Fbxo2 mediates clearance of damaged lysosomes and modifies neurodegeneration in the Niemann-Pick C brain

Abstract

A critical response to lysosomal membrane permeabilization (LMP) is the clearance of damaged lysosomes through a selective form of macroautophagy known as lysophagy. Although regulators of this process are emerging, whether organ- and cell-specific components contribute to the control of lysophagy remains incompletely understood. Here, we examined LMP and lysophagy in Niemann-Pick type C (NPC) disease, an autosomal recessive disorder characterized by the accumulation of unesterified cholesterol within late endosomes and lysosomes, leading to neurodegeneration and early death. We demonstrated that NPC human fibroblasts show enhanced sensitivity to lysosomal damage as a consequence of lipid storage. Moreover, we described a role for the glycan-binding F-box protein 2 (Fbxo2) in CNS lysophagy. Fbxo2 functions as a component of the S phase kinase-associated protein 1-cullin 1-F-box protein (SKP1-CUL1-SCF) ubiquitin ligase complex. Loss of Fbxo2 in mouse primary cortical cultures delayed clearance of damaged lysosomes and decreased viability after lysosomal damage. Moreover, Fbxo2 deficiency in a mouse model of NPC exacerbated deficits in motor function, enhanced neurodegeneration, and reduced survival. Collectively, our data identified a role for Fbxo2 in CNS lysophagy and establish its functional importance in NPC.

Keywords: Autophagy; Lysosomes; Neurodegeneration; Neuroscience.

Figure 1. I1061T NPC1 human fibroblasts are more sensitive to lysosomal damage.

(A) Primary human fibroblasts homozygous for WT NPC1 (Ctrl) or I1061T NPC1 (I1061T) were treated with Veh or indicated doses of LLOMe for 1 hour and stained for Gal3 to detect damaged lysosomes. Gal3 puncta per cell quantified below. (B) Ctrl and I1061T human fibroblasts were treated with Veh or 2 mM LLOMe for 1 hour and stained for Gal3 and LAMP-1. Colocalization was performed on 4 fields each from 3 independent experiments, with 100–200 cells per experiment. Data are shown as mean ± SEM from 3 independent experiments. *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001 by (A) 1-way ANOVA with Tukey’s multiple comparisons (B) t test [(A) F = 24.04, (B) F = 12.78]. Scale bar: 25 μm. NPC, Niemann-Pick disease type C; LLOMe, L-leucyl-L-leucine methyl ester; Ctrl, control; Gal3, galectin-3; Veh, vehicle.

Figure 2. Gal3 and LC3 puncta induced by lysosomal damage are cleared in I1061T NPC1 human fibroblasts.

(A) Ctrl and I1061T human fibroblasts were treated with Veh or 2 mM LLOMe for 1 hour and stained for Gal3 at indicated times after washout. Quantified at the right. (B) Ctrl and I1061T human fibroblasts were treated with Veh or 0.5 mM LLOMe for 1 hour and stained for LC3 at indicated times. LC3 puncta per cell quantified at the right. Data are shown as mean ± SEM from (A) 3 or (B) 4 independent experiments. **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by 2-way ANOVA with Holm-Šidák’s test. Scale bar: 25 μm. Gal3, galectin-3; NPC, Niemann-Pick disease type C; Ctrl, control; Veh, vehicle; LLOMe, L-leucyl-L-leucine methyl ester.

Figure 3. Increased lysosomal damage is dependent upon lipid storage.

(A) Seven-week-old WT and I1061T mice were treated with vinblastine for 2 hours. Liver was collected and stained for Gal3 and LAMP-2. Mander’s coefficient in I1061T liver: 0.76. Scale bar: 5 μm. (B) Ctrl and I1061T human fibroblasts were treated with Veh or 1 mM cyclodextrin (Cyclo) for 48 hours and then treated with 2 mM LLOMe for 1 hour and stained for Gal3. Quantified at the right. Scale bar: 25 μm. (C) Ctrl, I1061T, and 2 independent lines of NPC2 human fibroblasts were treated with Veh or 2 mM LLOMe for 1 hour and stained for Gal3. Quantified at the right. Scale bar: 25 μm. Data are shown as mean ± SEM from (B and C) 4 independent experiments. *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001 by (B) t test (C) 1-way ANOVA with Tukey’s multiple comparisons [(B) F = 2.845, (C) F = 30.99]. Gal3, galectin-3; Ctrl, control; Veh, vehicle; NPC, Niemann-Pick disease type C; LLOMe, L-leucyl-L-leucine methyl ester.

Figure 4. Increased lysosomal damage is not due to impaired clearance.

(A) To examine lysophagy progression, cells were treated with 30 μg/mL CHX and 2 mM LLOMe for 1 hour and CHX treatment continued for indicated times before lysates were collected. (B) WT and Atg5^–/– MEFs were treated as in A. LAMP-1 levels were analyzed and quantified at the right. (C) Ctrl and I1061T human fibroblasts were treated as in A. Gal3 levels were analyzed and quantified at the right. Data are shown as mean ± SEM from (B) 4 or (C) 3 independent experiments. *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001 by 1-way ANOVA with Tukey’s multiple comparisons [(B) F = 2.862, (C) F = 25.49]. CHX, cycloheximide; LLOMe, L-leucyl-L-leucine methyl ester; Ctrl, control.

Figure 5. Fbxo2 is the most highly expressed glycan-binding F-box protein in the brain.

(A) Ctrl patient fibroblasts were pretreated with DMSO or 1 μM MLN7243 for 4 hours and then treated as indicated with 30 μg/mL CHX and 2 mM LLOMe. LAMP-1 levels quantified at the right. (B) Allen Brain Atlas expression data of Fbxo2, Fbxo6, and Fbxo27 in mouse brain. (C) Relative expression of Fbxo2, Fbxo6, and Fbxo27 was determined in the cerebellum, cortex, and brainstem of WT and I1061T mice at 12 weeks by qPCR. N = 4–5 mice per genotype. Data are shown as mean ± SEM from (A) 4 independent experiments. *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001 by (A and C) 1-way ANOVA with Tukey’s multiple comparisons [(A) F = 2.803, (C) F = 78.88 (CB), F = 95.66 (CX), F = 128.7 (BS)]. Fbxo2, F-box protein 2; Ctrl, control; CHX, cycloheximide; LLOMe, L-leucyl-L-leucine methyl ester; CB, cerebellum; CX, cortex; BS, brainstem.

Figure 6. Fbxo2 localizes to damaged lysosomes.

(A) Ctrl and I1061T human fibroblasts were transfected with HA-FBXO2, then treated with Veh or 1 mM LLOMe for 1 hour and stained for HA and Gal3. Colocalization is indicated by yellow staining in merged image. Mander’s coefficients: 0.80 (Ctrl) and 0.81 (I1061T). (B) I1061T human fibroblasts were electroporated with HA-FBXO2 and, after 48 hours, treated with 2 mM LLOMe for 2 hours. Lysates were immunoprecipitated with either HA antibody or Ctrl IgG. Arrowheads at approximately 50kD and approximately 25kD indicate immunoglobulin heavy and light chains, respectively. Asterisk denotes a nonspecific band detected by the LAMP2 antibody. (C) WT primary cortical cultures were transfected with HA-FBXO2 and treated with Veh or 2 mM LLOMe for 1 hour on D9IV. Cell were stained for HA and NeuN. Fbxo2, F-box protein 2; Ctrl, control; Veh, vehicle; LLOMe, L-leucyl-L-leucine methyl ester; Gal3, galectin-3.

Figure 7. Fbxo2 mediates CNS lysophagy.

(A) Relative expression of Fbxo2, Fbxo6, and Fbxo27 was determined in WT and Fbxo2^–/– primary cortical cultures at D9IV by qPCR. (B) WT and Fbxo2^–/– primary cortical cultures were transfected with EGFP-hGal3 on D4IV, treated with Veh or 2 mM LLOMe for 1 hour, and stained at various times after washout. Percentage of cells with Gal3+ puncta was quantified. (C) WT and Fbxo2^–/– primary cortical cultures were treated on D9IV with 30 μg/mL CHX and 2 mM LLOMe for 1 hour, and CHX treatment continued for indicated times before lysates were collected. Gal3 levels quantified at the right. (D) WT and Fbxo2^–/– primary cortical cultures were treated with Veh or LLOMe (0.5 mM or 2 mM) for 1 hour, and cell viability was determined by XTT assay 4 hours after LLOMe washout. (E) Fbxo2^–/– primary cortical cultures were pretreated with Veh (-) or 100 μM E64D for 30 minutes and then treated with 2 mM LLOMe for 1 hour, and cell viability was determined by XTT assay 4 hours after LLOMe washout. Data are shown as mean ± SEM from (A, B, and D) 3, (C) 5, or (E) 4 independent experiments. *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001 by (A and D–E) 1-way ANOVA with Tukey’s multiple comparisons or (B and C) 2-way ANOVA with Holm-Šidák’s test [(A) F = 81.08, (D) F = 23.2, (E) F = 14.55]. Fbxo2, F-box protein 2; Veh, vehicle; LLOMe, L-leucyl-L-leucine methyl ester; Gal3, galectin-3.

Figure 8. Loss of Fbxo2 exacerbates the NPC disease phenotype.

(A) Age-dependent performance on balance beam. Mice were trained at 5 weeks and run every other week starting at 6 weeks. The average of 3 trials was taken and max time was set at 20 seconds. N = 5 males and 5 females per genotype. (B) Performance on accelerating rotarod from 4–40 rpm at 9 weeks. N = 5 males and 5 females per genotype. (C) Kaplan-Meier survival curves. N = 6–10 males and 6–10 females per genotype. (D) Relative expression of Fbxo2, Fbxo6, and Fbxo27 was determined by qPCR in 8 weeks brainstem. N = 4 mice per genotype. (E) Quantification of Purkinje cell density in lobules IV and V of midline cerebellar sections. N = 3–4 mice per genotype. (F) The relative abundance of p62 in brainstem from 8 weeks mice was determined by Western blot. N = 5 mice per genotype. Data are shown as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by (A) 2-way ANOVA, (B, E, and F) 1-way ANOVA or (C) log-rank (Mantal-Cox) test and Gehan-Breslow-Wilcoxon test with (A and E) Bonferroni’s or (B and F) Tukey’s multiple comparisons [(A) F = 26.88, (B) F = 63.59, (E) F = 19.49, (F) F = 11.64]. Scale bar: 25 μm. Fbxo2, F-box protein 2; NPC, Niemann-Pick disease type C.