Loss of TMEM106B exacerbates C9ALS/FTD DPR pathology by disrupting autophagosome maturation

Abstract

Disruption to protein homeostasis caused by lysosomal dysfunction and associated impairment of autophagy is a prominent pathology in amyotrophic lateral sclerosis and frontotemporal dementia (ALS/FTD). The most common genetic cause of ALS/FTD is a G4C2 hexanucleotide repeat expansion in C9orf72 (C9ALS/FTD). Repeat-associated non-AUG (RAN) translation of G4C2 repeat transcripts gives rise to dipeptide repeat (DPR) proteins that have been shown to be toxic and may contribute to disease etiology. Genetic variants in TMEM106B have been associated with frontotemporal lobar degeneration with TDP-43 pathology and disease progression in C9ALS/FTD. TMEM106B encodes a lysosomal transmembrane protein of unknown function that is involved in various aspects of lysosomal biology. How TMEM106B variants affect C9ALS/FTD is not well understood but has been linked to changes in TMEM106B protein levels. Here, we investigated TMEM106B function in the context of C9ALS/FTD DPR pathology. We report that knockdown of TMEM106B expression exacerbates the accumulation of C9ALS/FTD-associated cytotoxic DPR proteins in cell models expressing RAN-translated or AUG-driven DPRs as well as in C9ALS/FTD-derived iAstrocytes with an endogenous G4C2 expansion by impairing autophagy. Loss of TMEM106B caused a block late in autophagy by disrupting autophagosome to autolysosome maturation which coincided with impaired lysosomal acidification, reduced cathepsin activity, and juxtanuclear clustering of lysosomes. Lysosomal clustering required Rab7A and coincided with reduced Arl8b-mediated anterograde transport of lysosomes to the cell periphery. Increasing Arl8b activity in TMEM106B-deficient cells not only restored the distribution of lysosomes, but also fully rescued autophagy and DPR protein accumulation. Thus, we identified a novel function of TMEM106B in autophagosome maturation via Arl8b. Our findings indicate that TMEM106B variants may modify C9ALS/FTD by regulating autophagic clearance of DPR proteins. Caution should therefore be taken when considering modifying TMEM106B expression levels as a therapeutic approach in ALS/FTD.

Keywords: ALS/FTD; C9orf72; DPR; TMEM106B; autophagy.

FIGURE 1

Knockdown of TMEM106B leads to accumulation of RAN translated C9ALS/FTD DPR proteins. HeLa cells were treated with non-targeting control (siCtrl) or TMEM106B-targeted siRNA (siTMEM) for 4 days before being transfected with empty vector (EV) or with 45 uninterrupted sense GGGGCC repeats (G4C2 × 45, Sense) or 43 uninterrupted antisense CCCCGG repeats (C4G2 × 43, Anti) with V5 tags in all three frames downstream of the repeats. (A) The levels of RAN translated DPR proteins were determined by immunoblot using anti-V5 antibodies. α-Tubulin indicates equal loading of samples. DPR levels were normalized to α-Tubulin and presented relative to siCtrl (unpaired t-test: *P ≤ 0.05, **P ≤ 0.01; Sense RAN DPR N = 4, Antisense RAN DPR N = 3 experiments). Knockdown of TMEM106B was confirmed by immunoblot using a TMEM106B antibody. (B) RAN translation was confirmed by immunoblot using DPR-specific poly(GA), poly(GR), poly(GP), and poly(PR) antibodies.

FIGURE 2

Knockdown of TMEM106B leads to accumulation of exogenous and endogenous C9ALS/FTD DPR proteins. (A–C) HeLa cells treated with either siCtrl or siTMEM were transfected with empty vector (EV) or AUG-driven synthetic, codon-optimized, AcGFP1-tagged 6 repeat, or V5-tagged 36 or 100 repeat poly(PR) (A), poly(GR) (B) or poly(GA) (C) DPR constructs. Levels of AcGFP1- or V5-tagged DPRs and TMEM106B were determined on immunoblot. α-Tubulin indicates equal loading of samples. DPR levels were normalized to α-Tubulin and presented relative to siCtrl (unpaired t-test: **P ≤ 0.01, ***P ≤ 0.001; AcGFP1-PR6 N = 3, V5-PR36 N = 4, V5-PR100 N = 7, AcGFP1-GR6 N = 3, V5-GR36 N = 5, V5-GR100 N = 2, AcGFP1-GA6 N = 3 experiments). (D) Endogenous GP DPR levels were determined by MSD ELISA in C9ALS/FTD iAstrocytes (patient) transduced with non-targeting control (shNTC) or TMEM106b-specific shRNA (shTMEM) and iAstrocytes derived from a neurologically healthy control (ctrl) (mean ± SEM; one-way ANOVA with Fisher’s LSD test: **P ≤ 0.01, ****P ≤ 0.0001; N = 3).

FIGURE 3

Loss of TMEM106B inhibits autophagy. (A) HeLa cells were treated with non-targeting control (siCtrl) or TMEM106B-targeted siRNA (siTMEM) for 4 days after which they were left untreated (Ctrl) or were treated with 100 nM Bafilomycin A1 (BafA1) for 6 h. The levels of LC3-II and SQSTM1/p62 were determined on immunoblot. α-Tubulin indicates equal loading of samples. LC3-II and SQSTM1/p62 levels were normalized against α-Tubulin and are presented relative to the siCtrl/Ctrl sample (mean ± SEM; unpaired t-test: ***P ≤ 0.001; ****P < 0.0001; siCtrl/Ctrl, siTMEM/Ctrl, N = 8; siCtrl/BafA1, siTMEM/BafA1, N = 7 experiments). An irrelevant lane was removed between the siTMEM/Ctrl and siCtrl/BafA1 lanes. (B) HeLa cells were treated with non-targeting control (siCtrl) or TMEM106B-targeted siRNA (siTMEM) for 3 days before transfection with mCherry-EGFP-LC3b and empty vector (EV) or myc-TMEM106B/T185. 24 h after transfection the cells were fixed and the number of EGFP/mCherry-positive (EGFP + &mCherry +) autophagosomes and mCherry-only-positive (mCherry +) autolysosomes quantified. Counts are presented relative to the combined count of autophagosomes and autolysosomes (mean ± SEM; one-way ANOVA with Fisher’s LSD test: ns not significant, ****P ≤ 0.0001; N = 92 (EV), 156 (siTMEM/EV), 93 (siTMEM/TMEM) cells from 3 experiments). Knockdown efficiency is shown in Supplementary Figure 4.

FIGURE 4

Knockdown of TMEM106B affects lysosomal pH and hydrolase activity. (A) Representative images of HeLa cells loaded with LysoSensor^® Green DND-189 following treatment with either siCtrl or siTMEM. Quantification of LysoSensor^® Green DND-189 fluorescence intensity per cell normalized to the mean intensity of siCtrl cells (mean ± SEM; unpaired t-test: ****P ≤ 0.0001; N = 4 experiments). (B) Representative images of cresyl violet fluorescence in HeLa cells incubated with cathepsin B Magic Red substrate following treatment with either siCtrl or siTMEM. Quantification of cresyl violet fluorescence intensity per cell normalized to the mean intensity of siCtrl cells (mean ± SEM; unpaired t-test: **P ≤ 0.01; N = 2 experiments). Knockdown efficiency is shown in Supplementary Figure 4.

FIGURE 5

TMEM106B regulates Arl8b-mediated trafficking of lysosomes to the cell periphery. (A) HeLa cells were treated with either siCtrl or siTMEM and transfected as indicated with empty vector (EV) or myc-tagged TMEM106B/T185. Cells were immunostained for the endogenous lysosomal marker LAMP2A (green) and transfected myc-TMEM106B (magenta). Images were blinded and lysosomes were classified as clustered or dispersed based on the distribution of LAMP2A. The percentage of cells with clustered lysosomes in each condition are presented (mean ± SEM; one-way ANOVA with Fisher’s LSD test: ***P ≤ 0.001, ****P ≤ 0.0001; siCtrl/EV and siTMEM/EV, N = 7 experiments; siTMEM/TMEM, N = 6 experiments. Knockdown efficiency is shown in Supplementary Figure 4. (B) HeLa cells were treated with either siCtrl, siTMEM, or siTMEM together with a pool of Rab7A targeted siRNA (siRab7) and immunostained for endogenous LAMP2A (green). Images were blinded and lysosomes classified as clustered or dispersed based on the distribution of LAMP2A. The percentage of cells with clustered lysosomes in each condition are presented (mean ± SEM; one-way ANOVA with Fisher’s LSD test: **P ≤ 0.01; N = 2 experiments. Scale bar = 20 μm. Knockdown efficiency was determined by immunoblot of Rab7 using tubulin as loading control. (C) HeLa cells treated with either siCtrl or siTMEM and transfected as indicated with empty vector (EV) or mycDDK-tagged Arl8b (Arl8b-mycDDK) were immunostained for the endogenous lysosomal marker LAMP2A (green) and transfected Arl8b-mycDDK (magenta). Images were blinded and lysosomes were classified as clustered or dispersed based on the distribution of LAMP2A. The percentage of cells with clustered lysosomes in each condition are presented (mean ± SEM; one-way ANOVA with Fisher’s LSD test: ****P ≤ 0.0001; N = 3 experiments. Scale bar = 20 μm. Knockdown efficiency is shown in Supplementary Figure 4.

FIGURE 6

Restoring Arl8b-mediated lysosomal trafficking rescues impaired autophagy in TMEM106B-depleted cells. (A) HeLa cells treated with either siCtrl or siTMEM106B were transfected with empty vector (EV), myc-tagged TMEM106B/T185, or mycDDK-tagged Arl8b (Arl8b-mycDDK). Cells were immunostained for endogenous SQSTM1/p62 (magenta) and the transfected myc-tag (green). Accumulation of SQSTM1/p62 was quantified as the mean intensity per cell (mean ± SEM; one-way ANOVA with Fisher’s LSD test, ****P ≤ 0.0001; N = 907 (siCtrl/EV), 1,488 (siTMEM/EV), 456 (siTMEM/TMEM), 367 (siTMEM/Arl8b) cells from 3 experiments). Scale bar = 20 μm. (B) HeLa cells were treated with non-targeting control (siCtrl) or TMEM106B targeted siRNA (siTMEM) for 3 days before transfection with mCherry-EGFP-LC3b and empty vector (EV) or Arl8b-mycDDK. 24 h after transfection the cells were fixed and the number of EGFP/mCherry-positive autophagosomes and mCherry-only-positive autolysosomes quantified. Counts are presented relative to the combined count of autophagosomes and autolysosomes (mean ± SEM; one-way ANOVA with Fisher’s LSD test: ns not significant, ****P ≤ 0.0001; N = 92 (siCtrl/EV), 156 (siTMEM/EV), 166 (siTMEM/Arl8b) cells from 3 experiments; siCtrl/EV and siTMEM/EV are identical to Figure 3B). Knockdown efficiency is shown in Supplementary Figure 4.

FIGURE 7

Restoring Arl8b-mediated lysosomal trafficking rescues C9ALS/FTD DPR levels in TMEM106B-depleted cells. HeLa cells treated with either siCtrl or siTMEM106B were transfected with mCherry, myc-tagged TMEM106B/T185, or mycDDK-tagged Arl8b (Arl8b-mycDDK) together with control mCherry (mCh), V5-tagged 100 repeat poly(PR) (PR100) (A) or V5-tagged 100 repeat poly(GR) (GR100) (B). Cells were immunostained for V5 (green) and myc (magenta); mCherry fluorescence (magenta) was visualized when present. The mean V5-DPR intensity per cell was quantified for each condition and normalized to the mean intensity in siCtrl-treated cells (mean ± SEM; one-way ANOVA with Fisher’s LSD test, ****P ≤ 0.0001; (A) N = 419 (siCtrl/mCh), 448 (siTMEM/mCh), 496 (siTMEM/TMEM), 473 (siTMEM/Arl8b), (B) N = 341 (siCtrl/mCh), 354 (siTMEM/mCh), 350 (siTMEM/TMEM), 326 (siTMEM/Arl8b) cells from 3 experiments). Scale bar = 20 μm. Knockdown efficiency is shown in Supplementary Figure 4.