Inhibition of sphingolipid synthesis improves outcomes and survival in GARP mutant wobbler mice, a model of motor neuron degeneration

Abstract

Numerous mutations that impair retrograde membrane trafficking between endosomes and the Golgi apparatus lead to neurodegenerative diseases. For example, mutations in the endosomal retromer complex are implicated in Alzheimer's and Parkinson's diseases, and mutations of the Golgi-associated retrograde protein (GARP) complex cause progressive cerebello-cerebral atrophy type 2 (PCCA2). However, how these mutations cause neurodegeneration is unknown. GARP mutations in yeast, including one causing PCCA2, result in sphingolipid abnormalities and impaired cell growth that are corrected by treatment with myriocin, a sphingolipid synthesis inhibitor, suggesting that alterations in sphingolipid metabolism contribute to cell dysfunction and death. Here we tested this hypothesis in wobbler mice, a murine model with a homozygous partial loss-of-function mutation in Vps54 (GARP protein) that causes motor neuron disease. Cytotoxic sphingoid long-chain bases accumulated in embryonic fibroblasts and spinal cords from wobbler mice. Remarkably, chronic treatment of wobbler mice with myriocin markedly improved their wellness scores, grip strength, neuropathology, and survival. Proteomic analyses of wobbler fibroblasts revealed extensive missorting of lysosomal proteins, including sphingolipid catabolism enzymes, to the Golgi compartment, which may contribute to the sphingolipid abnormalities. Our findings establish that altered sphingolipid metabolism due to GARP mutations contributes to neurodegeneration and suggest that inhibiting sphingolipid synthesis might provide a useful strategy for treating these disorders.

Keywords: amyotrophic lateral sclerosis; myriocin; neurodegeneration; sphingolipid; wobbler mice.

Fig. 1.

The wobbler mutation of the GARP complex leads to missorting of proteins. (A) GOLGA2 staining in MEFs reveals no significant (ns) differences in the size of the Golgi compartment, whereas LAMP1 staining in wild-type and wobbler MEFs, followed by quantification of LAMP1 per cell (n ∼ 250 cells per condition), shows reduced numbers and larger size of lysosomes. Mean, SD ****P < 0.0001 by Mann–Whitney test. (Scale bar, 10 μm.) Analysis also revealed altered distribution of lysosomes depicted as distance from LAMP1 to GOLGA2 and between LAMP1 particles per cell area (n ∼ 20 cells per genotype) Mean, 95% CI ****P < 0.0001 by Mann–Whitney test. (B) TMT-based quantitative proteomics shows the enrichment of lysosomal proteins (orange) (P value = 2.98 × 10⁻³⁵) and (C) enrichment of sphingolipid metabolism enzymes (P value = 6.85 × 10⁻¹¹) in the Golgi-enriched fractions from wobbler cells, by Wilcoxon rank sum test. Lysosomal proteins (orange and green) are enriched in wobbler MEFs while Golgi proteins (purple) were relatively unchanged. Proteins labeled for imaging by immunofluorescence (A) or for Western blots (D and SI Appendix, Fig. S1F) are labeled (purple and green). Sphingolipid biosynthetic (blue) and catabolic (red) enzymes in the Golgi-enriched fraction shows many catabolic enzymes normally residing at lysosome (asterisk). (D) Western blot and Ponceau analysis of lysosomal proteins in Golgi-enriched fractions from wild-type and wobbler cells verifies enrichment of lysosomal proteins and reveals defects in CTSD processing. Actin is shown as a loading control in these fractions with enriched, but not purified, Golgi membranes.

Fig. 2.

The wobbler mutation of the GARP complex leads to accumulation of sphingolipid intermediate species in wobbler MEFs and spinal cords. (A) Lipidomics analysis of different classes of lipids shows a strong accumulation of sphingolipid species in wobbler MEF, that is rescued by myriocin treatment (n = 4 biological replicates). ***P < 0.001 and ****P < 0.0001 by one-way ANOVA with Bonferroni post hoc test. ChE: cholesterol esters, PE: phosphatidylethanolamine, PC: phosphatidylcholine. (B) Schematic representations of sphingolipid metabolism pathway, illustrating the effect of the GARP mutation, enzymes up-regulated and/or mislocalized from lysosome to Golgi and the changes in lipids upon myriocin treatment. Sphingolipid enzymes in the biosynthesis and degradation pathway but not identified by mass spectrometry in those experiments are shown. (C) Lipidomics analysis of different classes of lipids shows a strong accumulation in sphingolipid species in the spinal cord of wobbler mice (n = 4 to 6 biological replicates). *P < 0.05 and ****P < 0.0001 by unpaired t test. So: sphingosine, Sa: sphinganine, SM: sphingomyelin, CL: cardiolipin. No changes in phospholipids, PS and PE, ceramide (Cer) triglyceride (TG) or diacylglycerol (DG).

Fig. 3.

Inhibiting sphingolipid synthesis improves locomotor activity and grip strength in wobbler mouse. (A) Locomotion activity was measured in metabolic cages equipped with laser beams. Total activity over 6 h shown as area under the curve (AUC). ****P < 0.0001 by one-way ANOVA and Tukey’s multiple comparison (n = 4 to 8 per group) (B, C, and D) Boxplots representing the weekly monitoring of weight, wellness score, and grip strength for the untreated and myriocin-treated wobbler mice (n = 14 to 15 per genotype). Blue regions highlight the interquartile range for untreated wobbler controls at week 10. Scores for wild type marked by gray line (C and D). (E) After 10 wk of myriocin treatment, wobbler mice exhibit improved wellness index and grip strength. **P = 0.002 and P = 0.006 by Mann–Whitney test, for wellness index and grip strength, respectively. (n = 8 and 12 for untreated and myriocin-treated wobbler, respectively.)

Fig. 4.

Inhibiting sphingolipid synthesis improves neuropathology outcomes and increases the survival rate of wobbler mouse. (A) Hematoxylin and eosin staining of the ventral horn in the gray matter from lower cervical spinal cord shows the presence of different stages of dying motor neurons in wobbler mice, compared to wild-type mice, as illustrated in the Insets. (B) Reduced astrocyte gliosis in wobbler mice as revealed by GFAP staining in the ventral horn of the lower cervical spinal cord (black arrowheads show astrogliosis). (Scale bar, 50 μm.) (C) Quantification of the average number of astrocytes (GFAP-positive cells) in the ventral horn of the lower cervical spinal cord (one-way ANOVA P = 0.003, Mann–Whitney between untreated and treated wobbler **P = 0.0047). (D) Myriocin treatment improved the survival rate of wobbler mice during a 1-y treatment (log-rank test: P = 0.003) (n = 17 to 18 mice per group).