Mechanism-based rescue of Munc18-1 dysfunction in varied encephalopathies by chemical chaperones

Abstract

Heterozygous de novo mutations in the neuronal protein Munc18-1 are linked to epilepsies, intellectual disability, movement disorders, and neurodegeneration. These devastating diseases have a poor prognosis and no known cure, due to lack of understanding of the underlying disease mechanism. To determine how mutations in Munc18-1 cause disease, we use newly generated S. cerevisiae strains, C. elegans models, and conditional Munc18-1 knockout mouse neurons expressing wild-type or mutant Munc18-1, as well as in vitro studies. We find that at least five disease-linked missense mutations of Munc18-1 result in destabilization and aggregation of the mutant protein. Aggregates of mutant Munc18-1 incorporate wild-type Munc18-1, depleting functional Munc18-1 levels beyond hemizygous levels. We demonstrate that the three chemical chaperones 4-phenylbutyrate, sorbitol, and trehalose reverse the deficits caused by mutations in Munc18-1 in vitro and in vivo in multiple models, offering a novel strategy for the treatment of varied encephalopathies.

Fig. 1

Neuronal impairments in C. elegans expressing mutant UNC-18. a, b Locomotion of C. elegans. Body bends of indicated worm strains per 30 s were counted. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t test; n = 5 independent experiments on ten worms per experiment). c Scheme of the aldicarb assay. d Mutants display reduced acetylcholine release at the worm neuromuscular junction. Paralysis of young adult worms expressing WT or mutant unc-18 was measured 60 min after exposure to aldicarb (see Supplementary Fig. 2d for entire curves). Data are means ± SEM (**p < 0.01, ***p < 0.001 by Student’s t test; n = 6 independent experiments on 20 worms per experiment). e Heat-induced paralysis. Indicated worm strains were exposed to 37 °C over a period of 300 min, and paralysis was scored at indicated time points. Data are means ± SEM (***p < 0.001 by two-way ANOVA, compared to N2; ^###p < 0.001 by two-way ANOVA, compared to unc-18-WT in unc-18^−/−, n = 9–13 independent experiments on ten worms per experiment). f Worm traces after heat-induced paralysis. Plates were imaged after heat shock analysis in e. g–i Locomotion, acetylcholine release, and heat shock paralysis of CRISPR-edited C. elegans. Worms were assayed as in a–e. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t test in g and h and two-way ANOVA in i; n = 15–20 worms for g, n = 5–6 independent experiments on 20–25 worms per experiment for h, and n = 6 independent experiments on ten worms per experiment for i)

Fig. 2

Synaptic impairments in neurons expressing mutant Munc18-1. a Primary neurons were plated on a multi-electrode array. b–g Munc18-1 knockout neurons (c) or knockout neurons expressing Munc18-1b variants (d) were subjected to analysis of mean firing rate, burst frequency, burst duration, and network burst activity (b, e–g). Purple boxes in c and d indicate network activity. Data are means ± SEM (*^,#p < 0.05, **^,##p < 0.01, ***^,###p < 0.001 by Student’s t test; n = 4–6 independent experiments). h, i Uptake of synaptotagmin-1 antibody during high K⁺ stimulation. Neurons expressing cre recombinase and/or WT or mutant Munc18-1b were subjected to an antibody uptake assay. Endocytosed synaptotagmin-1 antibody was quantified by immunostaining (h), via counting the number of pixels > intensity of 15 (i). Data are means ± SEM (*^,#p < 0.05, **^,##p < 0.01 by Student’s t test; n = 4 independent experiments). Scale bar in h = 20 µm

Fig. 3

Increased turnover of Munc18-1 mutants in neurons. a Total protein levels of Munc18-1. WT and mutant Munc18-1b were expressed in primary neurons infected with lentiviral vectors expressing cre recombinase. Total protein levels were quantified by immunoblotting, normalized to the levels of the synaptic protein synaptophysin-1 (SypI). Data are means ± SEM (*p < 0.05, ***p < 0.001 by Student’s t test; n = 3 independent experiments). b Turnover of Munc18-1 by cycloheximide chase. Neurons as in a were subjected to a cycloheximide (CHX) chase experiment for the indicated time to stop protein translation. Remaining protein levels were quantified by immunoblotting, normalized to α-tubulin levels. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by two-way ANOVA; n = 3 independent experiments). c–f Turnover of Munc18-1 by Dendra2 photoconversion. HEK293T cells were transfected with WT or mutant Munc18-1b:Dendra2 fusion constructs. Two days after transfection, expressed Dendra2 was photoconverted. The green signal was quantified before and after photoconversion (c, d), and the red signal was quantified at 0, 3, and 24 h after photoconversion (e, f). Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t test in d and f, and by two-way ANOVA in e; n = 4–6 independent experiments). Scale bar in c = 50 µm

Fig. 4

Aggregation of Munc18-1 mutants. a Solubility of Munc18-1. Munc18-1 knockout neurons expressing WT or mutant Munc18-1b were solubilized in 0.1% Triton X-100 (TX). Equal volumes of soluble and insoluble fractions were analyzed by quantitative immunoblotting (α-tubulin = control). Data are means ± SEM (**p < 0.01, ***p < 0.001 by Student’s t test; n = 4 independent experiments). b Limited proteolysis. Neurons as in a were incubated with increasing concentrations of trypsin. Remaining protein levels were analyzed by quantitative immunoblotting (α-tubulin = control). Data are means ± SEM (***p < 0.001 by two-way ANOVA; n = 3 independent experiments). c Aggregation of mutant Munc18-1. Neurons as in a were analyzed for the subcellular localization of Munc18-1b by immunocytochemistry. Arrows depict aggregates. Scale bar = 20 µm. d Locomotion of C. elegans expressing GFP-tagged WT or G544D unc-18. Body bends per 30 s were counted. Data are means ± SEM (***p < 0.001 by Student’s t test; n = 5 independent experiments on ten worms per experiment). e Paralysis of WT worms (N2) or worms expressing GFP-tagged WT or G544D mutant unc-18 after 60 min exposure to aldicarb. Data are means ± SEM (***p < 0.001 by Student’s t test; n = 6 independent experiments on ten worms per experiment). f Lack of axonal localization of mutant UNC-18. C. elegans expressing WT::GFP or G544D::GFP were immobilized, and the ventral nerve cord was imaged (solid arrowheads = pairs of bigger puncta, broken arrowheads = single, smaller puncta. Scale bar = 10 µm). g Worm traces after the heat shock assay (Supplementary Fig. 8e). h, i Aggregation of mutant Munc18-1 in yeast. S. cerevisiae expressing GFP-tagged Munc18-1 variants were imaged 24 h after induction of protein expression (h) to quantify aggregation (i). Data are means ± SEM (n = 3 independent experiments). Scale bar = 5 µm. j Expression of mutant Munc18-1 in yeast. Munc18-1 levels were analyzed by measuring GFP fluorescence in a plate reader 24 h post induction. Data are means ± SEM (**p < 0.01, ***p < 0.001 by Student’s t test; n = 4 independent experiments)

Fig. 5

Dominant-negative activity of mutant Munc18-1 on wild-type Munc18-1. a Protein levels of Munc18-1 in primary neurons. Levels of GFP-tagged WT or mutant Munc18-1 and endogenous Munc18-1 in heterozygous Munc18-1 neurons were analyzed by quantitative immunoblotting, normalized to α-tubulin levels. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t test; n = 3 independent experiments). b Same as in a, except that myc-tagged Munc18-1 variants were co-expressed with HA-tagged WT Munc18-1 in Munc18-1 knockout neurons. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t test; n = 5–6 independent experiments). c Co-immunoprecipitation of mutant with wild-type Munc18-1. Lysates of HEK293T cells that were co-transfected with HA-tagged WT and myc-tagged mutant Munc18-1 were subjected to immunoprecipitation with an anti-c-myc antibody (IP) or no antibody (control). Precipitated myc- and HA-tagged Munc18-1 was analyzed with the respective input by immunoblotting. GAPDH served as control. d Solubility of Munc18-1. Primary neurons infected as in b were solubilized in 0.1% Triton X-100 (TX). Equal volumes of TX-soluble and -insoluble fractions were analyzed by quantitative immunoblotting. α-Tubulin served as control. Data are means ± SEM (*p < 0.05, ***p < 0.001 by Student’s t test; n = 5–7 independent experiments). e Locomotion of C. elegans. Body bends per 30 s were counted. Data are means ± SEM (*p < 0.05, **p < 0.01 by Student’s t test; n = 15 independent experiments on ten worms per experiment). f Heat-induced paralysis. Paralysis at 37 °C was scored at indicated time points. Data are means ± SEM (***p < 0.001 by two-way ANOVA, compared to N2; n = 3 independent experiments on ten worms per experiment). g Mutants display reduced acetylcholine release at the worm neuromuscular junction. Paralysis of N2 worms expressing WT or mutant unc-18 was measured after 60 min exposure to aldicarb. Data are means ± SEM (**p < 0.01, ***p < 0.001 by Student’s t test; n = 6 independent experiments on 20 worms per experiment)

Fig. 6

Chemical chaperones rescue mutant Munc18-1 deficits in neurons. a Total protein levels of Munc18-1. WT and mutant Munc18-1b were expressed in primary cortical neurons infected with lentiviral vectors expressing cre recombinase and Munc18-1b variants in the presence or absence of chemical chaperones. Total protein levels were quantified 7 days after infection by immunoblotting, normalized to the levels of the post-synaptic protein PSD-93. Data are means ± SEM (*p < 0.05, **p < 0.01 by Student’s t test; n = 5 independent experiments). b Triton X-100 solubility of Munc18-1. WT or mutant Munc18-1b were expressed as above. Seven days after infection, cells were solubilized in 0.1% Triton X-100 (TX). Equal volumes of soluble and insoluble fractions were separated by SDS-PAGE, and TX-soluble Munc18-1 was measured as percent of total Munc18-1 by quantitative immunoblotting. Solubility of PSD-93 served as control (Supplementary Fig. 13b). Data are means ± SEM (*p < 0.05, **p < 0.01 by Student’s t test; n = 5 independent experiments). c, d Uptake of synaptotagmin-1 antibody during high K⁺ stimulation. Primary cortical neurons infected at 6 DIV with lentivirus expressing cre recombinase and WT or mutant Munc18-1b were subjected to an antibody uptake assay at 13 DIV in the absence or presence of chemical chaperones. Endocytosed synaptotagmin-1 antibody was quantified by immunostaining (c), via counting the number of pixels > intensity of 15 (d). Data are means ± SEM (*^,#p < 0.05, **^,##p < 0.01, ***^,###p < 0.001 by Student’s t test; n = 4 independent experiments). Scale bar in c = 20 µm

Fig. 7

Chemical chaperones rescue deficits of mutant UNC-18 in C. elegans. a Locomotion of C. elegans. Body bends per 30 s were counted. Data are means ± SEM (**p < 0.01, ***p < 0.001 by Student’s t test; n = 10 independent experiments on ten worms per experiment). b Rescue of reduced acetylcholine release in worms expressing mutant UNC-18 variants. Young adult worms expressing WT or mutant unc-18 were exposed to aldicarb, and paralysis at 60 min was measured. Data are means ± SEM (^#p < 0.05, **^,#p < 0.01, ***p < 0.001, by Student’s t test; n = 6 independent experiments on 20 worms per experiment). c Heat-induced paralysis. Indicated worm strains were exposed to 37 °C over a period of 180 min, and paralysis was scored at indicated time points. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by two-way ANOVA, compared to control, n = 5–6 independent experiments on ten worms per experiment). d, e Rescue of reduced acetylcholine release and heat-induced paralysis in CRISPR/Cas9-generated P334L and R405H knock-in worms. Worms were analyzed as described in a–c. Data are means ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 by two-way ANOVA, compared to control, n = 4 independent experiments on ten worms per experiment for d, and n = 4 independent experiments on 20 worms per experiments for e). f Rescue of subcellular localization of UNC-18 in worms expressing mutant UNC-18. C. elegans expressing WT::GFP or G544D::GFP were immobilized, and the ventral nerve cord was imaged. Solid arrowheads point to pairs of bigger puncta, broken arrowheads to single, smaller puncta. Scale bar = 10 µm. g Rescue of reduced acetylcholine release in worms expressing GFP-tagged mutant UNC-18 variants. Experiments were performed as described under (b). Data are means ± SEM (*p < 0.05, ^##p < 0.01, ^###p < 0.001, by Student’s t test; n = 6 independent experiments on 20 worms per experiment)

Fig. 8

Model of Munc18-1 dysfunction in encephalopathies and rescue of deficits with chemical chaperones. Munc18-1-linked encephalopathies are caused by a dominant-negative disease mechanism. Heterozygous missense mutations in Munc18-1 cause a reduction in functional Munc18-1 levels significantly below 50%, due to accelerated degradation of misfolded mutant Munc18-1, aggregation of misfolded mutant Munc18-1 that is resistant to cellular clearance, and due to co-aggregation of WT Munc18-1. Chemical chaperones not only shift the unfolded–folded protein equilibrium significantly toward a folded state, but also result in an increase in total Munc18-1 levels. This overall increase in Munc18-1 levels and solubility is sufficient to rescue the Munc18-1-linked neuronal deficits in vitro and in vivo