Targeting the autophagy-NAD axis protects against cell death in Niemann-Pick type C1 disease models

Abstract

Impairment of autophagy leads to an accumulation of misfolded proteins and damaged organelles and has been implicated in plethora of human diseases. Loss of autophagy in actively respiring cells has also been shown to trigger metabolic collapse mediated by the depletion of nicotinamide adenine dinucleotide (NAD) pools, resulting in cell death. Here we found that the deficit in the autophagy-NAD axis underpins the loss of viability in cell models of a neurodegenerative lysosomal storage disorder, Niemann-Pick type C1 (NPC1) disease. Defective autophagic flux in NPC1 cells resulted in mitochondrial dysfunction due to impairment of mitophagy, leading to the depletion of both the reduced and oxidised forms of NAD as identified via metabolic profiling. Consequently, exhaustion of the NAD pools triggered mitochondrial depolarisation and apoptotic cell death. Our chemical screening identified two FDA-approved drugs, celecoxib and memantine, as autophagy activators which effectively restored autophagic flux, NAD levels, and cell viability of NPC1 cells. Of biomedical relevance, either pharmacological rescue of the autophagy deficiency or NAD precursor supplementation restored NAD levels and improved the viability of NPC1 patient fibroblasts and induced pluripotent stem cell (iPSC)-derived cortical neurons. Together, our findings identify the autophagy-NAD axis as a mechanism of cell death and a target for therapeutic interventions in NPC1 disease, with a potential relevance to other neurodegenerative disorders.

Fig. 1. Metabolic deficits underlying apoptotic cell death in respiring Npc1^-/- cells.

Immunoblot analyses for caspase-3 cleavage, LC3B lipidation and NPC1 expression (a), phase contrast images (b) and cytotoxicity assay results (c) of Npc1^+/+ and Npc1^-/- MEFs transduced with empty or NPC1 WT, after 72 h (a, b) or 96 h (c) culture in galactose medium. d, e, Metabolomics analyses on Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium for 48 h. d, Volcano plot representation of metabolites in a pairwise comparison of Npc1^-/- to Npc1^+/+. Thresholds (|Log₂(fold change)| > 1 and -log₁₀(adjusted P) > 1) are shown as dashed lines. e Heatmap and clustering representations of analysed metabolites. Metabolites that change significantly d are highlighted in red. f Measurement of NAD⁺ and NADH levels in Npc1^+/+ and Npc1^-/- MEFs after 60 h culture in galactose medium. Data are mean ± SEM of n = 3 biological replicates (a, c, f). P values were calculated by one-way ANOVA followed by multiple comparisons with the two-stage linear step-up procedure of Benjamini, Krieger and Yakutieli (a, c), multiple t-test with original FDR method of Benjamini and Hochberg (d) and unpaired two-tailed Student’s t-test (f). **P < 0.01; ***P < 0.001; with respect to Npc1^+/+ MEFs or between the indicated groups. Scale bar: 200 µm (b).

Fig. 2. Mitochondrial deficits in Npc1^-/- cells.

a Fluorescence microscopy images and quantification of mitophagy of Npc1^+/+ and Npc1^-/- MEFs expressing mt-mKeima after 24 h culture in galactose medium. b transmission electron micrographs and the proportion of mitochondria with abnormal morphology of Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium for 24 h. c–h Mitochondrial respiration and glycolytic function of Npc1^+/+ and Npc1^-/- MEFs were analysed by measuring oxygen consumption rate (OCR) or extracellular acidification rates (ECAR), respectively. ECAR and/or OCR and ATP production rates in whole cell after 24 h culture either in glucose (c, d) or galactose medium (g, h). OCR in permeabilised cells treated with either CI or CII substrates (e, f) was measured. Respiratory control ratios (RCR) indicate a capacity for substrate oxidation and ATP turnover (f). i Live fluorescence microscopy images of Npc1^+/+ and Npc1^-/- MEFs after 24 h culture in galactose medium, co-stained with MitoSOX and mitotracker green (MTG). j Confocal microscopy images of live Npc1^+/+ and Npc1^-/- MEFs after 60 h culture in galactose medium, co-stained with TMRM and MTG. Data are mean ± SEM of n = 3-4 biological replicates as indicated (a, b, i, j) or n = 4-14 technical replicates (c–h) as indicated. P values were calculated by unpaired two-tailed Student’s t-test (a, b, h–j) and the multiple t-test with the two-stage linear step-up procedure of Benjamini, Krieger and Yakutieli (d, f). *P < 0.05; **P < 0.01; ***P < 0.001; ns (non-significant) with respect to Npc1^+/+ MEFs. Scale bars: 20 µm (a, i, j); 500 nm (b).

Fig. 3. Autophagy inducers rescue NAD levels and cell death in Npc1^-/- cells.

a Schematic illustration of chemical screen. b, c Volcano-plot representation of chemical screen results. Highlighted in red are hit compounds that reduce luminescence significantly (log2(FC) < 0 and –log₁₀(adjusted P) > 1.3), and listed in (c). d, e Immunoblot analysis of autophagy flux in Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium supplemented with 10 µM celecoxib (Cele) (d) or 30 µM memantine (Mem) (e) in the presence or absence of 100 nM bafilomycin A1 (Baf_A1) for 24 h. f Schematic illustration of Halo processing assay. g, Halo processing assay in Npc1^-/- MEFs expressing Halo-GFP-LC3B cultured in galactose medium supplemented with Cele or Mem for 8 h. h, Fluorescence microscopy images and quantification of mitophagy of Npc1^+/+ and Npc1^-/- MEFs expressing mt-mKeima 24 h after culture in galactose medium supplemented with Cele or Mem. Measurement of NAD⁺ and NADH levels (i), ΔΨm (j), Phase-contrast images and immunoblot analysis for caspase-3 cleavage (k), and cytotoxicity assay (l) in Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium supplemented with Cele or Mem for 60 h (i, j), 72 h (k) or 96 h (l). Data are mean ± SEM of n = 3 biological replicates. P values were calculated by multiple t-test with FDR method of Benjamini and Hochberg (b, c) or by one-way ANOVA followed by multiple comparisons with the two-stage linear step-up procedure of Benjamini, Krieger, and Yakutieli (d, e, g-l). *P < 0.05; **P < 0.01; ***P < 0.001 with respect to untreated Npc1^-/- MEFs. Scale bars: 20 µm (h, j); 200 µm (k).

Fig. 4. NAD precursor supplementation acts downstream of autophagy dysfunction, restores NAD levels and suppresses apoptotic cell death in Npc1^-/- cells.

a Chemical structures of NAD precursors, nicotinamide (NAM), nicotinamide riboside (NR), dihydronicotinamide riboside (NRH), and nicotinic acid riboside (NAR) used in the study, and their conversion pathway into NAD(H) are shown. b Immunoblot analysis of autophagy flux in Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium supplemented with 5 mM nicotinamide (NAM), 2 mM nicotinamide riboside (NR), 300 µM reduced nicotinamide riboside (NRH) and 50 µM nicotinic acid riboside (NAR) in the presence or absence of 100 nM bafilomycin A1 (Baf_A1) for 24 h. c Halo processing assay in Npc1^-/- MEFs expressing Halo-GFP-LC3B cultured in galactose medium supplemented with the indicated NAD precursors for 8 h. d Fluorescence microscopy images and quantification of mitophagy of Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium supplemented with the indicated NAD precursors for 24 h. Measurement of NAD⁺ and NADH levels (e), ΔΨm (f), phase-contrast images and immunoblot analysis for caspase-3 cleavage (g), and cytotoxicity assay (h) in Npc1^+/+ and Npc1^-/- MEFs cultured in galactose medium supplemented with the indicated NAD precursors for 60 h (e, f), 72 h (g) or 96 h (h). Data are mean ± SEM of n = 3 (c–h) or 4 b biological replicates. P values were calculated by one-way ANOVA followed by multiple comparisons with the two-stage linear step-up procedure of Benjamini, Krieger, and Yakutieli. *P < 0.05; **P < 0.01; ***P < 0.001; ns (non-significant) with respect to untreated Npc1^-/- MEFs. Scale bars: 20 µm (d, f); 200 µm (g).

Fig. 5. Restoring autophagy-NAD axis promotes cell survival of NPC1 patient-derived primary fibroblasts.

a Overview of control and NPC1 patient-derived primary fibroblasts. b Immunoblot analyses of autophagy flux in control and NPC1 patient-derived primary fibroblasts cultured in galactose medium in the presence or absence of 100 nM Baf_A1 for 24 h. c Representative confocal microscopy images and quantification of the number of puncta of immunofluorescence analysis with LC3B antibody in control and NPC1 patients-derived fibroblasts cultured in galactose medium for 7 days. d FACS analysis on cellular ROS levels by H₂DCFDA staining in control and NPC1 patient-derived fibroblasts cultured in galactose medium for 7 days. e Fluorescence microscopy images of live primary fibroblasts in the same conditions as d, co-stained with MitoSOX and mitotracker green (MTG). f–h Metabolomics analyses on control and NPC1 patients-derived fibroblasts cultured in galactose medium for 7 days. f Principal component analysis (PCA) of metabolomics datasets. g Volcano plot representation of metabolites in a pairwise comparison of NPC1 to control fibroblasts. Thresholds (|Log₂(fold change)| > 1 and -log₁₀(adjusted P) > 1) are shown as dashed lines. h Heatmap and clustering representations of analysed metabolites. Metabolites that change significantly (e) are highlighted in red. Fluorescence microscopy images (i) and quantification of mitophagy (j) in control and patient-derived primary fibroblasts cultured in galactose medium supplemented with 10 µM Cele or 100 µM NRH for 24 h. k Measurement of NAD⁺ and NADH levels in control and patient-derived primary fibroblasts cultured in galactose medium supplemented with Cele or NRH for 7 days. l, m Cytotoxicity assay (k) and phase-contrast images (l) in control and patient-derived primary fibroblasts cultured in galactose medium supplemented with Cele or NRH for 7 days and challenged with 500 µM H₂O₂ in serum free medium for 2 h. Data are mean ± SEM of n = 3 cell lines per group (b–e, j, k, m). P values were calculated by unpaired two-tailed Student’s t-test (b–e), multiple t-test with the original FDR method of Benjamini and Hochberg (g), or by one-way ANOVA followed by multiple comparisons with the two-stage linear step-up procedure of Benjamini, Krieger and Yakutieli (j, k, m). *P < 0.05; **P < 0.01; ***P < 0.001 with respect to control (b–e) or untreated NPC1 primary fibroblasts (j, k, m). Scale bars: 20 µm (c, e, i); 200 µm (l).

Fig. 6. Restoring autophagy-NAD axis protects against cell death of NPC1 patient iPSC-derived cortical neurons.

a Overview of control and NPC1 patient-derived iPSCs. Immunofluorescence images (b) and quantification of LC3B and p62 puncta in TUJ1⁺ or MAP2⁺ cells (c) in control (Ctrl_#13) and NPC1 patient (NPC1-2_#26) iPSC-derived cortical neurons after 4 weeks of neuronal differentiation, where NPC1 neurons were treated with 10 µM Cele or 100 µM NRH for the last 6 days. d TMRE fluorescence intensity (percentage of pre- and post-FCCP treatment) for ΔΨm in Ctrl_#13 and NPC1-2_#26 iPSC-derived cortical neurons in the same conditions as (b, c). Fluorescence images (e) and quantification of TUNEL⁺ apoptotic nuclei in TUJ1⁺ cells (f), and cytotoxicity assay results (g) in Ctrl_#13 and NPC1-2_#26 iPSC-derived cortical neurons in the same conditions as (a, b). Data are mean ± SEM of n = 5 (c), 6 (d, treated conditions), 8 (d, untreated conditions), 9 (f) or 12 (g) biological replicates. P values were calculated by one-way ANOVA followed by multiple comparisons with the two-stage linear step-up procedure of Benjamini, Krieger, and Yakutieli. *P < 0.05; **P < 0.01; ***P < 0.001 with respect to untreated NPC1 mutant cells. Scale bars: 10 µm (b); 100 µm (e).

Fig. 7. Schematic representation of the present study.

Loss of function in NPC1 results in autophagy/mitophagy dysfunction leading to NAD depletion and subsequent mitochondrial depolarisation and apoptotic cell death. Autophagy inducers and NAD precursors (green) can ameliorate the phenotypes in mouse and human NPC1 model cells.