Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disorder characterized by a selective loss of motor neurons in the brain and spinal cord. Multiple toxicity pathways, such as oxidative stress, misfolded protein accumulation, and dysfunctional autophagy, are implicated in the pathogenesis of ALS. However, the molecular basis of the interplay between such multiple factors in vivo remains unclear. Here, we report that two independent ALS-linked autophagy-associated gene products; SQSTM1/p62 and ALS2/alsin, but not antioxidant-related factor; NFE2L2/Nrf2, are implicated in the pathogenesis in mutant SOD1 transgenic ALS models. We generated SOD1H46R mice either on a Nfe2l2-null, Sqstm1-null, or Sqstm1/Als2-double null background. Loss of SQSTM1 but not NFE2L2 exacerbated disease symptoms. A simultaneous inactivation of SQSTM1 and ALS2 further accelerated the onset of disease. Biochemical analyses revealed that loss of SQSTM1 increased the level of insoluble SOD1 at the intermediate stage of the disease, whereas no further elevation occurred at the end-stage. Notably, absence of SQSTM1 rather suppressed the mutant SOD1-dependent accumulation of insoluble polyubiquitinated proteins, while ALS2 loss enhanced it. Histopathological examinations demonstrated that loss of SQSTM1 accelerated motor neuron degeneration with accompanying the preferential accumulation of ubiquitin-positive aggregates in spinal neurons. Since SQSTM1 loss is more detrimental to SOD1H46R mice than lack of ALS2, the selective accumulation of such aggregates in neurons might be more insulting than the biochemically-detectable insoluble proteins. Collectively, two ALS-linked factors, SQSTM1 and ALS2, have distinct but additive protective roles against mutant SOD1-mediated toxicity by modulating neuronal proteostasis possibly through the autophagy-endolysosomal system.

Introduction

Amyotrophic lateral sclerosis (ALS) is a heterogeneous group of devastating and incurable neuromuscular disorders characterized by a selective loss of motor neurons in the brain and spinal cord (1). Most patients die of respiratory failure within 3 to 5 years from the onset of symptoms. Recent genetic and molecular studies on the ALS pathogenesis have unveiled dozens of ALS causative as well as associated genes involving various toxicity pathways, such as oxidative stress, endoplasmic reticulum stress, excitotoxicity, mitochondrial dysfunction, neural inflammation, protein misfolding and accumulation, and dysfunctional intracellular trafficking (2–4). Although defining the functional interaction networks among dozens of ALS-associated factors at molecular levels is quite important to understand the pathogenesis of ALS, molecular basis of the interplay between such multiple factors in vivo are still largely unknown.

Mutations in Cu/Zn superoxide dismutase (SOD1)-encoding gene, SOD1, account for an approximately 10% of familial and 1% of sporadic cases (4). To date, more than 160 different SOD1 mutations have been identified (http://alsod.iop.kcl.ac.uk/als). Transgenic mice expressing mutant SOD1 have been shown to develop selective motor neuron degeneration with accompanying the accumulation of protein aggregates in the spinal cord (5,6), recapitulating the spectrum of disease in patients with ALS and other motor neuron diseases (MNDs), and thus have widely been used in studies on the molecular pathogenesis of ALS/MNDs (6). Although the precise nature of the mutant SOD1-mediated toxicity has not been fully uncovered, a large number of studies using these animal models has supported the notion that a progressive disturbance of proper proteome homeostasis (aka proteostasis) by the mutant SOD1-meidated dominant toxic function is primarily implicated in motor neuron degeneration (7).

We have previously shown that loss of ALS2/alsin, a causative gene product of a juvenile autosomal recessive form of ALS/MNDs (8–10); amyotrophic lateral sclerosis 2 (ALS2), juvenile primary lateral sclerosis (PLSJ), and infantile-onset ascending hereditary spastic paralysis (IAHSP) (10), hinders the autophagy-endolysosomal system and accelerates disease progression in mutant SOD1 (SOD1H46R) transgenic mice (11). Notably, ALS2-deficient SOD1H46R mice show a massive accumulation of SQSTM1 (sequestosome 1/p62)-positive aggregates in the spinal cord (11), resembling a typical pathological feature of familial as well as sporadic ALS and ALS-frontotemporal dementia (ALS-FTD) (12). It has also been shown that misfolded mutant SOD1 is recognized by SQSTM1, sequestered to form inclusions, and destined for autophagic degradation via the binding of SQSTM1 to microtubule-associated protein 1 light chain 3 (LC3) (13). Since ALS2, as a guanine nucleotide exchange factor (GEF) for the small GTPase Rab5 (14), involves not only in macropinocytosis-associated endosome trafficking and fusion (15,16) and neurite outgrowth (16,17), but also in amphisomal dynamics by modulating the fusion between endosomes and autophagosomes (18,19), it is conceivable that the ALS2-mediated autophagy-endolysosomal system might play a crucial role in proteostasis, including those for SQSTM1, and its dysfunction may result in motor neuron degeneration in mutant SOD1-expressing mice.

Importantly, several genetic studies have revealed the presence of missense mutations in the SQSTM1 gene in patients with familial as well as sporadic ALS and ALS-FTD (20–26). SQSTM1 mutations have originally been identified in patients with Paget disease of bone (PDB) (27), and its gene product, SQSTM1, is known to regulate not only the selective-autophagy via association with LC3 and ubiquitinated misfolded proteins (28,29), but also the kelch-like ECH-associated protein 1 (KEAP1)-nuclear factor erythroid 2-like 2 (NFE2L2/Nrf2) antioxidant pathway by interacting with KEAP1 (30). Genetic inactivation of Sqstm1 in mice results in the accumulation of hyperphosphorylated tau and neurodegeneration (31), the accelerated presentation of aging phenotypes (32), the enhanced α-synuclein pathology (33), and the exacerbation (34) or amelioration (35) of polyglutamine (polyQ)-linked diseases. However, no clear experimental evidence showing that SQSTM1 links to the pathogenesis of ALS has been demonstrated thus far. It is also worth noting that the expression of both SQSTM1 (36) and ALS2 (37) are transcriptionally regulated by NFE2L2 under oxidative stress conditions. These findings prompted us to speculate that molecular networks consisting of SQSTM1, KEAP1/NFE2L2, and/or ALS2 might play crucial roles in the maintenance of motor neurons by regulating oxidative stress and proteostasis in vivo.

To prove such hypothesis, we used SOD1H46R mice, which exhibit a widespread axonal degeneration with slowly progressive motor neuron degeneration in the spinal cord (11,38), and generated and characterized SOD1H46R mice on a Nfe2l2-null (39), Sqstm1-null (28), or Sqstm1/Als2-double null background (28,40). We here revealed that two independent ALS-linked autophagy-associated gene products; SQSTM1 and ALS2, but not antioxidant-related factor; NFE2L2, are linked to the pathogenesis in a mutant SOD1 transgenic mouse ALS model. Our findings suggest that SQSTM1 and ALS2 have additive protective roles against mutant SOD1-mediated toxic insults possibly via modulating the autophagy-endolysosomal system.

Results

Loss of NFE2L2 does not affect lifespan in SOD1H46R mice

To investigate the effect of NFE2L2 expression on disease progression and survival in SOD1H46R mice, we generated C57BL/6N (B6) congenic mice with six different genotypes; Nfe2l2+/+ (wild-type), Nfe2l2+/-, Nfe2l2-/-, SOD1H46R, Nfe2l2+/-;SOD1H46R, and Nfe2l2-/-;SOD1H46R, by crossing female Nfe2l2+/- and male Nfe2l2+/-;SOD1H46R mice (38,39). The mutant alleles were transmitted in the expected Mendelian ratio (data not shown). Both Nfe2l2+/- and Nfe2l2-/- mice were viable and fertile with no evidences for abnormalities as observed for at least 32 weeks of age. Kaplan-Meier survival analysis revealed that loss of NFE2L2 showed no observable effects on lifespan in SOD1H46R mice (Figure 1A,

), consistent with previous studies in which SOD1G93A mice were used (41,42). These results suggest that NFE2L2-mediated antioxidant pathway as well as the transcriptional regulation of SQSTM1 and/or ALS2 by NFE2L2 play minor roles in the pathogenesis of ALS that is linked to the SOD1 mutations.
Figure 1.

Loss of SQSTM1, but not NFE2L2, accelerates mutant SOD1-mediated motor dysfunction in mice. (A) Survival curves for SOD1H46R (H46R; filled circle; n = 28), Nfe2l2+/-;SOD1H46R (H46R;Nfe2l2Het; open square; n = 65), and Nfe2l2-/-;SOD1H46R (H46R; Nfe2l2KO; gray triangle; n = 18) mice. Kaplan-Meier analysis identifies no significant differences between groups. (B) Survival curves for SOD1H46R (H46R; filled circle; n = 49), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; open square; n = 98), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; gray triangle; n = 46) mice. Kaplan-Meier analysis identifies significant differences between SOD1H46R and Sqstm1-/-;SOD1H46R mice (Log-rank test; P < 0.0001), and between Sqstm1+/-;SOD1H46R and Sqstm1-/-;SOD1H46R mice (Log-rank test; P < 0.0001). (C) Survival curves for SOD1G93A (G93A; filled circle; n = 30), Sqstm1+/-;SOD1 G93A (G93A;Sqstm1Het; open square; n = 73), and Sqstm1-/-;SOD1G93A (G93A;Sqstm1KO; grey triangle; n = 40) mice. Kaplan-Meier analysis identifies significant differences between SOD1G93A and Sqstm1-/-;SOD1G93A mice (Log-rank test; P = 0.0006), and between SOD1G93A and Sqstm1+/-;SOD1H46R mice (Log-rank test; P = 0.0048). (D) Growth curves for female mice [wild-type (WT) (filled circle; n = 12–19), Sqstm1+/- (Sqstm1Het; open square; n = 10–30), Sqstm1-/- (Sqstm1KO; open triangle; n = 17–34), SOD1H46R (H46R; filled inverted-triangle; n = 5–20), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; filled diamond; n = 7–24), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; open circle; n = 6–27)]. (E) Growth curves for male mice [WT (filled circle; n = 17–26), Sqstm1+/- (Sqstm1Het; open square; n = 32–37), Sqstm1-/- (Sqstm1KO; open triangle; n = 27–35), SOD1H46R (H46R; filled inverted-triangle; n = 6–22), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; filled diamond; n = 15–43), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; open circle; n = 11–30)]. (D and E) In either gender, age at which body weight loss begins in Sqstm1-/-;SOD1H46R mice is earlier than that in SOD1H46R mice (female; ***P < 0.0001, at 20–22 weeks, male; ***P < 0.0001, at 22 weeks). There are no differences in the mean values between SOD1H46R and Sqstm1+/-;SOD1H46R mice at any ages. The values for SOD1H46R-expressing mice (SOD1H46R, Sqstm1+/-;SOD1H46R, and Sqstm1-/-;SOD1H46R) later than 5 weeks of ages are all significantly lower than those for WT animals (levels of significance are not shown). Values are mean ± SD. Statistical significance is evaluated by ANOVA with Bonferroni’s post hoc test. (F) Changes in the balance beam test scores in wild-type (WT; filled circle), Sqstm1-/- (Sqstm1KO; open triangle), SOD1H46R (H46R; filled inverted triangle), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; open circle) mice. Values are means ± SEM [WT; n = 16 (female; n = 8, male; n = 8), Sqstm1KO; n = 16 (female; n = 8, male; n = 8), H46R; n = 17 (female; n = 9, male; n = 8), and H46R;Sqstm1KO; n = 20 (female; n = 10, male; n = 10)]. Statistical significance is evaluated by one-way ANOVA with Bonferroni’s post hoc test. There are significant differences between SOD1H46R and Sqstm1-/-;SOD1H46R mice (*P < 0.05, **P < 0.01, or ***P < 0.0001 at 18–22 weeks of ages).

Figure 1.

Loss of SQSTM1, but not NFE2L2, accelerates mutant SOD1-mediated motor dysfunction in mice. (A) Survival curves for SOD1H46R (H46R; filled circle; n = 28), Nfe2l2+/-;SOD1H46R (H46R;Nfe2l2Het; open square; n = 65), and Nfe2l2-/-;SOD1H46R (H46R; Nfe2l2KO; gray triangle; n = 18) mice. Kaplan-Meier analysis identifies no significant differences between groups. (B) Survival curves for SOD1H46R (H46R; filled circle; n = 49), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; open square; n = 98), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; gray triangle; n = 46) mice. Kaplan-Meier analysis identifies significant differences between SOD1H46R and Sqstm1-/-;SOD1H46R mice (Log-rank test; P < 0.0001), and between Sqstm1+/-;SOD1H46R and Sqstm1-/-;SOD1H46R mice (Log-rank test; P < 0.0001). (C) Survival curves for SOD1G93A (G93A; filled circle; n = 30), Sqstm1+/-;SOD1 G93A (G93A;Sqstm1Het; open square; n = 73), and Sqstm1-/-;SOD1G93A (G93A;Sqstm1KO; grey triangle; n = 40) mice. Kaplan-Meier analysis identifies significant differences between SOD1G93A and Sqstm1-/-;SOD1G93A mice (Log-rank test; P = 0.0006), and between SOD1G93A and Sqstm1+/-;SOD1H46R mice (Log-rank test; P = 0.0048). (D) Growth curves for female mice [wild-type (WT) (filled circle; n = 12–19), Sqstm1+/- (Sqstm1Het; open square; n = 10–30), Sqstm1-/- (Sqstm1KO; open triangle; n = 17–34), SOD1H46R (H46R; filled inverted-triangle; n = 5–20), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; filled diamond; n = 7–24), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; open circle; n = 6–27)]. (E) Growth curves for male mice [WT (filled circle; n = 17–26), Sqstm1+/- (Sqstm1Het; open square; n = 32–37), Sqstm1-/- (Sqstm1KO; open triangle; n = 27–35), SOD1H46R (H46R; filled inverted-triangle; n = 6–22), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; filled diamond; n = 15–43), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; open circle; n = 11–30)]. (D and E) In either gender, age at which body weight loss begins in Sqstm1-/-;SOD1H46R mice is earlier than that in SOD1H46R mice (female; ***P < 0.0001, at 20–22 weeks, male; ***P < 0.0001, at 22 weeks). There are no differences in the mean values between SOD1H46R and Sqstm1+/-;SOD1H46R mice at any ages. The values for SOD1H46R-expressing mice (SOD1H46R, Sqstm1+/-;SOD1H46R, and Sqstm1-/-;SOD1H46R) later than 5 weeks of ages are all significantly lower than those for WT animals (levels of significance are not shown). Values are mean ± SD. Statistical significance is evaluated by ANOVA with Bonferroni’s post hoc test. (F) Changes in the balance beam test scores in wild-type (WT; filled circle), Sqstm1-/- (Sqstm1KO; open triangle), SOD1H46R (H46R; filled inverted triangle), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; open circle) mice. Values are means ± SEM [WT; n = 16 (female; n = 8, male; n = 8), Sqstm1KO; n = 16 (female; n = 8, male; n = 8), H46R; n = 17 (female; n = 9, male; n = 8), and H46R;Sqstm1KO; n = 20 (female; n = 10, male; n = 10)]. Statistical significance is evaluated by one-way ANOVA with Bonferroni’s post hoc test. There are significant differences between SOD1H46R and Sqstm1-/-;SOD1H46R mice (*P < 0.05, **P < 0.01, or ***P < 0.0001 at 18–22 weeks of ages).

Loss of SQSTM1 shortens lifespans in SOD1H46R and SOD1G93A mice

To investigate the effect of SQSTM1 expression on disease progression and survival in mutant SOD1-expressing ALS mouse models, we generated B6 congenic lines with six different genotypes; Sqstm1+/+ (wild-type), Sqstm1+/-, Sqstm1-/-, SOD1H46R, Sqstm1+/-;SOD1H46R, and Sqstm1-/-;SOD1H46R, by crossing female Sqstm1+/- and male Sqstm1+/-;SOD1H46R mice (28,38). The mutant alleles were transmitted in the expected Mendelian ratio (data not shown). Both Sqstm1+/- and Sqstm1-/- mice were viable and fertile as observed for at least 32 weeks of age. Kaplan-Meier survival analysis revealed that Sqstm1-/-;SOD1H46R mice died significantly earlier than did either SOD1H46R or Sqstm1+/-;SOD1H46R mice (Figure 1B,

). In order to confirm the effects of SQSTM1 loss in the different type of an ALS mouse model, we produced SOD1G93A mice (5) on a Sqstm1-null background in similar manners. Similar to SOD1H46R mice, Sqstm1-/-;SOD1G93A mice had a shorter lifespan than did SOD1G93A, although the effects of SQSTM1 loss in SOD1G93A mice were much smaller than those in SOD1H46R mice (Figure 1B and C). It is noted that unlike in the case of SOD1H46R mice, Sqstm1+/-;SOD1G93A mice showed a comparable lifespan to Sqstm1-/-;SOD1G93A mice (Figure 1C, ). Since the effects of SQSTM1 loss were much stronger in SOD1H46R mice when compared to SOD1G93A mice, we chose SOD1H46R lines on a Sqstm1-null background in the following experiments, and characterized them in more detail to clarify the effects of SQSTM1 loss in the pathogenesis of mutant SOD1-mediated MNDs.

Loss of SQSTM1 accelerates body weight declines in SOD1H46R mice

We first investigated the changes in body weight, since the onset of disease in mutant SOD1 transgenic mice on a B6 background could be estimated by the reduction of body weight (43–45). Wild-type, Sqstm1+/-, and Sqstm1-/- mice showed constant increases in their body weight, while mice carrying the SOD1H46R transgene reached their maximum body weight at 14-16 weeks of age, and terminally decreased as disease symptoms progressed (Figure 1D and E). SOD1H46R-expressing mice had a significantly lower weight than those without SOD1 mutant throughout the entire experimental periods of ages as previously reported (45). Although Sqstm1-/- mice, in particular male ones, showed obese phenotypes, which was in line with previous findings (46), Sqstm1-/-;SOD1H46R mice had comparable body weights to either SOD1H46R or Sqstm1+/-;SOD1H46R mice prior to disease onset, demonstrating no evidence for obese phenotypes in SOD1H46R mice lacking SQSTM1. Notably, SQSTM1-deficient SOD1H46R (Sqstm1-/-;SOD1H46R) mice showed a marked and earlier decrease in their body weight compared to either SOD1H46R or Sqstm1+/-;SOD1H46R littermates (Figure 1D and E), suggesting that loss of SQSTM1 accelerates disease onset in SOD1H46R mice.

Loss of SQSTM1 aggravates motor dysfunction in SOD1H46R mice

We next assessed whether loss of SQSTM1 in SOD1H46R mice affects the course of motor deficits. To evaluate motor coordination and balance quantitatively, we performed balance beam test, with which the onset of disease was sensitively determined (47). Sqstm1-/-;SOD1H46R mice showed earlier motor dysfunction (∼ 17 weeks of age) than did SOD1H46R littermates (∼20 weeks of age) (Figure 1F), indicating an earlier disease onset in SQSTM1-deficient SOD1H46R mice.

Loss of SQSTM1 does not affect bone morphology

Since mutations in SQSTM1 were originally identified in patients with PDB (27,48) and also caused a PDB-like skeletal disorder in mice (49), we investigated whether loss of SQSTM1 resulted in abnormal skeletal phenotypes in SOD1H46R mice. Both X-ray analysis of the whole body and histological examinations of femoral long bones using hematoxylin and eosin (H&E) staining revealed no observable differences among mice with four distinct genotypes; wild-type, Sqstm1-/-, SOD1H46R, and Sqstm1-/-;SOD1H46R mice (

). Thus, loss of SQSTM1 does not affect bone tissues, at least, up to 20 weeks of age, where SOD1H46R mice explicitly show progressive motor dysfunction.

Loss of SQSTM1 in SOD1H46R mice results in premature degeneration of motor neurons in the spinal cord

To determine whether the accelerated motor deficits observed in SQSTM1-deficient SOD1H46R mice were associated with motor neuron degeneration, we conducted histological analyses using early-to-intermediate symptomatic mice (16-20 weeks of age). Although there were no evidences for the motor neuron loss, a wide-spread axonal degeneration in the spinal tracts of the lateral and ventral columns was evident in SOD1H46R mice. Remarkably, SQSTM1-deficient SOD1H46R mice, in particular, those at 20 weeks of age, exhibited premature degenerative phenotypes not only in the spinal tracts but also in anterior horn cells (shrunken motor neurons with disorganized nuclear envelope and perinuclear vacuoles) of the spinal cord (Figure 2). By contrast, no structural and histological abnormalities in the brain among all examined animals at 20 weeks of age were observed (

).
Figure 2.

SOD1H46R mice lacking SQSTM1 exhibit premature degenerative phenotypes in the spinal cord. Representative Nissl and toluidine blue (TB) stainings of the transverse section of lumbar spinal cord (L4-L5) from wild-type (WT, 16 weeks of age), Sqstm1-/- (Sqstm1KO, 16 and 20 weeks of ages), SOD1H46R (H46R, 16 and 20 weeks of ages), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO, 16 and 20 weeks of ages) mice. Images for Nissl staining of anterior horns (Anterior H.), and those for TB staining of ventral columns (Ventral C.), lateral columns (Lateral C.), and dorsal columns (Dorsal C.) are shown. Red arrowheads indicate degenerating motor neurons and axons. Axonal and motor neuron degeneration are most prominent in 20-week old Sqstm1-/-;SOD1H46R mice. Scale bars = 50 µm.

Figure 2.

SOD1H46R mice lacking SQSTM1 exhibit premature degenerative phenotypes in the spinal cord. Representative Nissl and toluidine blue (TB) stainings of the transverse section of lumbar spinal cord (L4-L5) from wild-type (WT, 16 weeks of age), Sqstm1-/- (Sqstm1KO, 16 and 20 weeks of ages), SOD1H46R (H46R, 16 and 20 weeks of ages), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO, 16 and 20 weeks of ages) mice. Images for Nissl staining of anterior horns (Anterior H.), and those for TB staining of ventral columns (Ventral C.), lateral columns (Lateral C.), and dorsal columns (Dorsal C.) are shown. Red arrowheads indicate degenerating motor neurons and axons. Axonal and motor neuron degeneration are most prominent in 20-week old Sqstm1-/-;SOD1H46R mice. Scale bars = 50 µm.

To investigate the histopathology in the spinal cord in more detail, we conducted an electron microscopic (EM) analysis (Figure 3). In contrast to wild-type and SOD1H46R mice at 20 weeks of age, in which there were no detectable abnormalities in some of motor neurons (Figure 3A and B), we observed a large number of degenerative motor neurons in Sqstm1-/-;SOD1H46R mice (Figure 3C and D). In particular, perinuclear condensation of disorganized cellular organelle, such as mitochondria, endoplasmic reticulum, and Golgi apparatus, due to loss of cytosolic compartment was prominently observed in shrunken neurons from Sqstm1-/-;SOD1H46R mice (Figure 3c1, d1–d3). Further, degenerating axons (Figure 3G), membrane saccules containing granular/osmiophilic aggregates and small membranous vesicles (Figure 3F and H), degenerating cellular remnants and/or debris (Figure 3I, i–1i3), axons containing osmiophilic and autophagosome-like vesicles (Figure 3J), and swollen axon accumulating granular aggregates and vesicles (Figure 3K and L) were observed in the spinal cord of SOD1H46R and Sqstm1-/-;SOD1H46R mice (Figure 3E–L). Disorganization of axonal microtubules was also evident (Figure 3L1). These data underscore the relevance of axonal dysfunction on the disease onset in SOD1H46R-expressing mice, presumably due to the impaired trafficking of proteins as well as endosomal/autophagosomal vesicles in the axons. Further, loss of SQSTM1 is likely to be associated with earlier motor neuron loss in the spinal cord, leading to an earlier disease onset and accelerated motor dysfunction in SQSTM1-deficient SOD1H46R mice.

Figure 3.

SOD1H46R mice lacking SQSTM1 show premature motor neuron and axonal degeneration in the spinal cord. (A–L) Electron micrographs of lumbar (L4-L5) spinal cords from wild-type (WT), SOD1H46R (H46R), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO) mice at 20 weeks of age. Representative images from anterior horns (Anterior H.) (A–D), lateral columns (Lateral C.) (E–I), and ventral columns (Ventral C.) (J–L) are shown. Higher magnification images for H46R;Sqstm1KO mouse (c1, d1–d3, i1-i3, and l1) are also shown. Degenerating motor neurons, which exhibit the condensed accumulation of disorganized cellular organelle, such as mitochondria, endoplasmic reticulum, and Golgi apparatus (C, c1, D, and d1-d3), degenerating axons (arrowheads, G and K), membrane saccules containing granular/osmiophilic aggregates and small membranous vesicles (arrowheads, F and H), degenerating cellular remnants and/or debris (I, i1-i3), axon containing osmiophilic and autophagosome-like vesicles (J), and swollen axon accumulating granular aggregates and vesicles (K, L, and l1), are shown. Scale bars are as indicated.

Figure 3.

SOD1H46R mice lacking SQSTM1 show premature motor neuron and axonal degeneration in the spinal cord. (A–L) Electron micrographs of lumbar (L4-L5) spinal cords from wild-type (WT), SOD1H46R (H46R), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO) mice at 20 weeks of age. Representative images from anterior horns (Anterior H.) (A–D), lateral columns (Lateral C.) (E–I), and ventral columns (Ventral C.) (J–L) are shown. Higher magnification images for H46R;Sqstm1KO mouse (c1, d1–d3, i1-i3, and l1) are also shown. Degenerating motor neurons, which exhibit the condensed accumulation of disorganized cellular organelle, such as mitochondria, endoplasmic reticulum, and Golgi apparatus (C, c1, D, and d1-d3), degenerating axons (arrowheads, G and K), membrane saccules containing granular/osmiophilic aggregates and small membranous vesicles (arrowheads, F and H), degenerating cellular remnants and/or debris (I, i1-i3), axon containing osmiophilic and autophagosome-like vesicles (J), and swollen axon accumulating granular aggregates and vesicles (K, L, and l1), are shown. Scale bars are as indicated.

Loss of SQSTM1 promotes the progressive accumulation of insoluble SOD1 in the spinal cord of SOD1H46R mice

We first confirmed that there were no significant differences in the expression levels of the SOD1H46R transcript (

) and soluble SOD1H46R protein (Figure 4A and B) between SOD1H46R and Sqstm1-/-;SOD1H46R mice, although the SOD1H46R transcript in SOD1H46R mice at 20 weeks of age was significantly lower than those at either 12 or 16 weeks of age (). The results indicate that the exacerbation of motor neuron degeneration observed in SQSTM1-deficient SOD1H46R mice is not simply due to the increased expression of the mutant SOD1 protein.
Figure 4.

SOD1H46R mice lacking SQSTM1 show a progressive accumulation of insoluble SOD1 in the spinal cord. (A) Western blot analysis of SOD1, SQSTM1, and ubiquitin (Ub) in the spinal cord from 12, 16, and 20 week-old mice with four distinct genotypes; wild-type (WT), Sqstm1-/- (Sqstm1KO), SOD1H46R (H46R), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO). Two fractions; 1% Triton X-soluble fraction (TX-soluble; left panels) and 1% Triton X-insoluble/5% SDS-soluble fraction (TX-insoluble; right panels) are analyzed. SOD1_mono and SOD1_HMW represent monomeric and high molecular-weight (aggregated) forms of SOD1, respectively. Ub_mono and Ub_HMW represent monomeric Ub and the polyubiquitinated proteins, respectively. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and β-actin are used for a loading control in TX-soluble and -insoluble fractions, respectively. The positions of size-markers are shown on the left. (B-D) Quantitative analysis of western blotting for SOD1, SQSTM1, and Ub. (B) Quantitation of TX-soluble monomeric SOD1 (SOD1_mono; upper panel), TX-insoluble monomeric SOD1 (SOD_mono; middle panel), and TX-insoluble high molecular-weight SOD1 (SOD1_HMW; lower panel). (C) Quantitation of TX-soluble (upper panel) and -insoluble (lower panel) SQSTM1. (D) Quantitation of TX-soluble (upper panel) and -insoluble (lower panel) polyubiquitinated proteins (Ub_HMW). Densitometric data for immunoreactive signals in TX-soluble and -insoluble fractions are normalized by the levels of GAPDH and β-actin, respectively. Values are mean ± SEM (n = 4) in an arbitrary unit relative to 12 week-old wild-type mice except for the case of SOD1, in which an arbitrary unit relative to SOD1H46R mice is used. Statistical significance is evaluated by two-way ANOVA with Bonferroni’s post hoc test (comparisons between different genotypes in the same age; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001) or one-way ANOVA with Bonferroni’s post hoc test (comparisons between different ages in the same genotype; §P < 0.05, §§P < 0.01, §§§P < 0.001, §§§§P < 0.0001).

Figure 4.

SOD1H46R mice lacking SQSTM1 show a progressive accumulation of insoluble SOD1 in the spinal cord. (A) Western blot analysis of SOD1, SQSTM1, and ubiquitin (Ub) in the spinal cord from 12, 16, and 20 week-old mice with four distinct genotypes; wild-type (WT), Sqstm1-/- (Sqstm1KO), SOD1H46R (H46R), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO). Two fractions; 1% Triton X-soluble fraction (TX-soluble; left panels) and 1% Triton X-insoluble/5% SDS-soluble fraction (TX-insoluble; right panels) are analyzed. SOD1_mono and SOD1_HMW represent monomeric and high molecular-weight (aggregated) forms of SOD1, respectively. Ub_mono and Ub_HMW represent monomeric Ub and the polyubiquitinated proteins, respectively. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and β-actin are used for a loading control in TX-soluble and -insoluble fractions, respectively. The positions of size-markers are shown on the left. (B-D) Quantitative analysis of western blotting for SOD1, SQSTM1, and Ub. (B) Quantitation of TX-soluble monomeric SOD1 (SOD1_mono; upper panel), TX-insoluble monomeric SOD1 (SOD_mono; middle panel), and TX-insoluble high molecular-weight SOD1 (SOD1_HMW; lower panel). (C) Quantitation of TX-soluble (upper panel) and -insoluble (lower panel) SQSTM1. (D) Quantitation of TX-soluble (upper panel) and -insoluble (lower panel) polyubiquitinated proteins (Ub_HMW). Densitometric data for immunoreactive signals in TX-soluble and -insoluble fractions are normalized by the levels of GAPDH and β-actin, respectively. Values are mean ± SEM (n = 4) in an arbitrary unit relative to 12 week-old wild-type mice except for the case of SOD1, in which an arbitrary unit relative to SOD1H46R mice is used. Statistical significance is evaluated by two-way ANOVA with Bonferroni’s post hoc test (comparisons between different genotypes in the same age; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001) or one-way ANOVA with Bonferroni’s post hoc test (comparisons between different ages in the same genotype; §P < 0.05, §§P < 0.01, §§§P < 0.001, §§§§P < 0.0001).

We next performed western blot analysis to identify the molecules that were associated with behavioural and pathological features observed in SQSTM1-deficient SOD1H46R mice. Progressive accumulation of insoluble high-molecular weight SOD1 (SOD1_HMW) and polyubiquitinated proteins (Ub_HMW) was observed in the spinal cord from mice expressing mutant SOD1 (Figure 4). Remarkably, a quantitative analysis of 4 independent animals with each respective genotype showed that the average level of SOD1_HMW in insoluble fractions from Sqstm1-/-;SOD1H46R mice at 20 week of age was significantly higher than that from SOD1H46R mice (Figure 4B). It was also noted that two glial intermediate filament proteins, vimentin and glial fibrillary acidic protein (GFAP), were both progressively increased in the spinal cord as the disease progressed (

). A quantitative reverse transcriptase PCR (qRT-PCR) revealed that although there was no transcriptional activation of SOD1, Sqstm1, Map1lc3a (LC3A), and Map1lc3b (LC3B), expressions of Nes, Gfap, and Aif1 were significantly up-regulated as the disease deteriorated (), indicating a progressive glial cell activation and neuronal inflammation in the spinal cord. By contrast, no increased levels of such insoluble proteins as well as transcripts in the cerebral cortex were observed (). Thus, adverse effects either by the systemic expression of mutant SOD1 or by loss of SQSTM1, or by both of them, may not widely occur throughout the central nervous system (CNS) and other tissues, but are rather confined to particular regions of the CNS such as the spinal cord.

Taken together with the fact that mutant SOD1 is degraded by both the proteasome and autophagy (13,50,51), impairment of the ubiquitin-proteasome system (UPS) and/or the autophagy-endolysosomal system might result in the accumulation of such insoluble proteins in SOD1H46R mice, particularly in those lacking SQSTM1.

Distinct localization of ubiquitin-positive aggregates in the spinal cord between SOD1H46R and SQSTM1-deficient SOD1H46R mice

To investigate the cellular localization of the MND-associated factors in the spinal cord, we performed immunohistochemical analysis. Higher levels of SOD1-immunoreactivities were seen in SOD1H46R-expressing mice (

). However, unlike in SOD1G93A mice (52), no obvious SOD1-positive aggregates and/or inclusions were detected, at least, in the samples from early-to-intermediate symptomatic SOD1H46R mice (Figure 5B and E, ), supporting the notion that the SOD1H46R mutant protein was less prone to form visible aggregates (11,53).
Figure 5.

Distinct localization of disease-associated proteins in the spinal cord between SOD1H46R and SQSTM1-deficient SOD1H46R mice. (A-C) Representative images of double immunostaining with ubiquitin (green) and SQSTM1 (red) (A), ubiquitin (green) and SOD1 (red) (B), and ubiquitin (green) and GFAP (red) (C) in the ventral horn of the lumbar spinal cord (L4-L5) from wild-type (WT, 1st row), SOD1H46R (H46R, 2nd row), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO, 3rd row) mice at 20 weeks of age. (D and E) Representative images of double immunostaining with MAP2 (green) and ubiquitin (red) (D), and SDO1 (green) and SQSTM1 (red) (E) in the ventral horn of the lumbar spinal cord (L4-L5) from SOD1H46R (H46R, upper row) and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO, lower row) mice at 20 weeks of age. The nuclei are counterstained with DAPI (Blue). Right columns display the merged images. It is notable that most ubiquitin-positive aggregates are observed in non-neuronal cells and/or extracellular spaces (white arrows) and are rarely seen in neurons of SOD1H46R mice, whereas SQSTM1-deficient SOD1H46R mice exhibit the preferential localization of ubiquitin-positive inclusions in soma of neurons (yellow arrowheads). Scale bars = 20 μm.

Figure 5.

Distinct localization of disease-associated proteins in the spinal cord between SOD1H46R and SQSTM1-deficient SOD1H46R mice. (A-C) Representative images of double immunostaining with ubiquitin (green) and SQSTM1 (red) (A), ubiquitin (green) and SOD1 (red) (B), and ubiquitin (green) and GFAP (red) (C) in the ventral horn of the lumbar spinal cord (L4-L5) from wild-type (WT, 1st row), SOD1H46R (H46R, 2nd row), and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO, 3rd row) mice at 20 weeks of age. (D and E) Representative images of double immunostaining with MAP2 (green) and ubiquitin (red) (D), and SDO1 (green) and SQSTM1 (red) (E) in the ventral horn of the lumbar spinal cord (L4-L5) from SOD1H46R (H46R, upper row) and Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO, lower row) mice at 20 weeks of age. The nuclei are counterstained with DAPI (Blue). Right columns display the merged images. It is notable that most ubiquitin-positive aggregates are observed in non-neuronal cells and/or extracellular spaces (white arrows) and are rarely seen in neurons of SOD1H46R mice, whereas SQSTM1-deficient SOD1H46R mice exhibit the preferential localization of ubiquitin-positive inclusions in soma of neurons (yellow arrowheads). Scale bars = 20 μm.

Most prominent feature found in this study was the accumulation of ubiquitin-positive aggregates and/or inclusions in the spinal cord of SOD1H46R-expressing mice. A large number of ubiquitin-positive aggregates were observed throughout white and grey matters of the spinal cord of both SOD1H46R and Sqstm1-/-;SOD1H46R mice at 20 weeks of age (

). The accumulation of such ubiquitin-positive aggregates in 20-week old mice was more prominent compared to those at 16 weeks of age (). SQSTM1-positive aggregates, which were partially colocalized with ubiquitin-positive signals, were also progressively increased in SOD1H46R mice (Figure 5A and E, ). Thus, overexpression of mutant SOD1 might result in the progressive impaired clearance of polyubiquitinated proteins as well as SQSTM1 in the spinal cord.

Next, we more closely analyzed the subcellular localization of those ubiquitin-positive aggregates in ventral horn, and revealed that most of large ubiquitin-positive aggregates observed in SOD1H46R mice at 20 weeks of age were localized to extracellular and/or non-neuronal DAPI-positive nuclei (Figure 5A–E; white arrows), but were rarely seen within motor neurons at this stage of disease. Occasionally, ubiquitin-positive signals were also observed in the nuclei of GFAP-positive astrocytes in SOD1H46R mice (Figure 5C). Remarkably, contrary to SOD1H46R mice, the cytoplasmic ubiquitin-positive inclusions in intact/survived motor neurons (Figure 5A–D; yellow arrowheads), including MAP2-positive neurons (Figure 5D), were more prominent in the same aged Sqstm1-/-;SOD1H46R mice. These results illustrate a distinct cellular and subcellular localization of ubiquitin-positive aggregates between SOD1H46R and Sqstm1-/-;SOD1H46R mice at 20 weeks of age, indicating that loss of SQSTM1 is associated with the impaired clearance of misfolded and/or aggregated proteins, unless otherwise to be degraded, in neurons, thereby accelerating motor neuron degeneration in the spinal cord.

Loss of SQSTM1 does not have a major impact on the autophagic flux in the spinal cord of SOD1H46R mice

Since SQSTM1 was believed to act as an adaptor and/or a cargo receptor in the autophagy-endolysosomal system (28), we investigated to what extent loss of SQSTM1 affected the autophagic flux in vivo using SOD1H46R or Sqstm1-/-;SOD1H46R mice expressing GFP-fused human-MAP1LC3B (microtubule-associated protein 1-light chain 3) (GFP-LC3) (54).

Prior to this analysis, we first investigated whether overexpression of GFP-LC3 affected the disease-onset and/or the course of disease progression in SOD1H46R mice. We generated B6 congenic mice with three different genotypes; GFP-MAP1LC3B, SOD1H46R;GFP-MAP1LC3B, and Sqstm1-/-;SOD1H46R;GFP-MAP1LC3B, by crossing female Sqstm1+/-;GFP-MAP1LC3B and male Sqstm1+/-;SOD1H46R mice, or female Sqstm1+/- and male Sqstm1+/-;SOD1H46R;GFP-MAP1LC3B mice. The mutant alleles were transmitted in the expected Mendelian ratio (data not shown). Kaplan-Meier survival analysis revealed that overexpression of GFP-LC3 in SOD1H46R mice did not affect their lifespans (

). Further, we observed neither beneficial nor detrimental effects of GFP-LC3 overexpression on the disease-onset and the course of disease progression in these mice (data not shown).

Next, we investigated the effect of SQSTM1 loss on the levels of a lipidated-form of LC3 (LC3-II), which was degraded by the autophagy-lysosomal system as did SQSTM1 (28). In SOD1H46R;GFP-MAP1LC3B mice, the levels of both SQSTM1 and LC3-II were significantly increased in insoluble fractions as the disease progressed (

). Similarly, LC3-II levels were increased in Sqstm1-/-;SOD1H46R;GFP-MAP1LC3B mice. By contrast, no increased levels of LC3-II in the cerebral cortex from both SOD1H46R;GFP-MAP1LC3B and Sqstm1-/-;SOD1H46R;GFP-MAP1LC3B mice were observed (). These results combined with the histological data indicate that the autophagy-endolysosomal system in the spinal cord is progressively impaired by the mutant SOD1-mediated toxic insults. Nonetheless, the LC3-II to LC3-I ratio (LC3-II/LC3-I), as a marker of the autophagosome formation, of Sqstm1-/-;SOD1H46R;GFP-MAP1LC3B mice was comparable to those of SOD1H46R;GFP-MAP1LC3B mice (), suggesting that loss of SQSTM1 did not seem to have a major impact on the autophagic flux in the spinal cord of mutant SOD1-expressing mice, at least, up to 20 weeks of age.

Simultaneous loss of SQSTM1 and ALS2 further accelerates disease phenotypes associated with mutant SOD1 expression in mice

We have previously demonstrated that loss of ALS2, a causative for a form of juvenile ALS (8,9), exacerbates motor dysfunction in SOD1H46R mice by disturbing endolysosomal trafficking (11). In this study, we demonstrated that motor dysfunction observed in a SOD1H46R-expressing ALS mouse model deteriorated by loss of other ALS-linked protein SQSTM1 (2026). These findings combined suggest that both ALS2 and SQSTM1 play neuroprotective roles against mutant SOD1-mediated insults via modulating the autophagy-endolysosomal system in vivo. However, it is still unknown whether and how SQSTM1 and ALS2 are functionally interrelated in the pathogenesis of mutant SOD1-mediated MNDs. In order to clarify the functional interaction network among different ALS-associated factors at molecular levels, we created and analyzed SQSTM1 and ALS2 double-knocked out mutant SOD1-expressing mice.

We generated B6 congenic mice with nine different genotypes; SOD1H46R, Sqstm1+/-;SOD1H46R, Sqstm1-/-;SOD1H46R, Als2+/-;SOD1H46R, Als2-/-;SOD1H46R, Als2+/-;Sqstm1+/-;SOD1H46R, Als2+/-;Sqstm1-/-;SOD1H46R, Als2-/-;Sqstm1+/-;SOD1H46R, and Als2-/-;Sqstm1-/-;SOD1H46R mice, by a series of crossings between animal pairs with appropriate genotypes. Both Als2-/- (40) and Als2-/-;Sqstm1-/- mice were viable and fertile as observed for at least 32 weeks of age. First, we compared the lifespans between SOD1H46R and three different types of heterozygous mice; Sqstm1+/-;SOD1H46R, Als2+/-;SOD1H46R, and Als2+/-;Sqstm1+/-;SOD1H46R, by Kaplan-Meier analysis. There were no significant differences between groups (Figure 6A,

), indicating that a partial loss of neither SQSTM1 nor ALS2, nor a combination of them, was sufficient to accelerate the mutant-SOD1 mediated MNDs in mice. Next, we compared the lifespans among animals with a combination of homozygous and heterozygous mice; SOD1H46R, Sqstm1-/-;SOD1H46R, Als2-/-;SOD1H46R, Als2+/-;Sqstm1-/-;SOD1H46R, Als2-/-;Sqstm1+/-;SOD1H46R, and Als2-/-;Sqstm1-/-;SOD1H46R. Although there were no differences between Sqstm1-/-;SOD1H46R and Als2+/-;Sqstm1-/-;SOD1H46R, and between Als2-/-;SOD1H46R and Als2-/-;Sqstm1+/-;SOD1H46R mice, the degree of the accelerated disease phenotypes by lack of SQSTM1 was much larger than those by lack of ALS2 (Figure 6B, ). Most importantly, a simultaneous inactivation of SQSTM1 and ALS2 further shortened their lifespans compared to mice with any other combinations of genotypes examined (Figure 6B, ). The results strongly indicate that loss of SQSTM1 and ALS2 has additive adverse effects on the disease phenotypes in mutant SOD1-expressing mice.
Figure 6.

Simultaneous loss of SQSTM1 and ALS2 accelerates disease phenotypes associated with mutant SOD1 expression in mice. (A) Survival curves for SOD1H46R (H46R; filled circle; n = 105), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; open square; n = 107), Als2+/-;SOD1H46R (H46R;Als2Het; open inverted triangle; n = 25), and Als2+/-;Sqstm1+/-;SOD1H46R (H46R;Als2Het;Sqstm1Het; gray diamond; n = 45) mice. Kaplan-Meier analysis identifies no significant differences between groups. (B) Survival curves for SOD1H46R (H46R; filled circle; n = 105), Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; gray triangle; n = 49), Als2+/-;Sqstm1-/-;SOD1H46R (H46R;Als2Het;Sqstm1KO; open circle; n = 22), Als2-/-;SOD1H46R (H46R;Als2KO; open square; n = 11), Als2-/-;Sqstm1+/-;SOD1H46R (H46R;Als2KO;Sqstm1Het; gray inverted triangle; n = 55), and Als2-/-;Sqstm1-/-;SOD1H46R (H46R;Als2KO;Sqstm1KO; filled diamond; n = 21) mice. Kaplan-Meier analysis identifies significant differences between SOD1H46R and Sqstm1-/-;SOD1H46R (Log-rank test; P < 0.0001), between SOD1H46R and Als2-/-;SOD1H46R (Log-rank test; P < 0.025), between SOD1H46R and Als2-/-;Sqstm1-/-;SOD1H46R (Log-rank test; P < 0.0001), between Sqstm1-/-;SOD1H46R and Als2-/-;SOD1H46R (Log-rank test; P < 0.0001), between Sqstm1-/-;SOD1H46R and Als2-/-;Sqstm1-/-;SOD1H46R (Log-rank test; P < 0.0001), and between Als2-/-;SOD1H46R and Als2-/-;Sqstm1-/-;SOD1H46R mice (Log-rank test; P < 0.0001), while no significant differences between Sqstm1-/-; SOD1H46R and Als2+/-;Sqstm1-/-; SOD1H46R, and between Als2-/-;SOD1H46R and Als2-/-;Sqstm1+/-;SOD1H46R mice are detected.

Figure 6.

Simultaneous loss of SQSTM1 and ALS2 accelerates disease phenotypes associated with mutant SOD1 expression in mice. (A) Survival curves for SOD1H46R (H46R; filled circle; n = 105), Sqstm1+/-;SOD1H46R (H46R;Sqstm1Het; open square; n = 107), Als2+/-;SOD1H46R (H46R;Als2Het; open inverted triangle; n = 25), and Als2+/-;Sqstm1+/-;SOD1H46R (H46R;Als2Het;Sqstm1Het; gray diamond; n = 45) mice. Kaplan-Meier analysis identifies no significant differences between groups. (B) Survival curves for SOD1H46R (H46R; filled circle; n = 105), Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO; gray triangle; n = 49), Als2+/-;Sqstm1-/-;SOD1H46R (H46R;Als2Het;Sqstm1KO; open circle; n = 22), Als2-/-;SOD1H46R (H46R;Als2KO; open square; n = 11), Als2-/-;Sqstm1+/-;SOD1H46R (H46R;Als2KO;Sqstm1Het; gray inverted triangle; n = 55), and Als2-/-;Sqstm1-/-;SOD1H46R (H46R;Als2KO;Sqstm1KO; filled diamond; n = 21) mice. Kaplan-Meier analysis identifies significant differences between SOD1H46R and Sqstm1-/-;SOD1H46R (Log-rank test; P < 0.0001), between SOD1H46R and Als2-/-;SOD1H46R (Log-rank test; P < 0.025), between SOD1H46R and Als2-/-;Sqstm1-/-;SOD1H46R (Log-rank test; P < 0.0001), between Sqstm1-/-;SOD1H46R and Als2-/-;SOD1H46R (Log-rank test; P < 0.0001), between Sqstm1-/-;SOD1H46R and Als2-/-;Sqstm1-/-;SOD1H46R (Log-rank test; P < 0.0001), and between Als2-/-;SOD1H46R and Als2-/-;Sqstm1-/-;SOD1H46R mice (Log-rank test; P < 0.0001), while no significant differences between Sqstm1-/-; SOD1H46R and Als2+/-;Sqstm1-/-; SOD1H46R, and between Als2-/-;SOD1H46R and Als2-/-;Sqstm1+/-;SOD1H46R mice are detected.

Loss of SQSTM1 reduces the accumulation of polyubiquitinated proteins in the spinal cord of mutant SOD1-expressing mice at the end-stage

To investigate the effect of ALS2 and SQSTM1 double-knocked out on the levels of MND-associated factors, including SOD1, SQSTM1, and ubiquitin, we performed western blot analysis of the spinal cord samples obtained from 12, 16 and 20 week-old mice with five distinct genotypes; wild-type, SOD1H46R, Sqstm1-/-;SOD1H46R, Als2-/-;SOD1H46R, and Als2-/-;Sqstm1-/-;SOD1H46R. As previously reported (11), loss of ALS2 in SOD1H46R mice facilitated the accumulation of insoluble SQSTM1 and Ub_HMW at 20 weeks of age when compared to SOD1H46R mice (Figure 7A,

). However, despite much more severe motor deficit in SQSTM1 single-knockout SOD1H46R than SOD1H46R mice, and in ALS2/SQSTM1 double-knockout SOD1H46R than ALS2 single-knockout SOD1H46R mice, there were no significant differences in the levels of Ub_HMW between groups, respectively ().
Figure 7.

Loss of SQSTM1 reduces the accumulation of polyubiquitinated proteins in the spinal cord of mutant SOD1-expressing mice at the end-stage. (A) Western blot analysis of ALS2, SOD1, SQSTM1, and ubiquitin (Ub) in the spinal cord from 20 week-old mice with four distinct genotypes; SOD1H46R (H46R), Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO), Als2-/-;SOD1H46R (H46R;Als2KO), and Als2-/-;Sqstm1-/-;SOD1H46R (H46R;Als2KO;Sqstm1KO), and from end-stage mice with six distinct genotypes; H46R (25 week), H46R;Sqstm1KO (20 week), H46R;Als2KO (23 week), H46R;Als2KO;Sqstm1KO (20 week), SOD1G93A (G93A) (31 week), and Sqstm1-/-;SOD1G93A (G93A;Sqstm1KO) (24 week). Two fractions; 1% Triton X-soluble (TX-sol; left panels) and 1% Triton X-insoluble/5% SDS-soluble (TX-insol; right panels) fractions are analyzed. SOD1_mono and SOD1_HMW represent monomeric and high molecular-weight (aggregated) forms of SOD1, respectively. Ub_mono and Ub_HMW represent monomeric Ub and the polyubiquitinated proteins, respectively. GAPDH and β-actin are used for a loading control in TX-soluble and -insoluble fractions, respectively. First and second lanes in each panel represent the spinal cord samples from 23 week-old Als2+/-;Sqstm1+/-;SOD1H46R (H46R;Als2Het;Sqstm1Het) mouse, which are used for loading controls to compare the signal intensities between blots. The positions of size-markers are shown on the left. (B, C) Quantitation of TX-insoluble high molecular-weight SOD1 (SOD1_HMW) (B) and TX-insoluble polyubiquitinated proteins (Ub_HMW) (C). Densitometric data for immunoreactive signals in TX-insoluble fractions are normalized by the levels of β-actin. Values for (B) and (C) are mean ± SEM (20 weeks; n = 4, end-stage; n = 3) in an arbitrary unit relative to 12 week-old wild-type and SOD1H46R mice, respectively. Average ages for mice at the end stage are 24.5 week (range; 23-25 week), 19.7 week (range; 19-20 week), 22.7 week (range; 21-24 week), and 19.7 week (range; 19-20 week), in H46R, H46R;Sqstm1KO, H46R;Als2KO, and H46R;Als2KO;Sqstm1KO, respectively. Statistical significance is evaluated by unpaired t-test (comparisons between between H46R and H46R;Sqstm1KO at 20 weeks of age; **P < 0.01), one-way ANOVA with Bonferroni’s post hoc test [comparisons between same-aged mice with different genotypes; H46R (20 week), H46R;Sqstm1KO (20 week), and H46R;Als2KO;Sqstm1 (end-stage/∼19.7 week), and comparisons between different genotypes at the end stage; *P < 0.05, **P < 0.01, ***P < 0.001], or one-way ANOVA with Bonferroni’s post hoc test (comparisons between different ages in the same genotype; §P < 0.05, §§P < 0.01).

Figure 7.

Loss of SQSTM1 reduces the accumulation of polyubiquitinated proteins in the spinal cord of mutant SOD1-expressing mice at the end-stage. (A) Western blot analysis of ALS2, SOD1, SQSTM1, and ubiquitin (Ub) in the spinal cord from 20 week-old mice with four distinct genotypes; SOD1H46R (H46R), Sqstm1-/-;SOD1H46R (H46R;Sqstm1KO), Als2-/-;SOD1H46R (H46R;Als2KO), and Als2-/-;Sqstm1-/-;SOD1H46R (H46R;Als2KO;Sqstm1KO), and from end-stage mice with six distinct genotypes; H46R (25 week), H46R;Sqstm1KO (20 week), H46R;Als2KO (23 week), H46R;Als2KO;Sqstm1KO (20 week), SOD1G93A (G93A) (31 week), and Sqstm1-/-;SOD1G93A (G93A;Sqstm1KO) (24 week). Two fractions; 1% Triton X-soluble (TX-sol; left panels) and 1% Triton X-insoluble/5% SDS-soluble (TX-insol; right panels) fractions are analyzed. SOD1_mono and SOD1_HMW represent monomeric and high molecular-weight (aggregated) forms of SOD1, respectively. Ub_mono and Ub_HMW represent monomeric Ub and the polyubiquitinated proteins, respectively. GAPDH and β-actin are used for a loading control in TX-soluble and -insoluble fractions, respectively. First and second lanes in each panel represent the spinal cord samples from 23 week-old Als2+/-;Sqstm1+/-;SOD1H46R (H46R;Als2Het;Sqstm1Het) mouse, which are used for loading controls to compare the signal intensities between blots. The positions of size-markers are shown on the left. (B, C) Quantitation of TX-insoluble high molecular-weight SOD1 (SOD1_HMW) (B) and TX-insoluble polyubiquitinated proteins (Ub_HMW) (C). Densitometric data for immunoreactive signals in TX-insoluble fractions are normalized by the levels of β-actin. Values for (B) and (C) are mean ± SEM (20 weeks; n = 4, end-stage; n = 3) in an arbitrary unit relative to 12 week-old wild-type and SOD1H46R mice, respectively. Average ages for mice at the end stage are 24.5 week (range; 23-25 week), 19.7 week (range; 19-20 week), 22.7 week (range; 21-24 week), and 19.7 week (range; 19-20 week), in H46R, H46R;Sqstm1KO, H46R;Als2KO, and H46R;Als2KO;Sqstm1KO, respectively. Statistical significance is evaluated by unpaired t-test (comparisons between between H46R and H46R;Sqstm1KO at 20 weeks of age; **P < 0.01), one-way ANOVA with Bonferroni’s post hoc test [comparisons between same-aged mice with different genotypes; H46R (20 week), H46R;Sqstm1KO (20 week), and H46R;Als2KO;Sqstm1 (end-stage/∼19.7 week), and comparisons between different genotypes at the end stage; *P < 0.05, **P < 0.01, ***P < 0.001], or one-way ANOVA with Bonferroni’s post hoc test (comparisons between different ages in the same genotype; §P < 0.05, §§P < 0.01).

We reasoned that comparisons between groups at the same chronological age were insufficient, because mice with different genotypes exhibited different disease severity; e.g., ALS2/SQSTM1 double-knockout SOD1H46R mice at 20 weeks of age already reached to the end-stage of disease, while other SOD1H46R expressing mice still were in the intermediate-stage. Therefore, we compared the levels of MND-associated factors at the end-stage of animals among different genotypes by western blot analysis. Loss of SQSTM1 increased the level of insoluble SOD1_HMW at 20 weeks of age, whereas no further accumulation occurred at the end-stage (Figure 7A and B). Remarkably, the level of insoluble Ub_HMW in SQSTM1 single-knockout SOD1H46R mice at the end-stage, which was almost comparable to that at 20 weeks of age, was much lower than that in SOD1H46R mice (Figure 7A and B). By contrast, ALS2-deficient SOD1H46R mice resulted in the progressive accumulation of insoluble SOD1_HMW as well as Ub_HMW, although their levels were comparable to those for ALS2/SQSTM1 double-knockout SOD1H46R mice at the end-stage (Figure 7A and B). Interestingly, SOD1G93A mice exhibited the lower levels of insoluble SOD1_HMW and Ub_HMW than did SOD1H46R mice (Figure 7A). Further, loss of SQSTM1 in SOD1G93A mice similarly showed a lower tendency in the level of insoluble Ub_HMW as did in SOD1H46R mice (Figure 7A). Incidentally, although levels of vimentin in SOD1G93A-expressing mice were lower than those of SOD1H46R mice, GFAP levels were similar, indicating that the degree of glial activation at the end-stage of animals with different genotypes were most comparable (

).

Taken together, unlike in the cases of ALS2 (11), loss of SQSTM1 rather reduces the accumulation of biochemically-detectable insoluble polyubiquitinated proteins in the spinal cord of mutant SOD1-expressing mice at the end-stage. If the accumulation of biochemically-detectable insoluble polyubiquitinated proteins were beneficial and/or protective rather than harmful to neurons by reducing their toxicities, loss of SQSTM1 could result in the increased level of poisonous products, thereby SQSTM1-deficient mutant SOD1-expressing mice exhibited severer motor degenerative phenotype.

Extensive accumulation of ubiquitin-positive aggregates in motor neurons of the spinal cord from SQSTM1- and ALS2/SQSTM1-deficient SOD1H46R mice at the end-stage

Since the levels of biochemically-detectable insoluble SOD1 and ubiquitinated proteins did not seem to correlate with disease severity, we next conducted immunohistochemical studies to see if the subcellular localization of the MND-associated factors in the anterior horn of the spinal cord was associated with disease severity in SOD1H46R mice and those lacking either SQSTM1, ALS2, or both of them at the end-stage.

Most ubiquitin-positive aggregates were observed in non-neuronal cells and/or extracellular spaces (white arrows), and were rarely seen in neurons of SOD1H46R mice both at 20 weeks of age and at the end-stage. However, SQSTM1-, ALS2-, and ALS2/SQSTM1-deficient SOD1H46R mice exhibited a preferential formation of ubiquitin-positive inclusions and/or vacuolar structures in soma of MAP2-positive neurons (yellow arrowheads) (Figure 8A and B,

). Enlarged ubiquitin-positive dendrites and/or axons (marked as asterisk) were occasionally observed in all the mice overexpressing SOD1H46R at the end-stage (Figure 8A and B, ). High-magnification images for MAP2-positive spinal neurons from Sqstm1-/-;SOD1H46R mice demonstrated that ubiquitin-positive aggregates were localized to perinuclear compartments (Figure 8C). Neurons containing such ubiquitin-positive aggregates were more frequently observed in SQSTM1- and ALS2/SQSTM1-deficient SOD1H46R mice at the end-stage when compared to ALS2-deficient SOD1H46R mice (Figure 8A and B, ). It was also noted that ubiquitin-positive extracellular large vacuolar structures were frequently observed in ALS2-deficient SOD1H46R mice (open arrowheads) (Figure 8A and B, ).
Figure 8.

Extensive accumulation of ubiquitin-positive aggregates and vesicular structures in neurons of the spinal cord from SQSTM1-, ALS2-, and ALS2/SQSTM1-deficient SOD1H46R mice. (A) Representative images of double immunostaining with ubiquitin (green) and SQSTM1 (red), and (B) those with MAP2 (green) and ubiquitin (red) in the ventral horn of the lumbar spinal cord (L4-L5) from SOD1H46R mice (H46R) at 20 weeks of age (1st row) and at the end stage (25 week) (2nd row), Sqstm1-/-;SOD1H46R mice (H46R;Sqstm1KO) at 20 week of age (3rd row) and at the end stage (21 week) (4th row), Als2-/-;SOD1H46R mice (H46R;Als2KO) at the end stage (25 week) (5th row), and Als2-/-;Sqstm1-/-;SOD1H46R mice (H46R;Als2;Sqstm1KO) mice at the end stage (20 week) (6th row), respectively. The nuclei are counterstained with DAPI (Blue). Right columns display the merged images. Most ubiquitin-positive aggregates are observed in non-neuronal cells and/or extracellular spaces (white arrows) and are rarely seen in neurons of SOD1H46R mice, whereas SQSTM1-, ALS2-, and ALS2/SQSTM1-deficient SOD1H46R mice exhibit the preferential localization of ubiquitin-positive inclusions and/or vacuolar structures in soma of neurons (yellow arrowheads). Enlarged ubiquitin-positive dendrites and/or axons (marked as asterisk) are occasionally observed in all the mice overexpressing SOD1H46R at the end stage. It is noted that ubiquitin-positive extracellular large vacuolar structures are also observed in ALS2-deficient SOD1H46R mice (open arrowheads). Scale bars = 20 μm. (C) High-magnification images for MAP2-positive spinal neuron containing perinuclear ubiquitin-positive vacuoles and/or aggregates from Sqstm1-/-;SOD1H46R mice (H46R;Sqstm1KO) at the end stage. Images are obtained through optical serial sectioning (thickness; 0.8 μm/slice) with a step-size of 0.61μm along the z-axis. Scale bars = 10 μm.

Figure 8.

Extensive accumulation of ubiquitin-positive aggregates and vesicular structures in neurons of the spinal cord from SQSTM1-, ALS2-, and ALS2/SQSTM1-deficient SOD1H46R mice. (A) Representative images of double immunostaining with ubiquitin (green) and SQSTM1 (red), and (B) those with MAP2 (green) and ubiquitin (red) in the ventral horn of the lumbar spinal cord (L4-L5) from SOD1H46R mice (H46R) at 20 weeks of age (1st row) and at the end stage (25 week) (2nd row), Sqstm1-/-;SOD1H46R mice (H46R;Sqstm1KO) at 20 week of age (3rd row) and at the end stage (21 week) (4th row), Als2-/-;SOD1H46R mice (H46R;Als2KO) at the end stage (25 week) (5th row), and Als2-/-;Sqstm1-/-;SOD1H46R mice (H46R;Als2;Sqstm1KO) mice at the end stage (20 week) (6th row), respectively. The nuclei are counterstained with DAPI (Blue). Right columns display the merged images. Most ubiquitin-positive aggregates are observed in non-neuronal cells and/or extracellular spaces (white arrows) and are rarely seen in neurons of SOD1H46R mice, whereas SQSTM1-, ALS2-, and ALS2/SQSTM1-deficient SOD1H46R mice exhibit the preferential localization of ubiquitin-positive inclusions and/or vacuolar structures in soma of neurons (yellow arrowheads). Enlarged ubiquitin-positive dendrites and/or axons (marked as asterisk) are occasionally observed in all the mice overexpressing SOD1H46R at the end stage. It is noted that ubiquitin-positive extracellular large vacuolar structures are also observed in ALS2-deficient SOD1H46R mice (open arrowheads). Scale bars = 20 μm. (C) High-magnification images for MAP2-positive spinal neuron containing perinuclear ubiquitin-positive vacuoles and/or aggregates from Sqstm1-/-;SOD1H46R mice (H46R;Sqstm1KO) at the end stage. Images are obtained through optical serial sectioning (thickness; 0.8 μm/slice) with a step-size of 0.61μm along the z-axis. Scale bars = 10 μm.

These results illustrate preferential and the increasing accumulation of ubiquitin-positive aggregates in neurons of the spinal cord in SOD1H46R mice lacking SQSTM1. Since SQSTM1 loss is more harmful to SOD1H46R mice than lack of ALS2 (Figure 6B), the selective accumulation of such aggregates in neurons might be more insulting than be the overall accumulation of biochemically-detectable insoluble proteins in the spinal cord, albeit the elevated levels of biochemically-detectable insoluble proteins themselves are also associated with disease severity. Thus, a simultaneous inactivation of SQSTM1 and ALS2 results in distinct but additive adverse effects on proteostasis, thereby further accelerating motor neuron degeneration in SOD1H46R mice.

Discussion

We here showed that two independent ALS-linked autophagy-associated gene products; SQSTM1 and ALS2, but not antioxidant-related factor; NFE2L2, were linked to the pathogenesis in a mutant SOD1 transgenic mouse ALS model. To our knowledge, this is a first report showing that loss of SQSTM1 affects the manifestation of ALS/MNDs in experimental animal models, and also that there is a functional interrelationship between SQSTM1 and ALS2 in vivo.

ALS2, a causative gene product of a juvenile autosomal recessive form of ALS/MNDs (8,9), acts as a GEF for the small GTPase Rab5 (14), and is involved in a variety of physiological processes including the fusion and trafficking of endosomes (14,55), macropinosomes (15,16), and autophagosomes (18,19), as well as the regulation of neurite outgrowth (16,17). It has also been reported that ALS2 has a neuroprotective function against MND-associated pathological insults, such as toxicity induced by mutant SOD1 in cultured cells (56,57). Indeed, we have previously shown that disturbance of the autophagy-endolysosomal system by loss of ALS2 exacerbates the mutant SOD1-mediated neurotoxicity by lowering the autophagic flux in mice (11). Most recently, it has been reported that loss of ALS2 impairs autophagy in corticospinal motor neurons in mice (58). Thus, ALS2 might play an important neuroprotective role via regulating the maturation and trafficking of autophagosomes in the autophagy-endolysosomal system in vivo (19) (Figure 9).

Figure 9.

Proposed model for the additive cellular functions of SQSTM1/p62 and ALS2/alsin in the autophagy-endolysosomal system. Expression of mutant SOD1 results in the increased levels of misfolded and/or oligomeric forms of mutant SOD1, thereby accelerating the accumulation of insoluble protein aggregates and/or inclusions. Although it is still controversial whether such protein aggregates have a toxic or protective role in the pathogenesis of ALS, the formation of them results from the imbalance between generation and degradation of misfolded proteins within neurons. The ubiquitin-proteasome system (UPS), the autophagy lysosomal system, secretory autophagy, and autophagy endo-lysosomal system are cooperatively responsible for the degradation of such disease-associated proteins, thereby reducing disease-associated toxicity. However, when the toxic insults overwhelm and/or disturb the competence of protein homeostasis, the accumulation of insoluble proteins as aggregates and/or inclusions emerges. SQSTM1/p62 and ALS2/alsin play distinct but cumulative roles in autophagy; SQSTM1, one of the adaptor and/or cargo receptors for autophagy, regulates a step of the cargo-loading to autophagosome. ALS2, an activator of small GTPase Rab5, regulates the fusion between endosome, macropinosome, and autophagosome, leading to the formation of amphisome. Thus, ALS2 plays a role at a step of the autophagosome maturation and facilitates the lysosomal degradation of cargo. A simultaneous inactivation of SQSTM1 and ALS2 exacerbates the disease symptoms by disturbing the protein degradation within neurons.

Figure 9.

Proposed model for the additive cellular functions of SQSTM1/p62 and ALS2/alsin in the autophagy-endolysosomal system. Expression of mutant SOD1 results in the increased levels of misfolded and/or oligomeric forms of mutant SOD1, thereby accelerating the accumulation of insoluble protein aggregates and/or inclusions. Although it is still controversial whether such protein aggregates have a toxic or protective role in the pathogenesis of ALS, the formation of them results from the imbalance between generation and degradation of misfolded proteins within neurons. The ubiquitin-proteasome system (UPS), the autophagy lysosomal system, secretory autophagy, and autophagy endo-lysosomal system are cooperatively responsible for the degradation of such disease-associated proteins, thereby reducing disease-associated toxicity. However, when the toxic insults overwhelm and/or disturb the competence of protein homeostasis, the accumulation of insoluble proteins as aggregates and/or inclusions emerges. SQSTM1/p62 and ALS2/alsin play distinct but cumulative roles in autophagy; SQSTM1, one of the adaptor and/or cargo receptors for autophagy, regulates a step of the cargo-loading to autophagosome. ALS2, an activator of small GTPase Rab5, regulates the fusion between endosome, macropinosome, and autophagosome, leading to the formation of amphisome. Thus, ALS2 plays a role at a step of the autophagosome maturation and facilitates the lysosomal degradation of cargo. A simultaneous inactivation of SQSTM1 and ALS2 exacerbates the disease symptoms by disturbing the protein degradation within neurons.

On the other hand, SQSTM1 has originally been identified as an interacting protein with the atypical protein kinases (aPKCs) (59,60). It acts as a scaffold protein that regulates not only the activation of NF-κB signalling through the binding to aPKCs, but also as an adaptor protein that is responsible for recognition and loading cargo, including misfolded proteins and damaged organelles, into autophagosomes for recycling the cellular components (61–64). Further, the accumulation of SQSTM1 by defective autophagy causes an increased interaction between SQSTM1 and KEAP1, resulting in a competitive inhibition of the binding between KEAP1 and NFE2L2, thereby activating NFE2L2 and its target antioxidant-related genes (30). It is worth noting that the expression of both SQSTM1 and ALS2 is transcriptionally upregulated by NFE2L2 under oxidative stress conditions (36,37), indicating that both SQSTM1 and ALS2 are among the NFE2L2-regulating genes. We here showed that MND symptoms resulted from the systemic expression of mutant SOD1 further deteriorated by loss of SQSTM1, but not by loss of NFE2L2. These findings suggest that the antioxidative stress function mediated by the SQSTM1-KEAP1-NFE2L2 axis (65), including NFE2L2-mediated ALS2 expression (37) as well as NFE2L2-mediated SQSTM1-feedback upregulation (66), may not play a major role in the pathogenesis in this disease model. Since phosphorylation of SQSTM1 underlies autophagy-mediated clearance of the polyubiquitinated proteins that has failed to be degraded by the UPS under neurodegenerative conditions (62), loss of SQSTM1 might result in disturbed clearance of misfolded proteins, which would otherwise be destined for degradation, by decrease in the selective autophagy.

The remarkable findings obtained from the present study include that motor dysfunction observed in mutant SOD1-expressing mice is deteriorated by loss of SQSTM1, and that a simultaneous inactivation of SQSTM1 and ALS2 further shortened the lifespans of mutant SOD1-expressing mice. We have previously shown that loss of ALS2 augments mutant SOD1-mediated neurotoxicity by lowering the autophagic flux, resulting in the disturbed fusion-mediated endolysosomal trafficking and/or autophagic clearance of misfolded protein aggregates (11). By contrast, we here showed that loss of SQSTM1 compromised autophagy-associated proteostasis with minimal influences on the autophagic flux. These results combined indicate that SQSTM1 and ALS2 exert additive and/or cooperative effects, though closely related but distinct molecular mechanisms, in the pathogenesis of this animal model. As previously reported, ALS2 plays a role in the regulation of endosome and autophagosome maturation (11,18,19), while SQSTM1 can mediate the recognition and loading of misfolded proteins into nascent autophagosomes (61–64). Thus, a simultaneous inactivation of them leads to not only cargo-loading deficits but also to incomplete maturation of autophagosomes. Collectively, we here proposed that two ALS-linked factors, SQSTM1 and ALS2, have distinct but additive protective roles against mutant SOD1-mediated toxicity by modulating neuronal proteostasis possibly through the autophagy-endolysosomal system (Figure 9).

It is speculated that persistent and long-term excess production of mutant SOD1 overwhelms the intracellular capacity for the clearance of misfolded proteins by the UPS and/or the autophagy-endolysosomal system, resulting in the accumulation of insoluble proteins in tissues (11). Indeed, we demonstrated that loss of SQSTM1 enhanced the accumulation of insoluble SOD1 in the spinal cord of mice at the intermediate stage of the disease, suggesting a decreased degradation of mutant SOD1 via the SQSTM1-mediated system, consistent with previous studies (13). However, regardless of the presence or absence of SQSTM1, the increased levels of insoluble polyubiquitinated proteins in mutant SOD1-expressing mice at the intermediate stage of disease were unaffected. Further, SQSTM1-deficient SOD1H46R mice rather showed the lower levels of insoluble polyubiquitinated proteins than did SOD1H46R mice at the end-stage, despite that cytoplasmic ubiquitin-positive aggregates were more prominent in SQSTM1-deficient SOD1H46R mice. These results suggest that the levels of biochemically-detectable insoluble polyubiquitinated proteins in the spinal cord do not always occur in parallel with those for immunohistochemically-visible ubiquitin-positive aggregates. Our findings showing that SOD1G93A mice, in which a prominent accumulation of SOD1 and ubiquitin-positive aggregates was frequently observed (67), exhibited the lower levels of biochemically-detectable insoluble SOD1 and polyubiquitinated proteins when compared to SOD1H46R mice support this notion.

One important question arising from this study is as to how autophagy dysfunction caused by loss of SQSTM1 exacerbates the disease symptoms, even though it suppresses the mutant SOD1-dependent accumulation of biochemically-detectable insoluble polyubiqutinated proteins. Differences in the subcellular distribution of them between SQSTM1-deficient SOD1H46R and SOD1H46R mice can explain such discrepancy. We here showed that most of the large ubiquitin-positive aggregates were localized to extracellular and/or non-neuronal nuclei, and were rarely seen within motor neurons in SOD1H46R mice, while cytoplasmic ubiquitin-positive inclusions in MAP2-positive neurons were much prominent in SQSTM1-deficient SOD1H46R mice. Recent studies have shown that intercellular propagation of misfolded SOD1 through exsosome-dependent as well as independent mechanisms is implicated in spreading the pathology in ALS (68,69). Further, another type of autophagy, termed “secretory autophagy”, facilitates unconventional secretion of aggregation-prone proteins (70). Thus, it is possible that loss of SQSTM1 could disrupt such unconventional secretory pathway, thereby disturbing the proper extracellular secretion of insoluble polyubiquitinated proteins, which in turn accelerates the intracellular accumulation of ubiquitin-positive aggregates in neurons. On the other hand, loss of ALS2 may predominantly impair the intracellular autophagy-endolysosomal pathway, but not such secretory pathway, thereby further enhancing the accumulation of biochemically-detectable insoluble SOD1 and polyubiquitinated proteins as well as extracellular and/or non-neuronal ubiquitin-positive aggregates. Assuming that the selective accumulation of such aggregates in neurons is more insulting than extracellular and/or non-neuronal aggregates (or the biochemically-detectable insoluble proteins), and that the SQSTM1-mediated system plays a crucial role in reducing and detoxifying such mutant SOD1-linked cellular insults by regulating the secretory autophagy, lack of SQSTM1 can be attributed to the increased levels of toxic molecules within neurons, thereby accelerating motor neuron degeneration in the spinal cord. Alternatively, it has recently been shown that SQSTM1 serves as a scaffold rather than a degrading cargo to regulate the signaling pathways towards either apoptosis or necroptosis via the autophagosome-resident SQSTM1-necrosome complex (71), and that loss of SQSTM1 tilts tumor necrosis factor α (TNFα)-dependent cell death from necroptosis to apoptosis in cells (71). Thus, loss of SQSTM1 may alter the fate of neurons resulting in accelerated cell death. However, as there are no direct evidences to support these notions thus far, further detailed studies will be required.

Another question arising from this study includes whether the effects of SQSTM1 loss are different among the mutations in SOD1. We have previously shown that loss of ALS2 exacerbates SOD1H46R-, but not SOD1G93A-linked disease phenotypes in mice (11), and that SOD1H46R and SOD1G93A exert distinct harmful effects on gross phenotypes in mice (45). In this study, we demonstrated that loss of SQSTM1 accelerated the onset of disease both in SOD1H46R and SOD1G93A mice, although its effect in SOD1G93A mice was much smaller than those in SOD1H46R mice. Further, unlike in the case of SOD1H46R mice, in which partial inactivation of SQSTM1 had no detectable impact on the phenotypes, heterozygous defects in SQSTM1 in SOD1G93A mice shortened their lifespan to the levels comparable to those for SQSTM1-knockout SOD1G93A mice. Consistently, it has previously been shown that not only homozygous but also heterozygous mutations in SQSTM1 aggravates motor phenotypes in a mouse model of spinal and bulbar muscular atrophy (SBMA) (34). Although the exact reason for these inconsistencies is still unclear, different pathogenic mechanisms that are associated with different mutations in SOD1 might be concerned. Indeed, although both mutations cause the selective and progressive motor neuron degeneration, the G93A mutation results in prominent mitochondrial vacuolar phenotypes in mice (72) with accompanying faster progression of disease symptoms (73), while the H46R mutation leads to a widespread axonal degeneration with preserved motor neurons in the spinal cord (11) and with accompanying much slower progression of disease symptoms (74). These findings indicate that effects of SQSTM1 loss on SOD1H46R and SOD1G93A mice are, at least, not identical, but still possibly overlapped with each other.

There are several issues that remain unanswered. First, although both SQSTM1 and ALS2 are implicated in the autophagy-endolysosomal system, it is still controversial whether mutant SOD1-linked toxicity itself activates or inhibits autophagy; one has reported that the autophagy-lysosomal system is hyperactive in motor neurons of presymptomatic SOD1G85R and SOD1G93A ALS mouse models (75), while the other has demonstrated that progressive lysosomal deficits accompanied by impaired autophagic degradation begin at asymptomatic stages in SOD1G93A mice (76). Second, the effects of SQSTM1 loss on neurodegeneration are still a matter of conjecture. In this study, we revealed that loss of SQSTM1 accelerated degeneration of motor neurons in the spinal cord. Consistent with our findings, absence of SQSTM1 also exacerbates the disease symptoms in mice overexpressing mutant α-synuclein as a Parkinson’s disease model (33), in mice with expanded poly-Q coded androgen receptor as a spinal and bulbar muscular atrophy (SBMA) model (34), and in flies expressing expanded poly-Q proteins (77). Contrary to these, it has also been reported that loss of SQSTM1 reduces nuclear inclusions and rather ameliorates disease phenotypes in Huntington's disease model mice (35). Third, despite that more than 20 mutations in SQSTM1 have been identified in patients with ALS and/or ALS-FTD thus far (20–26), the functional analysis for mutant SQSTM1 proteins has yet to be achieved. Forth, although recent genetic and molecular studies have unveiled that many ALS-causative factors; such as charged multivesicular body protein 2B (CHAMP2B) (78), valosin-containing protein (VCP) (79), optineurin (OPTN) (80), ubiquilin 2 (UBQLN2) (81), and TANK-binding kinase 1 (TBK1) (82,83), as well as other adaptor protein-linked mechanisms than SQSTM1; such as neighbor of BRCA1 gene 1 (NBR1), OPTN, UBQLN2, and nuclear domain 10 protein 52 (NDP52)-dependent ones (84–86) are implicated in proteostasis by regulating the UPS and/or the autophagy-endolysosomal system, the molecular links and their relationship among those factors are still unclear. Lastly, we cannot formally exclude a possibility that unidentified SQSTM1- and/or ALS2-mediated mechanisms other than autophagy as well as antioxidant pathways link to motor neuron degeneration. In order to address such daunting issues, future studies not only on the exact role of SQSTM1 and ALS2 in the pathogenesis of ALS/MNDs, but also on the molecular networks associated with other ALS-linked factors are warranted.

In conclusions, this study provides the first evidence showing that two independent ALS-linked autophagy-associated gene products; SQSTM1 and ALS2, but not antioxidant-related factor; NFE2L2, are linked to the pathogenesis in a mutant SOD1 transgenic mouse ALS model. Our findings indicate that SQSTM1 and ALS2 have additive effects on the protection against mutant SOD1-mediated toxic insults possibly via modulating the autophagy-endolysosomal system. Further characterization of these molecular networks will give us more clues to understanding the autophagy-endolysosomal process in the manifestation of ALS and other MNDs.

Materials and Methods

Antibodies

Primary antibodies used for western blot analysis included rabbit polyclonal anti-ALS2; MPF1012-1651 (1:3,000) (55), rabbit polyclonal anti-SOD1 (FL-154) (1:10,000-15,000, Santa Cruz), mouse monoclonal anti-ubiquitin (P4D1) (1:3,000, Santa Cruz), rabbit polyclonal anti-SQSTM1/p62 (1:2,000, MBL), rabbit polyclonal anti-LC3 (1:2,000, MBL), mouse monoclonal anti-vimentin (1:3,000, Sigma-Aldrich), rabbit polyclonal anti-GFAP (D1F4Q) (1:3,000, Cell Signaling), mouse monoclonal anti-GAPDH (1:10,000, MBL), and rabbit polyclonal anti-β-actin (1:1,000, Sigma-Aldrich) antibodies. Secondary antibodies included horseradish peroxidase (HRP)-conjugated anti-rabbit IgG (1:5,000, GE Healthcare Life Sciences) and anti-mouse IgG (1:5,000, Jackson) antibodies.

Antibodies used for immunohistochemistry included rabbit polyclonal anti-SOD1 (FL-154) (1:100, Santa Cruz), guinea pig polyclonal anti-SQSTM1/p62 (1:1,000, Progen), mouse monoclonal anti-ubiquitin (1B3) (1:200, MBL), rabbit polyclonal anti-ubiquitin (1:1,000, Dako), mouse monoclonal anti-MAP2 (1:500, Sigma-Aldrich), rabbit polyclonal anti-GFAP (Histofine) (1:1, Nichirei), Secondary antibodies included Alexa 594-conjugated goat anti-guinea pig IgG (1:500, Thermo Fisher Scientific), Alexa 594-conjugated anti-rabbit IgG (1:500, Invitrogen), and Alexa 488-conjugated anti-mouse IgG (1:500, Thermo Fisher Scientific) antibodies.

Animals

We used six congenic lines of genetically-engineered mice that were generated by backcrossing more than 10 generations with C57BL/6N (B6) mice; transgenic mice carrying the H46R mutation in the human SOD1 gene (SOD1H46R) (11,38,45), transgenic mice carrying the G93A mutation in the human SOD1 gene (SOD1G93A) (5), Nfe2l2 (nuclear factor, erythroid derived 2, like 2; NFE2L2) knockout mice (39), Sqstm1 (sequestosome 1; SQSTM1) knockout mice (28), Als2 (alsin) knockout mice (40,87), and transgenic mice carrying the GFP-fused human-MAP1LC3B cDNA (GFP-LC3) (54). It is noted that the mean lifespan of SOD1G93A mice in our colony is significantly extended compared to that of original line (5) due to differences in a genetic background (B6 background) (45), and possibly in a copy number of the transgene. We generated SOD1H46R transgenic mice on Nfe2l2-null, Sqstm1-null, and Als2-null backgrounds by crossing the SOD1H46R mouse line with Nfe2l2-/-, Sqstm1-/-, and Als2-/- mice, respectively. We also procuded SOD1G93A transgenic mice on a Sqstm1-null background, double transgenic line on a Sqstm1-null background (SOD1H46R;GFP-LC3;Sqstm1-/-), and SOD1H46R transgenic mice on a double knockout background (SOD1H46R; Als2-/-;Sqstm1-/-) by appropriately intercrossing between two or three lines. Four lines of mice; Nfe2l2-null, Sqstm1-null, Als2-null, and GFP-LC3, were viable and normally fertile, while only male mice of 2 lines of an ALS mouse model; SOD1H46R and SOD1G93A, were fertile. All of the offsprings were genotyped by PCR using genomic DNA extracted from ear tissues. Primers for Als2-knockout mice were as previously reported (40). Other primers used were as follows; SOD1; hSOD1_ex2L: 5’-TCAGAAACTCTCTCCAACTTTGC-3’, hSOD 1_ex2R : 5’-CAAGTATGGGTCACCAGCAC-3’, hSOD1_ex4L: 5’-GGCATCAGCCCTAATCCATC-3’, hSOD1_ex4R: 5’-CCGC GACTA ACAATCAAA GTG-3’, Nfe2l2; NRF2-5’: 5’-TGGACGGG ACT ATTG AAGGCTG-3’, LacZ: 5’-GCGGATTGACCGTAATGGGATAGG-3’, Nrf2 antisense: 5’-GCCGCCTT TTCAGTAGATGGAGG-3’ (39), Sqstm1; A170#1F: 5’-CTGCATGTCTTCTCCCATGAC-3’, A170#1R: 5’-TAGA TACCTAGGTGAGCTCTG-3’, A170#2F: 5’-CTTACG GGTCCTTTTCCCAAC-3’, A170#2R: 5’-TCCTCCTTG CCCAGAA GATAG-3’, GFP-LC3; GFP(LC3)1: 5’-TCCTGCTGG AGTTCGT GACCG-3’, LC3*rc3_modified: 5’-AGCCGTCT TCATCTCT CTCGC-3’ (54). Mice were housed at an ambient temperature of 22 ˚C with a 12 hr light-dark cycle. Food and water were fed ad libitum. Copy numbers of the SOD1H46R or SOD1G93A transgene genes, which affected the severity of motor dysfunction, remained unchanged (∼20 copies) within our mating cohorts (data not shown) (45). All animal experimental procedures were approved by The Institutional Animal Care and Use Committee at Tokai University.

Lifespan and behavioral analyses

The end-stage of mice carrying either SOD1H46R or SOD1G93A transgene was defined by the observations that mice showed complete paralysis of their hindlimb but were still able to crawl by their forelimb and eat/drink by themselves. Lifespan (endpoint) of such mice was determined by the observations that mice were unable to move by themselves. Motor coordination and balance was assessed by a balance beam test (47,87) using the fixed-stainless steel bar (45 cm long and 0.9 cm in diameter) at 12 weeks of age, and weekly thereafter until 28 weeks of age or the day at which mice were unable to stay on the bar. Each mouse was given five trials, and the maximum durations (up to 60 s) at which mice fell off from the bar were scored. Onset of disease was defined as the first day at which mice could not stay on the bar up to 50 sec for two consecutive weeks on the balance beam test. Since our previous study revealed that there were no gender differences of median survivals and onset of disease in either SOD1H46R or SOD1G93A mice on a B6 background (45), we here combined those data from both genders and analyzed.

Preparation of total RNA and protein samples

Brain and spinal cord tissues were weighed and homogenized in 2 weight-volume (mg/µL) of phosphate buffer saline (PBS). Twenty-five µL of the homogenate was used for RNA extraction using Sepazole RNA II super G (Nakarai Tesque), followed by the purification of total RNA by SV Total RNA Isolation System (Promega). The remaining homogenates were subjected to protein extraction. Seven weight-volume of Lysis buffer A [25mM Tris-HCl (pH 7.5), 50mM NaCl, 1%(w/v) Triton X-100 (TX), Complete Protease Inhibitor Cocktail (Roche)] with respective to a remaining volume of the tissue homogenates was added to the homogenates and mixed well followed by centrifugation at 23,000g for 20min at 4˚C. The supernatant was collected as a TX-soluble fraction. The insoluble pellet fraction was once washed with Lysis buffer A, resuspended with Lysis buffer B [25 mM Tris-HCl (pH 7.5), 50mM NaCl, 5% sodium dodecyl sulphate (SDS)], sonicated, and collected as TX-insoluble/SDS-soluble fraction. Protein concentration was determined by the Micro BCA system (Pierce).

Western blot analysis

Equal amount of protein (2 µg) from each fraction was subjected to SDS-polyacrylamide gel electrophoresis, and transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad). The membranes were blocked with 0.5% skimmed milk in TBST [20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Tween 20] for 1 hr at RT, and incubated with the primary antibody in TBST. After washing with TBST, membranes were incubated with HRP-conjugated secondary antibody. Signals were visualized by ImmobilonTM Western (Millipore) and detected by Ez-Capture MG (ATTO). The signal intensities were quantified by using CS Analyzer ver3 (ATTO).

Quantitative reverse transcriptase-PCR (qRT-PCR)

The qRT-PCR was performed on a 1 ng of total RNA using QuantiFastTM SYBR Green RT-PCR (Qiagen) with specific primers for SOD1, Sqstm1, Map1lc3a, Map1lc3b, Nes, Bysl, Gfap, Aif1, and Gapdh as previously reported (11,88). The levels of all transcripts were normalized for the Gapdh mRNA level in each sample.

Bone morphology and histology

Morphology and histology of bones from mutant mice were analyzed as previously described (89). In brief, a radiograph of animals was performed with a DHF-105CX medical X-ray apparatus according to the manufacture's operating manual (Hitachi Medical Corporation). For histological analysis, femoral long bones were fixed in 10% buffered formalin and decalcified in 10% formic acid. The decalcified tissues were then embedded in paraffin, and histological sections were stained with hematoxylin and eosin (H&E).

Histology and electron microscopic observations of the CNS

Mice were anaesthetized with 4% isoflurane by inhalation, and transcardially perfused with physiological saline containing 1,000U/ml heparin, followed by 2% paraformaldehyde (PFA)/2% glutaraldehyde (GA) in 0.1 M phosphate buffer (PB) (pH 7.2). Brain and spinal cord were removed and post-fixed with the same fixative for 12 hr at 4°C. Lumbar segment was dissected out and fixed with 2.5% GA for 2 hr at 4°C, followed by washing with 0.1 M PB (pH 7.3) and post-fixed in 1% osmium tetroxide in 0.05 M PB (pH 7.3). After dehydration in a graded alcohol, the tissues were embedded in the epoxy resin. Semi-thin sections (2 µm) of L4-L5 lumbar cord were stained with 0.3% toluidine blue (TB) and examined. Selected areas of the spinal cord were cut into ultrathin sections, and stained with uranyl acetate and lead citrate for ultrastructural examination using electron microscopes (JEM-1400, JEOL; HT7700, HITACHI). All images presented are representative of 2-3 animals examined in each group at each time point.

Histology and immunohistochemistry of the CNS

Mice were anaesthetized with 4% isoflurane by inhalation, and transcardially perfused with 4% PFA in 0.1 M PB (pH 7.2). Brain and spinal cord were removed and post-fixed for at least 48 hr in 4% PFA followed by paraffin embedding. Paraffin sections were sliced on microtome at a thickness of 6 µm for histopathological (Nissl) and fluorescent immunohistochemical evaluations. For immunohistochemistry, spinal cord sections were incubated in phosphate buffered saline (PBS, pH 7.4) with 5% normal goat serum (NGS) and 0.1% TX for 30 min at RT. For double-immunostaining, sections were incubated with primary antibodies in PBS containing 1.5% NGS and 0.05% TX overnight at 4°C. Sections were incubated with secondary antibody for 2 hr at RT. The controls for all immunostainings were performed simultaneously by omitting the primary antibody. Sections were coverslipped using the shield solution [0.01M PBS (pH 8.8), 10% glycerol, 1.25 mg/ml DABCO (Sigma)] with or without 4',6-diamidino-2-phenylindole dihydrochloride (DAPI) for nuclei counterstaining, and analyzed by a Zeiss Axioplan2 fluorescence microscope (Carl Zeiss), or Zeiss LSM700 laser scanning confocal microscope (Carl Zeiss) and processed by ImageJ 1.50v (NIH). All images presented are representative of 2–3 animals examined in each group at each time point.

Statistical analysis

Statistical significance was evaluated by ANOVA followed by appropriate post hoc tests for multiple comparisons between groups. Survival data were compared using Kaplan-Meier survival analysis with Log-rank (Mantel-Cox) test. All analyses were conducted using PRISM 5 (GraphPad). A P-value < 0.05 was considered as reaching statistical significance.

Supplementary Material

is available at HMG online.

Acknowledgements

We are grateful to all members of our laboratory at Tokai University School of Medicine for helpful discussion throughout this work. We thank Dr. Noboru Mizushima (The University of Tokyo) for the generous gift of GFP-LC3#53 transgenic mice, and Drs. Nahoko Fukunishi, Shunji Amano, Chisa Okada, and all other members of Support Center for Medical Research and Education (Tokai University) for their technical help.

Conflict of Interest statement. None declared.

Funding

This study was supported by a Grant-in-Aid Scientific Research (B) (23300129, 26290018 to S.H.) from the Japanese Society for Promotion of Science (JSPS), and the National Natural Science Foundation of China (NSFC) and JSPS Bilateral Joint Research Project (to S.H.). It was also partly supported by a Grant-in-Aid for Challenging Exploratory Research (24650189 to S.H.).

References

1
Robberecht
W.
Philips
T.
(
2013
)
The changing scene of amyotrophic lateral sclerosis
.
Nat. Rev. Neurosci
 .,
14
,
248
264
.
2
Pasinelli
P.
Brown
R.H.
(
2006
)
Molecular biology of amyotrophic lateral sclerosis: insights from genetics
.
Nat. Rev. Neurosci
 .,
7
,
710
723
.
3
Al-Chalabi
A.
Jones
A.
Troakes
C.
King
A.
Al-Sarraj
S.
van den Berg
L.H.
(
2012
)
The genetics and neuropathology of amyotrophic lateral sclerosis
.
Acta Neuropathol
 .,
124
,
339
352
.
4
Renton
A.E.
Chio
A.
Traynor
B.J.
(
2014
)
State of play in amyotrophic lateral sclerosis genetics
.
Nat. Neurosci
 .,
17
,
17
23
.
5
Gurney
M.E.
Pu
H.
Chiu
A.Y.
Dal Canto
M.C.
Polchow
C.Y.
Alexander
D.D.
Caliendo
J.
Hentati
A.
Kwon
Y.W.
Deng
H.X.
, et al.  . (
1994
)
Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation
.
Science
 ,
264
,
1772
1775
.
6
Kato
S.
(
2008
)
Amyotrophic lateral sclerosis models and human neuropathology: similarities and differences
.
Acta Neuropathol
 .,
115
,
97
114
.
7
Polymenidou
M.
Cleveland
D.W.
(
2011
)
The seeds of neurodegeneration: prion-like spreading in ALS
.
Cell
 ,
147
,
498
508
.
8
Hadano
S.
Hand
C.K.
Osuga
H.
Yanagisawa
Y.
Otomo
A.
Devon
R.S.
Miyamoto
N.
Showguchi-Miyata
J.
Okada
Y.
Singaraja
R.
, et al.  . (
2001
)
A gene encoding a putative GTPase regulator is mutated in familial amyotrophic lateral sclerosis 2
.
Nat. Genet
 .,
29
,
166
173
.
9
Yang
Y.
Hentati
A.
Deng
H.X.
Dabbagh
O.
Sasaki
T.
Hirano
M.
Hung
W.Y.
Ouahchi
K.
Yan
J.
Azim
A.C.
, et al.  . (
2001
)
The gene encoding alsin, a protein with three guanine-nucleotide exchange factor domains, is mutated in a form of recessive amyotrophic lateral sclerosis
.
Nat. Genet
 ,
29
,
160
165
.
10
Hadano
S.
Kunita
R.
Otomo
A.
Suzuki-Utsunomiya
K.
Ikeda
J.E.
(
2007
)
Molecular and cellular function of ALS2/alsin: implication of membrane dynamics in neuronal development and degeneration
.
Neurochem. Int
 .,
51
,
74
84
.
11
Hadano
S.
Otomo
A.
Kunita
R.
Suzuki-Utsunomiya
K.
Akatsuka
A.
Koike
M.
Aoki
M.
Uchiyama
Y.
Itoyama
Y.
Ikeda
J.E.
(
2010
)
Loss of ALS2/Alsin exacerbates motor dysfunction in a SOD1-expressing mouse ALS model by disturbing endolysosomal trafficking
.
PLoS ONE
 ,
5
,
e9805.
12
Al-Sarraj
S.
King
A.
Troakes
C.
Smith
B.
Maekawa
S.
Bodi
I.
Rogelj
B.
Al-Chalabi
A.
Hortobagyi
T.
Shaw
C.E.
(
2011
)
p62 positive, TDP-43 negative, neuronal cytoplasmic and intranuclear inclusions in the cerebellum and hippocampus define the pathology of C9orf72-linked FTLD and MND/ALS
.
Acta Neuropathol
 .,
122
,
691
702
.
13
Gal
J.
Strom
A.L.
Kwinter
D.M.
Kilty
R.
Zhang
J.
Shi
P.
Fu
W.
Wooten
M.W.
Zhu
H.
(
2009
)
Sequestosome 1/p62 links familial ALS mutant SOD1 to LC3 via an ubiquitin-independent mechanism
.
J. Neurochem
 .,
111
,
1062
1073
.
14
Otomo
A.
Hadano
S.
Okada
T.
Mizumura
H.
Kunita
R.
Nishijima
H.
Showguchi-Miyata
J.
Yanagisawa
Y.
Kohiki
E.
Suga
E.
, et al.  . (
2003
)
ALS2, a novel guanine nucleotide exchange factor for the small GTPase Rab5, is implicated in endosomal dynamics
.
Hum. Mol. Genet
 .,
12
,
1671
1687
.
15
Kunita
R.
Otomo
A.
Mizumura
H.
Suzuki-Utsunomiya
K.
Hadano
S.
Ikeda
J.E.
(
2007
)
The Rab5 activator ALS2/alsin acts as a novel Rac1 effector through Rac1-activated endocytosis
.
J. Biol. Chem
 .,
282
,
16599
16611
.
16
Otomo
A.
Kunita
R.
Suzuki-Utsunomiya
K.
Mizumura
H.
Onoe
K.
Osuga
H.
Hadano
S.
Ikeda
J.E.
(
2008
)
ALS2/alsin deficiency in neurons leads to mild defects in macropinocytosis and axonal growth
.
Biochem. Biophys. Res. Commun
 .,
370
,
87
92
.
17
Jacquier
A.
Buhler
E.
Schafer
M.K.
Bohl
D.
Blanchard
S.
Beclin
C.
Haase
G.
(
2006
)
Alsin/Rac1 signaling controls survival and growth of spinal motoneurons
.
Ann. Neurol
 .,
60
,
105
117
.
18
Otomo
A.
Kunita
R.
Suzuki-Utsunomiya
K.
Ikeda
J.E.
Hadano
S.
(
2011
)
Defective relocalization of ALS2/alsin missense mutants to Rac1-induced macropinosomes accounts for loss of their cellular function and leads to disturbed amphisome formation
.
FEBS Lett
 .,
585
,
730
736
.
19
Otomo
A.
Pan
L.
Hadano
S.
(
2012
)
Dysregulation of the autophagy-endolysosomal system in amyotrophic lateral sclerosis and related motor neuron diseases
.
Neurol. Res. Int
 .,
2012
,
498428
.
20
Fecto
F.
Yan
J.
Vemula
S.P.
Liu
E.
Yang
Y.
Chen
W.
Zheng
J.G.
Shi
Y.
Siddique
N.
Arrat
H.
, et al.  . (
2011
)
SQSTM1 mutations in familial and sporadic amyotrophic lateral sclerosis
.
Arch. Neurol
 .,
68
,
1440
1446
.
21
Rubino
E.
Rainero
I.
Chio
A.
Rogaeva
E.
Galimberti
D.
Fenoglio
P.
Grinberg
Y.
Isaia
G.
Calvo
A.
Gentile
S.
, et al.  . (
2012
)
SQSTM1 mutations in frontotemporal lobar degeneration and amyotrophic lateral sclerosis
.
Neurology
 ,
79
,
1556
1562
.
22
Hirano
M.
Nakamura
Y.
Saigoh
K.
Sakamoto
H.
Ueno
S.
Isono
C.
Miyamoto
K.
Akamatsu
M.
Mitsui
Y.
Kusunoki
S.
(
2013
)
Mutations in the gene encoding p62 in Japanese patients with amyotrophic lateral sclerosis
.
Neurology
 ,
80
,
458
463
.
23
Shimizu
H.
Toyoshima
Y.
Shiga
A.
Yokoseki
A.
Arakawa
K.
Sekine
Y.
Shimohata
T.
Ikeuchi
T.
Nishizawa
M.
Kakita
A.
, et al.  . (
2013
)
Sporadic ALS with compound heterozygous mutations in the SQSTM1 gene
.
Acta Neuropathol
 .,
126
,
453
459
.
24
Teyssou
E.
Takeda
T.
Lebon
V.
Boillee
S.
Doukoure
B.
Bataillon
G.
Sazdovitch
V.
Cazeneuve
C.
Meininger
V.
LeGuern
E.
, et al.  . (
2013
)
Mutations in SQSTM1 encoding p62 in amyotrophic lateral sclerosis: genetics and neuropathology
.
Acta Neuropathol
 .,
125
,
511
522
.
25
Chen
Y.
Zheng
Z.Z.
Chen
X.
Huang
R.
Yang
Y.
Yuan
L.
Pan
L.
Hadano
S.
Shang
H.F.
(
2014
)
SQSTM1 mutations in Han Chinese populations with sporadic amyotrophic lateral sclerosis
.
Neurobiol. Aging
 ,
35
,
e727729
e727726
.
26
Yang
Y.
Tang
L.
Zhang
N.
Pan
L.
Hadano
S.
Fan
D.
(
2015
)
Six SQSTM1 mutations in a Chinese amyotrophic lateral sclerosis cohort
.
Amyotroph. Lateral Scler. Frontotemporal Degener
 .,
16
,
378
384
.
27
Laurin
N.
Brown
J.P.
Morissette
J.
Raymond
V.
(
2002
)
Recurrent mutation of the gene encoding sequestosome 1 (SQSTM1/p62) in Paget disease of bone
.
Am. J. Hum. Genet
 .,
70
,
1582
1588
.
28
Komatsu
M.
Waguri
S.
Koike
M.
Sou
Y.S.
Ueno
T.
Hara
T.
Mizushima
N.
Iwata
J.
Ezaki
J.
Murata
S.
, et al.  . (
2007
)
Homeostatic levels of p62 control cytoplasmic inclusion body formation in autophagy-deficient mice
.
Cell
 ,
131
,
1149
1163
.
29
Pankiv
S.
Clausen
T.H.
Lamark
T.
Brech
A.
Bruun
J.A.
Outzen
H.
Overvatn
A.
Bjorkoy
G.
Johansen
T.
(
2007
)
p62/SQSTM1 binds directly to Atg8/LC3 to facilitate degradation of ubiquitinated protein aggregates by autophagy
.
J. Biol. Chem
 .,
282
,
24131
24145
.
30
Komatsu
M.
Kurokawa
H.
Waguri
S.
Taguchi
K.
Kobayashi
A.
Ichimura
Y.
Sou
Y.S.
Ueno
I.
Sakamoto
A.
Tong
K.I
., et al.  . (
2010
)
The selective autophagy substrate p62 activates the stress responsive transcription factor Nrf2 through inactivation of Keap1
.
Nat. Cell Biol
 ,
12
,
213
223
.
31
Ramesh Babu
J.
Lamar Seibenhener
M.
Peng
J.
Strom
A.L.
Kemppainen
R.
Cox
N.
Zhu
H.
Wooten
M.C.
Diaz-Meco
M.T.
Moscat
J.
, et al.  . (
2008
)
Genetic inactivation of p62 leads to accumulation of hyperphosphorylated tau and neurodegeneration
.
J. Neurochem
 .,
106
,
107
120
.
32
Kwon
J.
Han
E.
Bui
C.B.
Shin
W.
Lee
J.
Lee
S.
Choi
Y.B.
Lee
A.H.
Lee
K.H.
Park
C.
, et al.  . (
2012
)
Assurance of mitochondrial integrity and mammalian longevity by the p62-Keap1-Nrf2-Nqo1 cascade
.
EMBO Rep
 .,
13
,
150
156
.
33
Tanji
K.
Odagiri
S.
Miki
Y.
Maruyama
A.
Nikaido
Y.
Mimura
J.
Mori
F.
Warabi
E.
Yanagawa
T.
Ueno
S.
, et al.  . (
2015
)
p62 Deficiency Enhances alpha-Synuclein Pathology in Mice
.
Brain Pathol
 .,
25
,
552
564
.
34
Doi
H.
Adachi
H.
Katsuno
M.
Minamiyama
M.
Matsumoto
S.
Kondo
N.
Miyazaki
Y.
Iida
M.
Tohnai
G.
Qiang
Q.
, et al.  . (
2013
)
p62/SQSTM1 differentially removes the toxic mutant androgen receptor via autophagy and inclusion formation in a spinal and bulbar muscular atrophy mouse model
.
J. Neurosci
 ,
33
,
7710
7727
.
35
Kurosawa
M.
Matsumoto
G.
Kino
Y.
Okuno
M.
Kurosawa-Yamada
M.
Washizu
C.
Taniguchi
H.
Nakaso
K.
Yanagawa
T.
Warabi
E.
, et al.  . (
2015
)
Depletion of p62 reduces nuclear inclusions and paradoxically ameliorates disease phenotypes in Huntington's model mice
.
Hum. Mol. Genet
 .,
24
,
1092
1105
.
36
Ishii
T.
Itoh
K.
Takahashi
S.
Sato
H.
Yanagawa
T.
Katoh
Y.
Bannai
S.
Yamamoto
M.
(
2000
)
Transcription factor Nrf2 coordinately regulates a group of oxidative stress-inducible genes in macrophages
.
J. Biol. Chem
 .,
275
,
16023
16029
.
37
Taylor
R.C.
Acquaah-Mensah
G.
Singhal
M.
Malhotra
D.
Biswal
S.
(
2008
)
Network inference algorithms elucidate Nrf2 regulation of mouse lung oxidative stress
.
PLoS Comput. Biol
 .,
4
,
e1000166.
38
Sasaki
S.
Nagai
M.
Aoki
M.
Komori
T.
Itoyama
Y.
Iwata
M.
(
2007
)
Motor neuron disease in transgenic mice with an H46R mutant SOD1 gene
.
J. Neuropathol. Exp. Neurol
 .,
66
,
517
524
.
39
Itoh
K.
Chiba
T.
Takahashi
S.
Ishii
T.
Igarashi
K.
Katoh
Y.
Oyake
T.
Hayashi
N.
Satoh
K.
Hatayama
I.
, et al.  . (
1997
)
An Nrf2/small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through antioxidant response elements
.
Biochem. Biophys. Res. Commun
 .,
236
,
313
322
.
40
Hadano
S.
Benn
S.C.
Kakuta
S.
Otomo
A.
Sudo
K.
Kunita
R.
Suzuki-Utsunomiya
K.
Mizumura
H.
Shefner
J.M.
Cox
G.A
., et al.  . (
2006
)
Mice deficient in the Rab5 guanine nucleotide exchange factor ALS2/alsin exhibit age-dependent neurological deficits and altered endosome trafficking
.
Hum. Mol. Genet
 ,
15
,
233
250
.
41
Vargas
M.R.
Burton
N.C.
Kutzke
J.
Gan
L.
Johnson
D.A.
Schafer
M.
Werner
S.
Johnson
J.A.
(
2013
)
Absence of Nrf2 or its selective overexpression in neurons and muscle does not affect survival in ALS-linked mutant hSOD1 mouse models
.
PLoS ONE
 ,
8
,
e56625.
42
Guo
Y.
Zhang
Y.
Wen
D.
Duan
W.
An
T.
Shi
P.
Wang
J.
Li
Z.
Chen
X.
Li
C.
(
2013
)
The modest impact of transcription factor Nrf2 on the course of disease in an ALS animal model
.
Lab. Invest
 .,
93
,
825
833
.
43
Boillee
S.
Yamanaka
K.
Lobsiger
C.S.
Copeland
N.G.
Jenkins
N.A.
Kassiotis
G.
Kollias
G.
Cleveland
D.W.
(
2006
)
Onset and progression in inherited ALS determined by motor neurons and microglia
.
Science
 ,
312
,
1389
1392
.
44
Yamanaka
K.
Chun
S.J.
Boillee
S.
Fujimori-Tonou
N.
Yamashita
H.
Gutmann
D.H.
Takahashi
R.
Misawa
H.
Cleveland
D.W.
(
2008
)
Astrocytes as determinants of disease progression in inherited amyotrophic lateral sclerosis
.
Nat. Neurosci
 .,
11
,
251
253
.
45
Pan
L.
Yoshii
Y.
Otomo
A.
Ogawa
H.
Iwasaki
Y.
Shang
H.F.
Hadano
S.
(
2012
)
Different human copper-zinc superoxide dismutase mutants, SOD1G93A and SOD1H46R, exert distinct harmful effects on gross phenotype in mice
.
PLoS ONE
 ,
7
,
e33409
.
46
Harada
H.
Warabi
E.
Matsuki
T.
Yanagawa
T.
Okada
K.
Uwayama
J.
Ikeda
A.
Nakaso
K.
Kirii
K.
Noguchi
N
., et al.  . (
2013
)
Deficiency of p62/Sequestosome 1 causes hyperphagia due to leptin resistance in the brain
.
J. Neurosci
 ,
33
,
14767
14777
.
47
Tanaka
K.
Okada
Y.
Kanno
T.
Otomo
A.
Yanagisawa
Y.
Shouguchi-Miyata
J.
Suga
E.
Kohiki
E.
Onoe
K.
Osuga
H.
, et al.  . (
2008
)
A dopamine receptor antagonist L-745,870 suppresses microglia activation in spinal cord and mitigates the progression in ALS model mice
.
Exp. Neurol
 .,
211
,
378
386
.
48
Hocking
L.J.
Lucas
G.J.
Daroszewska
A.
Mangion
J.
Olavesen
M.
Cundy
T.
Nicholson
G.C.
Ward
L.
Bennett
S.T.
Wuyts
W.
, et al.  . (
2002
)
Domain-specific mutations in sequestosome 1 (SQSTM1) cause familial and sporadic Paget's disease
.
Hum. Mol. Genet
 .,
11
,
2735
2739
.
49
Daroszewska
A.
van 't Hof
R.J.
Rojas
J.A.
Layfield
R.
Landao-Basonga
E.
Rose
L.
Rose
K.
Ralston
S.H.
(
2011
)
A point mutation in the ubiquitin-associated domain of SQSMT1 is sufficient to cause a Paget's disease-like disorder in mice
.
Hum. Mol. Genet
 .,
20
,
2734
2744
.
50
Kabuta
T.
Suzuki
Y.
Wada
K.
(
2006
)
Degradation of amyotrophic lateral sclerosis-linked mutant Cu,Zn-superoxide dismutase proteins by macroautophagy and the proteasome
.
J. Biol. Chem
 .,
281
,
30524
30533
.
51
Urushitani
M.
Kurisu
J.
Tsukita
K.
Takahashi
R.
(
2002
)
Proteasomal inhibition by misfolded mutant superoxide dismutase 1 induces selective motor neuron death in familial amyotrophic lateral sclerosis
.
J. Neurochem
 .,
83
,
1030
1042
.
52
Bruijn
L.I.
Houseweart
M.K.
Kato
S.
Anderson
K.L.
Anderson
S.D.
Ohama
E.
Reaume
A.G.
Scott
R.W.
Cleveland
D.W.
(
1998
)
Aggregation and motor neuron toxicity of an ALS-linked SOD1 mutant independent from wild-type SOD1
.
Science
 ,
281
,
1851
1854
.
53
Strom
A.L.
Shi
P.
Zhang
F.
Gal
J.
Kilty
R.
Hayward
L.J.
Zhu
H.
(
2008
)
Interaction of amyotrophic lateral sclerosis (ALS)-related mutant copper-zinc superoxide dismutase with the dynein-dynactin complex contributes to inclusion formation
.
J. Biol. Chem
 .,
283
,
22795
22805
.
54
Mizushima
N.
Yamamoto
A.
Matsui
M.
Yoshimori
T.
Ohsumi
Y.
(
2004
)
In vivo analysis of autophagy in response to nutrient starvation using transgenic mice expressing a fluorescent autophagosome marker
.
Mol. Biol. Cell
 ,
15
,
1101
1111
.
55
Kunita
R.
Otomo
A.
Mizumura
H.
Suzuki
K.
Showguchi-Miyata
J.
Yanagisawa
Y.
Hadano
S.
Ikeda
J.E.
(
2004
)
Homo-oligomerization of ALS2 through its unique carboxyl-terminal regions is essential for the ALS2-associated Rab5 guanine nucleotide exchange activity and its regulatory function on endosome trafficking
.
J. Biol. Chem
 .,
279
,
38626
38635
.
56
Kanekura
K.
Hashimoto
Y.
Niikura
T.
Aiso
S.
Matsuoka
M.
Nishimoto
I.
(
2004
)
Alsin, the product of ALS2 gene, suppresses SOD1 mutant neurotoxicity through RhoGEF domain by interacting with SOD1 mutants
.
J. Biol. Chem
 .,
279
,
19247
19256
.
57
Kanekura
K.
Hashimoto
Y.
Kita
Y.
Sasabe
J.
Aiso
S.
Nishimoto
I.
Matsuoka
M.
(
2005
)
A Rac1/phosphatidylinositol 3-kinase/Akt3 anti-apoptotic pathway, triggered by AlsinLF, the product of the ALS2 gene, antagonizes Cu/Zn-superoxide dismutase (SOD1) mutant-induced motoneuronal cell death
.
J. Biol. Chem
 .,
280
,
4532
4543
.
58
Gautam
M.
Jara
J.H.
Sekerkova
G.
Yasvoina
M.V.
Martina
M.
Ozdinler
P.H.
(
2016
)
Absence of alsin function leads to corticospinal motor neuron vulnerability via novel disease mechanisms
.
Hum. Mol. Genet
 .,
25
,
1074
1087
.
59
Sanchez
P.
De Carcer
G.
Sandoval
I.V.
Moscat
J.
Diaz-Meco
M.T.
(
1998
)
Localization of atypical protein kinase C isoforms into lysosome-targeted endosomes through interaction with p62
.
Mol. Cell Biol
 .,
18
,
3069
3080
.
60
Ishii
T.
Warabi
E.
Siow
R.C.
Mann
G.E.
(
2013
)
Sequestosome1/p62: a regulator of redox-sensitive voltage-activated potassium channels, arterial remodeling, inflammation, and neurite outgrowth
.
Free Radic. Biol. Med
 .,
65
,
102
116
.
61
Moscat
J.
Diaz-Meco
M.T.
(
2009
)
p62 at the crossroads of autophagy, apoptosis, and cancer
.
Cell
 ,
137
,
1001
1004
.
62
Matsumoto
G.
Wada
K.
Okuno
M.
Kurosawa
M.
Nukina
N.
(
2011
)
Serine 403 phosphorylation of p62/SQSTM1 regulates selective autophagic clearance of ubiquitinated proteins
.
Mol. Cell
 ,
44
,
279
289
.
63
Chen
S.
Zhou
L.
Zhang
Y.
Leng
Y.
Pei
X.Y.
Lin
H.
Jones
R.
Orlowski
R.Z.
Dai
Y.
Grant
S.
(
2014
)
Targeting SQSTM1/p62 induces cargo loading failure and converts autophagy to apoptosis via NBK/Bik
.
Mol. Cell Biol
 .,
34
,
3435
3449
.
64
Lim
J.
Lachenmayer
M.L.
Wu
S.
Liu
W.
Kundu
M.
Wang
R.
Komatsu
M.
Oh
Y.J.
Zhao
Y.
Yue
Z.
(
2015
)
Proteotoxic stress induces phosphorylation of p62/SQSTM1 by ULK1 to regulate selective autophagic clearance of protein aggregates
.
PLoS Genet
 .,
11
,
e1004987
.
65
Filomeni
G.D
Zio
D.
Cecconi
F.
, (
2015
)
Oxidative stress and autophagy: the clash between damage and metabolic needs
.
Cell Death Differ
 .,
22
,
377
388
.
66
Katsuragi
Y.
Ichimura
Y.
Komatsu
M.
(
2015
)
p62/SQSTM1 functions as a signaling hub and an autophagy adaptor
.
Febs. J
 .,
282
,
4672
4678
.
67
Sasaki
S.
Warita
H.
Murakami
T.
Shibata
N.
Komori
T.
Abe
K.
Kobayashi
M.
Iwata
M.
(
2005
)
Ultrastructural study of aggregates in the spinal cord of transgenic mice with a G93A mutant SOD1 gene
.
Acta Neuropathol
 .,
109
,
247
255
.
68
Grad
L.I.
Yerbury
J.J.
Turner
B.J.
Guest
W.C.
Pokrishevsky
E.
O'Neill
M.A.
Yanai
A.
Silverman
J.M.
Zeineddine
R.
Corcoran
L
., et al.  . (
2014
)
Intercellular propagated misfolding of wild-type Cu/Zn superoxide dismutase occurs via exosome-dependent and -independent mechanisms
.
Proc. Natl. Acad. Sci. U. S. A
 ,
111
,
3620
3625
.
69
Silverman
J.M.
Fernando
S.M.
Grad
L.I.
Hill
A.F.
Turner
B.J.
Yerbury
J.J.
Cashman
N.R.
(
2016
)
Disease mechanisms in ALS: Misfolded SOD1 transferred through exosome-dependent and exosome-independent pathways
.
Cell Mol. Neurobiol
 .,
36
,
377
381
.
70
Ponpuak
M.
Mandell
M.A.
Kimura
T.
Chauhan
S.
Cleyrat
C.
Deretic
V.
(
2015
)
Secretory autophagy
.
Curr. Opin. Cell Biol
 .,
35
,
106
116
.
71
Goodall
M.L.
Fitzwalter
B.E.
Zahedi
S.
Wu
M.
Rodriguez
D.
Mulcahy-Levy
J.M.
Green
D.R.
Morgan
M.
Cramer
S.D.
Thorburn
A.
(
2016
)
The autophagy machinery controls cell death switching between apoptosis and necroptosis
.
Dev. Cell
 ,
37
,
337
349
.
72
Dal Canto
M.C.
Gurney
M.E.
(
1995
)
Neuropathological changes in two lines of mice carrying a transgene for mutant human Cu,Zn SOD, and in mice overexpressing wild type human SOD: a model of familial amyotrophic lateral sclerosis (FALS)
.
Brain Res
 .,
676
,
25
40
.
73
Rosen
D.R.
Siddique
T.
Patterson
D.
Figlewicz
D.A.
Sapp
P.
Hentati
A.
Donaldson
D.
Goto
J.
O'Regan
J.P.
Deng
H.X
., et al.  . (
1993
)
Mutations in Cu/Zn superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis
.
Nature
 ,
362
,
59
62
.
74
Aoki
M.
Ogasawara
M.
Matsubara
Y.
Narisawa
K.
Nakamura
S.
Itoyama
Y.
Abe
K.
(
1993
)
Mild ALS in Japan associated with novel SOD mutation
.
Nat. Genet
 .,
5
,
323
324
.
75
Bandyopadhyay
U.
Nagy
M.
Fenton
W.A.
Horwich
A.L.
(
2014
)
Absence of lipofuscin in motor neurons of SOD1-linked ALS mice
.
Proc. Natl. Acad. Sci. U. S. A
 ,
111
,
11055
11060
.
76
Xie
Y.
Zhou
B.
Lin
M.Y.
Wang
S.
Foust
K.D.
Sheng
Z.H.
(
2015
)
Endolysosomal deficits augment mitochondria pathology in spinal motor neurons of asymptomatic fALS mice
.
Neuron
 ,
87
,
355
370
.
77
Saitoh
Y.
Fujikake
N.
Okamoto
Y.
Popiel
H.A.
Hatanaka
Y.
Ueyama
M.
Suzuki
M.
Gaumer
S.
Murata
M.
Wada
K
., et al.  . (
2015
)
p62 plays a protective role in the autophagic degradation of polyglutamine protein oligomers in polyglutamine disease model flies
.
J. Biol. Chem
 .,
290
,
1442
1453
.
78
Skibinski
G.
Parkinson
N.J.
Brown
J.M.
Chakrabarti
L.
Lloyd
S.L.
Hummerich
H.
Nielsen
J.E.
Hodges
J.R.
Spillantini
M.G.
Thusgaard
T
., et al.  . (
2005
)
Mutations in the endosomal ESCRTIII-complex subunit CHMP2B in frontotemporal dementia
.
Nat. Genet
 .,
37
,
806
808
.
79
Johnson
J.O.
Mandrioli
J.
Benatar
M.
Abramzon
Y.
Van Deerlin
V.M.
Trojanowski
J.Q.
Gibbs
J.R.
Brunetti
M.
Gronka
S.
Wuu
J.
, et al.  . (
2010
)
Exome sequencing reveals VCP mutations as a cause of familial ALS
.
Neuron
 ,
68
,
857
864
.
80
Maruyama
H.
Morino
H.
Ito
H.
Izumi
Y.
Kato
H.
Watanabe
Y.
Kinoshita
Y.
Kamada
M.
Nodera
H.
Suzuki
H
., et al.  . (
2010
)
Mutations of optineurin in amyotrophic lateral sclerosis
.
Nature
 ,
465
,
223
226
.
81
Deng
H.X.
Chen
W.
Hong
S.T.
Boycott
K.M.
Gorrie
G.H.
Siddique
N.
Yang
Y.
Fecto
F.
Shi
Y.
Zhai
H
., et al.  . (
2011
)
Mutations in UBQLN2 cause dominant X-linked juvenile and adult-onset ALS and ALS/dementia
.
Nature
 ,
477
,
211
215
.
82
Cirulli
E.T.
Lasseigne
B.N.
Petrovski
S.
Sapp
P.C.
Dion
P.A.
Leblond
C.S.
Couthouis
J.
Lu
Y.F.
Wang
Q.
Krueger
B.J
., et al.  . (
2015
)
Exome sequencing in amyotrophic lateral sclerosis identifies risk genes and pathways
.
Science
 ,
347
,
1436
1441
.
83
Freischmidt
A.
Wieland
T.
Richter
B.
Ruf
W.
Schaeffer
V.
Muller
K.
Marroquin
N.
Nordin
F.
Hubers
A.
Weydt
P
., et al.  . (
2015
)
Haploinsufficiency of TBK1 causes familial ALS and fronto-temporal dementia
.
Nat. Neurosci
 .,
18
,
631
636
.
84
Nixon
R.A.
(
2013
)
The role of autophagy in neurodegenerative disease
.
Nat Med
 .,
19
,
983
997
.
85
Shaid
S.
Brandts
C.H.
Serve
H.
Dikic
I.
(
2013
)
Ubiquitination and selective autophagy
.
Cell Death Differ
 .,
20
,
21
30
.
86
Damme
M.
Suntio
T.
Saftig
P.
Eskelinen
E.L.
(
2015
)
Autophagy in neuronal cells: general principles and physiological and pathological functions
.
Acta Neuropathol
 .,
129
,
337
362
.
87
Hadano
S.
Yoshii
Y.
Otomo
A.
Kunita
R.
Suzuki-Utsunomiya
K.
Pan
L.
Kakuta
S.
Iwasaki
Y.
Iwakura
Y.
Ikeda
J.E.
(
2010
)
Genetic background and gender effects on gross phenotypes in congenic lines of ALS2/alsin-deficient mice
.
Neurosci. Res
 .,
68
,
131
136
.
88
Yoshii
Y.
Otomo
A.
Pan
L.
Ohtsuka
M.
Hadano
S.
(
2011
)
Loss of glial fibrillary acidic protein marginally accelerates disease progression in a SOD1(H46R) transgenic mouse model of ALS
.
Neurosci. Res
 .,
70
,
321
329
.
89
Abe
K.
Fuchs
H.
Lisse
T.
Hans
W.
Hrabe de Angelis
M.
(
2006
)
New ENU-induced semidominant mutation, Ali18, causes inflammatory arthritis, dermatitis, and osteoporosis in the mouse
.
Mamm. Genome
 ,
17
,
915
926
.

Supplementary data