Phosphorylation of α‑Synuclein Fibrils at S129 Changes DNAJB1 Binding as Probed by Solid-State NMR

Abstract

Amyloid fibrils formed by the protein α-synuclein are implicated in the pathogenesis of synucleinopathies. In addition to their rigid cross-β core, these fibrils have intrinsically disordered regions on their surface, which are important for interactions with other cellular components, such as chaperones. Chaperones play a vital role in preventing and reversing amyloid formation in neurodegenerative diseases. How they recognize misfolded proteins is an active field of research. DNAJB1 is a cochaperone that recognizes fibrils and recruits other chaperones such as Hsp70 and Apg2, which collectively disaggregate fibrils formed by α-synuclein, tau, and huntingtin. Because DNAJB1 was reported to bind the C-terminus of α-synuclein and S129 in this C-terminus is predominantly phosphorylated in patient-derived fibrils, we wanted to determine the effect of this post-translational modification on DNAJB1 binding. Using electron micrographs, NMR spectroscopy, and binding assays, we show that phosphorylation at S129 reduces the dynamics of the intrinsically disordered C-terminus of α-synuclein fibrils and increases the binding of DNAJB1 to this very C-terminus. MD simulations further suggest that the reduced dynamics is due to increased interaction of the phosphorylated C-terminus with the fibril core. DNAJB1 binds the exact same region at the C-terminus, indicating the phosphorylation at S129 might have a dual effect of reducing fibril surface dynamics and increasing chaperone recognition.

Introduction

Synucleinopathies including Parkinson’s disease (PD), dementia with Lewy bodies, and multiple system atrophy are neurodegenerative disorders characterized by the presence of Lewy bodies, which are primarily composed of α-synuclein (aSyn) fibrils. , The aSyn sequence is usually divided into three domains: the N-terminus (residues 1–60), which is of an intermediate flexibility in fibrils; , the NAC (nonamyloid-β component) domain (residues 61–94), which forms the majority of the fibril core; and the C-terminus (residues 95–140), which is negatively charged and dynamic. aSyn forms multiple fibril polymorphs in vivo and in vitro with distinct cytotoxicity and seeding profiles. − While the majority of structural studies have focused on the fibril core, the less-ordered and highly dynamic intrinsically disordered regions (IDRs) framing this core remain understudied. These IDRs form the surface of the fibrils that are exposed to the cellular components and interact with a range of proteins including chaperones such as protein isomerase 1, Hsp70, DNAJB1, , and lymphocyte-activation gene 3, , highlighting the significance of IDRs in neurodegenerative diseases.

Molecular chaperones play a key role in preventing protein misfolding, facilitating protein disaggregation and amyloid disassembly. Heat shock proteins (Hsp) are a family of highly conserved molecular chaperones, which can disassemble amyloid fibrils, formed by huntingtin, , tau, , and aSyn. Hsp70 has been identified in Lewy bodies, and aSyn overexpression in mice results in the upregulation of Hsp70 and Hsp40. Recent studies suggest that Hsp70, along with its cochaperone DNAJB1 and the nucleotide-exchange factor Apg2, can bind to both aSyn monomers and fibrils, facilitating fibril disaggregation in vitro. Notably, DNAJB1 specifically recognizes aSyn fibrils and directs Hsp70 to the fibrils to enable disaggregation. DNAJB1 belongs to the family of J-domain proteins (also known as Hsp40s) that are defined by a conserved J-domain that interacts with Hsp70. In addition, DNAJB1 has an N-terminal dimerization domain and two C-terminal domains, CTD I and CTD II, which are connected by a linker region and responsible for substrate recognition. , Specifically, DNAJB1 binds aSyn as a dimer through its CTD II domain. Using solution NMR, the specific binding site on aSyn monomers was mapped to residues 123–129 in the C-terminus.

Although various post-translational modifications (PTMs) of aSyn have been identified, such as phosphorylation, acetylation, , ubiquitination, SUMOylation, nitration, and O-GlcNAcylation, the most prevalent PTM in Lewy bodies is phosphorylation at residue S129, which is found in about 90% of aggregated aSyn. , aSyn phosphorylation at serine 129 serves as both a regulator of neuronal activity through fine-tuning synaptic function and a pathological marker of aSyn accumulation in synucleinopathies, making it a critical post-translational modification for understanding protein homeostasis mechanisms. Because monomeric pS129 does not readily aggregate, phosphorylation of aSyn at S129 (aSyn pS129) must occur post fibril formation.

Because pS129 is the dominant PTM in Lewy bodies, we wondered what this modification meant for DNAJB1 binding. Here, we are using binding assays, solid-state NMR, and molecular dynamics (MD) simulations to probe the interaction between DNAJB1 and both WT and pS129 aSyn fibrils.

Methods

Protein Expression and Purification

Expression of aSyn (Unlabeled)

Wild-type aSyn in the PRK172 plasmid was transformed into BL21­(DE3) cells. A single colony was isolated from an ampicillin (AMP)-resistant LB-agar plate and inoculated into a 50 mL primary LB culture with 100 μg/mL AMP overnight at 200 rpm and 30 °C. The next day, 7 mL of primary culture was centrifuged, and the cell pellet was resuspended in 7 mL of fresh LB and transferred to 1 L of LB medium containing 100 μg/mL AMP until the OD reached 0.6–0.7. Protein expression was induced with 0.5 mM isopropyl 1-thiol-β d-galactopyranoside (IPTG) for 18 h at 25 °C and 200 rpm. Once the expression was completed, cells were harvested at 3470 g for 20 min at 21 °C. Cell pellets were flash-frozen in liquid nitrogen and stored at −80 °C for further use.

Expression of ¹⁵N–¹³C aSyn and ¹⁵N aSyn

Uniformly labeled ¹³C–¹⁵N WT aSyn was expressed using the approach reported by Marley et al. Cells were transformed and inoculated in the same way as the unlabeled protein. Once OD reached 0.6–0.7, cells were centrifuged at 3470g for 20 min at 21 °C, the supernatant was discarded, and the cell pellets were washed in M9 wash buffer (47.7 mM Na₂HPO₄·7H₂O, 22.0 mM KH₂PO₄, 8.6 mM NaCl). The cells were centrifuged once again under the aforementioned conditions. The cell pellets were resuspended in M9 minimal media and incubated for 1 h at 25 °C and 200 rpm. This M9 minimal medium was prepared from M9 wash buffer with an additional 2 mM MgSO₄, 100 μM CaCl₂, 2% of solution C (v/v), 0.05% of vitamin stock (v/v), 0.1 μg/L of thiamin, 0.5 g/L ¹⁵N NH₄Cl, 4 g/L ¹³C glucose, and 100 μg/mL AMP. Solution C was 130 mM KOH, 157 mM MgCl₂, 23 mM CaCl, and 5% v/v Metals 44 solution. Metals 44 solution was composed of 8 mM K₂EDTA, 38 mM ZnCl₂, 25 mM FeCl₂, 9 mM MnCl₂, 150 μM (NH₄)₆Mo₇O₂₄, 920 μM CuCl₂, 852 μM Co­(NO₃)₂, and 1.8 mM boric acid. The vitamin stock contained 10 g/L nicotinic acid, 1 g/L p-aminobenzoic acid, and 0.1 g/L biotin. Protein expression was induced with 0.5 mM IPTG for 18 h at 25 °C and 200 rpm. Once the expression was completed, cells were harvested at 3470 g for 20 min at 21 °C. Cell pellets were frozen in liquid nitrogen and stored at −80 °C for further use.

Purification of aSyn

Protein pellets were frozen with liquid nitrogen and thawed three times. Thawed cells were resuspended in 10 mL of lysis buffer (500 mM NaCl, 100 mM Tris, 1 mM EDTA, and 10 mM β-mercaptoethanol, pH 8.0). One tablet of Pierce Protease Inhibitor EDTA-free was dissolved in the protein suspension followed by incubation on ice for 30 min. The mixture was centrifuged at 30,000 g for 20 min at 4 °C. The supernatant containing aSyn was acidified to pH 3.5 with 1 and 6 M HCl and then incubated on ice for 30 min. The suspension was then centrifuged at 30,000 g for 20 min at 4 °C. The supernatant was dialyzed against 4 L of 1% acetic acid in a 6–8 kDa dialysis membrane overnight. The mostly pure protein sample was then filtered through a 0.22 μm PES filter and loaded into a Jupiter 15 μm C4 300 Å, LC column 250 mm × 4.6 mm for hydrophobic interaction purification using an FPLC and was eluted with a gradient of buffer A (0.1% trifluoroacetic acid (TFA)/water) and increasing buffer B (0.1% TFA/acetonitrile). The protein fractions pooled from this reverse phase purification were dialyzed against water followed by dialysis against 20 mM Tris and 1 mM EDTA at pH 8 for anion exchange purification. The dialyzed protein was filtered, loaded to an anion exchange column (Cytiva HiTrap QXL) and eluted with an increasing NaCl gradient. SDS-PAGE was used to confirm the presence and purity of the protein. The pure fractions were pooled and dialyzed against 50 mM phosphate buffer pH 7.4 with 0.1 mM EDTA and 0.02% NaN₃ for further use.

Ulp1 Protease

The pET-28b plasmid with His-tagged Ulp1 protease catalytic domain (residue 403–621) (Addgene Plasmid no. 64697) was transformed into chemically competent BL21­(DE3) cells. Starter cultures were grown at 30 °C for 14–16 h and then expanded into 500 mL of LB medium cultures containing 50 μg/mL kanamycin. Cultures were then grown at 37 °C until an OD₆₀₀ of 0.6–0.8. Then, IPTG was added at a final concentration of 1 mM to induce protein expression. Cultures were incubated at 37 °C for an additional 4 h, after which cells were centrifuged at 3470 g for 20 min. Cell pellets were stored at −80 °C until further use. For purification, cells were resuspended in 20 mM Tris-HCl pH 8, 300 mM NaCl, and 10 mM imidazole containing 20% sucrose (w/v), 0.2% Triton X-100 (v/v), and 1 mM β-mercaptoethanol. The cells were then lysed using a QSonic Ultrasonic Sonicator and the cell lysate was centrifuged for 20 min at 39,191 g for 20 min. The supernatant was then incubated on His60 Ni Superflow resin (Takara Bio USA, Inc.) and shaken at 4 °C for 20–30 min. The column was then washed with 20 mM Tris-HCl at pH 8, 300 mM NaCl, and 50 mM imidazole containing 1 mM β-mercaptoethanol. Protein was eluted in 20 mM Tris-HCl pH 8, 300 mM NaCl, 300 mM imidazole, 1 mM β-mercaptoethanol, and dialyzed against 25 mM Tris-HCl pH 8.00, 250 mM NaCl, 1% Triton X-100 (v/v), and 50% glycerol (v/v) containing 0.5 mM dithiothreitol. After the dialysis, protein aliquots were stored at −80 °C.

DNAJB1

Plasmid containing DNAJB1 construct having an N-terminal His-SUMO tag (pET24 vector) was transformed into chemically competent BL21­(DE3) cells. Starter cultures containing 50 μg/mL kanamycin were grown for 14–16 h at 30 °C. The culture was then expanded into LB medium and grown at 37 °C to an OD₆₀₀ of 0.6–0.8, after which protein expression was induced using IPTG at a final concentration of 1 mM. The cell culture was incubated at 20 °C, 140 rpm for 20 h. The cells were harvested by centrifugation at 3470 g for 20 min and stored at −80 °C for further use. For purification, cells were resuspended in lysis buffer composed of 30 mM HEPES pH 7.4, 500 mM KCl, 5 mM MgCl₂, 30 mM imidazole, and 10% glycerol containing 2 mM phenylmethylsulfonyl fluoride (PMSF), 1 tablet of Pierce Protease Inhibitor EDTA-free, 10 units of Dnase I (Lucigen, Cat No. D9905K), and 1 mM β-mercaptoethanol. The cells were then lysed using a QSonic Ultrasonic Sonicator and cell lysate was centrifuged for 30 min at 48,384 g. The supernatant was added to His60 Ni Superflow resin and incubated on a shaker at 4 °C for 1 h. The column was then washed with high salt buffer (30 mM HEPES pH 7.4, 1 M potassium acetate (KAc), 5 mM MgCl₂, 25 mM imidazole, and 10% glycerol) followed by low salt buffer (30 mM HEPES pH 7.4, 50 mM KAc, 5 mM MgCl₂, 25 mM imidazole, and 10% glycerol) both containing 1 mM β-mercaptoethanol. The protein was then eluted with a buffer containing 30 mM HEPES pH 7.4, 100 mM KAc, 5 mM MgCl₂, and 300 mM imidazole and 1 mM β-mercaptoethanol. Ulp1 was added to the eluted DNAJB1 fusion protein at a concentration of 400 μg/mL and dialyzed against 30 mM HEPES at pH 7.4, 100 mM KAc, 5 mM MgCl₂, and 10% glycerol with 1 mM β-mercaptoethanol. After dialysis, the sample was centrifuged for 20 min at 2880 g using an Eppendorf A-4-44 rotor and the supernatant was then added to equilibrated His60 Ni Superflow column. The sample in the column was incubated at 4 °C with mild shaking for 20 min, after which the flowthrough, containing the cleaved chaperone, was collected.

Fibril Formation

After dialyzing in 50 mM phosphate buffer pH 7.4, 0.1 mM EDTA, and 0.02% NaN₃, monomeric aSyn was concentrated to 200–300 μM (as determined via its absorbance at 280 nm using an extinction coefficient of ε = 5960 M^–1 cm^–1). For fibrillization, aliquots of 500 μL at a protein concentration ranging from 200 to 300 μM were kept in 1.5 mL low binding tubes. Tubes were then sealed with parafilm to prevent evaporation and placed in a MATRIX Orbital Delta F1.5 thermomixer, shaking at 1000 rpm at 37 °C for 2 weeks. The concentration of monomer in aSyn fibrils was determined by subtracting the monomer concentration in the supernatant after fibrilization and ultracentrifugation (1 h, 163,202 g) from the initial monomer concentration.

Phosphorylation of aSyn Fibrils

To remove the fibrillization buffer, fibrils were washed twice with nuclease-free water followed by ultracentrifugation at 163,202 g, 21 °C, for 1 h. The fibrils were resuspended in phosphorylation buffer (20 mM HEPES, 10 mM MgCl₂, 1.09 mM ATP, 2 mM DTT, pH 7.4) and briefly bath-sonicated. PLK2 phosphorylation enzyme (Invitrogen, Catalog No. PV4204) was then added to the fibrils (1 μg of enzyme per 1.44 mg of protein). The reaction was carried out by incubating the samples at 30 °C with mild shaking for 36 h. The reaction was stopped by brief bath sonication and freezing of samples at −80 °C.

Phosphorylation of the fibrils was confirmed using Western blot with the Ep1536Y antibody (ab51253) against phosphorylated S129 aSyn and using the mass spectrometry techniques described below.

MALDI-TOF

Fibrils were dissolved in 1% DMSO in deionized water. Each of the solubilized fibril samples was diluted 1:1 in a MALDI matrix solution composed of 15 mg/mL sinapinic acid, 70% acetonitrile, and 0.1% trifluoroacetic acid. A 2 μL aliquot of this mixture was spotted onto a Bruker MTP 384 Target Plate Ground Steel, and the samples were dried under nitrogen gas to promote evaporation and crystallization.

The analysis was conducted using the Rapiflex TissueTyper MALDI-TOF mass spectrometer (Bruker Daltonik, Bremen, Germany), equipped with a smartbeam laser (Nd355 nm), and operated in linear positive mode to detect [M + H] species. An optimized acquisition method was employed to identify proteins across a 5–60 kDa mass range. Instrument calibration was performed just prior to each analysis using appropriate standards, including insulin ([M + H]+ 5734.52 Da), ubiquitin I ([M + H]+ 8565.76 Da), cytochrome C ([M + H]+ 12360.97 Da, [M+2H]­2+ 6180.99 Da), myoglobin ([M + H]+ 16952.30 Da, [M+2H]­2+ 8476.65 Da), and bovine albumin ([M+2H]­2+ 33216 Da, [M + H]+ 66431 Da).

Peak areas from the raw mass spectrometry data were integrated and extracted by using the FlexAnalysis software (Bruker Daltonik, Bremen, Germany). Data smoothing and background subtraction was applied to reduce spectral noise and enhance peak resolution, facilitating more accurate quantification.

Trypsin Digestion Followed by LC-MS

Each 5 μg protein sample (phosphorylated aSyn fibrils solubilized in 1% DMSO/water) was diluted in lysis buffer (12 mM sodium lauroyl sarcosine, 0.5% sodium deoxycholate, 50 mM triethylammonium bicarbonate (TEAB), Halt Protease, and Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific)), then reduced and alkylated with tris­(2-carboxyethyl)­phosphine (10 mM) and chloroacetamide (40 mM) for 15 min at 95 °C. To perform a buffer exchange, SP3 SpeedBead Magnetic Carboxylate-Modified Particles (250 μg, Cytiva, Cat# 65152105050250 and 45152105050250, 1:1 ratio) were added to each sample. Ethanol was then added to reach a final concentration of 60% and the samples were gently shaken for 15 min to allow for protein binding. This was followed by three washes with 80% ethanol. The proteins were eluted from the SP3 beads using 50 mM TEAB (50 μL) containing mass spectrometry-grade Trypsin/Lys-C Mix (1 μg, Promega, catalog no. V50711) and were digested overnight at 37 for 16 h.

An aliquot (5 μL) of each sample was injected onto a nanobore C18 HPLC column (Cat# DNV75500, Thermo Fisher Scientific, 75 μm × 50 cm), equilibrated in solvent A (water/acetonitrile/fluoroacetic acid (FA), 98/2/0.1, v/v/v), and eluted at a flow rate of 200 nL/min with an increasing concentration of solvent B (acetonitrile/water/FA, 80/20/0.1, v/v/v: min/% B; 0/5, 40/32, 50/56, 53/100, 56/100, 60/2, 70/2) using an Easy nLC-1000 system (Thermo Fisher Scientific). The effluent from the column was directed to a nanospray ionization source connected to a hybrid quadrupole-Orbitrap mass spectrometer (Q Exactive, Thermo Fisher Scientific) acquiring mass spectra in a data-dependent mode alternating between a full scan (m/z 350–1700, automated gain control (AGC) target 3 × 106, 100 ms maximum injection time, FWHM resolution 70,000 at m/z 200) and up to 15 MS/MS scans (quadrupole isolation of charge states 2–7, isolation window 1.5 m/z) with previously optimized fragmentation conditions (normalized collision energy of 28, AGC target 1 × 105, 100 ms maximum injection time, FWHM resolution 17,500 at m/z 200).

Raw proteomic data were searched against a custom database containing the α-synuclein amino acid sequence, the complete E. coli proteome, and common mass spec contaminants using SEQUEST-HT in Proteome Discoverer (Version 3.2, Thermo Fisher Scientific), which provided measurements of the relative abundance of the identified peptides. Common dynamic modifications for the search included: oxidation (+15.995) on M; deamidation (+0.984) on N/Q; acetyl (+42.011) on K; phospho (+79.966) on S, T, Y; and methyl (+14.016) on K, R. In this assay, the detection of b⁺ and y⁺ fragment ions from tryptic peptides enabled high-confidence peptide identifications (false discovery rate < 1%) corresponding to specific proteins in the database for each sample. Chromatographic peak areas were utilized to assess the relative abundance of each peptide and identified post-translational modifications within each sample.

Sedimentation Assay

Prior to the sedimentation assay, phosphorylated fibrils were ultracentrifuged at 163,202 g for 1 h to remove the PLK2 enzyme and the phosphorylation buffer components. The loss of monomeric protein in the supernatant was estimated by subtracting the monomeric concentration calculated for previous washes of WT fibrils. To measure DNAJB1 binding, WT and aSyn pS129 fibrils were incubated with DNAJB1 at different aSyn:DNAJB1 molar ratios in 30 mM HEPES buffer containing 0.02% NaN₃ at pH 7.4, overnight at room temperature with mild shaking. The samples were ultracentrifuged at 179,343 g to pellet down fibrils and bound DNAJB1, leaving the unbound DNAJB1 and aSyn monomers in the supernatant. The supernatant was then analyzed by using SDS-PAGE.

To determine dissociation constants, the assay was repeated with a constant amount of DNAJB1 (10 μM) and increasing concentrations of aSyn WT and aSyn pS129 fibrils at ratios ranging from 1:0 to 1:8. The amount of unbound DNAJB1 remaining in the supernatant was estimated by SDS-PAGE and normalized against the control DNAJB1 sample that was not incubated with fibrils. Using ImageJ, we quantified the band intensities of the unbound DNAJB1 in the supernatant. We determined the dissociation constants of both fibril types by fitting to a one-site binding model.

Electron Microscopy (EM)

Electron micrograph grids were prepared by pipetting 10 μL of fibril sample onto Formvar-carbon-coated copper mesh grids (Electron Microscopy Sciences), which were then incubated for 5 min. The excess sample was removed from the grids by using filter paper. The grids were stained three times with 10 μL of 1% uranyl acetate, with each incubation lasting 5 min. After each incubation, excess uranyl acetate was blotted off with a filter paper. The grids were then washed with filtered, deionized H₂O, blotted dry with a filter paper, and left to dry overnight.

The electron micrographs were taken on a Jeol JEM-1400 electron microscope, operating at 100 kV and equipped with a Gatan-Orius CCD camera. Micrographs were captured at various magnifications, as indicated. Fibril widths were measured manually in ImageJ. Statistical analyses were performed in Python using the SciPy library. Fibril widths were compared using a two-tailed independent samples t test.

NMR Spectroscopy

Fibril samples were packed into 1.6 mm Agilent/Varian MAS rotors using a home-build packing tool similar to the one described by Böckmann and co-workers. All solid-state NMR experiments were acquired on an Agilent DD2 600 MHz solid-state NMR spectrometer equipped with a T3 1.6 mm probe operating at 25 kHz magic angle spinning (MAS), unless otherwise specified, and a set temperature of 0 °C. ¹H, ¹³C, and ¹⁵N hard radio-frequency (rf) pulses had amplitudes of 200, 100, and 50 kHz, respectively. ¹H–¹³C cross-polarization (CP) experiments utilized the Hartmann–Hahn matching condition at 60 kHz and 85 kHz for ¹³C and ¹H, respectively, with a 10% amplitude ramp, and a contact time of 1 ms.

One-dimensional ¹³C and ¹⁵N CP and ¹H–¹³C refocused INEPT (insensitive nuclei enhanced by polarization transfer) experiments were recorded with a spectral width of 50 kHz, and 1000 complex time domain points. A recycle delay of 3 s was used with a total of 1024 scans. 140 kHz two-pulse phase modulation (TPPM) decoupling was applied during acquisition.

¹³C–¹³C DREAM (dipolar recoupling enhanced by amplitude modulation) spectra were recorded at a MAS rate of 30 kHz with the probe in double resonance mode (¹H and ¹³C channels), with spectral widths of 50 kHz in both dimensions. One thousand complex points were recorded in the direct dimension, and 1000 real points (TPPI) were recorded in the indirect dimension. A recycle delay of 3 s was applied, and 80 scans were accumulated for each t₁ increment. The DREAM mixing time ranged from 3.5 to 4.5 ms and used a tangential amplitude modulation (Δ/2π = 3 kHz and β/2π = 1.4 kHz around the HORROR [homonuclear rotational resonance] condition of 15 kHz).

¹H–¹⁵N 2D HSQC spectra were acquired with a spectral width of 10 kHz and 500 complex points in the direct ¹H dimension and 3 kHz spectral width and 60 complex points for the indirect ¹⁵N dimension. 128 scans per t₁ increment were acquired. The pulse sequence included a BIRD filter with a water relaxation delay ranging from 320 to 380 ms, which served to suppress the residual water signal. , WALTZ-16 decoupling with a rf-field strength of 10 kHz was applied on ¹⁵N. To obtain the intensity plots I/I ₀ from the HSQCs of the DNAJB1-bound and unbound spectra, raw peak intensities were normalized according to the number of scans of the HSQC and the ¹H–¹⁵N CP signal. I is the peak intensity of the bound spectrum, and I ₀ is the peak intensity of the unbound spectrum. The error bars were calculated as follows

$${\sigma }_{I/I_0}=\frac{I}{I_0}\sqrt{{\left(\frac{\sigma }{I}\right)}^2+{\left(\frac{{\sigma }_0}{I_0}\right)}^2}$$

where σ is the noise estimate of the bound spectrum and σ₀ the noise estimate of the unbound spectrum.

The HNCA and HNcoCA spectra were recorded as described previously using nonuniform sampling (NUS) with a sampling density of 60% in both ¹³C and ¹⁵N dimensions. A total of 32 t₁ increments were acquired and the NUS data were reconstructed using the IST algorithm.

2D ¹H–¹³C INEPT-HETCOR spectra of the samples were acquired with 72 scans using a refocused INEPT transfer. The direct ¹³C dimension had a spectral width of 50 kHz with 1000 points; and the indirect ¹H dimension had a spectral width of 10 kHz with 128 points. Five kHz WALTZ-16 decoupling was applied on ¹H during acquisition. 2D ¹³C–¹³C INEPT-TOCSY experiments were conducted with 256 scans, using a FLOPSY8 spin lock sequence for a mixing time of 17 ms. The direct dimension had a 50 kHz spectral width with 1000 complex points, and the indirect dimension had a 10 kHz spectral width with 75 complex points.

For an exhaustive list of experimental NMR parameters, please see Table S1.

Spectra were processed using nmrPipe version 10.9 and plotted with nmrglue version 0.9 using in-house python scripts. Resonance assignments and peak integration were done using the program CARA.

Coarse-Grained Simulations of the Long Fibril

Based on the cryo-EM structure 6OSJ (PDB ID), a 40-layer long fibril core with a total of 80 monomers was built through repeatedly aligning the bottom layer to the top layer of 6OSJ. The missing terminal residues were filled using Modeler followed by performing a very-fast-level refinement. The resulting fibril model was converted into the hybrid resolution (HyRes) − coarse-grained representation and used to initiate the 2 μs MD simulation. HyRes has been designed for accurate description of dynamic protein conformations and interactions, where the backbone is represented at the all-atom level and side chains are represented at intermediate resolution with up to 5 beads. It has proven highly effective in simulating dynamic protein interactions including phase transitions, making HyRes uniquely suitable for studying conformational dynamics of disordered tails of the fibril and how they mediate the interactions with other proteins.

All HyRes simulations were carried out on GPUs using OpenMM 8.2. Langevin dynamics was performed with a collision frequency of 0.1 ps^–1 and MD time step of 8 fs. The salt concentration was set at 150 mM and the simulation temperature was 298 K. Harmonic position restraints with a constant of 100 kJ/mol/nm were imposed on the fibril core to maintain the fibril structure. The production simulation was performed as a nonperiodic system. To sample the binding between the fibril and DNAJB1, the fibril was first fixed at the center of the simulation box (30³ nm³), and then four dimers of DNAJB1 (PDB ID 3AGX) were randomly distributed in the box. Two 4 μs NVT simulations of HyRes were performed for interaction with the WT and pS129 fibrils, respectively. The first 200 ns trajectories were excluded from subsequent analysis. Contact analysis was based on the Cα-Cα distance with the cutoff of 8 Å using MDAnalysis. Details of the long fibril model and associated simulation scripts can be found on GitHub at https://github.com/lslumass/tutorial/tree/main/long_fibril_2.

Results

DNAJB1 Preferably Binds aSyn Fibrils Phosphorylated at S129

Because phosphorylation at S129 is the most common PTM of aSyn fibrils, we wondered whether this modification changed the binding of DNAJB1 to aSyn fibrils. To test this hypothesis, we phosphorylated aSyn WT fibrils at S129 using PLK2 kinase. PLK2 predominantly targets S129 in soluble aSyn, and we expected it to be even more specific in fibrils because all other serines are part of the N-terminus and core (S9, S42, S87). We confirmed the presence of pS129 using Western blotting with the aSyn pS129-specific Ep1536Y antibody (Figure S1A). In addition, we estimated the degree of phosphorylation using MALDI-TOF (Figure S1B) and trypsin digestion followed by LC-MS (Supporting Information), which showed that about 63% of S129 is phosphorylated after 24 h. LC-MS also showed that the phosphorylation at the other serine and threonine residues was <1% (see Supporting Information).

We first quantified the binding affinity of WT and pS129 aSyn fibrils for DNAJB1 using a sedimentation assay. In this assay, WT and pS129 aSyn fibrils were incubated overnight at room temperature with DNAJB1 at a 3:1 (aSyn:DNAJB1) molar ratio (Figure A,B). Afterward, we sedimented fibrils by centrifugation and analyzed the supernatant via SDS-PAGE. The SDS-PAGE showed that DNAJB1 was pulled down by both WT and pS129 fibrils, indicated by a decrease in DNAJB1 band intensity. The DNAJB1 band in the supernatant was weaker in the pS129 fibrils compared to the WT sample, suggesting that DNAJB1 is preferentially pulled down by pS129 fibrils (Figure B). We confirmed the presence of DNAJB1 in the fibril pellets of both fibril types by solubilizing the pellets in 1% DMSO and analyzing them via SDS-PAGE (Figure S1C). Interestingly, Figure B also shows that pS129 fibrils have a lower equilibrium aSyn monomer concentration compared to that of WT fibrils with and without the addition of DNAJB1.

1.

DNAJB1 is preferentially pulled down by aSyn pS129 fibrils compared to WT fibrils. (A) Schematic of the sedimentation assay used to assess DNAJB1 binding to WT and pS129 fibrils. WT fibrils were phosphorylated at S129 using PLK2. Both fibril types were incubated with DNAJB1 at a 3:1 molar ratio (aSyn:DNAJB1). (B) SDS-PAGE of supernatants from the sedimentation assay. Bands corresponding to unbound DNAJB1 (40 kDa) and monomeric aSyn (14 kDa) are indicated. (C) Dissociation constants K _D were determined from titrations with a fixed DNAJB1 concentration and increasing fibril concentrations. pS129 fibrils bound DNAJB1 with a K _D of 1.90 μM, whereas WT fibrils bound with a weaker affinity (K _D = 58.14 μM).

To determine the dissociation constant (K _D), we repeated the sedimentation assay with varying concentrations of WT or pS129 fibrils (Figure S2). We used densitometry of these SDS-PAGEs to quantify unbound DNAJB1 and fit the binding curves to a one-site binding model to obtain dissociation constants. The K _D of WT fibrils was 58.1 μM while that of pS129 fibrils was significantly lower at 1.9 μM, indicating that DNAJB1 binds more strongly to pS129 fibrils (Figure C).

Phosphorylation Decreases the Fibril Width and C-Terminal Dynamics

Next, we wanted to determine the consequences of phosphorylation at S129 for the fibril morphology, structure, and dynamics. Using negative stained EM, we compared the morphology of aSyn fibrils before and after phosphorylation. As can be seen from Figure A, the overall appearance of the fibrils and their degree of bundling did not change with phosphorylation. However, our analysis of the average fibril width (Figure B) indicated a reduction in the fibril width from 11.21 ± 2.04 nm for WT fibrils to 9.15 ± 1.94 nm for pS129 fibrils (p value = 3.7193 × 10^–14).

2.

Phosphorylation at S129 reduces the fibril width and causes spectral broadening likely due to increased interaction of the C-terminus with the fibril core. (A) Negative stain EM images of WT (cobalt blue) and pS129 (teal) aSyn fibrils. Zoomed-in regions highlight intact fibrils with characteristic lengths and bundled arrangements. Scale bar denotes 100 nm. (B) Histograms of WT and pS129 fibril widths showing averages of 11.21 ± 2.04 and 9.15 ± 1.94 nm, for WT and pS129 fibrils, respectively. Fibril width measurements were obtained manually from EM images. (C) HSQC spectra of WT and pS129 aSyn fibrils. ¹H 1D slices for two cross-peaks illustrate the increase in line width and decrease in signal-to-noise ratio caused by phosphorylation. (D) Position distribution of residue S129 calculated from an 80mer HyRes simulation based on the 6OSJ aSyn fibril structure (see Figure S5). (E) Additional overlay of S129 position distribution with snapshots of the simulation illustrating the location of the fibril cores (gray), two conformation of the C-terminus (green) with S129 highlighted in red, one N-terminus shown in cyan, and lysine residues in the fibril core (residues 43, 45, 58, 60, 80, 96, and 97), which S129 interacts with, highlighted in blue. (F) The position distribution of pS129 in the corresponding phosphorylated fibril simulation indicates increased interaction between residue pS129 and the fibril core. (G) Radial density profile of residue 129 in WT and pS129 fibrils starting from the core center ((0, 0) point in panels (D) and (F)). pS129 shows both a higher density and closer location to the fibril center compared to S129.

To determine the consequences of phosphorylation on the fibril structure and dynamics, we recorded solid-state NMR spectra of both WT and pS129 fibrils. For these measurements, we produced uniformly labeled ¹³C–¹⁵N WT fibrils and then divided the sample into two equal batches to minimize variability in the fibril quality and quantity. One batch was subsequently phosphorylated with PLK2 while the other was kept as the nonphosphorylated control. We packed about 3.4 mg of wet fibril pellet for each sample into 1.6 mm Varian/Agilent MAS rotors for NMR analysis. First, we recorded 1D ¹³C spectra using both ¹H–¹³C CP and ¹H–¹³C refocused INEPT steps to create the initial ¹³C magnetization. Where CP is sensitive to the most static parts of the sample, usually the fibril core, INEPT spectra highlight the dynamic regions of the sample, which are the dynamic IDRs in the case of aSyn fibrils. , The 1D CP spectra of WT and aSyn fibrils were virtually identical, indicating little to no effect of phosphorylation on the structure of the fibril core (see Figure S3A,B), and the INEPT/CP intensity ratio of pS129 fibrils was significantly reduced compared to WT fibrils, as can be seen from Figure S4. This reduction in dynamics of the IDRs is even better illustrated by the HSQC spectra of WT and pS129 fibrils shown in Figure C. Here, WT fibrils give much more intense spectra with narrower lines and more cross-peaks compared with pS129 fibrils.

How does the phosphorylation of the C-terminus of aSyn fibrils reduce the dynamics of this region? To get a possible explanation of this effect, we ran HyRes , coarse grain MD simulations of aSyn fibril 80mers (40 monomers per protofilament) based on the 6OSJ (PDB ID) cryo-EM structure with the N- and C-terminal IDRs modeled (see Methods section). We ran two separate 2 μs MD simulations for WT and pS129 fibrils using the HyRes model (Methods; Figure S5). To examine the effect of phosphorylation on the position of residue 129, we plotted the spatial distributions of S129 and pS129 (Figure D,F). Figure E additionally overlays two snapshots of C-terminal conformations onto the density, illustrating the relative orientation of the N-terminus, fibril core, and C-terminus. This analysis revealed that S129 samples a broad, peanut-shaped distribution around the protofilament with an increased probability density at the lysine-rich interface between protofilaments. In contrast, pS129 exhibited both an enrichment at this interface and an additional binding pocket near the point where the C-terminus emerges from the fibril core. The radial density profile of residue 129 (Figure G) further supports this conclusion, showing that pS129 interacts more frequently and more closely with the fibril core than S129 in WT fibrils.

Effect of DNAJB1 Binding on the Fibril Core of aSyn

Phosphorylation at S129 produced changes in the dynamics of the C-terminal IDRs of aSyn fibrils, as reflected by line broadening and loss of cross-peaks in the 2D HSQC spectra and altered C-terminal contacts in MD simulations. Motivated by these observations, we investigated the preference of DNAJB1 for aSyn pS129 fibrils. As a first step, we characterized the morphology of both fibril types in the presence of DNAJB1 using EM. Following the sedimentation assay, EM grids were prepared for the aSyn WT and pS129 fibrils incubated with DNAJB1. The average fibril width of DNAJB1-bound aSyn WT was 9.9 ± 2.04 nm, which was significantly narrower than that of unbound WT fibrils (11.21 ± 2.04 nm, p value = 3.3992 × 10^–5). In contrast, the width of bound pS129 fibrils did not change with DNAJB1 binding 9.30 ± 1.94 nm (compared to 9.15 ± 1.94, p value = 0.21198) (see Figure A).

3.

DNAJB1 binding reduces the width of WT fibrils and the dynamics close to the fibril core while having no effect on the pS129 fibril core. (A) EMs of WT and pS129 fibrils bound to DNAJB1 in a 3:1 molar ratio. Scale bars denote 100 nm. The magnified regions show intact fibrils displaying their characteristic lengths and bundled organization. The histograms reveal no significant differences in fibril width upon the addition of DNAJB1 for pS129 fibrils, whereas the WT fibrils showed a statistically significant decrease in average fibril width (see Figure ). (B) Overlay of ¹³C–¹³C DREAM spectra of unbound and DNAJB1-bound aSyn WT fibrils (left) and that of the pS129 fibrils (right). Negative contour levels are shown in the colors indicated in the legend, while positive contours (primarily representing the diagonal) are plotted in black. The corresponding 1D slices at 21.0 ppm are shown at the top. For the WT fibrils, the bound and unbound spectra overlap almost completely, with the exception of three additional residues (Ala, Lys, and Ser) that appear only in the bound sample (circled in black). These residues are likely derived from the C-terminal edge of the fibril core, suggesting local rigidification upon DNAJB1 binding. In contrast, spectra of pS129 fibrils exhibit complete overlap between the bound and unbound states.

To characterize the DNAJB1–fibril interaction at the molecular level, we recorded solid-state NMR experiments on the following ¹³C–¹⁵N-labeled fibril samples: aSyn WT, aSyn WT with DNAJB1, aSyn pS129, and aSyn pS129 with DNAJB1. These samples were all prepared from the same batch of aSyn fibrils to limit variations in the fibril quality and quantity. For the DNAJB1-bound samples, we chose a molar ratio of 3:1 (aSyn:DNAJB1) because our sedimentation assays indicated a noticeable difference in DNAJB1 binding to both fibril types (Figure B). Initially, we recorded 1D ¹³C CP and INEPT experiments to assess the overall dynamics. The 1D CP overlays of bound and unbound WT and pS129 fibrils overlapped almost perfectly and had very similar overall intensities. However, we observed clear differences in the relative intensity of our INEPT spectra (Figure S3C,D). We illustrated this relative difference in intensity via INEPT/CP intensity ratios in Figure S4, which show that both phosphorylation and DNAJB1 binding result in a substantial decrease in intensity, whereas the combination of both leads to a subtle increase rather than an additional decrease in intensity.

To check whether DNAJB1 binding affects the fibril core, we recorded 2D ¹³C–¹³C DREAM experiments on all four fibril samples. DREAM spectra detect mostly rigid regions, such as the fibril core, whose dipolar couplings are not averaged out by molecular motions. The 2D DREAM spectra of DNAJB1-bound WT fibrils generally overlapped well with the corresponding spectrum of unbound WT fibrils except for the appearance of a few additional resonances, which we assigned to Ser, Ala, and Lys (Figure B) residues. Because our DREAM spectra nicely fit a previous solid-state NMR assignment of aSyn fibrils done by Verasdnock and co-workers (BMRB entry 25535), we were able to assign one resonance to S87 at the C-terminal edge of the fibril core and identify the residue types of the two remaining resonances as Ala and Lys. In contrast, the DREAM spectra of pS129 fibrils with and without DNAJB1 were nearly identical and included the additional resonances described above, so that they closely resembled those of WT fibrils bound to DNAJB1. Overall, the overlaid DREAM spectra of bound and unbound WT and pS129 fibrils were nearly identical, indicating that DNAJB1 binding primarily targets the C-terminus.

Effect of DNAJB1 on the C-Terminus of WT and pS129 Fibrils

To site-specifically characterize the dynamic residues of WT and pS129 fibrils and their interaction with DNAJB1, we first assigned the 2D ¹H–¹⁵N HSQC spectra. To do so, we recorded liquid-state NMR-type 3D HNCA and HNcoCA NMR experiments of ¹³C–¹⁵N-labeled WT aSyn fibrils (Figure S6). Using these spectra, we were able to assign 26 residues from the highly dynamic C-terminus of the fibril. Specifically, we were able to identify most of the nonproline residues starting from A107 until A140, the last residue of the C-terminus. The CA shifts of these residues were close to their random coil values, confirming that the C-terminus of aSyn stays intrinsically disordered after aggregation. We next assigned the HSQCs of the pS129 fibrils. Interestingly, the phosphorylation site at S129 was absent in the recorded HSQCs of pS129 fibrils (Figures C and B). Despite their reduced overall intensity, we were able to assign 23 residues by overlaying the HSQC spectrum of pS129 fibrils with the corresponding spectrum of WT fibrils (Figure A). The residues near the phosphorylation site, specifically D121, N122, and E126, could not be detected, and the residues D119, A124, and M127 showed a significant decrease in intensity (Figure B).

4.

DNAJB1 binding occurs at similar C-terminal residues in WT and pS129 fibrils. (A) ¹H–¹⁵N HSQC spectra of WT aSyn fibrils alone and bound to DNAJB1 at molar ratios of 5:1 and 3:1 (aSyn:DNAJB1). Resonance assignments are indicated. (B) Corresponding HSQC spectra of pS129 fibrils. (C) I/I ₀ plots of WT fibrils calculated from the spectra in panel (A) where I ₀ is the intensity of a cross-peak in the unbound spectrum and I the intensity of the corresponding spectrum of DNAJB1-bound fibrils. (D) Corresponding I/I ₀ plots of pS129 fibrils. The black horizontal line marks the ratio of 0.5 for reference. The spectra of WT aSyn fibrils generally show a decrease in intensity around S129 with increasing DNAJB1 concentration. However, residues close to the fibril core and toward the very C-terminus are unchanged or even show a modest increase in cross-peak intensity at a 5:1 molar ratio. Please note that the spectrum of unbound pS129 fibrils already resembles that of DNAJB1-bound WT fibrils. Here, DNJAB1 binding results in a pronounced increase in relative intensity close to the fibril core and toward the very C-terminus at a 5:1 ratio. Only at a molar ratio of 3:1 does the overall intensity of pS129 fibrils decrease both around residue 129 and around 113 close to the fibril core.

To map the DNAJB1 binding sites in both fibril types and to confirm the reproducibility between sample batches, we prepared a second batch of ¹⁵N-labeled aSyn for both fibril types for DNAJB1 titration experiments. The HSQC spectra of these fibrils in their unbound state were identical to those of the previous sets (see Figures C vs A–B and S7), highlighting the good reproducibility of our results. We then incubated these fibrils with DNAJB1 at molar ratios of (aSyn:DNAJB1) 5:1, 4:1, 3:1, and 2:1, guided by the sedimentation assay results, and then recorded ¹H–¹⁵N HSQC spectra of each of these titrations. We normalized the intensity of these HSQC spectra to the number of acquisitions and to the 1D ¹⁵N CP signal of the corresponding sample to account for differences in sample packing. The spectra of both DNAJB1-bound WT and pS129 fibrils showed clear changes in peak intensities compared to unbound fibrils. Specifically, Figure shows the spectra of unbound and bound at 5:1 and 3:1 ratios. Please see Figures S8, S10, S11, S12 for all spectra in this series and their corresponding intensity ratio plots (I/I ₀). Certain residues, particularly in phosphorylated fibrils, showed a relative increase in intensity when bound to DNAJB1. In both WT and pS129 fibrils at a 5:1 ratio, residues at the extreme ends of the C-terminus (A107 and A140) exhibited I/I ₀ values greater than 1. Additionally, in pS129 fibrils, residues D119, D121, and A124 showed a subtle increase, and E139 showed a more pronounced increase in intensity, consistent with increased dynamics as reflected by the INEPT/CP ratios previously shown in Figure S4. However, at a 3:1 ratio, we mostly observed a decrease in intensity with the exception of A140 in pS129 fibrils.

To identify significant decreases in intensity upon DNAJB1 binding, we applied a threshold of I/I ₀ ≤ 0.5 across all titration series. For WT fibrils, only S129 showed a marked loss of intensity at a 5:1 ratio. However, at a 3:1 ratio, a larger set of residues, including D119, D121, E123, A124, M127, S129, E130, Q134, D135, E137, and E139, exhibited significant reductions. The 4:1 and 2:1 titrations followed the same progressive loss of relative peak intensity (Figure S8).

To confirm these site-specific intensity changes, we recorded 2D ¹H–¹³C HETCOR and ¹³C–¹³C TOCSY spectra on the unbound and bound (3:1 aSyn:DNAJB1 molar ratio) ¹³C–¹⁵N aSyn WT fibrils (Figure S9). The assignment of these spectra was made by correlating CA chemical shifts from HNCA with the corresponding spin system observed in TOCSY and HETCOR. Consistent with HSQC spectra discussed above, DNAJB1 binding also resulted in a reduction in peak intensities for residues within the binding region, notably D121, N122, M127, and S129 (Figure S10).

Similar to the WT fibrils, only one residue in the pS129 fibrils (i.e., M127 near the phosphorylation site) fell below the 0.5 threshold in relative intensity at a 5:1 molar ratio. At the 5:1 ratio, we observed an overall increase in intensities with most of the residues being unaffected and a few residues at the edges of the C-terminal domain showing a slight increase in relative intensity compatible with an increase in mobility in the presence of DNAJB1. However, at higher DNAJB1 concentrations, additional residues (Q109, L113, D121, E123, A124, E130, and D135) dropped below the threshold, indicating a more frequent interaction with DNAJB1. The 4:1 and 2:1 titrations also followed the progressive loss of relative peak intensity for the same set of residues (Figures S11 and S12).

The HETCOR and TOCSY spectra of pS129 fibrils without DNAJB1 and at a aSyn:DNAJB1 ratio of 3:1 overlay very well, as shown in Figure S13. However, comparing these spectra to those of the unbound WT fibrils, we noticed that the same residues that were reduced in intensity upon DNAJB1 binding were absent or exhibited weak signals in both sets of pS129 spectra similar to the ¹H–¹⁵N HSQC experiments discussed above. A schematic of the most affected residues is shown in Figure A. In both WT and pS129 fibrils, only one residue fell below the intensity threshold at a 5:1 ratio, both close to the phosphorylation site (S129 in WT and M127 in pS129). At the 3:1 ratio, this effect expanded to multiple neighboring residues, 11 in WT and seven in pS129, distributed across the C-terminal region.

5.

Summary of the HSQC titration analysis and the proposed mechanism of DNAJB1 interaction with WT and pS129 fibrils from MD and NMR. (A) NMR analysis reveals that DNAJB1 primarily engages the dynamic C-terminus of aSyn in both fibril types. Residues perturbed by DNAJB1 are mapped onto the C-terminal sequence: in WT fibrils, the residue below the 0.5 threshold at 5:1 is highlighted with a gold rectangle, and residues affected at 3:1 are shaded in gold (top). In pS129 fibrils, residues affected at the respective molar ratios are marked in crimson red (bottom). (B) The schematic illustrates our proposed model: in pS129 fibrils, the C-terminus forms stronger intramolecular contacts with the fibril core, resulting in a more constrained unbound conformation. DNAJB1 binding is proposed to weaken or disrupt these contacts, thereby “pulling” the C-terminus away from the fibril surface. A similar effect may occur in WT fibrils, but it appears more pronounced in pS129 fibrils. The greater binding affinity of DNAJB1 for pS129 fibrils may stem from electrostatic complementarity introduced by the negative charge of phosphoserine, which facilitates stronger engagement with the C-terminal region.

To get more molecular insight into how pS129 increases DNAJB1 binding, we ran CG MD simulations using the HyRes model , (see Methods). Specifically, we simulated a fibril 80mer together with 4 DNAJB1 dimers for 4 μs with and without S129 phosphorylation. Figure A shows a snapshot of this simulation of a WT fibril interacting with DNAJB1 dimers.

6.

HyRes coarse-grained simulations of aSyn fibrils confirm increased DNAJB1 binding upon phosphorylation at S129. (A) A representative snapshot from a 4 μs HyRes simulation of an 80mer aSyn fibril with 4 dimers of DNAJB1 (in purple). (B) The probabilities of the number of bound DNAJB1 dimers, showing a shift toward more bound dimers in the case of pS129 fibrils. The error bar was estimated as the derivation of two blocks of trajectory. (C) Per residue DNAJB1 contact frequency at the C-terminus of WT and pS129 fibrils. Only DNAJB1/fibril contacts with duration >200 ps were included. Phosphorylation of S129 results in both a high contact frequency and a more pronounced binding around the phosphorylation site.

These simulations confirm two main findings from our binding and NMR studies: Figure B plots the probability of the fibril to bind 1–4 of the DNAJB1 dimers. Here, pS129 fibrils have a much higher probability to bind to all 4 DNAJB1 dimers, whereas WT fibrils are more likely to bind 1–3 dimers. Figure S14 shows the time traces of the radial distance from the center of mass of DNAJB1 to the axis of the aSyn fibrils from which these probabilities were calculated. These results are in line with the higher affinity for pS129 fibrils detected in our sedimentation assay. Figure C shows the frequency of a more than 200 ps long contact between one of the 4 DNAJB1 dimers and a specific residue in any of the aSyn monomers making up the fibril. Overall, the simulation confirms the binding of DNAJB1 to the C-terminus of aSyn in the fibrillar form. There is only negligible contact between DNAJB1 and the N-terminus, virtually no contact to the fibril core, and the vast majority of the contacts are made to the C-terminus (Figure S15A). The contact frequency plot in Figure C further highlights the increase in binding to the pS129 fibrils, with the very phosphorylation site being the most contacted residue. The corresponding aSyn fibril contacts to DNAJB1 are found throughout the CTD I and CTD II domains (Figure S15B). These simulations further support the observation of the dynamic nature of the DNAJB1 interaction with the C-terminus, with numerous residues involved in intermolecular contacts.

Discussion

Our findings highlight the impact of PTMs, particularly in the disordered regions of aSyn fibrils, on chaperone interactions. The aSyn pS129 PTM is especially noteworthy, as 90% of aggregated aSyn in Lewy bodies is phosphorylated at this site. Given that DNAJB1’s binding site at the flexible C-terminus coincides precisely with the location of the pS129 PTM, understanding how this phosphorylation event modifies the chaperone–fibril interaction becomes essential for elucidating how PTMs affect chaperone-mediated amyloid disaggregation mechanisms.

In our initial sedimentation assay, we observed that pS129 fibrils pulled down more DNAJB1 than WT fibrils. To quantify the difference in binding affinity, we determined the K _D of DNAJB1 for both fibril types (Figure C). Titrations with increasing concentrations of aSyn WT and pS129 fibrils revealed that the K _D of pS129 fibrils was ∼30-fold lower than that of WT fibrils, indicating a markedly higher affinity of DNAJB1 for phosphorylated fibrils. This enhanced affinity is likely driven by phosphorylation-mediated electrostatic interactions that strengthen chaperone engagement. Supporting this idea, Monistrol and co-workers reported increased DNAJB1–aSyn interactions under low ionic strength conditions. In addition, we detected a decrease in the pool of free monomeric aSyn upon phosphorylation at S129 (Figure B). This effect may arise from the additional negative charge introduced by phosphorylation, which could promote intermolecular interactions between the N-terminus of aSyn monomers and the C-terminus of fibrils. ,

We next investigated how phosphorylation alters the fibril structure and dynamics. Our electron micrographs showed minimal changes in fibril morphology, as expected since the PTM occurs at the C-terminus and phosphorylation was performed after fibril formation. However, we did observe a slight narrowing of our fibrils with phosphorylation. Our CP-based methods (1D CP, 2D DREAM) indicated little effect of DNAJB1 binding on the static fibril core of our fibrils, with the exception of a few additional immobilized residues at the C-terminal end of the fibril core of WT fibrils when binding DNAJB1. These residues were already static in all of our pS129 fibrils spectra. The 1D INEPT/CP ratios and our HSQC spectra supported these findings, showing that both pS129 and DNAJB1 binding substantially reduced the dynamics in the C-terminus of our fibrils. MD simulations provided a structural rationale for these observations, revealing an increased number of contacts of the C-terminus with the fibril core in pS129 compared with WT fibrils. This is likely due to the increasing electrostatic interactions between the negative charges of the phosphate group on the serine and lysines found throughout the fibril core (K43, K45, K58, K60, K80, K96, and K97). The substitution of the polar serine with an acidic phosphoserine similarly increases the electrostatic interactions between the fibril C-terminus and DNAJB1.

The interaction of the C-terminus with Lys residues in the aSyn fibril core was also a key mechanism in a recent study by Han and co-workers. They showed that aSyn fibril polymorphs with very similar core structures but a different extent of C-terminus–core interactions differed in their pathological transmission activity. Similarly, pS129 was shown to lower the seeding propensity of aSyn fibrils in cultured cells.

While our data generally confirm the DNAJB1 binding site previously derived from fluorescence and solution NMR measurements of monomeric aSyn (123–129), we show that phosphorylation at S129 not only increases the interaction but also slightly changes the nature of this interaction. Where DNAJB1 binds WT aSyn fibrils around the phosphorylation site S129, the HSQC spectra of pS129 fibrils also show a decrease in relative intensity closer to the fibril core. Our MD simulations corroborate these two main sites of DNAJB1 interaction; however, they show an increase in binding around S129 with phosphorylation in contrast to our experimental data that indicated more pronounced binding around L113. Interestingly, we observed a slight increase in relative intensity in our HSQC spectra at the edge of the binding site, especially for our pS129 fibrils at low DNAJB1 concentration (i.e., 5:1 molar ratio). This indicates that part of the C-terminus can become more mobile in the presence of DNAJB1. Regarding the aSyn fibril contacts of DNAJB1, our MD simulations show contacts throughout the CTD I and CTD II domains, which is compatible with a previous solution NMR study.

Recent work by Ghosh and co-workers demonstrated the importance of the C-terminal domain of aSyn fibrils for chaperone interactions. They investigated the chaperone domain BRICHOS, showing that it can prevent secondary nucleation by binding to the C-terminus of aSyn fibrils and removing loosely bound monomeric aSyn that could otherwise aggregate on the fibril surface. Our data indicate that DNAJB1 behaves similarly in the sense that it is not tightly bound to the C-terminus of aSyn fibrils, resulting in an effect on the entire fibril at substoichiometric ratios. However, our data indicate that DNAJB1 does not substantially increase the monomer concentration of aSyn in equilibrium with the fibrils, which could be due to the fact that DNAJB1 is a dimer that can potentially bind C-termini from fibrillar and monomeric aSyn at the same time.

Monistrol and co-workers recently reported cryo-EM electron densities of DNAJB1-bound aSyn fibrils. Their data indicate that DNAJB1 is packed with a repeat of about 40 Å along the fibril axis. Because our data indicate that DNAJB1 is not tightly bound to the C-terminus of aSyn fibrils, these DNAJB1 repeats would either rapidly form and disassemble or run together along the fibril axis.

Upon closer inspection, we found that the unbound pS129 HSQC spectra closely resembled those of WT fibrils bound to DNAJB1 in a 3:1 ratio. This similarity suggests that binding of DNAJB1 to WT fibrils perturbs the same residues affected by phosphorylation at S129 in unbound fibrils, likely through enhanced intramolecular contacts between the C-terminus and the fibril core. Consistent with this interpretation, the reduced C-terminal dynamics of pS129 fibrils coincides with the appearance of three additional resonances (S87, Lys, and Ala) from the C-terminal edge of the fibril core in the 2D DREAM spectra. These resonances were absent in WT fibrils but also appeared in the DNAJB1-bound WT spectra, indicating that both phosphorylation and DNAJB1 binding may reduce the dynamics of residues located at the C-terminal end of the fibril core.

Our combined NMR and MD data suggest that phosphorylation at S129 enhances both (i) the C-terminal contacts with the fibril core and (ii) the electrostatic complementarity with DNAJB1. We propose that in pS129 fibrils the C-terminus engages in stronger intramolecular contacts with the fibril core, resulting in a more restricted conformation in the unbound state. DNAJB1 binding may partially disrupt these contacts, effectively “pulling” the C-terminus away from the fibril surface (Figure B). A similar effect may occur in WT fibrils but appears more pronounced in pS129 fibrils. Together, these findings suggest that phosphorylation modulates chaperone engagement by tuning the conformational landscape of the C-terminus, which may ultimately influence the efficiency of fibril disaggregation.

We deposited the NMR chemical shifts of WT aSyn fibrils as BMRB ID 53372. We deposited all NMR spectra, EM images, and mass spectrometry data using the Zenodo repository (10.5281/zenodo.17190007).

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.5c01266.

• Identified aSyn peptides containing serine 129 (Table S2); complete analysis (XLSX)

• Additional SDS-PAGEs, mass spectrometry data, NMR spectra, NMR analyses, MD simulation snapshots, analysis of MD simulation, and experimental NMR parameters (PDF)

§.

S.P. and D.S.R. contributed equally to this work. S.P. and D.S.R. prepared the samples, conducted experiments, and wrote the manuscript. Q.Z. and G.V. prepared samples and conducted experiments. S.L. ran and analyzed MD simulations and prepared figures. J.C. conceptualized and guided the simulations and their analysis. A.B.S. conceived and guided the study and revised the manuscript.

The authors declare no competing financial interest.