4-Phenylbutyric Acid (4-PBA) Derivatives Prevent SOD1 Amyloid Aggregation In Vitro with No Effect on Disease Progression in SOD1-ALS Mice

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by the degeneration of motor neurons. Mutations in the superoxide dismutase (SOD1) gene, causing protein misfolding and aggregation, were suggested as the pathogenic mechanisms involved in familial ALS cases. In the present study, we investigated the potential therapeutic effect of C4 and C5, two derivatives of the chemical chaperone 4-phenylbutyric acid (4-PBA). By combining in vivo and in vitro techniques, we show that, although C4 and C5 successfully inhibited amyloid aggregation of recombinant mutant SOD1 in a dose-dependent manner, they failed to suppress the accumulation of misfolded SOD1. Moreover, C4 or C5 daily injections to SOD1G93A mice following onset had no effect on either the accumulation of misfolded SOD1 or the neuroinflammatory response in the spinal cord and, consequently, failed to extend the survival of SOD1G93A mice or to improve their motor symptoms. Finally, pharmacokinetic (PK) studies demonstrated that high concentrations of C4 and C5 reached the brain and spinal cord but only for a short period of time. Thus, our findings suggest that use of such chemical chaperones for ALS drug development may need to be optimized for more effective results.


Introduction
Amyotrophic lateral sclerosis (ALS) is a rapidly progressive and fatal neurodegenerative disease characterized by the degeneration of the upper and lower motor neurons (MNs) in the brain and spinal cord [1]. About 10% of ALS cases are familial, most of which are inherited in a dominant manner [2], and approximately 20% of these familial cases are attributed to mutations in the superoxide dismutase 1 (SOD1) gene [3]. To date, more than 180 different SOD1 variants have been associated with ALS [4].
In the current study, we investigated the potential therapeutic effect of C4 and C5 in the mutant SOD1 G93A mouse model of ALS, combining both in vitro and in vivo studies. We hypothesized that C4 and C5 chaperone activity could reduce the levels of mutant SOD1 misfolding and/or aggregation, which may result in improved motor function as well as the prolonged survival rate of the transgenic mice. Although C4 and C5 success-  19 F-PBA, C4, 19 F-C4, and C5.
In the current study, we investigated the potential therapeutic effect of C4 and C5 in the mutant SOD1 G93A mouse model of ALS, combining both in vitro and in vivo studies. We hypothesized that C4 and C5 chaperone activity could reduce the levels of mutant SOD1 misfolding and/or aggregation, which may result in improved motor function as well as the prolonged survival rate of the transgenic mice. Although C4 and C5 successfully suppressed recombinant mutant SOD1 G93A amyloid aggregation levels in vitro, C4 or C5 daily injections to adult mice commencing after disease onset did not significantly affect disease progression or survival of the treated SOD1 G93A mice.

C4 and C5 Chemical Chaperones Strongly Inhibited the Formation of SOD1 Amyloid Aggregation In Vitro
C4 and C5 chemical chaperones are derivatives of 4-PBA ( Figure 1) and were shown previously to inhibit protein aggregation both in vitro and in vivo in models of autism spectrum disorder (ASD) and C9orf72-related ALS [33,34]. In order to determine the effect of C4 and C5 on SOD1 G93A aggregation, we purified and incubated recombinant SOD1 G93A protein under aggregation-promoting conditions (i.e., in the presence of a reducing agent and a metal chelator), with increasing concentrations of C4 (1-20 mM) ( Figure 2A) and C5 (0.5-20 mM) ( Figure 2C). We measured the kinetics of amyloid aggregation by monitoring the fluorescence intensity signal of thioflavin-T (Th-T) at 485 nm over a period of about 70 h. Incubating SOD1 G93A alone resulted in an exponential increase in Th-T fluorescence intensity, indicating formation of amyloid fibrillar aggregates (Figure 2A,C). This was confirmed by transmission electron microscopy (TEM) imaging performed at the end of the incubation period (after 72 h), which revealed a fibrous aggregated formation ( Figure 2B,D). In contrast, incubating SOD1 G93A in the presence of C4 (Figure 2A,B) or C5 ( Figure 2C,D) reduced the fluorescent signal in a dose-dependent manner, suppressed the formation of fibrillar aggregates, and switched the aggregation pattern to an amorphous disordered one ( Figure 2B,D). The fluorescent signal measured during the incubation of C4 or C5 alone without recombinant SOD1 was undetectable in the Th-T analysis, indicating that they did not produce aggregates whatsoever (Figure 2A,C).

Daily Injection of C4 or C5 Chemical Chaperones Following Disease Onset Had No Effect on Disease Progression and Survival of Mutant SOD1 G93A Mice
To test the potential therapeutic effect of C4 and C5 in a mouse model of ALS, we used the most studied mutant SOD1 G93A model. We alleged that C5, as a α-isopropyl derivative of 4-PBA, would pass the BBB, as is the case with 4-PBA [35]. In order to determine whether C4 is indeed able to cross the BBB, its fluorinated derivative was synthesized as described [33]. Wild type mice were intraperitoneally (IP) injected with 50 mg of 19 Flabeled C4 in acetate buffer, and 30 min later the brains were harvested and analyzed by nuclear magnetic resonance ( 19 F-NMR). Two clear 19 F-NMR signals were observed in injected but not in the noninjected mouse brain tissue ( Figure S1A,B). Two 19 F-NMR peaks in −118.03 ppm and −118.26 ppm were obtained: one related to the 19 F-labeled 4-phenylbutyric acid ( Figure 1) and the other to 19 F-C4. We assume that the compound passed the BBB and was partially hydrolyzed ( Figure S1C). Then, based on this positive penetration outcome, transgenic SOD1 G93A mice (males and females) were IP injected daily, starting at postnatal day 104 (following disease onset), with either C4 (100 mg/kg, in acetate buffer), C5 (25 mg/kg, in PBS), or 200 µL acetate buffer (pH 5.6) as a control. Disease progression was monitored until the end stage by the neurological scoring system NeuroScore (NS), the inverted screen test, and the routinely used mice weight loss rate ( Figure 3A). C4 or C5 daily injections starting after disease onset did not significantly extend SOD1 G93A mice survival ( Figure 3B; Control: 171 d; C4: 176 d; C5: 165 d). Moreover, the treatment did not significantly affect either the mice clinical neuroscore ( Figure 3E) or their weight loss rate ( Figure 3D) and did not improve the mice motor function as analyzed by the inverted screen test ( Figure 3C). Since gender differences exist in this transgenic model,

Daily Injection of C4 or C5 Chemical Chaperones Following Disease Onset Had No Effect on Disease Progression and Survival of Mutant SOD1 G93A Mice
To test the potential therapeutic effect of C4 and C5 in a mouse model of ALS, we used the most studied mutant SOD1 G93A model. We alleged that C5, as a α-isopropyl derivative of 4-PBA, would pass the BBB, as is the case with 4-PBA [35]. In order to determine whether C4 is indeed able to cross the BBB, its fluorinated derivative was synthesized as C5 daily injections starting after disease onset did not significantly extend SOD1 G93A mice survival ( Figure 3B; Control: 171 d; C4: 176 d; C5: 165 d). Moreover, the treatment did not significantly affect either the mice clinical neuroscore ( Figure 3E) or their weight loss rate ( Figure 3D) and did not improve the mice motor function as analyzed by the inverted screen test ( Figure 3C). Since gender differences exist in this transgenic model, the males and females were also analyzed separately. Confirming our previous results, we were not able to detect any statistical effect on survival, neither in males nor in females ( Figure S2).   In order to determine whether the 4-PBA derivatives have any effect on motor neuron survival, immunoblot against ChAT, a motor neuron marker was performed. ChAT expression levels were reduced in mutant SOD1 G93A spinal cord compared to littermate non-transgenic controls ( Figure 4A), with daily injections of C4 or C5 starting after onset having no effect on motor neuron survival ( Figure 4A,C). Moreover, ALS pathology is accompanied by non-cell autonomous processes [36,37]. Specifically, studies report widespread activation as well as impaired function of astrocytes [9,38] and microglia in both ALS patients and mouse models [39]. Indeed, immunoblotting of spinal cords of mutant SOD1 G93A mice showed a strong increase in astrocyte (GFAP) and microglia (Iba1) activation ( Figure 4B,D,E) in their spinal cord with no difference in this activation pattern ( Figure 4B,D,E) following C4 or C5 administration compared with untreated mice. These results were confirmed by immunofluorescence of untreated and treated mutant SOD1 G93A spinal cord sections ( Figure S3).
In order to determine whether the 4-PBA derivatives have any effect on motor neuron survival, immunoblot against ChAT, a motor neuron marker was performed. ChAT expression levels were reduced in mutant SOD1 G93A spinal cord compared to littermate non-transgenic controls ( Figure 4A), with daily injections of C4 or C5 starting after onset having no effect on motor neuron survival ( Figure 4A,C). Moreover, ALS pathology is accompanied by non-cell autonomous processes [36,37]. Specifically, studies report widespread activation as well as impaired function of astrocytes [9,38] and microglia in both ALS patients and mouse models [39]. Indeed, immunoblotting of spinal cords of mutant SOD1 G93A mice showed a strong increase in astrocyte (GFAP) and microglia (Iba1) activation ( Figure 4B,D,E) in their spinal cord with no difference in this activation pattern (Figure 4B,D,E) following C4 or C5 administration compared with untreated mice. These results were confirmed by immunofluorescence of untreated and treated mutant SOD1 G93A spinal cord sections ( Figure S3).  (B) Quantification of ChAT level intensity in non-transgenic mice (control) and C4-and C5-treated and untreated SOD1 G93A mice. (C-E) Neuroinflammatory response was evaluated by immunoblot of lysates from spinal cords of non-transgenic and SOD1 G93A mice with anti-GFAP and anti-Iba1 antibodies (C). (D) Quantification of activated astrocytes (GFAP antibody, D) and activated microglia (Iba1 antibody, (E)) in non-transgenic and SOD1 G93A mice untreated or treated with C4 or C5. Ctrlnon-transgenic mouse; NI-SOD1 G93A noninjected; B-SOD1 G93A injected with acetate buffer. Quantification analysis was performed with Student's t-test. Bars represent mean ± SEM. Data indicate one representative experiment out of 3-5 independent experiments. n.s, nonsignificant; * p < 0.05; ** p < 0.01.

Daily Injection of C4 or C5 Chemical Chaperones Starting after Disease Onset Failed to Reduce the Accumulation of Misfolded SOD1
Immunoprecipitation with a conformational antibody that specifically recognizes misfolded SOD1, B8H10 [40,41], was used to determine the accumulation of misfolded SOD1 in spinal cords of treated and untreated SOD1 G93A mice ( Figure 5A). Mutant SOD1 G93A mice injected with C4 or C5 had similar accumulation of misfolded SOD1 in the lumbar spinal cord ( Figure 5B,C) and in the brain ( Figure S4) compared with that of SOD1 G93A mice injected with acetate buffer as control (bound fraction). Supporting these findings, end-stage mice were perfused with 4% PFA, and frozen sections of spinal cord were immunostained by the free-floating technique with the same conformational antibody against misfolded SOD1, B8H10, to compare the accumulation of misfolded SOD1 in the spinal cords of treated and nontreated SOD1 G93A mice ( Figure  S5). Misfolded SOD1 accumulation levels were not significantly reduced in the spinal cord of C4-( Figure S5A) or C5 ( Figure S5B)-injected mice compared to nontreated SOD1 G93A Supporting these findings, end-stage mice were perfused with 4% PFA, and frozen sections of spinal cord were immunostained by the free-floating technique with the same conformational antibody against misfolded SOD1, B8H10, to compare the accumulation of misfolded SOD1 in the spinal cords of treated and nontreated SOD1 G93A mice ( Figure S5). Misfolded SOD1 accumulation levels were not significantly reduced in the spinal cord of C4-( Figure S5A) or C5 ( Figure S5B)-injected mice compared to nontreated SOD1 G93A mice. Moreover, we have tested the ability of C4 ( Figure 5D,E) and C5 ( Figure 5F,G) to reduce misfolded SOD1 accumulation of recombinant SOD1 G93A protein. C4 and C5 were incubated with recombinant SOD1 G93A at different concentrations (C4, 1-50 mM; C5, 0.25-12.5 mM). Immunoprecipitation produced by the B8H10 antibody showed no difference in the accumulation of misfolded SOD1 following incubation with C4 ( Figure 5D,E) or C5 ( Figure 5F,G).

Daily Injection of C4 or C5 Chemical Chaperones Following Disease Onset Failed to Reduce Total Aggregate Formation of Mutant SOD1
In order to determine whether C4 or C5 had any effect on total protein aggregation in our treated mice, we separated the soluble and insoluble fractions from the spinal cords of the treated and nontreated mutant SOD1 G93A mice ( Figure 6A). Our analysis revealed no significant difference in SOD1 aggregation levels in the C4-or C5-injected mice compared to the untreated SOD1 G93A mice ( Figure 6B,C). In addition, we transfected SH-SY5Y neuronal cells with SOD1 WT and SOD1 G93A plasmids, followed by incubation with C4 at increasing concentrations (10-200 µM). We separated the soluble and insoluble fractions 72 h post-transfection and compared SOD1 aggregation levels in the insoluble fraction. Immunoblotting revealed no difference in SOD1 G93A aggregation levels following incubation with C4 at the tested concentrations ( Figure 6D,E).

High Levels of C4 and C5 Reached the Brain and Spinal Cord for a Short Period of Time
Finally, in order to determine whether sufficient amounts of both tested compounds indeed reach the brain and spinal cord, the classical PK experiment (single dose injection) was conducted, as described in "Materials and Methods". The obtained t 1/2 of tested compounds showed that C4 t 1/2 and C5 t 1/2 in serum were approximately 48 min and 20 min, respectively ( Figure S6A,D). The maximal amount of C4 in the brain was detected after 30 min: 0.189 ± 0.02 mM and after 3 h the concentration of C4 in the brain had dropped by 10-fold ( Figure S6B). Moreover, in the spinal cord after 30 min, the compound reached its peak concentration: 0.338 ± 0.02 mM, and similarly to the brain, by 3 h C4 concentration decreased by 10-fold ( Figure S6C). C5 reached the highest concentration in the brain immediately following administration (approximately after 6 min): 0.175 ± 0.043 mM, and by 30 min, the amount of the compound in the brain was negligible ( Figure S6E). Moreover, in the spinal cord, the compound showed a similar pattern of behavior: 0.136 ± 0.024 mM in the beginning, and after 30 min, only a minimal concentration of the compound was detected (17-fold lower compared to the starting level) ( Figure S6F).

Discussion
In this study, we evaluated the potential therapeutic effect of the chemical chaperones C4 and C5, two 4-PBA derivatives which exhibited positive outcomes when tested in several models of protein aggregation diseases [33,34], on the mutant SOD1 G93A model of ALS. We found that C4 and C5 strongly inhibited the formation of mutant SOD1 G93A amyloid aggregates in a dose-dependent manner in vitro by changing the aggregation pattern from amyloid aggregation to an amorphous less toxic one [42], as examined by the Th-T assay and confirmed by TEM imaging. However, daily injections of C4 or C5 into a SOD1 G93A mouse model, starting after disease onset, failed to significantly affect disease progression or survival of treated mice.
Correlating with these in vivo findings, was the observation that the accumulation of soluble misfolded SOD1 was not reduced in the spinal cord and brain of C4-and C5injected mice. Moreover, SOD1 aggregation in end-stage spinal cords was only slightly (but not significantly) reduced by C4 and C5 daily injections, as revealed by a soluble-insoluble assay and confirmed by immunofluorescence staining. Finally, our treatment had no effect on the neuroinflammatory response in SOD1 G93A mice lumbar spinal cord.
A key question in the field of protein misfolding-related disorders is whether the toxic species is the soluble misfolded form of the protein or the formation of insoluble cytoplasmic inclusions. Specifically, the benefits and harmful processes associated with the formation of insoluble aggregates are still being investigated [43]. Traditionally, the presence of protein inclusions in neurodegenerative diseases has been related to the cell's failure to refold misfolded proteins by chaperone activity [44]. However, other findings highlight the possibility that inclusion formation is not necessarily pathological [45]. Amyloid fibril formation is an intrinsic property of proteins in general [46,47], and there is accumulating evidence that this property may serve as a protective response, essential for several biological activities [45,48]. For example, protein aggregation into insoluble deposits was reported as a protective mechanism against oxidative stress [44] to allow efficient cell cycle restart after stress [49]. Recently, SOD1 insoluble aggregate formation was suggested as a protective mechanism to reduce the amount of toxic SOD1 trimers [50,51]. In addition, it was shown that inclusion body formation can function as a coping response to toxic mutant huntingtin [52]. Moreover, it was suggested that the aggregation process itself is related to toxicity, and that a common mechanism of toxicity is involved in several aggregation-related disorders [43]. In light of these findings, we raise the question of whether reducing the solubility of mutant proteins through aggregation may be part of the cell's protective strategy, and whether the soluble misfolded SOD1 form is indeed the most toxic. Since the inhibition of SOD1 amyloid aggregation observed in vitro could not be replicated in our treated mice, we cannot make any definitive claim on this issue.
In an attempt to explain the lack of in vivo effect of the compounds on SOD1 G93A pathogenesis, it is worth considering the possibility that C4 or C5 injection starting at an earlier time point, prior to disease onset, might have been more beneficial. The 4-PBA was already tested in the SOD1 G93A ALS mouse model [53][54][55]. These studies showed that treating mice prior to manifestation of clinical symptoms resulted in extended survival rate, in addition to improved body weight loss rate as well as improved motor function [53,55]. Supporting this notion is our in vitro Th-T assay, where C4 and C5 suppressed SOD1 amyloid aggregation before its formation. Starting with C4 or C5 injections at the presymptomatic stage would likely suppress the formation of amyloid aggregates and thus potentially affect disease progression.
We hypothesized that, when compared to 4-PBA, C4 and C5 may be effective in inhibition of aggregation at lower, more therapeutically relevant, concentrations. However, our Th-T results, accompanied by TEM imaging, revealed a complete suppression of SOD1 G93A amyloid aggregates during incubation with relatively high concentrations of the compounds, suggesting that the synthesis of these new 4-PBA derivatives failed to achieve therapeutic outcomes at low dosages, as we had expected. Supporting these findings was a soluble-insoluble assay, where the incubation of SOD1 G93A -expressing human neuronal cells with lower concentrations of C4 failed to reduce SOD1 aggregation levels.
Furthermore, previous findings from SOD1 G93A mice show that an effective therapeutic dosage was achieved by the administration of 200-400 mg/kg per day of 4-PBA [53][54][55]. Here, we aimed to test whether the new synthesized 4-PBA derivatives would present positive outcomes when administrating significantly lower dosages (100 mg/kg per day for C4 and 25 mg/kg per day for C5). Important to mention is that, even such high doses of both compounds that resulted in very high concentrations in the brain and spinal cord (mM range), although for a very short time, did not provide the expected outcome on disease progression and survival of SOD1 G93A . Thus, increasing C4 and C5 stability might have resulted in slower disease progression.
In addition, we tested the potential therapeutic effect of C4 and C5 on the early onset and very aggressive SOD1 G93A model of ALS [56]. Other SOD1 ALS-related mutations might present different outcomes. Likewise, C4 or C5 may act as chaperones for other ALS-related protein aggregates, such as ALS models presenting cytosolic TDP-43 inclusions, which may also involve other pathological mechanisms.
Riluzole [57,58] and edaravone [57,59] are the only two drugs approved by the FDA as therapeutic agents in ALS to date; however, their mechanism of action is not fully understood. Riluzole, a glutamate antagonist, appears to reduce damage to MNs by averting excitotoxicity [60,61]. Although riluzole was approved in 1995 based on clinical trials [58,62], more recent studies testing its therapeutic effect in currently available ALS-relevant mouse models revealed that it failed to improve the lifespan of the treated mice [63]. Edaravone is an antioxidant with a free radical-scavenging activity which successfully reduced oxidative stress and improved the motor performance of SOD1 G93A mice [59,[64][65][66]. Both riluzole and edaravone are mildly effective, prolonging some patients' survival by up to 2-3 months, and have a beneficial effect only when taken at the first few months after diagnosis [67]. Combining 4-PBA treatment with riluzole [55] or with the antioxidant AEOL 10150 [54] was reported to improve disease outcome and extended mice survival rate [54,55]. Thus, combining the chaperone activity of C4 or C5 with other known therapeutic strategies might be more effective in eliminating mutant SOD1 toxicity.
In conclusion, our findings suggest that the use of such chemical chaperones alone may not be realistic due to their high and barely tolerable active doses and bad PK parameters. Thus, the chemical chaperone-based strategy for ALS drug development may need to be optimized for more effective results.

Animals and Injection Protocol
Altogether, 48 B6 background (C57BL/6J TgN-SOD1-G93A-Gur; SOD1 G93A ) female and male mice were used for the experiments. The treatment protocol was approved by the Animal Care and Use Committee of Ben-Gurion University of the Negev, as required by Israeli legislation. Mice received 200 µL final volume of intraperitoneal (IP) injections every 24 h. C4-treated mice received 100 mg/kg/day diluted in acetate buffer, pH 5.6. C5-treated mice received 25 mg/kg/day diluted in PBS, pH 7.3. The control group included noninjected mice and mice injected only with acetate buffer.
All behavioral tests and body weight measurements were conducted twice a week. The inverted screen test [68] and the neurological score [69] were assessed as described previously.

19 F-NMR
A solution of compound 4 (labeled by the fluorine atom: 19 F-C4), Figure S1 (50 mg, final concertation: 0.34 M), was prepared in acetate buffer, pH = 5.6. The acidic pH of the formulation was used to ensure the solubility of the compound as a quaternary ammonium salt in the blood. Thirty minutes post-IP administration of the compound, mice were euthanized, the brain was reperfused by saline, isolated, and homogenized. The entire brain homogenate was mixed with D 2 O and the 19 F-NMR analysis was conducted as described [34].

Transmission Electron Microscopy (TEM)
TEM imaging was performed by the Nano-Fabrication Center team at Ben-Gurion University of the Negev, as described previously [21]. Briefly, at the end of the ThT aggregation assay (after~70 h), 2.5 µL samples were deposited on a carbon-coated copper 300 grid. After 1 min, the excess liquid was carefully blotted onto filter paper, which was then dried at ambient temperature for 1 min. Uranyl acetate (5 µL, 2%) was added to the grid, and after 1 min, the excess of the salt solution was carefully removed with a filter paper. The imaging was performed using a ThermoFisher Scientific (FEI, Waltham, MA, USA) Talos F200C transmission electron microscope operating at 200 kV. The images were taken with Ceta 16M CMOS camera at various magnifications (100-500 nm), depending on the size of the fibril aggregates. The visible features were sensitive to the electron bean exposure, indicating their organic origin.
Cells were transfected using TurboFect TM transfection reagent (Thermo, Waltham, Massachusetts, USA) according to manufacturer's protocol. Briefly, cells were seeded 0.8 × 10 5 in 2 mL media (DMEM, 10% FBS, 2 mM L-glutamine, 100 units/mL ampicillin and 0.1 mg/mL streptomycin) in a 60 mm petri dish. Keeping a ratio of DNA:TurboFect TM (1:2), 3 µg of DNA and 6 µL TurboFect TM were dissolved in 200 µL of DMEM, mixed by vortex, and incubated for 25 min at room temperature. The mixture was added dropwise to the preseeded cells and incubated for 48 h at 37 • C in a humidified, 5% CO 2 incubator.

Cell Lysis and Protein Extraction
Protocol was performed on ice. Briefly, media was removed, and cells were washed twice with 0.1 M PBS, and then lysed in 1 mL of ice-cold soluble buffer (0.1 M PBS, 1% Triton X, 5 mM EDTA, 10% Glycerol, 1 mM PMSF, 0.5% PI) with 20 min incubation at 4 • C. Cells were then detached using a cell scraper, collected into an Eppendorf, homogenized at 4000 RPM for 30 s, and centrifuged at 17,000× g for 30 min at 4 • C. The supernatant (soluble fraction) was collected and stored at −20 • C until use. The pellet was resuspended in 1 mL ice-cold soluble buffer (0.1 M PBS, 1% Triton X, 5 mM EDTA, 10% Glycerol) and centrifuged at 17,000× g at 4 • C for 30 min. Next, the pellet (insoluble fraction) was resolved using 400 µL of 8 M urea and sonication for 1 h at 4 • C. Protein concentration of soluble fractions was measured by the Bradford method using bovine serum albumin as standard, and the protein concentration of insoluble fractions was measured by the protein determination (BCA) kit (Cayman Chemical).

Tissue Harvesting and Protein Extraction
Brain/spinal cord (SC) tissues were dissected out, cut in half, and homogenized on ice in 3 volumes of ice-cold homogenization buffer (150 mM NaCl, 20 mM Tris-HCl pH 7.5, 1 mM PMSF, 1% triton, 1% PI (APExBio, Houston, TX, USA), 0.5% sodium deoxycholate, 0.1% SDS). Homogenates were centrifuged at 5000× g at 4 • C for 30 min, the supernatant (cytosolic fraction) was collected, and then stored at −80 • C until use. Protein concentration was measured by the Bradford method using bovine serum albumin as standard.

Immunoprecipitation (IP) Assay
Brain/SC tissue extracts (100 µg), or recombinant SOD1 G93A incubated with C4 (1-50 mM) or C5 (0.25-12.5 mM) for 1 h at 37 • C, were solubilized in IP buffer (0.5 M NaCl, 50 mM Tris (pH 7.4), 0.5% Nonidet P-40) and incubated at 4 • C overnight with B8H10 antibody (MediMabs, Montreal, Quebec), previously crosslinked to Dynabeads TM protein G (Thermo, Waltham, MA, USA) according to the manufacturer's instructions. The beads were magnetically isolated and, after crosslinking with the antibody, were washed three times with PBST (0.1 M PBS with 0.02% Tween20). After magnetic separation, unbound fractions were withdrawn for immunoblotting analysis. Beads were washed three times with IP buffer and the bound fractions were eluted by boiling for 5 min at 95 • C in X2 sample loading buffer.

Soluble-Insoluble Separation Assay
Spinal cord tissues were dissected out, cut in half, homogenized on ice in 200 µL of ice-cold soluble buffer (5 mM EDTA, 1 mM PMSF, 1% triton, 1% PI (APExBio, Houston, TX, USA), 0.1 M PBS), and incubated at 4 • C for 2 h while rotating. Homogenates were then centrifuged at 17,000× g at 4 • C for 30 min and the supernatant (soluble fraction) was collected and stored at −80 • C until use. The pellet was resuspended with 1 mL ice-cold soluble buffer (5 mM EDTA, 1% triton, 0.1 M PBS) and centrifuged at 17,000× g at 4 • C for 30 min. Next, the pellet (insoluble fraction) was resolved using 300 µL of 8 M urea and sonicated for 2 h at 4 • C. Protein concentration of soluble fractions was measured by the Bradford method using bovine serum albumin as standard, and protein concentration of insoluble fractions was measured by the protein determination (BCA) kit (Cayman Chemical).

Pharmacokinetic (PK) Study
A total of 26 mice (8-9 weeks old) were used in this study. The animals were randomly assigned to the treatment groups before the pharmacokinetic study. Six sampling time points (0, 0.5, 1, 3, 8, 24 h) were set for the experiment. Each of the time point treatment group included 3 animals. Test compounds: C4 (100 mg/kg) and C5 (25 mg/kg) were injected IP. Mice were sacrificed by cervical dislocation, and after that, the blood samples were collected by cardiac puncture, settled for 20 min, and then centrifuged for 10 min at 3000× g 4 • C degrees. Samples were snap-frozen and stored at −70 • C until subsequent analysis. An amount of 200 µL of acetonitrile was added to 50 µL of serum, and after centrifugation, the supernatant was used for mass spectroscopy analysis. Brains were harvested and lysed in 500 µL PBS using 18 G needle, and then 1 mL of acetonitrile was added. After the addition of the acetonitrile, the samples were homogenized again with the same needle. Spinal cords (SCs) were harvested and lysed in 250 µL PBS using 18 G needle, and then 500 µL of acetonitrile was added. After the addition of the acetonitrile, the samples were homogenized again with the needle. All procedures were conducted on ice. LC/MS from Agilent Technologies (Santa Clara, CA, USA) was used for PK analysis. Data were processed using mass L-ynX ver. 4.1 calculation and deconvolution software (Waters Corp., Milford, MA, USA). Spiking and calibration curve were generated using serum of the nontreated mice.

Immunofluorescence
Mice were anesthetized via inhalation of 1 mL isoflurane, followed by perfusion with 50 mL 0. and mounted on slides using Immu-Mount TM mounting solution (Thermo), dried at room temperature overnight, and stored at 4 • C until imaging. Images were acquired on a NIKON C2Plus laser unit dock to a Nikon Eclipse Ti unit of the confocal microscope by using 10× and 20× objectives and 60× oil immersion objective. Scanning settings were reused across the samples.

Statistical Analysis
Quantification of band intensity across experiments was done using Evolution-Capt Edge software (version 18,08, Vilber, Collegien, France). The data was transferred and statistically analyzed using OriginPro software (2021, OriginLab, Northampton, MA, USA). Values are reported throughout as mean ± SEM. After confirming a normal distribution by the Shapiro-Wilk normality test, a one-way ANOVA was performed to compare the database between the 3 experimental groups. For nonlinear data, a Kruskal-Wallis test was performed instead. Significance was set at a confidence level of 0.05.
ThT aggregation assay analysis was done using GraphPad Prism (La Jolla, CA, USA).