Species-specific differences in NPC1 protein trafficking govern therapeutic response in Niemann-Pick type C disease

Abstract

The folding and trafficking of transmembrane glycoproteins are essential for cellular homeostasis and are compromised in many diseases. In Niemann-Pick type C disease, a lysosomal disorder characterized by impaired intracellular cholesterol trafficking, the transmembrane glycoprotein NPC1 misfolds due to disease-causing missense mutations. While mutant NPC1 has emerged as a robust target for proteostasis modulators, drug development efforts have been unsuccessful in mouse models. Here, we demonstrated unexpected differences in trafficking through the medial Golgi between mouse and human I1061T-NPC1, a common disease-causing mutant. We established that these distinctions are governed by differences in the NPC1 protein sequence rather than by variations in the endoplasmic reticulum-folding environment. Moreover, we demonstrated direct effects of mutant protein trafficking on the response to small molecules that modulate the endoplasmic reticulum-folding environment by affecting Ca++ concentration. Finally, we developed a panel of isogenic human NPC1 iNeurons expressing WT, I1061T-, and R934L-NPC1 and demonstrated their utility in testing these candidate therapeutics. Our findings identify important rules governing mutant NPC1's response to proteostatic modulators and highlight the importance of species- and mutation-specific responses for therapy development.

Keywords: Lysosomes; Neuroscience; Protein misfolding; Protein traffic; Therapeutics.

Figure 1. I1061T-NPC1 undergoes species-specific trafficking.

(A) Human fibroblasts, mouse fibroblasts, and human Hap1 cells were treated with cycloheximide for the indicated times, and WT- or I1061T-NPC1 half-life was analyzed by Western blot and quantified. (B) Cell lysates from WT- or I1061T-NPC1 human fibroblasts, mouse fibroblasts, and human Hap1 cells were incubated with no treatment (NT) or digested with EndoH (E) or PNGase F (P) and then subjected to Western blot and quantified at right. (C) Lysates from human WT-NPC1 (Hu-WT), human I1061T/I1061T-NPC1 (Hu-I10), and mouse I1061T/I1061T-NPC1 (Ms-I10) fibroblasts were analyzed for total NPC1 by Western blot and quantified. (D and E) Lysates from a control cell line (CTRL; Coriell, GM08399) or a panel of I1061T/I1061T patient fibroblasts (Coriell, GM18453; NPC27, NPC4, NPC47, NPC24, NPC91, and NPC90) were analyzed for (D) total NPC1 protein or (E) incubated with NT or E and quantified. Data are shown as the mean ± SEM from (A) 6 to 8 or (C–E) 3 to 4 independent experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by (A) 2-way ANOVA with Bonferroni’s multiple comparison test, (B) t test, or (C–E) 1-way ANOVA with Tukey’s post hoc test. (A) F = 0.6, 9.0, 6.0; df = 3. (B) t = 13.1, 0.5, 11.3; df = 6, 6, 4. (C–E) F = 420, 46.3, 30.5; df = 2, 7, 7. See complete unedited blots in the supplemental material.

Figure 2. Mouse I1061T-NPC1 retains function.

(A and B) Lysates from primary mouse fibroblasts and human Hap1 cells with WT-NPC1, I1061T-NPC1, or Null-NPC1 were analyzed by Western blot for total NPC1 and quantified. (C) Cholesterol storage was measured by filipin (cyan) staining in WT-NPC1, I1061T-NPC1, and NPC1-null human Hap1 cells or mouse fibroblasts. WGA (orange) was used to label plasma membranes. Scale bar: 40 μm. Filipin intensity was normalized relative to WT and quantified. Data are shown as the mean ± SEM from 3 to 4 independent experiments. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by ANOVA with Tukey’s post hoc. (A and B) F = 282.0, 43.5, 282.0; df = 2. (C) F = 23.4; 666.0; df = 2.

Figure 3. NPC1 protein sequence drives differential trafficking.

(A and C) NPC1-deficient mouse fibroblasts were electroporated with (A) human WT, human I1061T, and mouse I1061T and (C) human R934L and mouse R934L, and total NPC1 was analyzed by Western blot and quantified. (B) Mouse fibroblasts and human Hap1 cells deficient in NPC1 were electroporated with GFP, human WT-NPC1, human I1061T-NPC1, or mouse I1061T-NPC1 plasmids. Lysates were incubated with no treatment (NT) or digested with EndoH (E) or PNGase F (P) and then subjected to Western blot and quantified. (D) NPC1-deficient mouse fibroblasts were electroporated with human WT-NPC1, human R934L-NPC1, or mouse R934L-NPC1 plasmids, and lysates were treated as in B and quantified. (E) Hu-I1061T, Mu-I1061T, or Mu-I1061T containing the human glycosylation sites (Ms-I1061T+HuGlycans) were overexpressed and digested with E or P and quantified. Data are shown as the mean ± SEM from indicated number of independent experiments. *P ≤ 0.05, ***P ≤ 0.001, ****P ≤ 0.0001 by (A, B, D, and E) ANOVA with Tukey’s post hoc test or (C) t test. (A and C) n = 3, F = 6.7, t = 3.8, df = 2,4. (B) Hap1, Mef; n = 4,4; F = 37.8, 104.5; df = 2. (D) n = 4; F = 95.5; df = 2, (E) n = 5, F = 248.2, df = 2. See complete unedited blots in the supplemental material.

Figure 4. Species- and mutation-specific response to DHBP.

(A) Primary human and mouse I1061T/I1061T or human R934L/R1077X fibroblasts were treated with 5 μM DHBP or vehicle for 5 days and analyzed by (A) Western blot for total NPC1 (quantified at right) or (B) incubated with no treatment (NT) or digested with EndoH (E) or PNGase F (P) to assess trafficking (quantified at right). Data are shown as the mean ± SEM from indicated number of independent experiments. ***P ≤ 0.001, ****P ≤ 0.0001 by t test. (A) Human I1061T, mouse I1061T, human R934L; n = 4, 4, 5; t = 7.1, 1.0, 3.8; df = 6, 6, 8; (B) Human I1061T, mouse I1061T, human R934L; n = 3, 3, 3, 3, 6, 6; t = 3.2, 5.7, 0.4, 1.0, 2.3, 1.1. df = 4, 4, 4, 4, 10, 10. See complete unedited blots in the supplemental material.

Figure 5. Mutation-specific response to DHBP.

I1061T/I1061T or NPC1-null primary human and mouse fibroblasts or R934L/R1077X human fibroblasts were treated with vehicle or 5 μM DHBP for 5 days. Cells were fixed and unesterified cholesterol was labeled with filipin (cyan) and quantified. Scale bar: 50 μm. Data are shown as the mean ± SEM from indicated number of independent experiments. ***P ≤ 0.001, by t test. Human I1061T, mouse I1061T, mouse null, R934L; n = 4–8; t = 6.6, 2.0, 1.9, 3.6; df = 6, 7, 12, 6.

Figure 6. Creation of human iPSC NPC1 mutants.

(A) Workflow describing the generation of human iPSC lines. (B and C) Lysates were collected from WT, I1061T, and R934L-NPC1 iPSCs and analyzed for (B) total NPC1 (quantified at right) or (C) incubated with no treatment (NT) or digested with EndoH (E) or PNGase F (P) (quantified at right). Data are shown as the mean ± SEM from 4 independent experiments. ***P ≤ 0.001, ****P ≤ 0.0001 by ANOVA with Tukey’s post hoc. (B) F = 111.6, df = 2. (C) F = 152.5; df = 2.

Figure 7. Mutation-specific response to dantrolene in iPSC-derived neurons.

(A) Schematic depicting iNeuron generation and confirmation. (B) Isogenic WT, I1061T, and R934L iNeurons were treated with vehicle (Veh) or dantrolene (10 μM) for 5 days. Lysates were incubated with no treatment (NT) or digested with EndoH (E) or PNGase F (P) (quantified at right). Data are shown as the mean ± SEM from 3 to 4 independent experiments. **P ≤ 0.01, ****P ≤ 0.0001 by t test or ANOVA with Tukey’s post hoc. (B) I1061T, t = 5.3, F = 75.1, df = 6,9; R934L, t = 1.1, 0.2, df = 4.

Figure 8. Dantrolene corrects lipid imbalances in I1061T-NPC1–induced neurons.

Isogenic WT, I1061T, or R934L iNeurons were treated with vehicle (Veh) or dantrolene (dan; 10 μM) for 5 days. (A and B) Cells were stained for TuJ1 (green), and unesterified cholesterol was labeled with filipin (cyan). Filipin intensity/TuJ1 quantified at right. (C) LC-MS analysis of relative unesterified cholesterol (area under the curve [AUC]). (D) Cells were stained for GM2 (green). Nuclei were stained with DAPI (blue). Percentage of cells with GM2 puncta are quantified. (E) LC-MS analysis of relative GM2 (34:1) (AUC). Scale bars: 30 μm. Data are shown as the mean ± SEM from (A) 8, (B and D) 4, or (C and E) 3 independent experiments. **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by ANOVA with Tukey’s post hoc or t test. (A) F = 17.9, DF = 2; (B) t = 0.4 DF = 6; (C) F = 0.3, DF = 2; (D) F = 46.47, DF = 2; (E) F = 0.5, DF = 2.