Loss of C9orf72 Enhances Autophagic Activity via Deregulated mTOR and TFEB Signaling

Abstract

The most common cause of the neurodegenerative diseases amyotrophic lateral sclerosis and frontotemporal dementia is a hexanucleotide repeat expansion in C9orf72. Here we report a study of the C9orf72 protein by examining the consequences of loss of C9orf72 functions. Deletion of one or both alleles of the C9orf72 gene in mice causes age-dependent lethality phenotypes. We demonstrate that C9orf72 regulates nutrient sensing as the loss of C9orf72 decreases phosphorylation of the mTOR substrate S6K1. The transcription factor EB (TFEB), a master regulator of lysosomal and autophagy genes, which is negatively regulated by mTOR, is substantially up-regulated in C9orf72 loss-of-function animal and cellular models. Consistent with reduced mTOR activity and increased TFEB levels, loss of C9orf72 enhances autophagic flux, suggesting that C9orf72 is a negative regulator of autophagy. We identified a protein complex consisting of C9orf72 and SMCR8, both of which are homologous to DENN-like proteins. The depletion of C9orf72 or SMCR8 leads to significant down-regulation of each other's protein level. Loss of SMCR8 alters mTOR signaling and autophagy. These results demonstrate that the C9orf72-SMCR8 protein complex functions in the regulation of metabolism and provide evidence that loss of C9orf72 function may contribute to the pathogenesis of relevant diseases.

Fig 1. C9orf72 knockout mice show age-dependent lethality.

A) Schematic representation of the generation of the C9orf72 KO mice. Diagram depicts the mouse C9orf72 wild-type allele, C9orf72 targeting cassette, and the KO allele after SOX-Cre recombination. B) Immunoblot analysis of C9orf72 -/- and wild-type littermate brain homogenates. C9orf72 antibody detects a band at 55 kDa not present in C9orf72-/- lysates. C) Survival analysis of C9orf72-/-, C9orf72+/-, and wild-type mice. Lifespans were monitored and plotted using Kaplan-Meyer curve. This analysis revealed a loss of C9orf72 causes a significant decrease in survival when compared with wild-type littermates (n = 74, C9orf72-/-; n = 79, C9orf72+/-; n = 31, wild-type, *p<0.005).

Fig 2. Loss of C9orf72 decreases mTOR activation.

A) Immunoblot analysis of mTOR activity after starvation and amino acid stimulation. HEK293T cells were transfected with control or C9orf72 shRNA for 72 hours before treatment. Cells were starved for 50 min before being supplemented with amino acids for 10–20 min prior to lysate collection. The mTOR activity was assessed by immunoblotting for the phosphorylation of its downstream substrate S6K1. Arrow points to C9orf72 and asterisk indicates a cross-reacting band. B) Quantification of p-S6K1 (T389) levels after starvation and amino acid stimulation in control and C9orf72 knockdown cells from three independent experiments. Knockdown of C9orf72 significantly decreased p-S6K1 levels compared with cells treated with control shRNA (n = 3, *p<0.5). C) Immunoblot analysis of mTOR activity in C9orf72-/- MEFs after starvation and amino acid stimulation. C9orf72-/- and wild-type cells were starved for 50 min using EBSS and supplemented with amino acids for 10 min prior to lysate collection. C9orf72-/- MEF cells show a decrease in p-S6K1 when compared with wild-type cells. Student’s t test is used and data is presented as mean ± SEM.

Fig 3. Loss of C9orf72 increases TFEB in the nucleus.

A) Western analysis of TFEB levels after knockdown of C9orf72 in HEK293T cells. HEK293T were transfected with GFP-tagged TFEB and C9orf72 shRNA or control shRNA and then the lysates were analyzed using an antibody to GFP. B) Quantification of GFP-TFEB levels after knockdown of C9orf72 in HEK293T cells. Loss of C9orf72 significantly increases GFP-TFEB protein levels (n = 3, *p<0.05). C) Representative image of the nuclear enrichment of GFP-TFEB after knockdown of C9orf72 in HEK293T cells. D) Quantification of GFP-TFEB localization. Loss of C9orf72 increases the number of cells that show strong nuclear levels of GFP-TFEB (n = 4 control groups of 277 cells and n = 3 groups of 384 cells for C9orf72 in two independent trials, *p<0.05). E) Western blot analysis of nuclear fractionation of GFP-TFEB in HEK293T cells. Equal amounts of proteins from nuclear and cytoplasmic fractions (a ratio of approximately 6:1 in nuclear to cytoplasmic parts) were loaded. The GFP-TFEB levels were significantly increased in the nucleus upon C9orf72 knockdown. PARP and Caspase-3 were used as the loading control for the nuclear and cytoplasmic fractions, respectively. F) Quantification of nucleocytoplasmic distribution of GFP-TFEB fluorescence in C9orf72-/- and wild-type MEF cells (n = 74 for control and n = 57 for C9orf72-/- in four independent trials, ***p<0.0001). G) Quantification of LysoTracker stained vesicles in in C9orf72-/- and wild-type MEF cells (n = 40 for control and n = 64 for C9orf72-/- in three independent trials, ***p<0.0001). H) Representative images of the nuclear enrichment of GFP-TFEB (green) and the increased LysoTracker vesicles (red) in C9orf72-/- MEF cells. I) Analysis of the levels of TFEB and its downstream targets in C9orf72 KO mice. Immunoblot analysis of C9orf72-/- mouse and wild-type littermate brain homogenates shows a dramatic increase of TFEB in C9orf72 KO animals compared with wild-type control animals. TFEB targets LAMP1 and LAMP2 were also increased in the KO mice. Scale bars: 10 μm. Student’s t test is used and data is presented as mean ± SEM.

Fig 4. Loss of C9orf72 increases starvation-induced autophagy flux.

A) Representative image of western blot analysis of LC3 in nutrient deprivation (ND) conditions. Wild-type or C9orf72-/- MEF cells were treated with or without Bafilomycin in nutrient deprivation conditions for three hours, and the LC3I and LC3II levels were determined by western blotting. B) Quantification of western blot analysis of LC3II as in A). LC3II level was significantly increased in C9orf72-/- cells treated with Bafilomycin compared with wild-type cells (n = 3, *p<0.05). C) Representative live cell images of RFP-Rab7/GFP-LC3 co-localization in C9orf72-/- MEFs. RFP-Rab7 and GFP-LC3 were transfected in wild-type or C9orf72-/- cells and treated with Bafilomycin in nutrient deprivation conditions. D) Quantification of GFP-LC3 intensity in C) (n = 3 independent trials with a total of 125 cells, *p<0.05). E) Quantification of an autophagy flux index in C) defined as the difference in the volumes of LC3-positive vesicles before and after Bafilomycin treatment (n = 3 independent trials with a total of 125 cells, *p<0.05). F) Quantification of the fraction of LC3/Rab7 colocalizing vesicles in C) (n = 3 independent trials with a total of 125 cells, *p<0.05). Scale bar: 10 μm. Student’s t test is used and data is presented as mean ± SEM.

Fig 5. p62 is decreased in C9orf72 knockout mouse brains.

A) Schematic representation of C9orf72 mouse tissue analysis. C9orf72-/- or wild-type mice were fed normal chow or amino acid-deficient chow (low- protein diet) for 4 weeks prior to harvesting tissue. B) Immunohistochemistry analysis of p62 in mouse brain. Brain of mouse on low protein diets were harvested and stained with p62 antibodies. p62 is decreased in C9orf72-/- brains compared to wild type brain, although it is expressed in the same region of brain. No p62 aggregates detected. C) Immunoblot analysis of brain homogenates from C9orf72 KO and wild-type animals fed control or low protein diet. C9orf72 KO mice show a decrease in p62 levels compared with wild-type littermates under the low protein diet condition. D) Quantification of p62 levels in brain homogenates derived from C9orf72 KO and wild-type animals fed control or low-protein diet. C9orf72 KO mice show a slight but significant decrease in p62 levels when compared with wild-type littermates under the low protein diet condition (n = 3, *p<0.05). Student’s t test is used and data is presented as mean ± SEM.

Fig 6. C9orf72 and SMCR8 form a stable protein complex.

A) Schematic representation of SILAC mass spectrometry screen for C9orf72 interacting proteins. Metabolically labeled HEK293T cells expressing Flag-tagged C9orf72 Isoform A (heavy) or non-transfected control cells (light) were incubated with anti-Flag-conjugated beads and the resulting eluents were pooled, resolved on an SDS-PAGE gel, trypsin-digested and analyzed by LC-MS/MS. B) Immunoblot analysis of C9orf72 immunoprecipitates. HEK293T cells were either mock-transfected or transfected with C9orf72-Flag, and the protein immunoprecipitated from the cell lysates using anti-Flag-conjugated beads. The resulting eluents were probed for Flag (to detect C9orf72) and SMCR8. SMCR8 is only present in C9orf72-positive immunoprecipitates and does not bind to beads alone. C) Immunoblot analysis of SMCR8 immunoprecipitates. HEK293T cells were transfected with C9orf72-Flag and the resulting lysates were incubated with SMCR8 antibody and protein A Sepharose beads and the resulting immunoprecipitates analyzed via immunoblotting. SMCR8 antibody, but not control IgG, immunoprecipitated C9orf72-Flag. D) Analysis of SMCR8 levels in C9orf72 KO mice. Immunoblot analysis of C9orf72-/- mouse and wild-type littermate brain homogenates shows the lack of SMCR8 protein in C9orf72 KO animals compared with wild-type control animals. E) Analysis of C9orf72 levels in SMCR8 KO cells. Immunoblot analysis of SMCR8 KO HAP1 cell lysates shows a decrease in C9orf72 levels compared with control HAP1 cells. Arrow points to C9orf72 and asterisk indicates a cross-reacting band. F) Analysis of SMCR8 levels after exogenous C9orf72 expression. HEK293T cells were either mock-transfected or transfected with C9orf72-Flag and the indicated proteins analyzed via immunoblotting. G) Analysis of C9orf72 turnover after SMCR8 co-expression. HEK293T cells were co-transfected with C9orf72-V5 and mCherry-SMCR8, or mCherry control, expression constructs. Approximately 48 hours after transfection, cells were treated with cycloheximide, collected at the indicated time points, and analyzed via immunoblotting. Overexpression of SMCR8 dramatically stabilizes C9orf72. H) Quantification of the results in panel F from four independent experiments. Overexpression of C9orf72 significantly increased SMCR8 levels (n = 4, *p<0.05). Student’s t test is used and data is presented as mean ± SEM.

Fig 7. SMCR8 regulates mTOR signaling and autophagy.

A) Immunoblot analysis of mTOR activity after starvation and amino acid stimulation. HAP1 control and SMCR8 KO cells were starved for 50 minutes and supplemented with amino acids for 10–20 min before lysate collection. The mTOR activity was assessed via immunoblotting for the phosphorylation of its downstream substrate S6K1. B) Quantification of p-SK61 (T389) levels after starvation and amino acid stimulation in control and SMCR8 knockout cells from three independent experiments. Knockout of SMCR8 significantly decreased p-S6K1 levels compared with wild-type cells (n = 3, *p<0.05, **p<0.005). C) Immunoblot analysis of LC3 levels after shRNA-mediated knockdown of SMCR8 in HEK293T cells. HEK293T cells were transfected with SMCR8 or control shRNA and lysates collected 72 hours after transfection. The indicated proteins were detected by immunoblotting of the lysates using an antibody against LC3. D) Quantification of the LC3II to LC3I ratio obtained after shRNA-mediated knockdown of SMCR8 in HEK293T cells from three independent experiments. Knockdown of SMCR8 significantly decreased the ratio of LC3II to LC3I compared with control cells (n = 3, *p<0.05). E) Immunoblot analysis of LC3 levels before and after autophagy induction with nutrient starvation. HEK293T cells were transfected with SMCR8 shRNA or scrambled shRNA control. Approximately 72 hours after transfection, cells were treated with starvation medium (EBSS) with and without Bafilomycin for 2 hours and the resulting lysates were analyzed via immunoblotting for LC3. Student’s t test is used and data is presented as mean ± SEM.