SOD1 mutations disrupt redox-sensitive Rac regulation of NADPH oxidase in a familial ALS model

Abstract

Neurodegeneration in familial amyotrophic lateral sclerosis (ALS) is associated with enhanced redox stress caused by dominant mutations in superoxide dismutase-1 (SOD1). SOD1 is a cytosolic enzyme that facilitates the conversion of superoxide (O(2)(*-)) to H(2)O(2). Here we demonstrate that SOD1 is not just a catabolic enzyme, but can also directly regulate NADPH oxidase-dependent (Nox-dependent) O(2)(*-) production by binding Rac1 and inhibiting its GTPase activity. Oxidation of Rac1 by H(2)O(2) uncoupled SOD1 binding in a reversible fashion, producing a self-regulating redox sensor for Nox-derived O(2)(*-) production. This process of redox-sensitive uncoupling of SOD1 from Rac1 was defective in SOD1 ALS mutants, leading to enhanced Rac1/Nox activation in transgenic mouse tissues and cell lines expressing ALS SOD1 mutants. Glial cell toxicity associated with expression of SOD1 mutants in culture was significantly attenuated by treatment with the Nox inhibitor apocynin. Treatment of ALS mice with apocynin also significantly increased their average life span. This redox sensor mechanism may explain the gain-of-function seen with certain SOD1 mutations associated with ALS and defines new therapeutic targets.

Figure 1. ALS-associated SOD1 mutations activate cellular Nox activity.

(A) Rate of NADPH-dependent O₂^•– production by total endomembranes isolated from the brain, spinal cord, and liver of nontransgenic or transgenic mice overexpressing SOD1^WT or SOD1^G93A (mean ± SEM; n = 3 per group). (B) DHE fluorescent detection of O₂^•– in lumbar spinal cord sections from nontransgenic and transgenic mice overexpressing SOD1^WT or SOD1^G93A. DAPI staining shows cell nuclei in each section. (C) Rate of NADPH-dependent O₂^•– production in total endomembranes isolated from SH-SY (neuronal) or MO59J (glial) cells at 48 hours following infection with adenoviral vectors expressing LacZ, SOD1^WT, SOD1^L8Q, or SOD1^G93A. (D) Cell death was quantified in SH-SY and MO59J cells using trypan-blue exclusion at 72 hours after infection with the indicated adenoviral vectors. (E) Using conditions specified in C and D, the rate of NADPH-dependent O₂^•– production and cell death was assessed in the presence or absence of the Nox inhibitor apocynin (100 μM). Values are mean ± SEM (n = 6 per group). (F) Assessment of GTP-bound Rac1 (activated form) in spinal cord lysates from 2 nontransgenic or 3 SOD1^G93A transgenic mice (120 days old) and from MO59J cells overexpressing WT or mutant SOD1 proteins (at 36 hours after adenoviral infection). The 2 left lanes are controls for the Rac activation assay, in which nontransgenic spinal cord lysates were preincubated with the indicated non-hydrolysable guanine nucleotide analogs.

Figure 2. Rac1 binds to SOD1 in a redox-dependent manner.

(A) Rac1 IP from heart, kidney, liver, and/or brain tissue of SOD1^+/+ or SOD1^–/– mice followed by Western blotting (WB) for SOD1 and Rac1. (B) In vitro IP of purified His-tagged Rac1 and Cdc42 in the presence of purified bovine SOD1 followed by Western blot for SOD1, Rac1, and Cdc42. The His-tagged GTPases were preloaded with the indicated nucleotide analogs prior to incubation with SOD1. (C) In vitro IP of purified His-tagged Rac1 in the presence of purified native, demetalated, or remetalated bovine SOD1 followed by Western blot for SOD1 and Rac1. The His-tagged Rac1 was preloaded with the indicated nucleotide analogs prior to incubation with SOD1. Additionally, untreated His-tagged Rac1 and 300 μM DTT prereduced His-tagged Rac1 were used for in vitro pulldown assays with each of the 3 forms of SOD1. (D) His-tagged Rac1 was prereduced (300 μM DTT), loaded with GTPγS, and treated with the indicated concentrations of H₂O₂ before performing pulldown assays with SOD1. (E) The indicated concentrations of DTT were added to the 300 pM H₂O₂-treated His-tagged Rac1 sample in D, and pulldown assays were performed with SOD1. (F) Schematic of GST-Rac1 deletion mutants used to define the SOD1 binding domain and in vitro IP of various GST-tagged Rac1 deletion mutants in the presence of purified bovine SOD1. GST-Rac1 fusion construct numbers above correspond with lane numbers below. Top lanes are Western blot for SOD1 following IP of GST; bottom lanes are Coomassie-stained gel of the purified fusion peptides used for IP.

Figure 3. SOD1 activates Rac1 and Nox2, a function dysregulated in certain SOD1 ALS mutants.

(A and B) Rac1 GTPase assays were performed in the presence or absence of purified (A) bovine SOD1 or (B) E. coli MnSOD and/or p29-GAP. His-tagged Rac1 was preloaded with γP³²-GTP, and aliquots of the reaction were analyzed at various time points by TLC for GTP hydrolysis by assessing percent ³²Pi released from Rac1. (C) GTPase assay for Rac1 and Cdc42 in the presence or absence of bovine SOD1 and/or p29-GAP. (D) Rac1 GTPase assay in the presence or absence of bovine SOD1 and/or X/XO (100 mU; final concentration 100 μM). (E) Pulldown assays of GTPγS-loaded His-tagged Rac1 in the presence of bovine SOD1 with or without 15-minute exposure to X/XO-derived ROS. Data are representative of at least 3 independent experiments. (F) Endosomes were isolated from PMDFs using Iodixanol density gradient fractionation, and fractions were characterized by Western blot for Nox2^gp91phox, Rac1, SOD1, and EEA1. (G) Lucigenin assays were used to assess the rate of NADPH-dependent O₂^•– production in fraction 10 vesicles from Nox2 WT and KO PMDFs in the presence or absence of 2.5 μM bovine SOD1 and/or 10 μM DPI, a general Nox inhibitor (n = 6). (H) Vesicular fractions from PMEFs heterozygous for a gene disruption of mouse SOD1 were assessed for NADPH-dependent O₂^•– production in the presence or absence of exogenously added 2.5 μM bovine SOD1 or 2.5 μM E. coli MnSOD (n = 3). (I) Coomassie-stained SDS-PAGE of purified bacterially expressed human SOD1 proteins. Bovine SOD1 was used as a reference control and migrates faster than human SOD1 (12). The Cu/Zn content of each SOD1 protein is shown. SOD activity gel of the bacterial purified SOD1 proteins is shown at bottom. (J) In vitro IP of purified prereduced His-tagged Rac1-GTPγS in the presence of the indicated human SOD1 proteins. Rac1/SOD1 complexes were then divided into 2 parts; 1 half was treated with X/XO-derived ROS for 15 minutes at room temperature prior to IP of the His-tag. Following IP, Western blots for SOD1 and Rac1 were performed. Long and short exposures of the SOD1 blot are shown to demonstrate enhanced binding of each of the mutant forms of SOD1 to Rac1. (K) Rac1 GTPase assays were performed using native bovine SOD1 or the indicated bacterially expressed purified human SOD1 proteins. The molar ratio of Rac1/SOD1 is indicated. (L) Time course of NADPH-dependent O₂^•– generation by isolated PMEF endosomes in the presence or absence of 1 μM human SOD1^WT or SOD1^L8Q.

Figure 4. Treatment with the Nox inhibitor apocynin increases lifespan and slows disease progression in mice hemizygous for the SOD1^G93A transgene.

(A) Kaplan-Meier survival curve for mice treated with indicated doses of apocynin in their water beginning at 14 days of age. n is shown along with median survival time (arrowhead) for each group. Survival differences were significant in all between-group comparisons (log-rank test). (B) Age of disease onset, as determined by a 10% weight loss from peak body weight, for the various doses of apocynin. (C) Dose affect of apocynin treatment on survival index, measured as time from disease onset (as determined by weight loss) until clinical death. (D) Rate of NADPH-dependent O₂^•– production in total endomembranes isolated from lumbar spinal cords of end-stage SOD1^G93Atransgenic mice (~120 days of age) that were either untreated or treated with apocynin (300 mg/kg) in the drinking water for 5 days prior to harvesting spinal cords (n = 5 per group). (E) DHE fluorescence was assessed in lumbar spinal cord sections from 2 mice evaluated in D. (F) Survival data of male and female mice at the indicated apocynin dose. Mice treated for eye infections with antibiotics are marked as squares; those unsuccessfully treated that died from eye infections are marked by “X” within the square. Circles denote animals that never contracted eye infection. n and mean survival time is indicated for each group. Data in B–D are mean ± SEM.

Figure 5. Redox-sensor model for SOD1-mediated regulation of Nox2 ROS production through Rac.

Under reducing conditions SOD1 is bound to Rac-GTP and stabilizes Rac activation by inhibiting intrinsic and GAP-mediated GTP hydrolysis. Increased Rac-GTP levels lead to activation of Nox2 and production of O₂^•–. O₂^•– generated by the Nox2 complex is converted to H₂O₂ by SOD1 or through spontaneous dismutation. As the local concentration of H₂O₂ rises, oxidation of Rac leads to the dissociation of SOD1. With SOD1 no longer bound to Rac-GTP, hydrolysis to Rac-GDP occurs more quickly, leading to inactivation of the Nox2 complex. SOD1 can then recycle to repeat the process as Rac/Nox2 is reactivated. Through this mechanism, we propose that SOD1 can sense the local concentration of ROS at sites of Rac/Nox2 complex activation and control the activity of the complex. In certain ALS mutants of SOD1, redox-dependent dissociation of SOD1 from Rac1 is impaired, leading to sustained activation of Rac1-GTP and higher levels of Nox2 activation.