Structurally targeted mutagenesis identifies key residues supporting α -synuclein misfolding in multiple system atrophy

Abstract

Multiple system atrophy (MSA) and Parkinson's disease (PD) are caused by misfolded $\alpha$ -synuclein spreading throughout the central nervous system. While familial PD is linked to several point mutations in $\alpha$ -synuclein, there are no known mutations associated with MSA. Our previous work investigating differences in $\alpha$ -synuclein misfolding between the two disorders showed that the familial PD mutation E46K inhibits replication of MSA prions both in vitro and in vivo, providing key evidence to support the hypothesis that $\alpha$ -synuclein adopts unique strains in patients. Here, to further interrogate $\alpha$ -synuclein misfolding, we engineered a panel of cell lines harboring both PD-linked and novel mutations designed to identify key residues that facilitate $\alpha$ -synuclein misfolding in MSA. These data were paired with in silico analyses using Maestro software to predict the effect of each mutation on the ability of $\alpha$ -synuclein to misfold into one of the reported MSA cryo-electron microscopy conformations. In many cases, our modeling accurately identified mutations that facilitated or inhibited MSA replication. However, Maestro was occasionally unable to predict the effect of a mutation on MSA propagation in vitro, demonstrating the challenge of using computational tools to investigate intrinsically disordered proteins. Finally, we used our cellular models to determine the mechanism underlying the E46K-driven inhibition of MSA replication, finding that the E46/K80 salt bridge is necessary to support $\alpha$ -synuclein misfolding. Overall, our studies use a structure-based approach to investigate $\alpha$ -synuclein misfolding, resulting in the creation of a powerful panel of cell lines that can be used to interrogate MSA strain biology.

Fig. 1.. Cryo-EM structures of patient-derived and synthetic α-synuclein fibrils.

The cryo-EM structures of $\mathrm{\alpha }$-synuclein fibrils resolved from (A) MSA and (B) Parkinson’spatient samples, as well as recombinant fibrils in the (C) ‘rod’ and (D) ‘twister’ conformations^– are shown. Residues that can be mutated in familial PD cases are shown in blue, with the exception of E46. Residues targeted with novel mutations designed to interfere with $\mathrm{\alpha }$-synuclein misfolding are shown in purple. The E46 and K80 residues, which often form a salt bridge that stabilizes a Greek key motif, are shown in red. PDB IDs are noted above each structure.

Fig. 2.. MSA prions differentially replicate in cells expressing PD-causing α-synuclein mutations.

HEK293T cells expressing (A) WT $\mathrm{\alpha }$-synuclein or (B-I) Parkinson’s-causing $\mathrm{\alpha }$-synuclein mutations fused to YFP were used to determine the effect of each mutation on MSA prion replication in vitro. Each cell line was infected with $\mathrm{\alpha }$-synuclein prions isolated from 4 control patient samples (C1, C2, C6, and C7) or 5 MSA patient samples (MSA2, MSA3, MSA6, MSA7, and MSA12) by phosphotungstic acid precipitation (x 10⁶ arbitrary units [A.U.]). (A-C) Consistent with previous findings, (A & C) MSA samples replicated in cells expressing full-length WT $\mathrm{\alpha }$-synuclein ($\mathrm{\alpha }$-syn140-YFP) and the A30P and A53T double mutation ($\mathrm{\alpha }$-syn140*A30P,A53T-YFP), (B) whereas the E46K mutation ($\mathrm{\alpha }$-syn140*E46K-YFP) blocked MSA propagation. (D-I) More recently identified PD-causing mutations were tested to determine the effect of each on MSA propagation. The (E) H50Q ($\mathrm{\alpha }$-syn140*H50Q-YFP), (F) G51D ($\mathrm{\alpha }$-syn140*G51D-YFP), and (H) A53V mutations ($\mathrm{\alpha }$-syn140*A53V-YFP) facilitate MSA propagation, whereas the (D) A30G ($\mathrm{\alpha }$-syn140*A30G-YFP) and (G) A53E mutations ($\mathrm{\alpha }$-syn140*A53E-YFP) ablate replication. (I) The T72M mutation ($\mathrm{\alpha }$-syn140*T72M-YFP) blunted MSA prion replication compared to the $\mathrm{\alpha }$-syn140-YFP cell line. (* = P < 0.05)

Fig. 3.. Using cryo-EM structures to inform α-synuclein mutation design reveals patient-to-patient differences in strain biology.

HEK293T cells expressing $\mathrm{\alpha }$-synuclein mutations designed to interfere with protein misfolding into either the rod (PDB ID: 6CU7), twister (PDB ID: 6CU8), or MSA filaments (PDB IDs: 6XYO, 6XYP, & 6XYQ) were used to determine the effect of single residue substitutions on MSA prion replication in vitro. Each cell line was infected with $\mathrm{\alpha }$-synuclein prions isolated from 4 control patient samples (C1, C2, C6, and C7) or 5 MSA patient samples (MSA2, MSA3, MSA6, MSA7, and MSA12) by phosphotungstic acid precipitation (x 106 arbitrary units [A.U.]). (A) The E61Q ($\mathrm{\alpha }$-syn140*E61Q-YFP) and (B) V66F ($\mathrm{\alpha }$-syn140*V66F-YFP) mutations were designed to preferentially interfere with $\mathrm{\alpha }$-synuclein misfolding into the twister conformation. While the E61Q mutation facilitated MSA replication, successful propagation in cells expressing the V66F mutation varied by sample. (C-E) The G36K, V37F, and V55Y mutations were designed to disrupt the protofilament interface in the MSA cryo-EM structures. (C) Infection in cells expressing the G36K mutation ($\mathrm{\alpha }$-syn140*G36K-YFP) also showed patient-to-patient variability whereas the (D) V37F mutation ($\mathrm{\alpha }$-syn140*V37F-YFP) facilitated replication and (E) the V55Y mutation ($\mathrm{\alpha }$-syn140*V55Y-YFP) prevented replication of all but one MSA patient sample sample. (F & G) The V74I and V74P mutations were selected based on the hypothesis that they would interfere with the twister conformation more than the MSA structures. (F) The V74I mutation ($\mathrm{\alpha }$-syn140*V74I-YFP) blocked MSA prion replication. (G) By comparison, the V74P mutation ($\mathrm{\alpha }$-syn140*V74P-YFP) blunted MSA replication without complete inhibition. (* = P < 0.05)

Fig. 4.. MSA prion replication requires the formation of the E46/K80 salt bridge.

HEK293T cells expressing $\mathrm{\alpha }$-synuclein mutations at lysine 80 were used to determine the requirement of the E46/K80 salt bridge for MSA prion replication in vitro. Each cell line was infected with $\mathrm{\alpha }$-synuclein prions isolated from 4 control patient samples (C1, C2, C6, and C7) or 5 MSA patient samples (MSA2, MSA3, MSA6, MSA7, and MSA12) by phosphotungstic acid precipitation (x 10⁶ arbitrary units [A.U.]). (A) The K80E mutation ($\mathrm{\alpha }$-syn140*K80E-YFP) blocked MSA prion replication. (B & C) The K80N ($\mathrm{\alpha }$-syn140*K80N-YFP) and K80Q mutations ($\mathrm{\alpha }$-syn140*K80Q-YFP) exerted variable effects on MSA prion replication, with the K80N mutation having a stronger blunting effect overall than the K80Q mutation. (D) The K80W mutation ($\mathrm{\alpha }$-syn140*K80W-YFP) also prevented MSA prion replication. (* = P < 0.05)

Fig. 5.. The E46K,K80E double mutation does not rescue MSA prion replication in vitro.

HEK293T cells expressing the E46K & K80E double mutation in $\mathrm{\alpha }$-synuclein was used to determine if residue swapping can rescue MSA prion propagation. The cells were infected with $\mathrm{\alpha }$-synuclein prions isolated from 4 control patient samples (C1, C2, C6, and C7) or 5 MSA patient samples (MSA2, MSA3, MSA6, MSA7, and MSA12) by phosphotungstic acid precipitation (x 10⁶ arbitrary units [A.U.]). However, the MSA patient samples could not replicate in the $\mathrm{\alpha }$-syn140*E46K,K80E-YFP cells.