Spastin MIT Domain Disease-Associated Mutations Disrupt Lysosomal Function

Abstract

The hereditary spastic paraplegias (HSPs) are genetic motor neuron diseases characterized by progressive degeneration of corticospinal tract axons. Mutations in SPAST, encoding the microtubule-severing ATPase spastin, are the most common causes of HSP. The broad SPAST mutational spectrum indicates a haploinsufficiency pathogenic mechanism in most cases. Most missense mutations cluster in the ATPase domain, where they disrupt the protein's ability to sever microtubules. However, several putative missense mutations in the protein's microtubule interacting and trafficking (MIT) domain have also been described, but the pathogenicity of these mutations has not been verified with functional studies. Spastin promotes endosomal tubule fission, and defects in this lead to lysosomal enzyme mistrafficking and downstream lysosomal abnormalities. We investigated the function of three disease-associated spastin MIT mutants and found that none was able to promote normal endosomal tubule fission, lysosomal enzyme receptor trafficking, or lysosomal morphology. One of the mutations affected recruitment of spastin to endosomes, a property that requires the canonical function of the MIT domain in binding endosomal sorting complex required for transport (ESCRT)-III proteins. However, the other mutants did not affect spastin's endosomal recruitment, raising the possibility of pathologically important non-canonical roles for the MIT domain. In conclusion, we demonstrate that spastin MIT mutants cause functional abnormalities related to the pathogenesis of HSP. These mutations do not directly affect spastin's microtubule-severing capacity, and so we identify a new molecular pathological mechanism by which spastin mutations may cause disease.

Keywords: CHMP1B; ESCRT; IST1; endosome tubule fission; hereditary spastic paraplegia.

Figure 1

Disease-associated spastin MIT mutants cannot support normal endosomal tubule fission. (A) Schematic diagram of spastin's domain structure, annotated with the position of artificial (above) and disease-associated (below) mutants used in this study. HR, hydrophobic region; MIT, MIT domain; MTB, microtubule binding domain; AAA, ATPase domain. (B–F) Wild-type HeLa cells or cells expressing the siRNA-resistant spastin constructs indicated were subjected to mock transfection or siRNA knockdown (KD) of endogenous spastin, then fixed and labeled for endogenous SNX1. Insets show higher magnification views of the boxed areas. Scale bar = 10 μm. (G) The percentage of cells with the longest SNX1 tubule >2 μm was counted (n = 100 cells per condition). Mean and SEM are shown for three biological repeats. P-values were calculated using one-way ANOVA with Bonferroni's post hoc multiple comparison test. In addition to the p-values shown, the M87-wild-type KD value was significantly different from all other KD values, with a significance of at least p < 0.05. (H) Depletion of spastin and expression of siRNA resistant constructs was confirmed by Western blotting with the antibodies indicated. GAPDH labeling is shown to verify equal sample loading.

Figure 2

Endosome to Golgi M6PR traffic is disrupted by MIT domain mutations in spastin. (A–J) Wild-type HeLa cells or cells expressing the siRNA-resistant spastin constructs indicated were subjected to mock transfection or siRNA-mediated KD of endogenous spastin, then processed for immunofluorescence microscopy and labeled with antibodies to M6PR and the lysosomal marker LAMP1. Insets show higher magnification views of the boxed areas. Scale bar = 10 μm. (K) Co-localization between M6PR and LAMP1 was quantified by calculating the Pearson's correlation coefficient for red and green pixels in each cell using Volocity software (20 cells per condition quantified in each experiment). Mean and SEM are shown for n = 4 experiments. P-values were calculated using paired two-tailed t-tests.

Figure 3

Spastin MIT domain mutants cannot support normal lysosomal morphology. (A–E) Wild-type HeLa cells or cells expressing the siRNA-resistant spastin constructs indicated were subjected to mock transfection or siRNA-mediated KD of endogenous spastin, then processed for immunofluorescence microscopy and labeled for LAMP1. Insets show higher magnification views of the boxed areas. Scale bar = 10 μm. (F) The diameter of the largest lysosome was measured in 100 cells per condition. The mean and SEM of the percentage of cells with largest lysosome >1.8 μm is shown (n = 4 biological repeats). P-values were calculated by one-way ANOVA with Bonferroni's post hoc multiple comparison test. In addition to the p-values shown, the M87 wild-type KD value was significantly different from all other KD values with a significance of at least p < 0.01. (G) Wild-type HeLa cells or HeLa cells expressing the siRNA-resistant spastin constructs indicated were subjected to mock transfection or siRNA-mediated KD of endogenous spastin. Electron micrographs of representative endolysosomal structures are shown. Note the presence of highly abnormal structures in each of the cells expressing mutant spastin. Scale bar = 500 nm.

Figure 4

MIT mutant forms of spastin show differential recruitment to VPS4-E235Q endosomes. GFP-VPS4-E235Q was transiently transfected into cell lines stably expressing myc-tagged wild-type spastin (A) or spastin mutants (B–F), and then cells were fixed and labeled with an anti-myc antibody. Insets show higher magnification views of the boxed areas. (G) Co-localization between GFP-VPS4-E235Q and spastin proteins was estimated by calculating the Pearson's correlation coefficient for red and green pixels in each cell, using Volocity software (n = 4 experiments, 15 cells per condition quantified in each experiment). Mean and SEM are shown, p-values vs. M87 wild-type spastin were calculated using one-way ANOVA with Dunnett's multiple comparison posttest. Scale bar = 10 μm.

Figure 5

Yeast two-hybrid analysis of spastin MIT domain binding to CHMP1B and IST1. (A–D) Interactions between the combinations of MIT domain, CHMP1B, and IST1 bait and prey vectors indicated in each panel. In each case, the first column shows the growth of diploid colonies on media lacking leucine and tryptophan, to select for the presence of both bait and prey vectors, while columns 2–6 show the growth of serial dilutions of diploid colonies on media lacking leucine, tryptophan, and histidine, which also require the activation of the His reporter gene that occurs on interaction of bait and prey proteins. In (B,D), empty vector controls test for auto-activation of by the MIT mutant proteins under investigation. Note that the N184T mutant causes auto-activation, and so binding results for this mutant were uninterpretable.

Figure 6

Assessment of interactions between spastin MIT domains and MIM-regions of CHMP1B and IST1 by GST pull-down. Purified proteins consisting of GST alone, GST fused to the wild-type spastin MIT domain, or to spastin MIT domains harboring the sequence changes indicated were used in in vitro GST pull-down experiments vs. myc- and 6His-tagged fragments of CHMP1B (A) or IST1 (B) that incorporate the spastin-binding MIM regions. In each set of experiments, the upper panels show Coomassie-stained gels, while the bottom panel shows corresponding anti-myc immunoblotting.