¹³C and ¹⁵N Resonance Assignments of Alpha Synuclein Fibrils Amplified from Lewy Body Dementia Tissue

Abstract

Fibrils of the protein α-synuclein (Asyn) are implicated in the pathogenesis of Parkinson Disease, Lewy Body Dementia, and Multiple System Atrophy. Numerous forms of Asyn fibrils have been studied by solid-state NMR and resonance assignments have been reported. Here, we report a new set of ¹³C, ¹⁵N assignments that are unique to fibrils obtained by amplification from postmortem brain tissue of a patient diagnosed with Lewy Body Dementia.

Biological Context

Alpha-synuclein (Asyn) is a 140-residue protein found in the presynaptic termini of neurons in the brain (Clayton and George 1999). While the exact function of this protein in the brain remains elusive (Lautenschlager, Stephens et al. 2018), the aggregation of Asyn in the form of fibrils has been a pathological hallmark of Parkinson Disease (PD), Lewy Body Dementia (LBD) and Multiple System Atrophy (MSA), all which can be classified as α-synucleinopathies. Dementia occurs frequently in PD, sometimes beginning at approximately the same time as motor symptoms (Dementia with Lewy bodies or DLB), or up to 20 years after motor symptoms begin (PD with dementia or PDD). The term LBD encompasses the spectrum of clinical presentations classified as DLB and PDD.

Asyn fibrils have been reported to have a range of tertiary and quaternary structures (Tuttle, Comellas et al. 2016, Schweighauser, Shi et al. 2020), contributing to an emerging understanding of the precise relationships of in vitro and ex vivo conditions. The in vitro structures depend on various factors such as mutations (Polymeropoulos, Lavedan et al. 1997, Kruger, Kuhn et al. 1998, Zarranz, Alegre et al. 2004, Comellas, Lemkau et al. 2011, Lemkau, Comellas et al. 2012, Lemkau, Comellas et al. 2013, Khalaf, Fauvet et al. 2014), cytosolic components (Guilarte 2010, Shin and Chung 2012, Kwakye, Paoliello et al. 2015), lipids (Bodner, Dobson et al. 2009, Bodner, Maltsev et al. 2010, Jakubec, Barias et al. 2021, Mahapatra, Mandal et al. 2021), and metals (Uversky, Li et al. 2001). Growing evidence indicates that distinct polymorphs are associated with pathologic Asyn accumulation in α-synucleinopathies, as detailed by previously reported ex vivo studies (Schweighauser et al. 2020, Yang et al. 2022) Therefore, structural determination of these fibrils is vital for advancing the understanding of disease etiology, and to aid the development of polymorph-specific clinical diagnostic tools and novel therapeutics.

We isolated insoluble Asyn fibrils from postmortem LBD tissue. Then, to analyze LBD fibril structure by SSNMR, we amplified the Asyn fibril seeds using uniform [¹³C, ¹⁵N] labeled wild-type Asyn. Here, we report the ¹³C and ¹⁵N chemical shifts for the amplified Asyn fibrils from an LBD autopsy case. The spectra of these postmortem seeded fibrils exhibit resonances that differ from those of previously reported in vitro fibril preparations. These findings demonstrate a different arrangement of β-strands, supporting the hypothesis that fibril structure is directly linked to disease phenotype.

Methods and experiments

Protein expression and purification

Expression of uniform [¹³C, ¹⁵N] labeled wild-type Asyn was carried out in E. coli BL21(DE3)/pET28a-AS in modified Studier medium M (Studier 2005). The labeling medium contained 3.3 g/L [¹³C]glucose, 3 g/L [¹⁵N]ammonium chloride, 11 mL/L [¹³C, ¹⁵N]Bioexpress (Cambridge Isotope Laboratories, Inc., Tewksbury, MA), 1 mL/L BME vitamins (Sigma), and 90 μg/mL kanamycin. After a preliminary growth in medium containing natural abundance (NA) isotopes, the cells were transferred to the labeling medium at 37 °C to an OD₆₀₀ of 1.2, at which point the temperature was reduced to 25 °C and protein expression induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and grown for 15 h to a final OD₆₀₀ of 4.1 and harvested.

Protein purification was done as described previously (Barclay, Dhavale et al. 2018). Briefly, cells were lysed chemically in the presence of Turbonuclease (Sigma) to digest nucleic acids. Purification began with heat denaturation of the cleared lysate, followed by ammonium sulfate precipitation (Kloepper, Woods et al. 2006). The resolubilized protein was bound to QFF anion exchange resin (GE Healthcare Life Sciences, Marlborough, MA) and eluted using a linear gradient of 0.2–0.6 M NaCl. Fractions containing Asyn monomer, which eluted at about 0.3 M NaCl, were pooled, concentrated, and run over a 26/60 Sephacryl S-200 HR gel filtration column (GE Healthcare Life Sciences) equilibrated in 50 mM Tris-HCl, 100 mM NaCl, pH 8 buffer. Fractions were pooled, concentrated to ~20 mg/mL Asyn, and dialyzed at 4 °C into 10 mM Tris-HCl pH 7.6, 50 mM NaCl, 1 mM DTT, and stored at a concentration of ~14 mg/mL at −80 °C until use. Yields were 95 mg purified AS protein/L growth medium for the uniform [¹³C, ¹⁵N] labeled monomer.

Preparation of Insoluble fraction seeds from LBD, MSA and control postmortem tissue

The protocol to sequentially extract human postmortem brain tissue was adapted from Appel-Cresswell et al². Briefly, gray matter dissected from tissue was sequentially homogenized in four buffers (3 ml/g wet weight of tissue) using Kimble Chase Konte^™ dounce tissue grinders (KT885300-0002). In the first step, 300mg of dissected grey matter tissue was homogenized using 20 strokes of Pestle A in High Salt (HS) buffer (50 mM Tris-HCl pH 7.5, 750 mM NaCl, 5 mM EDTA plus Sigma P2714 Protease Inhibitor (PI) cocktail). The homogenate was centrifuged at 100,000 ×g for 20 min at 4 °C and the pellet was homogenized in the next buffer using 20 strokes of Pestle B. Extractions using Pestle B were performed in HS buffer with 1% Triton X-100 with PI, then HS buffer with 1% Triton X-100 and 1 M sucrose, and with 50 mM Tris-HCl, pH 7.4 buffer. In the final centrifugation, the resulting pellet was resuspended in 50 mM Tris-HCl, pH 7.4 buffer (3 ml/g wet weight of tissue). The aliquots of insoluble fraction were stored at −80 °C until use. Similar extraction protocol was followed for LBD, MSA and control cases.

Amplification of isotopically labelled LBD fibrils from LBD insoluble fraction seeds

We amplified LBD-fibrils from gray matter dissected from the caudate region. We incubated insoluble fraction seeds with an Asyn monomer preparation containing isotopically labeled Asyn monomer supplemented with control fraction. The control fraction preparation was derived from E.Coli transformed with an empty expression vector, and was purified with the same protocol as the natural abundance Asyn monomer (JBC, 2017 V292, Pg9034). Asyn monomer and control fraction was filtered through a 50k MWCO Amicon Ultra centrifugation filter (Millipore, UFC805204) before use, to remove any preformed aggregates.

Insoluble fraction (10 μL) containing 3.3 μg wet wt. of tissue was brought to a final volume of 30 μL by addition of 20 mM Tris-HCl, pH 8.0 plus 100 mM NaCl buffer (fibril buffer) in a 1.7mL microcentrifuge tube. The insoluble fraction was sonicated for 2 min at amplitude 50 in a bath sonicator (Qsonica model Q700) with a cup horn (5.5 inch) attachment at 4 °C. 1.5 μL of 2 % Triton X-100 was added to the sonicated seeds. To this mixture, 50k Amicon ultra filtered isotopically labelled Asyn monomer was added to a final concentration of 2 mg/mL in a final volume of 100 μL. This mixture underwent quiescent incubation at 37 °C for 3 days, completing the 1^st round of sonication plus incubation. After the first round, the mixture was sonicated at 1 min at amplitude 50, and then an additional 300 μL of 2 mg/mL Asyn monomer was added. The mixture underwent quiescent incubation at 37 °C for 2 days (2^nd round). Then, sonication for 1 min at amplitude 50 and quiescent incubation for 2 days was repeated for the third round, followed by sonication for 1 min at amplitude 50 and quiescent incubation for 3 days for the 4^th round. At the end of 4^th round, LBD-amplified fibrils were stored at 4 °C until use.

Further expansion of the LBD-amplified fibrils was performed by centrifuging 60 μL of 4^th round LBD-amplified fibrils at 21,000 xg for 15 min at 4 °C. The pellet was resuspended in 100 μL of fibril buffer and sonicated for 1 min at amplitude 50. To this mixture, Asyn monomer was added to a final concentration of 2 mg/mL in a final volume of 400 μL in fibril buffer. This mix was quiescently incubated at 37 °C for 2 days (5^th round). At the end of 5^th round, samples were centrifuged at 21,000 xg for 15 min at 4 °C and the top 300 μL of spent Asyn monomer was moved to a separate tube. The pellet was resuspended by trituration and sonicated for 1 min at amplitude 50. After sonication, the previously removed 300 μL of 5^th round monomer was added back. Next, an additional 2.5 mg/mL of Asyn monomer was added to bring the total volume to 800 μL. This mixture was incubated at 37 °C for 2 days to complete 6^th round of incubation. The increased monomer concentration (2.5 mg/mL instead of 2 mg/mL) was calculated based on the average decrease in free Asyn monomer due to its incorporation into amplified fibrils. The 6^th round fibrils were stored at 4 °C until use.

Solid-state NMR spectroscopy

Magic-angle spinning (MAS) SSNMR experiments were performed at magnetic field of 11.7 T (500 MHz ¹H frequency) or 17.6 T (750 MHz ¹H frequency) using Agilent Technologies VNMRS spectrometers. Spinning was controlled with a Varian MAS controller to 11,111 ± 30 Hz or 22,222 ± 15 Hz (11.7 T) and 16,667 ± 15 Hz or 33,333 ± 30 Hz (17.6 T). All experiments were done with a variable-temperature (VT) airflow setting of 0 °C, primarily to keep samples cool from RF and MAS heating, without freezing out molecular motions. The 11.7 T magnet was equipped with a 1.6 mm HCDN T3 probe (Varian), with pulse widths of about 1.8 μs for ¹H and ¹³C, and 3.2 μs for ¹⁵N. The 17.6 T magnet was equipped with a HXYZ T3 probe (Varian) tuned to HCN triple resonance mode with pulse widths of about 1.9 μs for ¹H, 2.6 μs for ¹³C, and 3.0 μs for ¹⁵N. All experiments utilized ¹H-¹³C or ¹H-¹⁵N tangent ramped CP (Metz, Wu et al. 1994) and ~100 kHz SPINAL-64 decoupling during evolution and acquisition periods (Fung, Khitrin et al. 2000, Comellas, Lopez et al. 2011). Where applicable, SPECIFIC CP was used for ¹⁵N-¹³Cα and ¹⁵N-¹³C’ transfers (Baldus, Petkova et al. 1998), ¹³C-¹³C homonuclear mixing was performed using DARR (Takegoshi, Nakamura et al. 2001). Chemical shifts were externally referenced to the downfield peak of adamantane at 40.48 ppm (Morcombe and Zilm 2003). NUS schedules using biased exponential sampling were prepared using the nus-tool application in NMRbox (Maciejewski, Schuyler et al. 2017). Data conversion and processing was done with NMRPipe (Delaglio, Grzesiek et al. 1995). NUS data was first expanded with the nusExpand.tcl script in NMRPipe, converted, and processed using the built-in SMILE reconstruction function (Ying, Delaglio et al. 2017). Peak picking and chemical shift assignments were performed using NMRFAM-Sparky (Lee, Tonelli et al. 2015).

Extent of assignments and data deposition

Chemical shift assignments were performed for the LBD Asyn fibrils amplified using uniform [¹³C, ¹⁵N] labeling (uCN). Resonance assignments were determined using 2D ¹³C-¹³C, 2D ¹⁵N-¹³Cα, 2D ¹⁵N-¹³C’, 3D ¹⁵N-¹³Cα-¹³CX, 3D ¹⁵N-¹³C’-¹³CX, 3D ¹⁵N-¹³C’-¹³Cα and 3D ¹³Cα-¹⁵N-¹³C’ data sets following standard procedures (Comellas and Rienstra, 2013; Higman, 2018). The 2D ¹⁵N-¹³C’ (Figure 1A) and ¹⁵N-¹³Cα (Figure 1B) spectra serve as structural fingerprints for the fibril, with a focus on the backbone atoms. Critically, these highlight the benefit of adding a ¹⁵N dimension to disambiguate shifts, particularly for the Ala and Gly regions, which comprise 30% of the primary sequence between residues 30 and 100, as well as for key core residues like V71, V74, T75 and V77. Overall, there appears to be a predominant defined conformation that likely displays some localized heterogeneity. Figure 2 demonstrates representative strips corresponding to assignments from 3D ¹⁵N-¹³Cα-¹³CX, 3D ¹⁵N-¹³C’-¹³CX, and 3D ¹³Cα-¹⁵N-¹³C’ from T72 to A76.

Figure 1:

Backbone chemical shift assignments demarked on a ¹⁵N-¹³C’ (A) and a ¹⁵N-¹³Cα (B) spectra.

Figure 2:

Example backbone assignment strip from T72 to A76 demonstrating connectivity and sidechain assignments from a 3D ¹⁵N-¹³Cα-¹³CX correlation (black), a 3D ¹⁵N-¹³C’-¹³CX correlation (red), and a 3D ¹³Cα-¹⁵N-¹³C’correlation (blue).

Comparison to previously structurally characterized fibril

We compare the chemical shift assignments of the postmortem seeded LBD fibrils to the extensively studied in vitro fibril preparation (PDB: 2N0A). Figure 3 summarizes the differences in the chemical shift values that lead us to the conclusion that the postmortem seeded LBD fibrils presented herein are wholly different than the in vitro fibril preparations. This is based chiefly on two observations from extensive assignments of both fibril types. First, due to differences in dynamics or line broadening attributed to multiple confirmations, amino acid sequences that can be unambiguously sequentially assigned within the primary structure of Asyn are different. We discount the possibility of multiple confirmations being present as there is a single set of resonances assignable to each visible peak within the spectrum with no unambiguous peaks assignable to consecutive stretches from the 2D and 3D datasets. Secondly, unambiguous chemical shifts that are common between in vitro and postmortem seeded fibrils display significant chemical shift differences, with an average of ~1.9 ppm, when scaled with respect to ¹⁵N and ¹³C chemical shift values.

Figure 3:

Secondary structure and chemical shift comparison between postmortem seeded LBD and PDB ID: 2N0A. Top: Predicted secondary structure of the LBD case versus the structure of the in vitro fibril form from residues 28 to 99. The primary structure is labeled under the secondary structure for reference. Bottom: Plot of the residue specific chemical shift RMS differences of the LBD postmortem seeded case to the in vitro 2N0A case. RMS $\Delta \delta$ was calculated as $\Delta \delta ={\{{\sum }_{\mathrm{i}}[{({\Delta \delta }_{\text{Ci}})}^2+{(0.4\times {\Delta \delta }_{\mathrm{N}})}^2]∕\mathrm{n}\}}^{1∕2}$, where $\mathrm{i}$ refers to $\alpha$, $\beta$, $\gamma$, $\delta$, and $\epsilon$ and $\mathrm{n}$ is the number of assignments used to perform the calculation.

Data availability

Resonance assignments are available through the Biological Magnetic Resonance Data Bank (http://bmrb.io), Accession Number 51678.