An enhanced integrated stress response ameliorates mutant SOD1-induced ALS

Abstract

Varied stresses to cells can lead to a repression in translation by triggering phosphorylation of eukaryotic translation initiator factor 2α (eIF2α), which is central to a process known as the integrated stress response (ISR). PKR-like ER-localized eIF2 kinase (PERK), one of the kinases that phosphorylates eIF2α and coordinates the ISR, is activated by stress occurring from the accumulation of misfolded or unfolded proteins in the endoplasmic reticulum (ER). Mutant Cu/Zn superoxide dismutase (mtSOD1) is thought to cause familial amyotrophic lateral sclerosis (FALS) because it misfolds and aggregates. Published studies have suggested that ER stress is involved in FALS pathogenesis since mtSOD1 accumulates inside the ER and activates PERK leading to phosphorylated eIF2α (p-eIF2α). We previously used a genetic approach to show that haploinsufficiency of PERK significantly accelerates disease onset and shortens survival of G85R mtSOD1 FALS transgenic mice. We now show that G85R mice that express reduced levels of active GADD34, which normally dephosphorylates p-eIF2α and allows recovery from the global suppression of protein synthesis, markedly ameliorates disease. These studies emphasize the importance of the ISR, and specifically the PERK pathway, in the pathogenesis of mtSOD1-induced FALS and as a target for treatment. Furthermore, the ISR may be an appropriate therapeutic target for sporadic ALS and other neurodegenerative diseases since misfolded proteins have been implicated in these disorders.

Figure 1.

Plots of disease onset (A), the end of early disease (B), and survival (C) as well as diagrams of the duration (with standard deviation) of the early phase (D) and late phase (E) of disease in G85R/GADD34^+/ΔC (n = 19) versus G85R transgenic mice (n = 20). In the Y axis of (A)–(C), 1 refers to the total number of animals. ** P < 0.001.

Figure 2.

Neuropathological and immunohistochemical studies of the anterior horn of the lumbar spinal cord of G85R, G85R/GADD34^+/ΔC and littermate non-transgenic mice at varying ages. Sections of the anterior horn of the lumbar spinal cord were stained with: (A) Nissl for MNs, anti-GFAP antibody for astrocytes and anti-Iba1 antibody for microglia. (B) Bar diagrams show mean ± standard deviation calculated by counting cells in 15–20 sections from the lumbar spinal cord anterior horn of four mice from each of the groups (G85R, G85R/GADD34^+/ΔC and littermate non-transgenic mice) at varying ages. The scale bar = 50 µm. ** P < 0.001.

Figure 3.

mtSOD1 in the anterior horn of the lumbar spinal cord of G85R, G85R/GADD34^+/ΔC and littermate non-transgenic mice at varying ages. In this figure, ‘Wk’ refers to the early phase of disease, while ‘End’ refers to end stage; the approximate ages of the mice are noted. (A) Representative sections of the anterior horn of the lumbar spinal cord were stained with anti-human SOD1 antibody to identify aggregates. (B) Bar diagram shows the mean ± standard deviation of mtSOD1 aggregates/mm². Note that there are so few aggregates in the anterior horn of control and G85R/GADD34^+/ΔC mice at 345 days that the bars are not visible. The number of aggregates in G85R mice at end stage (345 days) was statistically significantly increased compared with G85R/GADD34^+/ΔC mice at end stage (420 days). *P < 0.05. (C) Representative western blot of homogenates of the lumbar spinal cord of G85R and G85R/GADD34^+/ΔC mice sacrificed at different times of disease, and processed as described in Materials and Methods. The samples were treated without β-mercaptoethanol before boiling and then loaded on 15% SDS polyacrylamide gels, electrophoresed, blotted and then immunostained using a human-specific anti-SOD1 antibody [to detect the monomeric and high molecular weight (HMW) forms] and an anti-β-tubulin antibody (as a loading control). (D) Representative western blot of homogenates of the lumbar spinal cord of G85R and G85R/GADD34^+/ΔC mice that were harvested and processed as described above; however, the samples were diluted 1:5 and treated with β-mercaptoethanol before boiling to identify total SOD1. (E) Bar diagram shows total SOD1 in the lumbar spinal cord of G85R and G85R/GADD34^+/ΔC mice at different ages and stages of disease. Each bar, which represents data from four animals, shows the mean ± standard deviation of the ratio of SOD1 signal to β-tubulin signal, displayed as a fraction of the value of the G85R mice at 330 days, which is set at 1. There was a statistically significant greater amount of total SOD1 in the lumbar spinal cord of G85R mice at 345 days (end stage) compared with G85R/GADD34^+/ΔC mice at 380, 400 and 420 days (end stage). *P < 0.05, ** P < 0.001.

Figure 4.

Detection of markers of the ISR in G85R and G85R/GADD34^+/ΔC mice. (A) Representative western blots from differently aged G85R, G85R/GADD34^+/ΔC, GADD34^+/ΔC and control non-transgenic littermate mice immunostained with antibodies to phosphorylated (p)-eIF2α, ATF4 and CHOP. Anti-β tubulin antibody was used as a loading control. ‘Pre’ refers to prior to disease onset, ‘Wk’ refers to the early phase of disease, while ‘End’ refers to end stage; the approximate ages of the mice are provided. (B) Bar diagrams showing quantitation of UPR markers from western blots of homogenates of the lumbar spinal cord from G85R (n = 4) and G85R/GADD34^+/ΔC (n = 4) mice at end stage of disease. The mean ± standard deviation is shown. The value of the specific ISR marker detected in G85R mice was arbitrarily set as 1. * P < 0.05, ** P < 0.001.