A Novel Acylaminoimidazole Derivative, WN1316, Alleviates Disease Progression via Suppression of Glial Inflammation in ALS Mouse Model

Amyotrophic lateral sclerosis (ALS) is an adult-onset motor neuron degenerative disease. Given that oxidative stress and resulting chronic neuronal inflammation are thought to be central pathogenic, anti-oxidative agents and modulators of neuronal inflammation could be potential therapies for ALS. We report here that the novel small molecular compound, 2-[mesityl(methyl)amino]-N-[4-(pyridin-2-yl)-1H-imidazol-2-yl] acetamide trihydrochloride (WN1316) selectively suppresses oxidative stress-induced cell death and neuronal inflammation in the late-stage ALS mice. WN1316 has high blood-brain-barrier permeability and water solubility, and boosts both neuronal apoptosis inhibitory protein (NAIP) and NF-E2-related factor 2 (Nrf2) which governed glutathione (GSH)-related anti-oxidation pathway protecting motor neurons against oxidative injuries. Post-onset oral administration of low dose (1–100 µg/kg/day) WN1316 in ALS(SOD1H46R) and ALS(SOD1G93A) mice resulted in sustained improved motor function and post onset survival rate. Immunohistochemical analysis revealed less DNA oxidative damage and motor neuronal inflammation as well as repression of both microgliosis and astrocytosis, concomitant down regulation of interleukin-1β and inducible nitric oxide synthase, and preservation of the motoneurons in anterior horn of lumbar spinal cord and skeletal muscle (quadriceps femoris). Thus, WN1316 would be a novel therapeutic agent for ALS.


Introduction
Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disorder characterized by a selective loss of upper and lower motor neurons, leading to death within 3-5 years. Approximately 90% of ALS cases are sporadic, while 10% of patients are familial ALS (FALS) [1]. Since Cu/Zn superoxide dismutase (SOD1) was originally identified as a causative gene of FALS [2], several causative genes including FUS, TARDBP, OPTN, and ALS2 have been identified [3]. Notwithstanding these discoveries, the exact mechanism of motor neuron death in both familial and sporadic ALS is still unclear. Moreover, there remains a pressing need for an effective ALS therapy drug given the modest impact riluzole [4,5].
Oxidative stress is being increasingly implicated in the neuronal inflammation, observed in ALS pathogenesis [6], and may possibly be a target for the development of novel therapeutic agents for ALS. The Neuronal apoptosis inhibitory protein (NAIP), which is a founding member of inhibitory of apoptosis proteins (IAPs), selectively suppresses oxidative stress-induced cell death [7], and overexpression of NAIP by adenovirus-mediated gene transfer has been shown to protect hippocampal neurons from ischemic injury [8]. Based on the endogenous NAIP functions, we previously developed NAIP-based drug screening (NAIP-ELISA) system [9], and demonstrated that the NAIP upregulating compounds exerted not only selective anti-oxidative stress activity in vitro [9,10] but also in vivo efficacy in a transgenic ALS mouse model carrying the H46R mutation in SOD1 gene [ALS(SOD1 H46R )] [10,11]. Although NAIP-ELISA system is promising strategy for the development of novel ALS treatment drugs, there is room for improvement of system in the point of optimization for time and cost.
In addition, we have recently started using in conjunction with a NAIP-ELISA system, a quantitative structure-activity relationship (QSAR) [12] used in ligand-based virtual screening [13], so that a modified algorithm for the relationship between chemical struc-ture of NAIP upregulation compounds (selectively suppress oxidative stress induced cell death) and their anti-oxidative cell death activity in vitro (termed NAIP-QSAR) was applied for the compound screening in silico. By using the NAIP-QSAR, we screened approximately 3,260,000 commercially available compound library, and resulted in the identification of N-(4-(2pyridyl)(1,3-thiazol-2-yl))-2-(2,4,6-trimethylphenoxy) acetamide (CPN-9), which showed a selective neuroprotection against oxidative-stress induced-cell death [14]. Moreover, the post-onset administration of CPN-9 to ALS(SOD1 H46R ) mice sustained motor functions and slowed disease progression [14]. Although CPN-9 acts as a potent remedy in ALS mice, it has also pharmacological weakness, such as a hardly water soluble, very poor blood-brain-barrier (BBB) permeability and a considerable toxicity. Thus, further modification and/or optimization of CPN-9 as a mother compound enables us to develop novel ALS therapeutic drugs.
Here we addressed this challenge with an in silico drug designing (termed virtual compounds), an in silico drug screening of virtual compounds by an algorithm of quantitative relationship between chemical structure of the compound and its anti-oxidative neuronal cell death activity (termed AOND-QSAR) [14], a chemical synthesis of virtual compounds, an in vitro assay of quantitative anti-oxidative neuronal cell death activity of the compounds, and an efficacy test in ALS mice.
In this study, we have successfully identified quite novel small molecular compound by performing in silico drug designing using CPN-9 as a primary mother compound, AOND-QSAR in silico screening, chemical synthesis and in vitro quantitative anti-oxidative neuronal cell death assay. The top hit compound was chosen as a secondary mother compound, and it was taken into in silico redesigning, modified AOND-QSAR screening, chemical synthesis and in vitro anti-oxidative neuronal cell death assay. Among hit compounds, top 4 compounds were selected, analyzed a specificity of in vitro anti-oxidative neuronal cell death activity, pharmacokinetics and toxicity in vitro as well as in vivo, and tested the efficacy in ALS(SOD1 H46R ) and ALS(SOD1 G93A ) mice. Among top 4 compounds, 2-[mesityl(methyl)amino]-N-[4-(pyridin-2-yl)-1H-imidazol-2-yl] acetamide trihydrochloride (WN1316) (Molecular weight: 458.82) was a compound possessing characteristics of highly soluble in water and prominent BBB permeability. In vitro study revealed that WN1316 exerted selectively anti-oxidative property by activating the NF-E2-related factor 2 (Nrf2)-antioxidant response element signaling pathway and upregulation of Nrf2, ATF3, p62, p21, glutathione (GSH) and NAIP.
On the other hand, to evaluate the in vivo efficacy of WN1316, we initiated a post onset-intragastric administration of WN1316 in ALS mice. The WN1316 administration, 1-100 mg/kg body weight/once daily per os, well sustained motor functions and prolonged the post-onset survival interval of not only ALS(-SOD1 H46R ) mice but also ALS(SOD1 G93A ) mice. Further, the 10 mg/kg body weight/day WN1316 treatment ameliorated motor neuron degeneration, gliosis and oxidative damage in the anterior horn of lumbar spinal cord in ALS(SOD1 H46R ) mice. Concomitantly, inflammatory factors, interleukin-1b (IL-1b) and inducible nitric oxide synthase (iNOS), diminished dramatically. The in vivo effective low dose WN1316 also efficiently suppressed the axonal degeneration and muscle fiber denervation in ALS(SOD1 H46R ) mice. Besides the highly potent at low dose in ALS mice, pharmacokinetic study in non-transgenic mouse (C57BL/6N) indicated that WN1316 was highly BBB permeable compound with wide safety range. Taken together, WN1316 might represent a potential therapeutic drug for the ALS treatment.

Identification of a Novel Acylaminoimidazole Derivative, WN1316, Selectively Suppresses Oxidative Stress Induced Cell Death
Recently, we identified a small-molecule CPN-9 which selectively suppressed oxidative stress-induced cell death via activation of the Nrf2 pathway [14]. In the search for novel ALS therapeutic agents with superior aqueous solubility, cytotoxicity and BBB permeability, further modification and/or optimization of CPN-9 as a mother compound was undertaken. We thus conducted the in silico design of approximately 1,390 novel molecules based on the chemical structure of a primary mother compound CPN9 which were then subjected AOND-QSAR and drug-likeness screening according to four criteria described in methods. Sixty-five molecules which best fulfill these four criteria were identified as active compounds and then broken down into 8 clusters; 3 to 4 representative virtual compounds from each cluster, 27 compounds in all, were next chemically synthesized. We then performed the anti-oxidative stress-induced neuronal cell death (AOND) analysis on these compounds, and identified a top-hit compound (WN0247) ( Figure S1).
We again virtually designed molecules based on the chemical structure of WN0247, followed by AOND-QSAR screening, and then chemically synthesized 14 novel compounds ( Figure S1). The biological activities of these compounds were tested by the AOND analysis; four active compounds, WN1303R, WN1312, WN1315 and WN1316, which showed marked cytoprotection against oxidative stress-induced cell death in neuronal SH-SY5Y cells superior to the mother compound CPN-9, were identified ( Figure 1A).
To establish whether the suppression of oxidative stress-induced cell death by WN compounds was similar to the mother compound CPN-9, we evaluated their effectiveness against various cytotoxins. All showed selective protection against the oxidative stressors, menadione, a-naphthoquinone and 6-hydroxydopamine (6-OHDA), but not the non-oxidative stressors, staurosporine and okadaic acid, in differentiated SH-SY5Y cells ( Figure 1C and S2). Among them, WN1316 conferred the greatest protection against oxidative stress-induced damage, and the most suppression of both reactive oxygen species (ROS) generation and lipid peroxide formation. Menadione increased the level of intracellular NOH and lipid peroxidation ( Figure 1D, a,c and e) in HeLa cells, while the WN1316 treatment markedly reduced their levels ( Figure 1D, b, d and f). Thus, of the four candidates, WN1316 was the most potent suppresser of highly toxic ROS generation in cultured cells.

WN1316 is a Highly BBB Permeable Compound
We next studied the pharmacokinetics of the four WN compounds in wild type mice (C57BL/6N). To measure WN compound concentration in brain and serum (termed the area under the curve: AUC0-8), 50 mg of each WN compound in distilled water was orally administered to mouse of 8 weeks of age (n = 4). The mice tolerated the WN compounds with no apparent impact. WN1316 revealed the highest brain AUC0-8, about 39fold of the level observed with CPN-9, and notably approximately 4-fold higher than serum levels (Table 1, Figure 1B). The clearance of WN1316 in both brain and serum was 8 h. differentiated SH-SY5Y cells for longer than 8 h was required for AOND activity ( Figure 1E). In another way, incubation of the cells with WN1316 for 2 to 4 h followed by 6 to 4 h of a chase incubation without the compound was essential to achieve a maximum cytoprotection against 60 mM menadione mediated oxidative stress induced cell death ( Figure 1F). These results indicate that WN1316 activates and/or stimulates de novo synthesis of the endogenous proteins conferring selective protection from oxidative stress-induced damages rather than acting as a radical scavenger. Effect of compounds against oxidative stress-induced cell death. Differentiated SH-SY5Y cells were treated with compounds or dimethyl sulfoxide (DMSO). After 24 h, cells were exposed to 60 mM menadione for 4 h. The cell viability was measured by AlamarBlue assay. The cell viability was expressed as a fold induction after the treatment with compounds compared with vehicle control (DMSO). Data are expressed as mean 6 SD (n = 4). *p,0.001 by one-way ANOVA with Dunnett's post hoc test compared with DMSO-treated control. (B) Chemical structure of WN1316; 2-[mesityl(methyl)amino]-N-[4-(pyridin-2-yl)-1H-imidazol-2-yl] acetamide trihydrochloride. (C) Effect of WN1316 against various cytotoxins. Differentiated SH-SY5Y cells were pretreated with 6 mM WN1316 or DMSO for 8 h at 37uC. The appropriate amount of each cytotoxin, which includes free radical generating compounds [menadione, a-naphtoquinone, 6-hydroxydopamine (6-OHDA)], protein kinase inhibitor (staurosporine), or phosphatase inhibitor (okadaic acid), was added, and incubated for another 4 to 6 h as described in Material & Methods. The cell viability was measured by AlamarBlue assay. Data are expressed as mean 6 SD (n = 4). *p,0.001 by one-way ANOVA with Dunnett's post hoc test compared with DMSO-treated control. (D) Representative image of the total intracellular ROS (a and b), ?OH (c and d), and the intracellular 4-HNE (e and f) levels in HeLa cells after oxidative insult. Cells were treated with DMSO (a, c, and e) or 10 mM WN1316 (b, d, and f) for 8 h and then loaded with 10 mM 29,79-dichlorodihydrofluorescein diacetate (H 2 DCF-DA) (a and b) or 100 mM hydroxyphenyl fluorescein (HPF) (c and d) for 30 min, followed by the exposure to 60 mM menadione for 1 h or 0.1 mM FeSO 4 /0.2 mM H 2 O 2 for 1 h, respectively. For immunocytochemistry, cells were treated with DMSO or 10 mM WN1316 for 8 h, followed by the exposure to 60 mM menadione for 1 h, and immunostained with anti-4-HNE antibody (green). The fluorescent images were obtained by using a confocal microscope. Scale bars indicate 10 mm. (E) Treatment time of WN1316 on the cytoprotection. Differentiated SH-SY5Y cells were incubated with 6 mM WN1316 for the indicated times, and then treated with 60 mM menadione for 4 h. The cell viability was determined by AlamarBlue. Data are expressed as mean 6 SD (n = 4). *p,0.001 by one-way ANOVA with Dunnett's post hoc test compared with DMSO-treated control. (F) Treatment time of WN1316 on the cytoprotection. Differentiated SH-SY5Y cells were incubated with 6 mM WN1316 for the indicated times, and then treated with 60 mM menadione for 4 h. The cell viability was determined by AlamarBlue. Data are expressed as mean 6 SD (n = 4). *p,0.001 by one-way ANOVA with Dunnett's post hoc test compared with DMSO-treated control. doi:10.1371/journal.pone.0087728.g001
To further establish the WN1316 activation of the Nrf2 pathway, knockdown experiments of Nrf2 were conducted in SH-SY5Y cells. Nrf2-knockdown dramatically reduced HO-1, NQO1 and GCLM protein levels as well as Nrf2 protein induced by the WN1316 treatment compared to the control ( Figure 2E). Further, WN1316-mediated AOND activity in SH-SY5Y cells was also reduced by Nrf2 knockdown, but not control (siNC) ( Figure 2F), implicating the Nrf2 signaling pathway in WN1316mediated cytoprotection.

WN1316 Requires Intracellular ROS for the Activation of Nrf2 Signaling Pathway
Given that N-acetylcysteine (NAC), ROS scavenger, downregulates the expression of Nrf2-regulated anti-oxidant molecule genes [14,17,18], we tested whether NAC moderates WN1316mediated Nrf2 activation. NAC treatment abrogated the WN1316 induction of Nrf2, ATF3, HO-1, NQO1, GCLM, p62 and p21 mRNAs and proteins ( Figure S3B and C). Concomitantly, WN1316 activity against oxidative stress-induced cell death was drastically suppressed by the NAC treatment ( Figure S3D). In a second approach, modulation of NADPH oxidase, a membranebound enzyme, that catalyzes the production of superoxide from oxygen and NADPH [19] was undertaken. Apocynin, an NADPH oxidase inhibitor, did not affect the AOND activity of WN1316 ( Figure S3D). These results suggest that the activation of Nrf2 signaling mediated by WN1316 requires intracellular ROS, but not those produced by NADPH oxidase.
Cytoplasmic Nrf2 forms a complex with Keap1 and is degraded under the basal condition, while ROS catalyzes dissociation of Nrf2 from the complex, resulting in nuclear translocation of Nrf2 [15,16]. Although WN1316 both in the presence and absence NAC did not affect Keap1 mRNA levels, the WN1316-dependent degradation of Keap1 protein was blocked by NAC ( Figure S3B and C). These results suggest that the activation of Nrf2 by WN1316 may be associated with the inhibition of Keap1mediated Nrf2 degradation in addition to the upregulation of Nrf2 mRNA/protein.

WN1316 Modulates GSH Levels
Since glutathione (GSH) is one of the most abundant antioxidants that scavenge reactive oxygen species, we evaluated total cellular GSH levels in differentiated SH-SY5Y cells with/ without WN1316 treatment. WN1316 revealed a dose dependent AOND activity and a concomitant upregulation of total GSH level, although a higher concentration than 20 mM WN1316 militates against cytoprotection ( Figure S4). These results implicate GSH modulation in WN1316 cytoprotection.

WN1316 Upregulates NAIP
NAIP is a founding member of IAP family of proteins and selectively suppresses oxidative stress-induced cell death. THP-1 cells which unlike SH-SY5Y encoded an intact NAIP gene (data not shown) were exposed to WN1316. WN1316 dose-dependent upregulation of NAIP in THP-1 cells was observed and was not affected by NAC treatment ( Figure 3A, B and C). Nrf2 does not impact NAIP gene transcription and biological function, suggesting that NAIP may function independently of Nrf2 in WN1316 mediated cytoprotection.

WN1316 Slows Disease Progression in ALS(SOD1 H46R ) Mice
To assess the in vivo efficacy of WN1316, we conducted a daily post-onset administration of WN1316 [0 mg/kg (vehicle), 1 mg/kg, 10 mg/kg and 100 mg/kg body weight] to ALS(SOD1 H46R ) mice. The average day of disease onset was 125.162.4 days (n = 104). At 22 weeks of age (late symptomatic stage), a majority of the vehicletreated mice showed a complete paralysis of hind limbs, and thus did not manifest the feet-clasping phenotype upon the tail suspension and rearing behavior ( Figure 4A, a and b). In contrast, the WN1316-treated mice at the same age still showed a feetclasping phenotype and rearing activity ( Figure 4A, c and d). Despite these significant effects, WN1316 did not affect mouse weight in this study ( Figure 4B).

WN1316 Preserves Motor Performance in ALS(SOD1 H46R ) Mice
To evaluate whether WN1316 preserves motor function in ALS(SOD1 H46R ) mice, we conducted balance beam, vertical pole, grip strength and footprint tests, as well as cage activity and rearing performance analyses. The balance beam test revealed that the WN1316 (1 , 100 mg/kg body weight/day) treatment significantly extended the time post onset that the mice were able to stay on the bar compared with vehicle-treated group ( Figure 4C).
In contrast, there was no observable difference in the ability to perform vertical pole climbing and grip strength between the WN1316-treated and vehicle-treated groups at 17 weeks of age (late pre-symptomatic stage) ( Figure 4D and E). However, at 21-22 weeks of age (late symptomatic stage), WN1316-treated groups showed a significant preservation of motor performance compared with vehicle-treated group ( Figure 4D and E).
All mice at 18 weeks (early symptomatic stage) showed almost normal ambulation on footprint analysis ( Figure 4F). At 21 weeks of age (late symptomatic stage), vehicle-treated mice showed an abnormal gait with reduced foot stride length and dragging of the  legs, while WN1316-treated mice showed a short but still steady gait ( Figure 4F and G). Furthermore, at the late symptomatic stage, the spontaneous motor activity of WN1316-treated groups was sustained in rearing performance compared with vehicletreated group ( Figure 4I), while there was a trend towards sustained cage activity in WN1316-treated groups compared with vehicle-treated group ( Figure 4H). These results demonstrate that lower doses of WN1316 (1 mg to 100 mg/kg/day) effectively sustains the motor function in ALS(SOD1 H46R ) mice and suggest that is effective in vivo.

WN1316 Exerts a Beneficial Effect on Motor Function in ALS(SOD1 G93A ) Mice
Next, to investigate the in vivo efficacy of WN1316 in another ALS(SOD1) mouse model, ALS(SOD1 G93A ) mouse, we conducted a daily post-onset oral administration of WN1316 (10 mg/kg body weight/day) to ALS(SOD1 G93A ) mice and carried out the analysis of motor performance. The average day of disease onset was 97.765.3 days (n = 15). No impact of WN1316 including weight gain or loss were observed ( Figure  S5A). ALS(SOD1 G93A ) mice treated with WN1316 showed a sustained motor performance compared to vehicle-treated ALS(SOD1 G93A ) mice at 20 weeks of age (late symptomatic stage) by footprint analysis ( Figure S5C and D). The WN1316treated groups also showed a trend towards sustained motor performance in the balance beam test and a higher spontaneous motor activity in both cage activity and rearing performance than vehicle-treated group in a late symptomatic stage ( Figure  S5B, E and F). These results support the in vivo efficacy of WN1316 in ALS mouse models.

WN1316 Treatment does not Alter SOD1 Expression
To ensure that the WN1316 did not alter the expression of the human SOD1 transgene in ALS(SOD1 H46R ) mice, Western blot analysis using tissue extracts from the lumbar spinal cord in vehicle-treated and 10 mg/kg WN1316-treated groups and agematched non-transgenic (non-Tg) littermates at 21-22 weeks of age was performed. There were no significant differences of the protein levels of mutant SOD1 in both the NP-40-soluble and the NP-40-insoluble/SDS-soluble fractions between vehicle-and WN1316-treated groups ( Figure S6).

WN1316 Represses Motor Neuron Loss in the Spinal Cord
To investigate whether WN1316 impacts motor neuron counts, lumbar sections (L4-L5) were immunostained using anti-choline acetyltransferase (ChAT) (for motor neurons) antibody. At 21-22 weeks of age, the vehicle-treated ALS(SOD1 H46R ) mouse showed a massive loss of ChAT-positive motor neurons ( Figure 5B) compared with non-Tg littermates ( Figure 5A), whereas ChATpositive motor neurons of 10 mg/kg WN1316-treated ALS(-SOD1 H46R ) mouse were relatively spared ( Figure 5C). The number of ChAT-positive motor neurons within the hatched region shown in Figure 5D was counted. At 21-22 weeks of age, the loss of ChAT-positive motor neurons was 35.5% in vehicletreated and 15.2% in WN1316-treated ALS(SOD1 H46R ) mice when compared with non-Tg littermates ( Figure 5E) consistent with sparing of the motor neurons.

WN1316 Prevents the Activation of Microglia and Astrocytes in Spinal Cord
Spinal cord glial cells activation is a histopathological hallmark both of human and murine ALS [20,21,22]. To analyze the impact of WN1316 on glial cell activation at sites of motor neuron loss in ALS(SOD1 H46R ) mice, immunostaining with anti-glial fibrillary acidic protein (GFAP) (for astrocytes) and anti-ionized calcium binding adaptor molecule 1 (Iba-1) (for activated microglia) antibodies using lumbar sections (L4-L5) was performed. At 21-22 weeks of age, marked immunoreactivities both for GFAP and Iba-1 were observed in lumbar spinal cord anterior horn of vehicle-treated ALS(SOD1 H46R ) mice ( Figure 6A, b and e), but not age-matched non-Tg ( Figure 6A, a and d). WN1316 (10 mg/kg)-treated mice exhibited significant reduction of the GFAP and Iba-1 immunoreactivity ( Figure 6A, c and f) and preservation of anterior horn cells ( Figure 6A, i) compared with vehicle-treated mice ( Figure 6A, h). These results reveal that the WN1316 alleviates both astrocytosis and microgliosis in the spinal cord of ALS(SOD1 H46R ) mice in the late symptomatic stage.

WN1316 Suppresses the Production of Inflammatory Factors
To examine whether WN1316 modulates the production of proinflammatory mediator IL-1b and inflammatory marker iNOS, we performed immunohistostaining for IL-1b and Western blotting for iNOS. A marked IL-1b immunoreactivity was detected in astrocytes and partially microglia, but not in neuron, in lumbar spinal cord anterior horn of vehicle-treated ALS(-SOD1 H46R ) mice at 21-22 weeks of age compared to non-Tg littermates ( Figure 6B). Importantly, 10 mg/kg WN1316-treated ALS(SOD1 H46R ) mice revealed that the IL-1b immunoreactivity diminished dramatically in astrocytes ( Figure 6B). We also detected an increased level of iNOS in the extract from lumbar spinal cord of vehicle-treated ALS(SOD1 H46R ) mice at 21-22 weeks of age compared to those in non-Tg littermates ( Figure 6C), whereas the WN1316 treatment decreased iNOS generation in ALS(SOD1 H46R ) mice ( Figure 6C).

WN1316 Mitigates the Ventral Motor Neuroaxonal Degeneration in ALS(SOD1 H46R ) Mice
It has been suggested that oxidative stress plays a major role in axonal degeneration [24]. To assess the effect of WN1316 on axonal degeneration, we counted large myelinated axons in the L5 ventral root. At 21-22 weeks of age, vehicle-treated ALS(-SOD1 H46R ) mice showed a marked loss of large myelinated axons compared to non-Tg littermates, while 10 mg/kg WN1316-treated ALS(SOD1 H46R ) mice showed the preservation of large myelinated axons ( Figure 7B and C).

WN1316 Reduces Myofibrosis and Myonecrosis of the Skeletal Muscle in ALS(SOD1 H46R ) Mice
To evaluate the effect of WN1316 on skeletal muscle integrity, the morphology of muscle fibers in quadriceps femoris muscle was examined. Transverse quadriceps femoris muscle of vehicletreated ALS(SOD1 H46R ) mice at 21-22 weeks of age demonstrated a severe abnormal pattern of fiber distribution and fiber type grouping (myonecrosis and myofibrosis) compared with agematched non-Tg mouse ( Figure 7D). However, the WN1316 treatment preserved the skeletal muscle integrity in the ALS(-SOD1 H46R ) mice ( Figure 7D). These results demonstrate that WN1316 alleviates disease progression in the ALS mouse models.

Discussion
Although the molecular mechanism of selective motor neuron loss in ALS is still unclear, substantial evidence points towards oxidative stress and neuronal inflammation playing a crucial role in ALS pathogenesis [6]. Employing NAIP-ELISA [9] and NAIP-QSAR [14], we have previously identified CPN-9 as potential novel ALS therapeutic agent. To overcome CPN-9's innate pharmacological weakness (e.g. poor solubility, poor BBB permeability and cytotoxicity), we used it as a mother compound to identify WN1316 as a first-in-class potential ALS therapy.
Previously, we showed that CPN-9 mediated the activation of Nrf2 signaling cascade [14]. In general, Nrf2 defends the cells against oxidative stress-induced cell death [25,26] by dissociation of cytoplasmic Nrf2-Keap 1 complexes followed by nuclear translocation of Nrf2 [25,26,27,28,29], In this study, we have demonstrated that WN1316 acts as Keap1 inhibitor and selectively suppresses oxidative stress-induced cell death via the activation of the Nrf2 signaling pathway. WN1316 also significantly increased both mRNA and protein levels of Nrf2-related genes including p21 and p62 with attendant increased cell viability. In addition, the generation of mitochondrial ROS (mitROS) is also associated with the activation and stabilization 12 mM WN1316 or DMSO in the presence or absence of 1 mM NAC for 24 h. Cell lysates were analyzed anti-NAIP antibody. b-tubulin was used as a loading control for protein. Densitometric data for NAIP bands was normalized by the level of b-tubulin. Data are expressed as mean 6 SEM (n = 3). * p,0.05 or ** p,0.01 by Student's t-test. (C) Effect of NAC on the WN1316-mediated anti-oxidative stress activity in THP-1 cells. Cells treated with 12 mM WN1316 or DMSO in the presence or absence of 1 mM NAC were exposed to 80 mM menadione, and then the cell viability was measured by AlamarBlue. Data are expressed as mean 6 SD (n = 4). *** p,0.0001 by Student's t-test. doi:10.1371/journal.pone.0087728.g003 of Nrf2 [30]. We also showed that WN1316 exerted cytoprotective activity against oxidative injury through the activation of Nrf2 signaling pathway in a mitROS-dependent but not NADPH oxidase-derived ROS-dependent manner. In addition, it has been reported that Cu 2+ enhances induction of HO-1 by 2-tert-butyl-1,4-hydroquinone (tBHQ), which is a known inducer of Nrf2 [31]. In the present study, we demonstrated that WN1316 activity in vitro required Cu 2+ (data not shown).
WN1316 also upregulates NAIP independently from Nrf2 activation pathway ( Figure S7). Mitochondrial, not cytoplasmic, Ca 2+ stimulates NAIP upregulation (data not shown). Thus, WN1316 primarily affects mitROS and Ca 2+ generation and achieves a selective protection against oxidative injury by a complementary and parallel courses of NAIP-and Nrf2-antioxidation pathways in vitro ( Figure S7).
To date, there are no effective therapeutic interventions that markedly improve the quality of life for ALS patients. In this study, we sought to evaluate the in vivo efficacy of WN1316 in both ALS(SOD1 H46R ) and ALS(SOD1 G93A ) mice. WN1316 did not affect weight in mice ( Figure 4B). Autopsy of WN1316-treated mice also exhibited no pathological abnormalities of organs such as heart and kidney (data not shown), indicating that the doses of WN1316 in our study are not harmful to mice. We have demonstrated that the post-onset administration of WN1316 preserved motor performance and improved post-onset survival rate, 20-25% prolongation in ALS(SOD1 H46R ) and 14% prolongation in ALS(SOD1 G93A ) mice. The WN1316 treatment also delayed motor neuron loss in the spinal cord accompanied by the suppression of glial cells activation in the late symptomatic stage ALS(SOD1 H46R ) mice. Further, WN1316 transiently upregulates an mRNA level of Nrf2-related gene HO-1 in a wild type mouse spinal cord, implying WN1316 actions in vivo are analogous to those observed in vitro. Recent studies have suggested that Nrf2dependent oxidative cytoprotection is mediated in part by ATF3 modulation of GSH levels [32], and that GSH suppresses the production of proinflammatory factors [33]. Indeed, we have demonstrated that the WN1316 treatment augmented the level of ATF3 protein and GSH in human neuronal cells in vitro and down-regulated the level of IL-1b and iNOS in the lumbar spinal cord of ALS(SOD1 H46R ) mice. Moreover, we have confirmed NAIP, which suppresses neuronal cell death induced by oxidative injury [9], is upregulated in WN1316-treated cells in vitro. Thus, it is conceivable that WN1316 protects from motor neuron loss via the combined pathway involving the enhancement of NAIP-and Nrf2-mediated cytoprotection and the intracellular GSH augmentation in ALS(SOD1 H46R ) mice ( Figure S7).
Inflammation in ALS is characterized by the massive activation of astrocytes and microglia [34,35] elevating the production of inflammatory markers and proinflammatory cytokines [36,37]. Several lines of evidence have reported that astrocytes have an impact on motor neuron degeneration [38,39] and microglia contributes to neuronal damage in neurodegenerative diseases [37,40]. Further, the link between oxidative stress and inflammatory response in ALS has been suggested [6,41]. In this study, we demonstrated that WN1316 suppressed both microgliosis and astrocytosis while concomitantly curbing IL-1b and iNOS generation in the spinal cord of ALS(SOD1 H46R ) mice. In addition, the WN1316 reduced the level of DNA oxidation in neurons, astrocytes and microglia of the spinal cord in the late symptomatic stage ALS(SOD1 H46R ) mice. Although the molecular base of the WN1316 effect on astrocytes and microglia remains elusive, WN1316 attenuates a gliosis and motoneuronal inflammation by inhibiting the production of inflammatory factors, resulting in the reduction of oxidative damage in the spinal cord and the rescue of motor neurons and ventral motor axons.
There are many reports of small compounds that show an in vivo effect with measure of concentration in the animal studies of ALS [42,43]. In our study, the effective in vivo oral dose of WN1316 is extremely low ranging 1 , 100 mg/kg/day. WN1316 concentration in brain and blood of non-transgenic control C57BL/6N mouse (oral administration of 50 mg/kg body weight) at the point of Cmax/AUC0-8 were 42.1 mM/38.965.7 mg?h/g and 9.78 mM/13.660.7 mg?h/ml, respectively. It is obvious that WN1316 is highly BBB permeable and biologically active compound, despite the fact that WN1316 in rat/human sera complexes with protein (unpublished data). Oddly, it appears that the in vitro effective WN1316 concentration is higher than the in vivo case. The difference in effective concentration between WN1316 (1 mg/kg, 10 mg/kg and 100 mg/kg)-treated mice at 18 and 21 weeks of age. Cumulative data counting for 2 consecutive nights are shown as Box-whisker plots (AU; arbitrary unit, each group; n = 17). * p,0.05 by non-parametric ANOVA (Kruskal-Wallis) with Dunn's post hoc test. doi:10.1371/journal.pone.0087728.g004 in vitro and in vivo studies may reflect chaperon-mediated WN1316 transport ( Figure S7), the identification of which is currently underway.
The development of novel and effective ALS drugs requires testing in animal models and pathological studies. Guidelines for the preclinical animal study in ALS recommends administration of candidate agents at symptomatic phase as is the case in the clinical setting [44]. In accordance with this guideline, we conducted postonset oral administration of WN1316 to SOD1 ALS mouse model, demonstrating that WN1316 remarkably ameliorated clinical symptoms. To further verify the WN1316 efficacy, it might be ideals that WN1316 is tested in another ALS models besides SOD1 mutations, but any animal or cell models suit to our experimental condition are hardly available at this moment. However, it is noteworthy that WN1316 shows analogous in vivo effect in different lines of ALS mouse model, ALS(SOD1 H46R ) and ALS(SOD1 G93A ) mice. That is, our results prove the drug efficacy of WN1316 in the preclinical animal study. Although the mechanism of WN1316 action is not yet fully understood, WN1316 is a promising candidate as a novel therapeutic agent for the treatment of ALS. The preclinical safety tests and chemistry, manufacturing and control (CMC) of WN1316, and a drug formulation for Phase1 clinical test are presently ongoing.

In silico Screening
In silico screening was applied to screen the virtual molecules designed in silico. A group of drugs with similar pharmacological function usually share common chemical characteristics. In silico screening was configured employing the following four criteria; 1) Quantitative structure-activity relationship (QSAR) equation derived using the chemical structure and the biological (antioxidative stress-induced neuronal cell death: AOND) activities of active molecules (AOND-QSAR) [14], 2) CNS drug-likeness  which are specific chemical properties shared among drugs acting upon the central nervous system (CNS), 3) presumptive bloodbrain barrier permeability, and 4) pharmacophoric analysis comprised of an ensemble of steric and electronic features to ensure the optimal supramolecular interaction with a specific biological target and to trigger its biological response. A number of compounds were synthesized employing these criteria, and assessed for activity in an AOND assay.
In silico virtual molecule designing 1 st novel hit compound and the in silico screening as stated above, and AOND analysis were conducted. All calculations were performed by the use of the software system MOE (Molecular Operating Environment; Chemical Computing Group Inc., Montreal, Quebec, Canada).

Anti-oxidative Stress-induced Neuronal Cell Death (AOND) Analysis
Human neuroblastoma SH-SY5Y cells (ATCC, CRL-2266) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (HyClone, Thermo Fisher Scientific Inc.), 100 U/ml penicillin, and 100 mg/ml streptomycin at 37uC in 5% CO 2 . For differentiation, SH-SY5Y cells were seeded at a density of 0.75610 4 cells per well in 96-well plates (Primaria, BD Falcon) for 1 day, and cultured in the medium containing 10 mM all-trans retinoic acid for additional 5 days.
SH-SY5Y cells were plated at density of 0.75610 4 cells/well in 96-well plates and treated with various concentrations of compound dissolved in dimethyl sulfoxide (DMSO) or vehicle control (DMSO), and cultured for another 24 h. Then, the cells were exposed to various concentrations of cytotoxins, including oxidative stressors [menadione, 10 to 80 mM for 4 to 6 h; anaphthoquinone, 12.5 mM for 4 h; 6-hydroxydopamine (6-OHDA), 400 mM for 4 h], protein kinase inhibitor (staurosporine, 25 mM for 6 h) and phosphatase inhibitor (okadaic acid, 1 mM for 6 h), in fresh medium. Cell viability was measured by AlamarBlue assay (Invitrogen) as described previously [9].

Detection of Intracellular ROS
HeLa cells (ATCC, CCL-2) cultured on poly-L-lysine-coated coverslips were treated with WN1316 (10 mM) or vehicle (DMSO) for 8 h and incubated with 10 mM H 2 DCF-DA for 30 min. Cells were then exposed to 60 mM menadione for 1 h. Confocal fluorescence images were obtained using Zeiss LSM 700 confocal microscope.
To detect the hydroxyl radical (NOH), we used hydroxyphenyl fluorescein (HPF) as a specific probe. In brief, HeLa cells were treated with WN1316 (10 mM) or vehicle (DMSO) for 8 h and then incubated with 100 mM HPF for 30 min at 37uC in Hanks's buffer (Invitrogen). After washing with phosphate-buffered saline (PBS), the reaction buffer (0.1 mM FeSO 4 , 0.2 mM H 2 O 2 ) was added, and cells were incubated for 1 h at 37uC to generate NOH. Confocal fluorescence images were obtained using the lambda mode of Zeiss LSM 510 Meta confocal microscope and analytical software.

Reverse Transcriptase-polymerase Chain Reaction (RT-PCR)
Total RNA was extracted from cells using the TRIzol reagent (Invitrogen). Isolated total RNA was purified using the SV Total RNA Isolation System (Promega) according to the manufacturer's instruction. RT

Quantitative Reverse Transcriptase-PCR (qRT-PCR)
Wild type mice at 8 weeks of age (male, C57BL/6N) were administered per os with either vehicle (physiological saline) or WN1316 (100 mg/kg body weight). After 1 h, 3 h, 6 h and 9 h, mice were transcardially perfused with physiological saline containing 10% heparin under anesthesia with 2% isoflurane by inhalation, and the lumbar spinal cord was dissected. Total RNA was extracted from lumbar spinal cord using Sepasol-RNA I Super G (Nacalai Tesque), and purified by SV Total RNA Isolation System (Promega) according to manufacturer's instructions. qRT-PCR was performed with QuantiFast TM SYBR Green RT-PCR Kit (QIAGEN) on CFX96 (BIO RAD). The PCR primers (QuantiTect Primer Assays) for HO-1 (Mm_Hmox1_1_SG) and PPIA (Mm_Ppia_1_SG) were purchased from QIAGEN. The levels of all transcripts were normalized for the PPIA mRNA level in each sample. Relative expression was calculated using the DDCt method (Livak 2001).

RNA Interference
Expression of Nrf2 was inhibited with Hs_NFE2L2_7 HP Validated FlexiTube siRNA (QIAGEN). Silencer Negative Control #1 (NC) siRNA was used as a non-targeting siRNA (AM4635, Ambion). SH-SY5Y cells at 50-60% confluence were transfected with NC and Nrf2 siRNA (5 nM) using Lipofectamine RNAi-MAX (Invitrogen). After 24 h and 48 h, cells were subjected to the isolation of RNA and protein, respectively. The degree of Nrf2 expression inhibition was evaluated by RT-PCR and Western blotting.

Antioxidant Treatment
Differentiated SH-SY5Y cells (0.75610 4 cells/well) were treated with 10 mM WN1316 or DMSO in the presence or absence of Nacetyl cysteine (NAC) (1 to 2.5 mM) or apocynin (20 to 80 mM) for 3 h (for RT-PCR) or 8 h (for Western blotting and cell viability analysis).
Human monocytic leukemia cell line THP-1 (ATCC, TIB-202) was cultured in RPMI-1640 (Wako) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 mg/ml streptomycin at 37uC in 5% CO 2 . Approximately 1610 5 cells, which were pretreated with 1 to 16 mM WN1316 or DMSO in the presence or absence of 1 mM NAC for 24 h, were plated on a 96-well plate and exposed to 80 mM menadione for 16 h. Cell viability was measured by AlamarBlue assay. For Western blotting, cell lysates were prepared by using M-PER mammalian protein extraction reagent (Thermo scientific) according to standard procedure.

Measurement of Intracellular Oxidized and Reduced Glutathione
Differentiated SH-SY5Y cells were treated with various concentrations of WN1316 (1-20 mM) or vehicle control (DMSO). After 8 h incubation, the cellular oxidized (GSSG) and reduced glutathione (GSH) levels were measured using the GSH/GSSG-Glo TM Assay kit (Promega) according to the manufacturer's instructions. GSH and GSSG levels were normalized to protein concentrations.

Animals
Transgenic mice carrying the H46R mutation in the human SOD1 gene; SOD1 H46R were used as a familial ALS mouse model. The ALS(SOD1 H46R ) transgenic mouse exhibits a slow disease progression with widespread axonal degeneration in the spinal cord when compared with ALS(SOD1 G93A ) transgenic mice (Sasaki 2007, Tanaka 2011Tanaka , 2008. Since the genetic background is one of the important factors modulating disease phenotypes in mutant SOD1 transgenic mice, we first generated a congenic line of ALS(SOD1 H46R ) transgenic mice by backcrossing more than 15 generations to B6 mice.
The lines were maintained as hemizygotes by mating ALS(-SOD1 H46R ) and ALS(SOD1 G93A ) males with B6 females. Offspring were genotyped by PCR using genomic DNA from tail tissue. Mice were housed at 23uC with a 12 h light/dark cycle, in which water and food was available ad libitum. The condition of each animal was monitored weekly initially and then daily after disease onset. The lifespan was determined by the days from disease onset to the endpoint, in which mouse lost active and/or spontaneous movement. All animal experimental procedures were approved by the Institutional Animal Care and Use Committee at Tokai University.

Pharmacokinetics of WN-compounds in Mice
Wild type mice at 8 weeks (male, C57BL/6N) were used for the analysis of CPN-9, WN1303R, WN1312, WN1315 and WN1316 pharmacokinetics. CPN-9 was suspended in 0.5% carboxymethyl cellulose (CMC)-Na (carmellose sodium, Maruishi Pharmaceutical Co., Ltd.), while WN1303R, WN1312 and WN1316 were dissolved in Milli-Q purified water (MQ water) and WN1315 was dissolved in 0.25 N HCl. CPN-9 (100 mg/kg body weight) and the four WN-compounds (50 mg/kg body weight) were injected intragastrically (i.g.). After 15 min, 30 min, 1 h, 2 h, 4 h, 6 h, and 8 h, blood was collected from heart of each mouse anesthetized with 2% isoflurane by inhalation or 4% halothane in a mixture of N 2 O/O 2 (70:30). Mice were thereafter perfused transcardially with physiological saline containing 10% heparin under anesthesia, and the whole brain was removed. Each blood sample was centrifuged at 1,0006g for 10 min at 4uC, and the serum fraction was collected. The brain was frozen in liquid nitrogen, and was cryopreserved until use.
We mixed serum with deproteinization buffer (0.05 M H 2 SO 4 , 2.5 M K 2 CO 3 , and diethylester), and collected the organic layer after centrifugation at 1,5006g for 10 min. The organic layer was mixed with 0.05 M H 2 SO 4 , and then, after centrifugation at 1,5006g for 10 min, the water layer was collected. After the addition of methanol, each serum derived sample was subjected to HPLC.
Mouse whole brain was homogenized with acetonitrile and left at 280uC for 10 min. The supernatant was collected after centrifugation at 10,0006g for 10 min at 4uC and was dried under vacuum. The dried sample was re-dissolved in methanol and mixed with two volumes of 10 mM phosphate buffer (pH7.0), after which the supernatant was collected by centrifugation at 10,0006g for 10 min at 4uC and then left at 280uC for 10 min. To further remove debris, the supernatant was filtered (0.45 mm filter), centrifuged (at 10,0006g for 10 min at 4uC), filtered (0.22 mm filter) and finally centrifuged again. Each sample solution from brain was subjected to HPLC.

Administration of the Compound
WN1316 was dissolved in MQ water, and diluted with physiological saline. ALS(SOD1 H46R ) transgenic mice were given WN1316 via gastric tube (CAT.No.4202, Fuchigami) at a dose of either 1 mg, 10 mg or 100 mg/5 ml/kg body weight. ALS(-SOD1 G93A ) transgenic mice were given 10 mg/kg WN1316. Animals simultaneously treated with physiological saline (vehicle group, 5 ml/kg) were used as controls. We commenced daily compound administration when mice showed signs of motor dysfunction (deemed age of onset) between 121d and 130d in ALS(SOD1 H46R ) transgenic mice and between 88d and 106d in ALS(SOD1 G93A ) transgenic mice, and continued until their endpoint.

Assessment of the Onset of the Symptom
To determine the onset of motor dysfunction in ALS(-SOD1 H46R ) transgenic mice, we adopted a balance beam test as previously described in detail [10]. To determine the onset of motor dysfunction in ALS(SOD1 G93A ) transgenic mice, we adopted both balance beam and grip strength tests. Grip strength test was assessed using a traction meter (BrainScience Idea Co., Ltd.). Each mouse was given three trials, and the highest limb strength score was recorded. For ALS(SOD1 G93A ) transgenic mice, grip strength tests were started at 11 weeks of age and performed twice a week until an over 20% reduction compared with the initial limb strength score at 11 weeks of age was observed. Finally, a grade 3 in the balance beam test [10] and over 20% reduction of limb strength score were defined as the sign of the disease onset in ALS(SOD1 G93A ) transgenic mice.

Observation of Gross Phenotypes in Mice
The gross phenotypes were measured blindly. The body weight of each mouse was measured weekly initially and then daily after disease onset. Mice were assessed for gross behavior every week from 12 weeks of age in ALS(SOD1 H46R ) mice and from 11 weeks of age in ALS(SOD1 G93A ) mice until the endpoint. In particular, hind limb movement and rearing behavior of each animal were monitored weekly from 12 weeks of age until the end stage. The days from the disease onset to the endpoint were counted individually as the survival interval.

Assessment of Motor Function
We assessed motor performance and coordination, and balance in the mice by the balance-beam test, vertical pole test, and footprint analysis as previously described in detail [10] and grip strength test as described above. In brief, the balance-beam test was performed every day after the onset until mice were unable to stay on the bar. Each mouse was given a maximum of five vertical pole trials using the following two criteria; 1) a cutoff of 45cm and 2) the task was terminated when mouse reached the vertical ascending distance of 45 cm. The highest vertical ascending distance was scored. The footprint trial was conducted at 18 and 21 weeks of age in ALS(SOD1 H46R ) mice and at 12, 16, and 20 weeks of age in ALS(SOD1 G93A ) mice. Grip strength test of ALS(SOD1 H46R ) mice was started at 17 weeks of age and performed once a week.
To evaluate the spontaneous motor activities in mice, we conducted cage activity and rearing performance tests by using SUPERMEX with infrared ray sensor monitor (Muromachi Kikai). Both cage activity and rearing performance were monitored for 2 consecutive nights starting at 18 weeks and 21 weeks of age in ALS(SOD1 H46R ) mice and at 13 weeks, 17 weeks and 21 weeks of age in ALS(SOD1 G93A ) mice. The cumulative counts of cage activity and rearing performance for the darkperiod (12 h; 19:00-7:00) were analyzed. The ALS(SOD1 H46R ) mice that accidentally died or showed unmeasurable scores on vertical pole test, grip strength test and/or stride length at 21-22 weeks of age (late-symptomatic stage) were omitted from the statistical analyses.
For the preparation of paraffin-embedding tissue, lumbar segment was post-fixed with the same fixative for 48 h at 4uC and embedded in paraffin. Serial transverse sections (6 mm thickness) of lumbar segment (L4-L5) were sliced and stained with hematoxylin and eosin (H&E) for histopathological evaluation.
For the preparation of frozen-embedding tissue, lumbar segment was post-fixed with 4% PFA in 0.1 M PB (pH 7.2) for 24 h at 4uC, and then processed in 30% sucrose in 0.1 M PB (pH 7.2) containing 0.1% NaN 3 for 96 h at 4uC (for cryoprotection) followed by frozen-embedding. Ten mm frozen sections were cut on a cryostat and used for fluorescent immunohistochemistry.
For double-immunostaining using frozen section, sections from late-symptomatic mice were washed with PBS, and treated with PBS containing 3% normal horse serum (NHS) (Vector Laboratories)/0.3% Triton X-100 for 1 h at room temperature. The sections were incubated with anti-IL-1b antibody in PBS containing 0.01% Triton X-100 for 48 h at 4uC, and then were incubated with CF TM 488A donkey anti-goat IgG (H+L) (1:400, Biotium) in PBS containing 0.01% Triton X-100 for 1 h at room temperature. Next, the sections were incubated with anti-GFAP, anti-Iba-1 or anti-MAP2 antibody in PBS containing 0.01% Triton X-100 overnight at 4uC, and then were incubated with CF TM 555 goat anti-rabbit IgG (H+L) (1:400, Biotium) in PBS containing 0.01% Triton X-100 for 1 h at room temperature. Fluorescent signals were observed under a light microscope (BIOREVO BZ-9000, KEYENCE).

Quantitative Analysis of the Number of Motor Neurons
Sections (6 mm thickness) of L4-L5 lumbar segment from latesymptomatic mice were immunostained with anti-ChAT (see above), a marker of motor neurons, and observed under a light microscope (BIOREVO BZ-9000, KEYENCE). A total of 10 representative images of every tenth serial section throughout L4-L5 segment were analyzed. A size of neuron (cross-sectional area of each soma) was also determined by utilizing BZ-II Analyzer (KEYENCE). ChAT positive neurons with soma sizes larger than 200 mm 2 in the anterior horn located within ventral half of the gray matter of the spinal cord were defined as motor neurons (see Figure 5D).

Preparation of Cell Extracts and Tissue Extracts
SH-SY5Y cells (3610 5 cells) were washed twice with ice-cold PBS and incubated in 1 ml of ice-cold 10% trichloroacetic acid for 30 min. Cells were centrifuged at 22,0006g for 15 min at 4uC and lysed by pipetting in 80 ml of 9 M urea containing 2% Triton X-100 and 1% dithiothreitol. After the addition of 20 ml of 10% lithium dodecyl sulfate and 2 ml of 2 M Tris base, cell lysates were sonicated until their viscosity disappeared.
Nuclear extracts were prepared using a nuclear extraction kit (Active Motif) according to the manufacturer's instruction. Protein concentration was quantified by Pierce 660 nm Protein Assay kit (Thermo Fisher Scientific Inc.) using bovine serum albumin as a standard.
At 21-22 weeks of age (late-symptomatic stage), mice were anesthetized with 2% isoflurane by inhalation and transcardially perfused with physiological saline containing 10% heparin, and the lumbar spinal cord was removed. We homogenized tissues in lysis buffer A [50 mM Tris-HCl (pH7.5), 150 mM NaCl, 1% NP-40, Complete Protease Inhibitor Cocktail (Roche), protein phosphatase inhibitor cocktail (Nacalai tesque)]. After centrifugation at 22,0006g for 30 min at 4uC, we collected the resulting supernatant as a NP-40-soluble fraction. The insoluble pellet fraction was washed with lysis buffer A three times, and then suspended with lysis buffer B [50 mM Tris-HCl (pH7.5), 150 mM NaCl, 5% (w/v) sodium dodecyl sulfate (SDS)], sonicated, and left for 30 min at room temperature. After the centrifugation at 22,0006g for 30 min, the supernatant was collected as a NP-40insoluble/SDS-soluble fraction. We determined protein concentration by the Pierce 660 nm Protein Assay system (Thermo Fisher Scientific).

Western Blotting
For Western blotting using cell extracts, equal amounts of protein were electrophoretically separated on 5-20% SDSpolyacrylamide gel (ATTO) and transferred onto polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories). Membrane was blocked with Blocking-One reagent (Nacalai Tesque) for 1 h at room temperature and then incubated with the appropriate primary antibody in Can Get Signal Solution 1 (TOYOBO) overnight at 4uC. The membrane was washed with TBST [20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% (w/v) Tween-20] three times, and incubated with horseradish peroxidaseconjugated secondary antibody in Can Get Signal Solution 2 (TOYOBO) for 1 h at room temperature. After washing with TBST, signals were visualized by Immobilon Western HRP Substrate (Millipore).
For Western blotting using tissue extracts, protein samples (2 mg) were electrophoretically separated on 5-20% SDS-polyacrylamide gel (ATTO) and transferred on to a PVDF membrane. The membrane was blocked with TBST containing 5% skimmedmilk for 1 h at room temperature and then incubated with the appropriate primary antibody in TBST containing 1% skimmedmilk overnight at 4uC. After washing with TBST buffer, the membrane was incubated with the peroxidase-conjugated secondary antibody in TBST containing 1% skimmed-milk for 1 h at room temperature. After washing with TBST, signals were visualized by Immobilon Western HRP substrate (Millipore).

Ventral Root Analysis
Under anesthesia with 2% isoflurane by inhalation, mice were transcardially perfused with physiological saline containing 10 U/ ml heparin, followed by 2% PFA/2% glutaraldehyde (GA) in 0.1 M PB (pH 7.2). Spinal cord was removed, and L5 ventral root was dissected and fixed with the same fixative for 48 h at 4uC and with 2.5% GA for 4 h at 4uC, followed by washing with 0.1 M PB (pH 7.2), and then post-fixed with 1% osmium tetroxide in 0.05 M PB (pH 7.2) for 2 h at 4uC. After dehydration in graded alcohol and in acetone, the tissues were embedded in epoxy resin. Semithin sections (2 mm thickness) of L5 ventral root were stained with 0.5% toluidine blue. The number of myelinated axons was observed under a light microscope (BIOREVO BZ-9000, KEY-ENCE). Axons over 4 mm in diameter per whole L5 ventral root were counted by utilizing BZ-II Analyzer (KEYENCE).

Morphological Analysis
Mice were euthanized by cervical dislocation after anesthesia with 2% isoflurane by inhalation. Fresh mouse muscle (quadriceps femoris) at a late-symptomatic stage (21-22 weeks of age) was harvested and flash frozen on 2-methylbutane cooled in liquid nitrogen. Transverse 10 mm sections were processed H&E and NADH histochemistry according to standard procedures.

Statistical Analysis
Data in this study were expressed as mean 6 SD or mean 6 SEM. Statistical analyses were conducted using PRISM 5 (GraphPad). Statistical significance was evaluated by ANOVA followed by appropriate post hoc test for multiple comparisons between groups and by unpaired Student's t-test for comparisons between the data of two groups. Survival data were compared using Kaplan-Meier survival analysis with Log-rank test. We considered p-values ,0.05 to be statistically significant. Figure S1 Workflow for in silico screening combined with in vitro drug screening of anti-oxidative stress compounds. In the first stage, we performed virtual designing of novel molecules based on the chemical structure of CPN-9 we identified previously [14]. For optimization of virtual designed molecules and CNS drugs, we defined the four criteria. According to these criteria, we selected active molecules, synthesized them, and then identified hit compound via the analysis of anti-oxidative stress cell death (AOSCD). In the second stage, we again virtually designed novel molecules based on the chemical structure of hit compound, and synthesized them. The AOSCD analysis was performed to select the promising compounds. (TIF) Figure S2 Effect of WN compounds against various cytotoxins. Differentiated SH-SY5Y cells were pretreated with 40 mM CPN-9, 10 mM WN1303R, 6 mM WN1312, 30 mM WN1315, 6 mM WN1316 or DMSO for 8 h at 37uC. The appropriate amount of each cytotoxin, which includes free radical generating compounds (a-naphthoquinone, 12.5 mM for 4 h; 6-OHDA, 400 mM for 4 h), protein kinase inhibitor (staurosporine, 25 mM for 6 h), and phosphatase inhibitor (okadaic acid, 1 mM for 6 h), was added, and incubated for another 4 to 6 h. The cell viability was measured by AlamarBlue assay. Data are expressed as mean 6 SD (n = 4). *p,0.001 by one-way ANOVA with Dunnett's post hoc test compared with DMSO-treated control. (TIF) Figure S3 NAC decreases the level of WN1316-induced Nrf2-regulated antioxidant proteins. (A) Effect of WN1316 on nuclear translocation of Nrf2. Differentiated SH-SY5Y cells were pre-incubated with 40 mM CPN-9, 10 mM WN1316, or DMSO for 3 h, and nuclear fractions were prepared using a nuclear extraction kit (Active Motif). The fractions were subjected to Western blot analysis using anti-Nrf2 antibody. Lamin B1 was used as an internal control for nuclear fraction. (B and C) Effect of NAC on the expression of mRNAs and proteins, including Nrf2, ATF3, Keap1, HO-1, NQO1, GCLM, p62 and p21, in WN1316treated cells. Differentiated SH-SY5Y cells were treated with 10 mM WN1316 or DMSO in the presence or absence of 2.5 mM NAC for 3 h. Expressions of the above-mentioned mRNAs and proteins were analyzed by RT-PCR (B) and Western blotting (C), respectively. PPIA was used as a control for RT-PCR. b-tubulin was used as a loading control for protein. (D) Effect of NAC and apocynin (APO) on the WN1316-mediated anti-oxidative stress activity. Differentiated SH-SY5Y cells were treated with 8 mM WN1316 or DMSO for 8 h in the presence or absence of NAC or APO, and then exposed to 60 mM menadione for 4 h. The cell viability was determined by AlamarBlue. Data are expressed as mean 6 SD (n = 4). *p,0.001 by one-way ANOVA with Dunnett's post hoc test compared with WN1316-treated cells.  Figure S7 Potential mechanism of WN1316-mediated neuroprotection against oxidative stress-induced cell death. In vitro pathway, WN1316 exerts neuroprotective potency against oxidative stress-induced cell death via the upregulation of endogenous NAIP and the activation of Nrf2 signaling cascade in an intracellular ROS-dependent manner and in a positive feedback manner between Nrf2-Keap1 complex and p62/p21. In vivo pathway, post-onset administration of WN1316 to ALS mouse models slows disease progression via the suppression of glial activation and neuronal inflammation by inhibiting the generation of inflammatory factor, the reduction of oxidative damage, and the protection of motor neurons and ventral motor axons. (TIF)