Genome-wide screen reveals that loss of DBT protects cells against toxicity of proteasomal inhibition.

(A) Workflow of the CRISPR screen in RPE1 cells, which were transduced with a lentiviral GeCKO sgRNA library and selected for the sgRNA expression and then survival after treatments with the proteasome inhibitor MG132. Individual surviving cell colonies were collected for sequencing and subsequent analysis. (B) Left: The cytotoxicity analysis of WT and DBT KO RPE1 cells treated with MG132 at different doses for 96 h (n = 3). Right: The time course analysis of MG132-induced cytotoxicity in the WT and DBT KO cells (n = 3). (C) Immunoblot analysis of WT RPE1, DBT KO, and DBT’ cells. The DBT’ cells expressed an engineered DBT cDNA that resisted DBT-targeted Cas9 cleavage and rescued the DBT expression in the KO cells. (D) Cell viability was measured by Calcein-AM staining in WT RPE1, DBT KO, and DBT’ cells treated with MG132 (2 μM, 96 hr). Scale bar, 100 μm. (E) Quantification of the cell viability measured by Calcein-AM staining in (D) (n = 9). (F) Left: Immunoblot analysis of RPE1 cells transfect with DBT shRNAs and non-targeting control shRNAs. Right: Quantification of the cell viability under treatment with MG132 (2 μM, 48 hr), as measured by Calcein-AM staining (n = 4). (G) Immunoblotting and quantification of cleaved PARP as an MG132-induced cell death marker (n = 4). (H) Immunoblotting and quantification of cleaved Caspase 3 as an MG132-induced cell death marker (n = 3). Error bars represent means ± SEM. *p ≤ 0.05; **p ≤ 0.01; ****p ≤ 0.0001.

Loss of DBT decreased accumulation of ubiquitinated proteins upon proteasomal inhibition.

(A) WT and DBT KO RPE1 cells treated with MG132 (2 μM, 48 hr) or the DMSO solvent control were analyzed for accumulation of ubiquitinated proteins upon proteasomal inhibition with denaturing SDS-PAGE. The bar graph represents quantification of the high-molecular-weight poly-ubiquitinated proteins (n = 4). (B) The cells treated with MG132 (2 μM, 72 hr) or the DMSO solvent control were analyzed for the levels of ubiquitinated protein with immunostaining. The bar graph represents quantification of the anti-ubiquitin immunofluorescent signals (n = 4 independent groups, each consisting of 7 cells). Scale bar, 10 μm. (C) The cells treated with MG132 (2 μM, 72 hr) or the DMSO solvent control were stained with dye that detects protein aggregates. The bar graph represents quantification of the ProteoStat signals (n = 7 biological replicates of DBT KO cells and 3 replicates of WT control cells). Scale bar, 20 μm. Error bars represent means ± SEM. “n.s.”, no significance; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001.

Loss of DBT preserves autophagic activities under proteasomal inhibition.

(A) Immunoblot analysis of LC3II levels in WT and DBT KO RPE1 cells treated with MG132 (2 μM, 48 hr) and with or without Baf A1 (Bafilomycin A1, 100 nM, 4 hr). (B) Quantification of the LC3II levels in (A) (n = 4). (C) Quantification of the autophagic flux as measured by the ratios of LC3II levels before and after the Baf A1 treatment (n = 4). (D) Immunoblot analysis of p62 in WT and DBT KO cells treated with MG132 (2 μM, 48 hr) with or without Baf A1 (Bafilomycin A1, 100 nM, 4 hr). (E) Quantification of the p62 levels in (D) (n = 3). (F) Quantification of the autophagic flux as measured by the ratios of p62 levels before and after the Baf A1 treatment (n = 3). (G) Cell viability analysis with crystal violet staining was performed on WT and DBT KO RPE1 cells treated with MG132 and with or without Baf A1 (n = 6). Error bars represent means ± SEM. “n.s.”, no significance; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001.

Loss of DBT activates AMPK under proteasomal inhibition through energy regulation.

(A) Intracellular BCAA levels were measured in WT and DBT KO RPE1 cells treated with MG132 (2 μM, 48 hr) (n = 6). (B) Intracellular ATP/ADP ratios were measured in WT and DBT KO RPE1 cells treated with MG132 (2 μM, 48 hr) (n = 5). (C) Activation of AMPK in DBT KO RPE1 cells treated with MG132 (2 μM, 48 hr), as indicated by the increase in the levels of phosphorylated AMPK (n = 3). (D) The knockdown of AMPK by specific shRNAs abolished the protective effects of loss of DBT against MG132-induced toxicity in DBT KO RPE1 cells, as indicated by the cell viability measured with crystal violet staining (n = 4). Error bars represent means ± SEM. “n.s.”, no significance; *p ≤ 0.05; ****p ≤ 0.0001.

AMPK downstream signaling enhances autophagy upon DBT deficiency and proteasomal inhibition.

(A) Immunoblot analysis of the autophagy marker LC3II in DBT KO RPE1 cells after the knockdown of AMPK using shRNAs versus non-targeting control shRNAs (CTRL), under MG132 treatment conditions (2 μM, 48 hr), with or without the Baf A1 treatment. The autophagic flux is measured by calculating the ratios of LC3II protein levels with Baf A1 treatment to those without the Baf A1 treatment (n = 3). (B) Immunoblot analysis of the AMPK downstream effectors that regulate mTOR activities, including ULK1 and TSC2, in WT and DBT KO RPE1 cells with or without treatment with MG132 (2 μM, 48 hr). The activities of these regulators are quantified by measuring the levels of phosphorylation of ULK1-S371, TSC2-S1387, and AMPK-T172 (n = 3). (C) Immunoblot analysis of the mTOR downstream marker S6K and its phosphorylation in DBT KO RPE1 cells after the knockdown of the negative regulator of mTOR, TSC1, using shRNAs versus control shRNAs (CTRL), under MG132 treatment conditions (2 μM, 48 hr) (n = 3). (D) Immunoblot analysis of LC3II and quantification of the autophagic flux in DBT KO RPE1 cells after the knockdown of TSC1 under MG132 treatment conditions (2 μM, 48 hr) with or without the Baf A1 treatment (n = 3). (E) Cell viability analysis with crystal violet staining of WT and DBT KO RPE1 cells after the knockdown of TSC1 using shRNAs versus control shRNAs (CTRL), under MG132 treatment conditions (2 μM, 48 hr) (n = 4). Error bars represent means ± SEM. “n.s.”, no significance; *p ≤ 0.05; **p ≤ 0.01; ****p ≤ 0.0001.

Loss of DBT protects against proteotoxicity of mutant TDP-43 in mammalian neurons and Drosophila models.

(A) Cell toxicity of ALS-linked TDP-43M337V expressed in WT and DBT KO RPE1 cells, with Dendra2 as a control, as measured by Calcein-AM staining (n = 4). Scale bar, 300 μm. (B) Neuronal toxicity of TDP-43M337V expressed in mouse ES cell-differentiated motor neurons with or without DBT shRNA-mediated knockdown, compared to that of GFP control, as measured by Calcein-AM staining (n = 9). Scale bar, 300 μm. (C) TDP-43M337V protein steady-state levels are significantly lower in DBT KO RPE1 cells than in WT control cells, as measured by immunoblot analysis (n = 3). (D) The half-life of TDP-43M337V protein as measured in cycloheximide chase assays is significantly shorter in the DBT KO cells than in WT RPE1 control cells (n = 3 independent experiments; p = 0.0326). (E) Immunoblot analysis of the autophagy marker LC3II and quantification of the autophagic flux in WT and DBT KO RPE1 cells transfected with myc-TDP-43M337V with or without the Baf A1 treatment. The autophagic flux is measured by calculating the ratios of LC3II protein levels with Baf A1 treatment to those without the Baf A1 treatment (n = 3). (F) The reduction of DBT by RNAi or CRISPR led to strongly suppressed eye degeneration phenotypes in the TDP-43M337V fly strain when compared with the control Luc RNAi (CTRL). The eye degeneration phenotypes were quantified by measuring the pigment content in adult eyes (n = 3 independent groups with each containing fly heads from 4 males and 4 females). Scale bar, 100 μm. Error bars represent means ± SEM. *p ≤ 0.05; ***p ≤ 0.001; ****p ≤ 0.0001.

DBT is abnormally upregulated in ALS patient neurons.

(A) Immunoblot analysis of human spinal cord tissues from ALS patients and non-neurological controls (n = 24 ALS cases and 8 non-neurological control cases). (B) Fluorescent immunostaining against DBT in the spinal cords from ALS patients and an age-matched control cases indicates that the accumulation of DBT in the patient’s neurons as identified by their morphological characteristics (n = 23 neurons from 3 ALS cases and n = 24 neurons from 3 control cases). Arrows point to representative motor neurons. Scale bar, 50 μm. Error bars represent means ± SEM. **p ≤ 0.01.