Metal-free superoxide dismutase forms soluble oligomers under physiological conditions: A possible general mechanism for familial ALS

Abstract

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disorder selectively affecting motor neurons; 90% of the total cases are sporadic, but 2% are associated with mutations in the gene coding for the antioxidant enzyme copper–zinc superoxide dismutase (SOD1). The causes of motor neuron death in ALS are poorly understood in general, but for SOD1-linked familial ALS, aberrant oligomerization of SOD1 mutant proteins has been strongly implicated. In this work, we show that wild-type human SOD1, when lacking both its metal ions, forms large, stable, soluble protein oligomers with an average molecular mass of ≈650 kDa under physiological conditions, i.e., 37°C, pH 7.0, and 100 μM protein concentration. It further is shown here that intermolecular disulfide bonds are formed during oligomerization and that Cys-6 and Cys-111 are implicated in this bonding. The formation of the soluble oligomers was monitored by their ability to enhance the fluorescence of thioflavin T, a benzothiazole dye that increases in fluorescence intensity upon binding to amyloid fibers, and by disruption of this binding upon addition of the chaotropic agent guanidine hydrochloride. Our results suggest a general, unifying picture of SOD1 aggregation that could operate when wild-type or mutant SOD1 proteins lack their metal ions. Although we cannot exclude other mechanisms in SOD1-linked familial ALS, the one proposed here has the strength of explaining how a large and diverse set of SOD1 mutant proteins all could lead to disease through the same mechanism.

Protein oligomerization, aggregation, and formation of insoluble amyloid deposits commonly are observed in neurodegenerative diseases, but the factors initiating and modulating the abnormal protein–protein interactions that lead to oligomerization remain elusive (1, 2). Metal ions frequently have been implicated in these phenomena, but how exactly they are involved remains unclear (3). Over 114 different variants of human copper–zinc superoxide dismutase (Cu₂Zn₂SOD1) have been linked to the neurodegenerative disease familial ALS (FALS) by a gain-of-function mechanism (4–6). Although the exact cellular sites and mechanisms of toxicity are unknown, aberrant SOD1 protein oligomerization has been strongly implicated in disease causation (7, 8). Several recent publications have presented compelling evidence that abnormal disulfide cross-linking of ALS-mutant SOD1 plays a role in this oligomerization, and disulfide-linked SOD1 multimers have been detected in neural tissues of SOD1-ALS transgenic mice that are presumed to be components of higher-molecular-weight species or intermediates in their formation (7, 9–11).

Wild-type (WT) human SOD1 is an exceptionally stable protein in its holo form and, although some of the ALS-mutant SOD1 proteins are severely destabilized by their mutations, others largely retain the stability of WT SOD1 (4). In the fully demetallated (apo) states, some of the ALS-mutant SOD1 proteins actually are more stable than apo WT SOD1 (4). WT human SOD1 contains four cysteines, Cys-6, Cys-57, Cys-111, and Cys-146. In the mature protein, a disulfide bond, a structural feature that is unusual for an intracellular protein located in the highly reducing environment of the cell, links Cys-57 and Cys-146.

Abnormal protein inclusions have been observed in neural tissue of most of the SOD1-ALS transgenic mouse lines, and it has been shown in several cases that these inclusions are amyloid-like, based on their reactivity with fluorescent dyes used to visualize the amyloid inclusions in brains of Alzheimer's patients (12). A few of the isolated ALS-mutant SOD1 proteins have been induced to oligomerize in vitro to form amyloid-like aggregates, but the conditions used in each case were far from physiological, i.e., either very low pH (13) or extensive metal-catalyzed oxidation (14), suggesting that oligomerization occurred only if the SOD1 protein was substantially damaged or unfolded.

The current study was undertaken to determine (i) the propensity for oligomerization of disulfide-intact, WT human SOD1 under the relatively mild conditions likely to be encountered by the protein in vivo and (ii) the role of metallation in hindering or promoting such oligomerization. We report that WT human SOD1 apoprotein, with its intrasubunit disulfide bonds intact, forms soluble thioflavin T (ThT)-positive, high-molecular-weight oligomeric assemblies upon incubation in solution at conditions very close to physiological, i.e., at pH 7.0, 37°C, and at typical cellular concentrations; that these assemblies are remarkably stable, persisting in the soluble state for months; that they are linked by noncovalent protein–protein interactions and further stabilized by intermolecular disulfide-bond cross-links involving Cys-6 and Cys-111; and that metallation with copper and zinc, or even with zinc alone, totally suppresses this oligomerization. These results indicate that metal-free WT human SOD1 protein is prone to aggregation under relatively mild conditions, even when the intrasubunit disulfide bond is intact, suggesting that the gain of toxic function of SOD1 in FALS, and possibly even in sporadic ALS (SALS), may be related to the inability of this protein to achieve or to maintain the metallated state.

Results

SOD1 has been estimated to be present in cells at concentrations of ≈100 μM (15–17). To mimic as closely as possible the conditions experienced by the protein in vivo, we prepared 100 μM protein solutions at pH 7.0 and incubated them at 37°C. Changes in secondary structure were monitored by CD spectroscopy, and the degree of oligomerization state was monitored through gel-filtration chromatography and fluorescence of ThT, a benzothiazole dye that exhibits increases in fluorescence intensity upon binding to amyloid fibers (18, 19). Solutions of fully metallated Cu₂Zn₂SOD1 were unchanged after incubation at 37°C for more than 1 month, and the same was true for the copper-free derivative, E₂Zn₂SOD1 (E denotes empty). In contrast, the behavior of the completely metal-depleted protein was dramatically different. Incubated in the presence of ThT, it showed a progressive increase in fluorescence, with a fast process in the first 230 h of incubation, which slowed down after 1 month and then reached a plateau in the following 4 months (Fig. 1). No turbidity was detected during the entire time period.

Fig. 1.

Formation of ThT-binding structures occurs concomitantly with disappearance of free cysteines when apo SOD1 is incubated at 37°C. (A) Left axis: Fluorescence caused by ThT binding to SOD1 [presented as arbitrary units (a.u.)] for fully metallated SOD1 (●), apo SOD1 (▲), copper-free SOD1 (○), apo AS SOD1 (□), and fully metallated AS SOD1 (■) during the incubation of the samples at 37°C. Right axis: Change in fluorescence of AMS bound to the free cysteines of the apo SOD1 WT sample (△) during its incubation at 37°C. (B) ThT-binding fluorescence of the apo SOD1 protein over 7 months of incubation.

The detection of ThT-binding fluorescence is a clear indication of the formation of oligomeric assemblies, and it raises the possibility that these assemblies may share structural features with amyloid (18). The rate of ThT fluorescence enhancement depended strongly on temperature: the higher the temperature, the faster the rate of aggregation. Different pH conditions also affected the reaction: all other things being equal, the rate doubled at pH 5 relative to pH 7 (data not shown). This increase in rate may be attributable to the diminished repulsive negative charge as the solution pH approaches the pI of the protein (4.8).

After incubation at 37°C for 5 days or more, apo SOD1 samples at pH 7 showed a heterogeneous population of states for SOD1, ranging from dimers to a broad band of high-molecular-weight species, as shown by gel-filtration chromatography. Analysis of the fractions by multiangle light-scattering [supporting information (SI) Fig. 5] showed that the broad high-molecular-weight band consisted of a mixture of species having an average molecular weight that increased with the incubation time (SI Table 1). After 7 months of incubation, the molecular mass ranged between 430 and 750 kDa, with an average value of ≈650 kDa (≈25–45 SOD1 monomers). CD analyses of the high-molecular-weight fraction indicate that the oligomeric assemblies have a significantly higher percentage (58%) of β secondary structure than the freshly prepared apo sample (39%) (Fig. 2).

Fig. 2.

CD analysis. CD spectra of the fully metallated WT SOD1 protein (——), the apo WT SOD1 (– – –), and the aggregated species eluted from the chromatographic column of the incubated apo WT SOD1 (· · · · · ·). The higher percentage of β secondary structure in the apo WT SOD1 sample after incubation is evidenced by the increased contribution to the overall CD spectrum of the negative band at 218 nm, which is characteristic of β-sheet structure.

The oligomeric material itself, although clearly soluble, did not ionize in electrospray-ionization mass spectrometry (ESI-MS) under denaturing, nonreducing conditions (Fig. 3 and SI Fig. 6). However, covalent SOD1 dimers and trimers were detected, which was suggestive of covalent intermediates formed in the initial stages of aggregation (Fig. 3). Addition of the reducing agent DTT to all of the experimental samples and to the controls, followed by analysis under the same conditions, resulted in the appearance of a strong signal for monomeric SOD1 and the disappearance of the signals for any higher-order species, suggesting that intermolecular disulfide bonds are a structural property of the aggregates (SI Fig. 6). This conclusion is supported further by the observations that incubation of apo WT SOD1 in the presence of a reducing agent, either tris(2-carboxyethyl)phosphine hydrochloride (TCEP) or DTT, resulted in no enhanced fluorescence in the presence of ThT (SI Fig. 7A) and only the monomeric state was observed through gel-filtration chromatography (SI Fig. 7B), as is expected for disulfide-reduced apo SOD1 (20).

Fig. 3.

Size-exclusion chromatography and mass spectrometric detection of covalent oligomers of WT SOD1 during aggregation. Nondenaturing size-exclusion chromatography was used to analyze a sample of SOD1 apo protein after a 5-day incubation at 37°C. The sample exhibits a broad peak centered ≈34 min, corresponding to a molecular mass of ≈400 kDa (labeled O, oligomer), with geometry suggestive of a heterogeneous range of species, some as large as 1 MDa. The void volume is labeled V. Control SOD1 is dimeric (42 min, chromatogram not shown), and a population is seen in the 5-day sample (labeled Dn, noncovalent dimer). The 5-day sample also shows evidence of intermediate absorbance peaks at 39 min (labeled T, trimer) and 40 min (labeled Dc, covalent dimer) that might be intermediates in the oligomerization process. Fractions (O, T, Dc, and Dn) subsequently were analyzed by ESI-MS under denaturing conditions before or after reduction (SI Fig. 6 and Inset). (Inset) The ESI mass spectrum recorded under nonreducing denaturing conditions showing a covalent trimeric form of SOD1 from the 39-min fraction above. Shown is the covalent profile of the molecule after zero-charge deconvolution with monomeric and trimeric forms of the protein. Reduction of all four fractions (O, T, Dc, and Dn) results in detection of monomeric species only (see SI Fig. 6 for the complete experiment).

Free-cysteine content in apo SOD1 can be monitored by changes in 4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid (AMS) fluorescence on its modification of free thiols. Freshly prepared SOD1 has two free cysteines per monomer. The number of free thiols was observed to decrease simultaneously with protein aggregation (monitored by ThT fluorescence) (Fig. 1), with the two processes following the same time course: they are fast initially, and they essentially plateau after 9 to 10 days. When the two free cysteines were mutated to noncysteine residues, either as single mutations (data not shown) or as a double mutation (C6A/C111S SOD1, hereafter called AS SOD1) (Fig. 1), no aggregate formation was observed during incubation; only dimeric species were detected, even after 7 months of incubation. This behavior is a demonstration of the involvement of both the two free cysteines, Cys-6 and Cys-111, in the mechanism of aggregation, through formation of disulfide bonds. The data on these SOD1 mutants further indicate that the disulfide cysteines, Cys-57 and Cys-146, need not be involved in the initiation of oligomer formation.

The involvement of Cys-6 and Cys-111 in promoting oligomerization of SOD1 also was evidenced in the expression behavior of WT and mutant SOD1 in Escherichia coli, where SOD1 is produced mainly in the demetallated form (SI Table 2). WT SOD1 and a large number of ALS-related SOD1 mutant proteins containing both Cys-6 and Cys-111 always were extracted in the insoluble fraction unless high concentrations of reducing agent were added. By contrast, when Cys-6 and/or Cys-111 were mutated, SOD1 was found in the soluble fraction after cell osmotic shock (21), suggesting that demetallated WT and ALS-related SOD1 mutants form disulfide-linked oligomers also in vivo, but only when both Cys-6 and Cys-111 are present.

When the chaotropic agent guanidine hydrochloride (Gdn·HCl), which disrupts hydrogen-bond networks that stabilize protein secondary and tertiary structures, was added to a solution of soluble oligomerized apo WT SOD1, the ThT-binding fluorescence was quenched in 15 min, whereas gel filtration of the resulting solution showed that high-molecular-weight species still were present. We attribute the loss of ThT-binding ability to disruption of the tertiary structure of the assemblies; we also attribute the fact that the SOD1 oligomeric state persists to covalent SS bonds between cysteines of the SOD1 subunits, similarly to what was observed in ESI-MS analysis under denaturing conditions (see above and SI Fig. 6).

Discussion

Human SOD1 is characterized by an unusual combination of structural features: a strong homodimeric quaternary structure; an eight-stranded β-barrel motif in each subunit; a strong intrasubunit disulfide bond between a highly conserved pair of cysteines, namely Cys-57 and Cys-146; the presence of metal ions (copper and zinc) with specific structural and enzymatic roles; and two free cysteines per subunit (Cys-6 and Cys-111). Maturation of the protein in vivo, to attain the correctly folded quaternary structure and enzymatic activity, requires formation of the disulfide bond, dimerization, and acquisition of copper and zinc ions. The lack of one or more of these features has a major impact on the properties of SOD1. For example, the apoprotein is dimeric when the disulfide bond is intact, but it is monomeric when the disulfide bond is reduced. By contrast, the disulfide-reduced SOD1 protein is dimeric when zinc ions are bound (22). Our results, described here, demonstrate that the dimeric SOD1 apoprotein, even with its intrasubunit disulfide bonds intact, has a marked tendency to form soluble high-molecular-weight oligomers under conditions very close to physiological. This oligomerization occurs in vitro as well as in vivo in E. coli, provided Cys-6 and Cys-111 are present. Metallated forms of the SOD1 protein do not oligomerize under the same conditions.

Although the absence of metal ions in the SOD1 apoprotein is clearly a major factor determining the propensity of the protein to oligomerize, formation of intermolecular disulfide bonds also plays an important role, as demonstrated by the absence of aggregation of the apoprotein under reducing conditions. Interestingly, disulfide-bond involvement in the formation of amyloid fibrils previously has been reported for other neurodegenerative diseases (23, 24). In SOD1, we identify Cys-6 and Cys-111 as responsible for this process because single point mutation of either one of them prevents aggregation. We therefore hypothesize a mechanism of aggregation in which the SOD1 apoprotein, which is known to be destabilized and highly flexible because of the lack of metal ions, can sample a broad range of conformations, some of which cause the two free cysteines to be exposed to the solvent. Indeed, it has been shown previously in an engineered SOD1 monomer, artificially stabilized through two mutations at the subunit–subunit interface, that the absence of metals, and particularly of the structural metal zinc, causes the protein to be quite mobile and that several loops, including loop IV and the electrostatic loop VII, are characterized by a molten globule state within which a quite broad range of conformations are sampled (25). This flexibility is expected to make the two free cysteines more solvent-accessible relative to the dimeric, metallated form. Once the free cysteines are solvent-accessible, high-molecular-weight oligomers readily may be assembled through disulfide-bond formation and noncovalent interactions between β-sheets (Fig. 4). It is not clear at this stage whether the driving force for oligomerization is primarily noncovalent protein–protein interactions, which then are stabilized further by covalent disulfide cross-linking, or the opposite. The binding of ThT to the oligomeric assemblies and their increased percentage of β secondary structure are consistent with the possibility that these assemblies are held together by hydrogen bonds and electrostatic interactions similar to those occurring in amyloid fibrils. In the absence of additional structural information, however, this conclusion remains speculative because ThT binding is not totally specific to amyloid (26). It also should be noted that this study supports a critical role for both Cys-6 and Cys-111 in the initiation of the aggregation process and intermolecular disulfide formation and does not exclude the possibility of subsequent involvement of Cys-57 and Cys-146 via disulfide exchange reactions as the aggregation proceeds.

Fig. 4.

Proposed mechanism for SOD1 aggregation. Possible mechanism for in vivo formation of soluble oligomers that occurs when apo WT SOD1 protein is kept close to physiological conditions (37°C, 100 μM, and pH 7) for an extended period. The dark gray shapes represent free cysteines (Cys-6 and Cys-111), and the black triplets of parallel bars represent amyloid-like arrangements of hydrogen bonds. In the absence of metal ions, SOD1 proteins form abnormal disulfide cross-links and noncovalent associations with other SOD1 monomers or dimers.

The key factors in aggregation appear to be lack of metals and the presence of the free cysteines, factors that may be exacerbated in ALS mutant proteins. A large number of ALS-related SOD1 mutants have a decreased metal affinity and, indeed, some fail to be metal-reconstituted even in vitro (27). Also for WT, the protein is produced in an immature form that then needs several steps to reach the final active form. Among the various posttranslational modifications SOD1 must undergo, metallation is a complex process that involves several steps and interactions with other proteins. Metal uptake might be altered by misregulation of any of these steps by external factors. Therefore, when some SOD1 is present in the immature, demetallated state, aggregation could start, giving rise to amyloid-like soluble oligomers. We also must take into account the fact that SOD1 is present in both cytoplasm and mitochondria, and that its metal ions are acquired independently in the two cell compartments. These two cell compartments have quite different redox properties that further could modulate aggregation in vivo (28). This finding also is consistent with a recent suggestion of Wang et al. (7): they report that non-native intermolecular disulfide bonds help to stabilize the aggregates in vivo and that the redox state of the cell may play a role in aggregation.

The oligomers observed in this study appear to us to be likely candidates for the toxic species that cause SOD1-linked ALS. The mechanism discussed here for SOD1 aggregation is quite general because it does not depend on a specific mutation. It provides a unified picture of SOD1 oligomerization, as well as an in vitro system in which to study this process.

Materials and Methods

Sample Preparation.

WT SOD1, expressed in yeast, was purified in the fully metallated form (29). N^α-acetylation (alanine), as expected for human SOD1, was confirmed by mass spectrometric analysis. WT SOD1 also was expressed in the E. coli BL21(DE3) strain grown in LB medium. The protein was isolated by osmotic shock in a 20 mM Tris/5 mM DTT buffer at pH 8. After incubation for 30 min at 37°C, followed by centrifugation at 165,000 × g for 20 min, the protein was purified from the supernatant following a reported procedure (30) modified by the addition of 1 mM DTT to each chromatographic buffer. The protein obtained with this procedure contained substoichiometric amounts of the metal ions (SI Table 2). Mutations were performed with a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA).

The demetallated (apo) protein was prepared according to previously published protocols (31). Zinc reconstitution was carried out as described in ref. 32. Metal content of the various forms of SOD1 was checked by inductively coupled plasma mass spectrometry (ICP-MS) using a Thermo Jarrell Ash (Franklin, MA) Atomscan 25 sequential inductively coupled spectrometer (SI Table 3).

Spectroscopic Characterization.

Protein samples were 100 μM in SOD1 concentration (as dimer) in 20 mM phosphate buffer at pH 7. The protein was incubated at 37°C to mimic physiological conditions. Optical and fluorescence spectroscopies, coupled with gel-filtration chromatography, were used to monitor the formation of oligomeric species at these sample conditions.

Far-UV CD spectra (190–250 nm) of SOD1 were recorded on a JASCO (Cremella, Italy) J-810 spectropolarimeter. A cell with a path length of 1 mm was used for the measurement, and the parameters were set as follows: bandwidth, 2 nm; step resolution, 1 nm; scan speed, 20 nm/min; and response time, 2 s. Each spectrum was obtained as the average of four scans. The protein concentration typically was ≈8–10 μM. Before the calculation of the mean residue molar ellipticity, all of the spectra were corrected by subtracting the contributions from the buffer. Spectra then were smoothed by using adjacent averaging or fast Fourier transform filtering. Quantitative estimations of the secondary structure contents were made by using the DICROPROT software package (33).

Fluorescence was followed with ThT, a dye that binds to amyloid-like structures (18). Free ThT has excitation and emission maxima at 350 and 450 nm, respectively. However, upon binding to amyloid-like oligomers, the excitation and emission wavelengths change to 450 and 485 nm, respectively. Fifty-four-microliter aliquots of sample were added to 646 μl of a 215 μM ThT solution in a 20 mM phosphate buffer at pH 7. The solution fluorescence emission was measured, over time of incubation, with a Cary 50 Eclipse spectrophotometer (Varian, Palo Alto, CA) supplied with a single-cell Peltier thermostated cell holder regulated at 37°C. The background fluorescence spectrum of the buffer was subtracted. The excitation wavelength was 446 nm (bandwidth, 10 nm), and the emission was recorded at 480 nm (bandwidth, 10 nm). Fluorescence intensity at 483 nm was plotted against the time of incubation.

Turbidity was measured at 400 nm to detect possible formation of insoluble precipitate. Solution turbidity was measured as apparent absorbance at 400 nm by using a Cary UV-visible spectrophotometer. Experiments were performed by diluting 120 μl of the incubation SOD1 stock into 280 μl of 20 mM phosphate buffer at pH 7. A 1-cm quartz cuvette was used, and the instrumental detection limit was 0.001 at 400 nm.

ESI-MS.

SOD1 samples were analyzed by nondenaturing size-exclusion chromatography (G4000 SW; Tosoh Biosciences, Tokyo, Japan; 20 mM potassium phosphate, pH 7.0; 250 μl/min; 24°C). Fractions (50–100 μl) were dried in a Savant SpeedVac and redissolved in 10 μl of water. For samples to be reduced, 0.33 M DTT was added (5 μl of 1 M solution in water), and the samples were incubated at room temperature for 15 min. For ESI-MS under denaturing conditions, the salt-containing samples were diluted with formic acid (90 μl of 90% formic acid; Fisher Scientific, Fair Lawn, NJ; ACS reagent grade), vortex-mixed for 1 min, and immediately injected onto an HPLC system for online liquid-chromatography mass spectrometry (LC-MS). A gel-filtration column [SW2000 XL; Tosoh Biosciences; chloroform/methanol/1% formic acid in water, 4/4/1 (vol/vol/vol); 250 μl/min; 40°C] was used to desalt the sample and provide highly denaturing conditions (7, 34). Column eluent was directed to an ion-spray source of a triple-quadrupole mass spectrometer (API III+; PE Sciex/Applied Biosystems, Foster City, CA) tuned and calibrated as described in ref. 35, yielding mass accuracy of 0.01%.

Monitoring SOD1 Aggregation by Gel Filtration and Light-Scattering.

One-hundred-microliter aliquots of the incubated protein at 37°C periodically were taken and analyzed by gel filtration on a Superdex 75 HR 10/30 column (Amersham Biosciences) at room temperature (Fig. 5B). The column was preequilibrated with 20 mM potassium phosphate (pH 7.0), and the flow rate was 0.6 ml/min. The chromatograms, which monitored the species formed during incubation, were obtained by monitoring the absorbance at 280 nm. The Superdex 75 HR 10/30 column was connected to a light-scattering spectrometer. The online multiangle light-scattering (MALS) detector (DAWN EOS; Wyatt Technology, Santa Barbara, CA) and differential refractive index (DRI) detector (Optilab DRI; Wyatt Technology) setup was used to measure the light scattered as a function of angle and absolute protein concentration of fractions eluting from the size-exclusion chromatography column (SI Fig. 5A). The Zimm approximation was used in Astra software (Wyatt Technology) to estimate molar mass. Data were fit by using a first-order polynomial. The analysis was performed for each of the 100-μl aliquots periodically taken from the incubation batch so as to monitor the increase in molecular weight of the soluble species formed during aggregation (SI Table 1).

Free-Thiol Quantification.

Estimation of free thiols during aggregation was performed by AMS modification. AMS has high water solubility and readily is conjugated to free thiols. An 800-μl sample of the reaction mixture in 20 mM phosphate buffer (pH 7) containing 7.5 μM protein, 250 μM AMS, and 1% SDS was incubated for 30 min at 37°C to complete the reaction. The excess AMS was removed by dialysis. Reacted aliquots were taken from an apo WT SOD1 sample (100 μM) along the incubation at 37°C. AMS is a stilbene derivate that shows a typical UV absorption at ≈328 nm and emission maximum at 408 nm. Fluorescence measurements (excitation, 322 nm; emission, 406 nm; excitation/emission slits, 10) were performed to monitor the change in the number of free thiols during aggregation. The calibration curve for the free-cysteine quantification was prepared by using freshly prepared apo SOD1 WT at different concentrations as a standard. MALDI analysis proved that there is a maximum of two free cysteines per apo SOD1 monomer. Aliquots taken from the incubation stock at the same time as those used for the AMS test were reacted with ThT, and the fluorescence of the solution was measured to monitor sample aggregation.

Abbreviations

ALS

amyotrophic lateral sclerosis

SOD1

copper–zinc superoxide dismutase

FALS

familial ALS

ThT

thioflavin T

ESI-MS

electrospray-ionization mass spectrometry

AMS

4-acetamido-4′-maleimidylstilbene-2,2′-disulfonic acid.