely with the pK_a of H50 (pK_a = 6.52) and aSyn’s isoelectric point (pI = 4.7), supporting the importance of both histidine and acidic residue’s role in aSyn fibrillization propensity. The first fibrillization rate transition point is remarkable, close to the pH ranges that are observed in neurons during hyperactivity or diseased conditions.^24–26

Taking advantage of the residue-specific sensitivity of ¹H,¹⁵N-HSQC NMR, we then examined pH-induced changes on the intrinsically disordered monomeric form of aSyn. It is important to note that the HSQC NMR experiments were performed under quiescent conditions and within 1 h of sample preparation without the use of detergents or lipids, precluding the potential for fibril formation or α-helical membrane-bound structures. Figures 1b(blue) and 2a illustrate the HSQC NMR spectra and residue assignments for aSyn at pH 7.4, recapitulating spectra in previous studies.^9,22,23 CSP of the amide resonances corresponding to basic and acidic residues are known to be sensitive to pH changes in HSQC NMR.^22,23 Figure 1b–g demonstrate the significant structural changes that occur to aSyn throughout the pH titration range (4.5, 5, 5.5, 6, 6.5, 7, and 7.4; ±0.1). Large-scale pH-induced structural perturbations are observed at all acidic pH values relative to pH 7.4. Throughout the pH titration, we observe the most significant ΔCSP for the H50 resonance (Figure 1b,c), where most of the change occurs between pH 7.4 and pH 6.0. Across the full pH titration, the histidine amide resonance moved from 8.30 to 8.53 ppm (Δδ = 0.23 ppm) along ¹H axis and 124.7−123.1 ppm (Δδ = 1.6 ppm) along the ¹⁵N axis. Indeed, when we look at the overall CSP profile (Figure 1g), it is dominated by the H50 perturbation, although additional, nonionizable residues near H50 also display significant perturbation at pH 6 (Figure 1g; T44, E46, V48, and V52), suggesting that the ionization of H50 induces a sufficient change in the local chemical environment to result in a structural change capable of increasing aSyn’s aggregation propensity, as observed in our ThioT experiments.

Figure 2.

HSQC overlay spectra of (a) aSyn (monomer) (red, Apo, pH 7); (b) aSyn (monomer) Apo (Red) and EGCG (green, 5 eqv); and (c) aSyn (monomer) Apo (Red) and Baicalin (black, 5 eqv); pH of the samples was measured prior to NMR acquisition.

aSyn’s acidic residues also undergo dramatic ΔCSP across the pH titration, with the greatest perturbations occurring near aSyn’s pI, between pH 5.0 and 4.5 (in particular residues E20, E83, D119, D121, E126, E137). Interestingly, this effect is predominantly observed with the C-terminal acidic residues more so than those located in the N-terminal domain and was previously described at a more acidic pH of 3.0.²³ The pH transition with the largest ΔCSP for these acidic residues coincides with the second increase in fibrillization rate. This suggests that the degree of ionization required to induce an aggregation-prone aSyn monomer conformation is closer to aSyn’s pI, and the corresponding structural changes are localized to C-terminal residues.

Multiple nonacidic residues (both polar and nonpolar) were also significantly impacted by pH changes. These include V3, G7, V52, G84, N103, Q109, L113, S129, and A140. All of these residues, except G7, are adjacent to at least one acidic residue, whereas N103, L113, and S129 (Figure 1d and Supporting Information Figure 1) precede 2 acidic residues. These shifts could be attributed to the electronic effects caused by changes in the ionization states of these acidic residues or due to changes in the local environment or structure. Threonine residues displayed a modest pH-dependent CSP. Among the aromatic residues, Y125 and Y136 showed appreciable movement (Figure 1e and Supporting Information Figure 1) of the amide resonance—which was not surprising as these residues are within the acidic C-terminal tail—whereas both phenylalanine residues displayed no resolvable ΔCSP.

These large pH-driven effects on aSyn HSQC spectrum highlight the need for caution while investigating small-molecule interactions. Many small molecules are formulated with acidic salts that could overwhelm the buffering capacity of samples at the concentrations typically used in HSQC NMR experiments. Evaluation of small molecules−protein interactions should control for these effects to minimize the potential for interpreting false-positive CSP. Below, we evaluate a series of known small-molecule binders to aSyn as well as novel compounds recently published by our group.¹²

Small-Molecule Binding with Monomeric α Synuclein Using HSQC NMR Spectroscopy.

Using rigorously controlled pH experimental conditions in HSQC NMR spectros-copy—including micro-pH electrode monitoring of NMR sample pH and H50’s resonance as an internal pH calibration point—we tested a series of compounds previously investigated as aSyn binders (EGCG, Baicalin, and DOPA); see Supporting Information Figure 2.

EGCG has been extensively reported in the literature to bind the monomeric or low-molecular-weight aSyn species using various analytical approaches, including HSQC NMR. We did not observe any significant shift at H50 with the addition of 5 molar equivalents (eqv) of EGCG at pH 7.0, indicating a consistent pH between apo (Figure 2a) and with added EGCG samples (Figure 2b and Supporting Information Figure 4). However, there was significant CSP for numerous other residues (see Supporting Information Figure 3a). These shifts across the protein indicate a nonspecific binding pattern with EGCG, as was previously described.⁹ Baicalin is another small molecule that is a known modulator of aSyn aggregation.^27–31 We observed modest changes in the amide chemical shifts, along with signal attenuation of several residues without perturbation of H50 (Figure 2c and Supporting Information Figures 3b and 5), indicating small-molecule binding with no pH effect.

pH-Driven Artifacts in HSQC NMR Spectroscopy Result in False-Positive Interpretation of Small-Molecule Binding.

We next evaluated DOPA-induced effects on aSyn’s HSQC NMR spectra. When we applied rigorous control for the pH of our samples, we did not observe any significant changes in CSP at physiological pH 7.0, even with 5 eqv of DOPA (Figure 3a,d). This is in stark contrast to results from a previous HSQC NMR study,⁹ where DOPA−aSyn interactions induced a significant CSP at H50 as well as throughout the C-terminal tail. Closer inspection of that DOPA titration CSP profile⁹ shows a large initial CSP at H50 (0.2 DOPA eqv), followed by increasing CSP at the C-terminal tail with increasing DOPA concentrations, a trend that mirrors the pH-induced CSP profile observed here (Figure 1g).

Figure 3.

pH effects on DOPA: HSQC overlay with aSyn (red, Apo, pH 7) and (a) DOPA (cyan, pH 7, 5 eqv). (b) Apo aSyn pH comparison (red, pH 7; blue, pH 5). (c) aSyn (Apo, blue, pH 5) and DOPA (green, pH 5, 5 eqv). CSP of APO vs +DOPA ((d) pH 7.0; (e) pH 5.0). The dashed line represents perturbation greater than the average plus one standard deviation. pH values of the samples were measured prior to NMR acquisition.

Commercially available DOPA is formulated as a hydrochloride (HCl) salt, and the acid component can significantly alter the pH of the solution. It is important to note that differences in buffering solutions would result in different magnitudes of pH change due to small molecule additions. In the previous DOPA−aSyn study, a Gly-NaOH buffer (ideal buffering range: pH 8.6−10.6) was used, whereas in this study, we employed phosphate buffer (ideal buffering range: pH 5.8− 8.0). Without pH correction in our phosphate-buffered samples, the addition of 5 eqv of DOPA-HCl resulted in a significant pH change (ΔpH = −2.0, and a corresponding H50 shift of ~0.05 ppm). A direct comparison of apo aSyn at pH 7.0 and pH 5.0 (Figure 3b) highlights the extent of resonance shift that is observable from only this pH effect. These large changes could easily mask or overwhelm small-molecule binding effects, highlighting that for a more accurate evaluation of ligand−aSyn interactions, a closer control of experimental conditions is essential.

When we compare apo aSyn at pH 5.0 to DOPA at pH 5.0 (Figure 3c,e and Supporting Information Figure 6), we observe significant DOPA-induced CSP for multiple resonances that are not due to pH effects (e.g., no discernible H50 Δδ ppm). Although these CSPs are significantly reduced relative to the pH effect CSP, there are perturbations not observed in the pH 7.0 Apo vs DOPA pH 7.0 overlay (Figure 3a) that suggests DOPA interacts with aSyn at pH 5.0, and the pH-controlled conditions provide an artifact-free mapping of DOPA interactions with aSyn.

In a previous study, Herrera et al. reported that both the ¹²⁵YEMPS¹²⁹ region in the acidic C-terminal tail and E83 within the NAC domain are essential for DOPA−aSyn interactions. They hypothesized that it is a long-range electrostatic coupling that would result in remodeling of the NAC domain, thereby inhibiting fibrillization.³² Although at physiological pH we did not observe DOPA-induced CSP on monomeric aSyn, under acidic conditions (pH 5.0), there were significant CSPs distributed well beyond ¹²⁵YEMPS¹²⁹ and E83 residues. (Figure 3d and Supporting Information Figure 5).

Taken together, these results highlight the significant challenge of potential false-positive interpretation of small-molecule binding to aSyn via HSQC NMR if pH effects are not considered, especially for compounds complexed in acidic salts.

Evaluating Direct Target Engagement of Novel Small-Molecule Inhibitors of aSyn Aggregation via HSQC NMR.

We recently reported three hit compounds, Demeclocycline HCl (DEM), Ro90–7501 (RO), and (±)-Bay K 8644 (BAY) that were identified from the LOPAC library using our FRET biosensor assay.¹² In that study, we were unable to conclude that these compounds interacted with monomeric aSyn; however, they were able to induce changes in aSyn fibrillization and oligomerization. Here, we examined DEM, BAY, and RO as well as a negative control compound from that study (hydrochlorothiazide, HCT) using HSQC NMR under controlled pH conditions to evaluate direct compound interactions.

For the control compound HCT (Figure 4a), we do not observe CSP upon the addition of 5 eqv of HCT at pH 7.0. In contrast, DEM (at 5 eqv)—also formulated as an HCl salt did demonstrate an initial pH drop, and when corrected, we observed a modest modulation of the amide resonances (Figure 4b). These effects were not localized to any specific region but were globally distributed (Supporting Information Figure 6). This global attenuation could be due to induced oligomerization, which agrees with our previous findings for DEM.¹² The addition of BAY (5 eqv) resulted in a local attenuation of signal intensity at residues S9, A17, G31, V48, and A53 (Figure 4c and Supporting Information Figure 7), suggesting specific interactions that span residues across the N-terminal domain. Experiments with RO were run at 1 eqv due to compound solubility issues at higher concentrations. At this lower concentration under tight pH control, RO resulted in a clear shift of the H50 resonance as well as perturbations at G51 and V52, suggesting the possibility of a potential binding pocket for the small molecule. (Figure 4d and Supporting Information Figure 8).

Figure 4.

HSQC overlay spectra of aSyn (monomer) (red, Apo) and (a) HCT (blue); (b) DEM (green); (c) BAY (magenta); and (d) RO (black). All NMR spectra were collected at 288 K with 5 eqv of small molecule (HCT, DEM, BAY) or 1 eqv (RO). pH values of the samples were measured prior to NMR acquisition.

This evidence of direct ligand−aSyn interaction for DEM, BAY, and RO allows us to hypothesize that the previously observed increase in aSyn oligomers could result from three possible pathways: (1) altered monomer conformation biased toward an oligomer-favorable state; (2) direct cross-linking of monomers; or (3) shielding of charged residues facilitating aggregation (similar to the effects of high ionic strength buffers or low pH). Although further investigation to resolve the details of these interactions will be the focus of future studies, these results all clearly indicate direct target engagement of these physiologically active compounds.

Conclusions.

aSyn is known to be highly sensitive to solution conditions, with numerous studies describing the influence of acidic pH and salt concentration on the rate of fibrillation and formation of larger aggregates.^15,17,33,34 Multiple studies have reported significant differences in conformation proximal to acidic residues (particularly in the highly charged C-terminal domain) when aSyn was studied at acidic pH (3.0−3.5) versus physiological pH (7.4).^22,23 Here, we have shown that small molecules can themselves, due to their formulations as acidic salts, cause a significant shift in the pH.

Failure to account for this pH shift can lead to false-positive interpretations of direct ligand-induced effects through both the magnitude and location of observed CSP. DOPA’s interactions with aSyn provide a case study for the need for stringent pH control in HSQC NMR experiments. The important relationship between DOPA and aSyn was first identified in cellular and in vivo studies that have correlated aSyn expression, mutation, and misfolding/oligomerization with DOPA regulation.^35–41 Direct DOPA−aSyn interactions have been suggested through in vitro studies of DOPA-induced aSyn oligomerization or DOPA-facilitated di-tyrosine crosslinking.^41–45 However, our new observation that DOPA’s interactions with aSyn are more subtle than previously reported⁹ and manifest only under acidic conditions raises interesting questions on the MOA of DOPA-induced aSyn pathology in vivo. Perhaps germane, aSyn is enriched at synaptic terminals and is known to regulate synaptic vesicle (SV) fusion.^46–52 SVs have reduced pH (pH 5.7−6.0; depending on neuron type) relative to the neuronal bouton (both cytosolic and intracellular).^53–56 Overexpression and misfolded aSyn can induce defects in SV recycling, DOPA transport and reuptake, and membrane permeabilization.^{39–41,57–59} α-synucleinopathies likely result from a gain of toxic function of misfolded aSyn. We speculate that SV permeabilization due to misfolded aSyn oligomers would expose enriched aSyn (extravesicular) and DOPA (SV lumen) concentration in an acidic environment, resulting in DOPA-induced off-pathway aSyn oligomers.

For future drug discovery campaigns that utilize HSQC NMR to characterize ligand−aSyn interactions, it is essential that tight pH control is implemented to minimize false-positive interpretations of direct ligand interactions. By eliminating the pH-induced CSP, direct ligand−aSyn interactions can be more accurately mapped to specific protein regions. The example with DOPA suggests that ligand−aSyn HSQC NMR experiments should be run throughout the range of physiological pH values that aSyn has been shown to exist to provide insight into where in the lifecycle of aSyn can a ligand modulate aSyn’s structure and function.

MATERIALS AND METHODS

Protein Expression and Purification.

aSyn (WT) with ¹⁵N enrichment was accomplished using an Escherichia coli bacterial strain BL21 (DE3). Isolation and purifications were conducted as described previously.⁶⁰

Compounds and Reagents.

Compounds and their respective pK_a values used in this study: epigallocatechin gallate (EGCG; Tocris #4524; pK_a 7.99), baicalin (Sigma #572667; pK_a1 4.21, pK_a2 8.56), dopamine hydrochloride (DOPA; Sigma #PHR1090; pK_a 8.93), hydrochlorothiazide (HCT; Sigma #H4759; pK_a 7.9), Ro90–7501 (RO; Sigma #R0529; pK_a 5.93, predicted via Chemaxon), demeclocycline hydrochloride (DEM; Sigma #PHR1735; pK_a 2.94), and (±)-Bay K 8644 (BAY; Millipore #196878-M; pK_a 3.96, predicted via Chemaxon). All compounds were prepared at 10 mM stock concentrations in DMSO-D⁶ (Cambridge Isotope Laboratories #DLM-10-S-10) prior to dilution for aggregation or HSQC NMR experiments.

Monomer Aggregation of α-Synuclein under Different pH Conditions.

Unlabeled WT aSyn protein aliquots (purified as described previously⁶¹) were quick-thawed from −80 °C and centrifuged at 21,000g for 30 min. Samples were prepared at 8 different pH levels (7.4, 7, 6.5, 6, 5.5, 5, 4.5, and 4) by adjusting with 0.1 M HCl. pH was verified for each sample using a Cole-Parmer P200. Thioflavin-T (ThioT) was added to all samples for a final aSyn monomer concentration of 105 μM (1.5 mg/mL) and 25 μM of ThioT. A volume of 100 μL of each pH condition was then pipetted into a black nonbinding 96-well plate with a single 3 mm borosilicate glass bead in triplicate. The 96-well plate was covered with an optically transparent plastic film to prevent sample evaporation. Plates were incubated at 37 °C under continuous orbital shaking. Fluorescence was measured using a Spectramax i3x plate reader (Molecular Devices) every 15 min with a 440/10 nm excitation and 490/10 nm emission. Kinetic fluorescence data points were processed by first subtracting the average of the PBS-only wells (with matched pH). Next, the data were normalized to the maximum value for each well, yielding kinetic curves ranging from 0 to 1.

NMR Spectroscopy.

¹H,¹⁵N-HSQC NMR data were acquired on an 850 MHz Bruker Avance III equipped with a 5 mm triple resonance cryoprobe at 288 K in pH 6 phosphate buffer containing 5 mM phosphate, 25 mM NaCl, 5% D₂O, and 0.01% DSS at the Minnesota NMR (MNMR) center at the University of Minnesota. Proteins were uniformly enriched with ¹⁵N, as described before.⁶² DSS (4,4-dimethyl-4-silapentane-1-sulfonic acid) was used as an internal reference for spectra calibration, and assignments were made by analogy as per published reports deposited in BMRB (BMRB accession number 16543). NMR spectra were processed using Topspin 4.0.8 and analyzed with Sparky 3.115. To investigate residue-specific structural perturbations due to pH changes in the α synuclein protein, CSPs (chemical shift perturbations) were calculated using the equation CSP = √(Δ δH_N)² + (0.1Δ δN_H)², where Δ δH_N and δN_H are the difference in the chemical shift of proton and nitrogen, respectively.⁶³ HSQC spectra were collected with 100 μM uniformly ¹⁵N-labeled protein samples with 256 data points and with 8 scans.

Data Availability Statement

The data sets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.