Neurons dispose of hyperactive kinesin into glial cells for clearance

Abstract

Microtubule-based kinesin motor proteins are crucial for intracellular transport, but their hyperactivation can be detrimental for cellular functions. This study investigated the impact of a constitutively active ciliary kinesin mutant, OSM-3CA, on sensory cilia in C. elegans. Surprisingly, we found that OSM-3CA was absent from cilia but underwent disposal through membrane abscission at the tips of aberrant neurites. Neighboring glial cells engulf and eliminate the released OSM-3CA, a process that depends on the engulfment receptor CED-1. Through genetic suppressor screens, we identified intragenic mutations in the OSM-3CA motor domain and mutations inhibiting the ciliary kinase DYF-5, both of which restored normal cilia in OSM-3CA-expressing animals. We showed that conformational changes in OSM-3CA prevent its entry into cilia, and OSM-3CA disposal requires its hyperactivity. Finally, we provide evidence that neurons also dispose of hyperactive kinesin-1 resulting from a clinic variant associated with amyotrophic lateral sclerosis, suggesting a widespread mechanism for regulating hyperactive kinesins.

Keywords: Autoinhibition; Conformation; Glia; Hyperactive; Kinesin.

Figure 1. A hyperactive OSM-3 kinesin disrupted cilia.

(A) Domain architecture of the full-length OSM-3 kinesin motor protein. The motor domain (blue), coiled coils (yellow), hinge regions (h1 and h2) and tail domain (gray) are indicated. The residue numbers of each domain are shown below the bar graph. (B) An autoinhibition model of OSM-3 kinesin motor protein. Cargo attachment or the G444E mutation in h2 region converts the full-length OSM-3 from its intramolecular folded and autoinhibited conformation to an extended and active conformation that leads to an abnormal processive movement without cargo loading in vitro. (C) Overlay of the elution profiles of WT (line in black) and G444E mutant (line in blue) OSM-3. The left shoulder of the OSM-3 WT overlaps with the elution profile of the OSM-3 G444E. The molar masses determined from the expected mass and the MALS fit are shown on the left. (D) The SDS–PAGE analyses to identify the elution peaks shown in (C). Protein constituents were determined by subsequent liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis. (E) Schematic of the C. elegans amphid and phasmid cilia. Each cilium contains a ciliary base, a middle segment (m.s.), and a distal segment (d.s.). The dashed boxes are enlarged on the right. (F) Ciliary localization of GFP-tagged wild-type (WT) OSM-3 or OSM-3-G444E protein in corresponding knock-in (KI) animals. Magenta shows mScarlet-tagged endogenous dynein-2 motor protein CHE-3. Arrows indicate the ciliary base. Arrowheads indicate the junctions between the middle and distal ciliary segments. Scale bars, 5 μm. (G) Representative transmission electron microscopy (TEM) images of the distal and middle ciliary segment from WT and osm-3(p802, null) or osm-3(sa125, G444E) mutant animals. The asterisks denote the presence of cilia in the WT amphid channel, while the cilia are absent in the osm-3(p802, null) or osm-3(sa125, G444E) alleles. Left and middle panels: overviews of the amphid channel. Right panels: the images of a single cilium. Scale bars, 500 nm (left and middle); 100 nm (right). Red arrowheads indicate ectopic singlet microtubules. (H) Representative fluorescence images of amphid and phasmid cilia. WT OSM-3 was expressed under the control of the ciliated sensory neuron-specific promoter Pdyf-1 in OSM-3-G444E::GFP KI animals. Arrows indicate the ciliary base. Arrowheads indicate the junctions between the middle and distal ciliary segments. Scale bar, 5 μm. (I) Quantification of cilium length in WT, osm-3(G444E) and the rescue strains. Green asterisk represent comparisons between WT and osm-3(G444E) animals; purple asterisk represents comparisons between osm-3(G444E) and the rescue strain. N = 20–53, number of worms analyzed. n.s. not significant; ***P < 0.001 by one-way ANOVA using BH method to adjust P values. Data are mean ± SD. Source data are available online for this figure.

Figure 2. OSM-3CA is absent from cilia but forms granules in axons or disposed through membrane abscission at the tips of neurites as ejecta.

(A) Representative fluorescence images of the endogenous OSM-3-G444E that abnormally accumulate between two pharynxes compared to WT OSM-3. The white dotted lines indicate the center between the two pharynxes. Scale bar, 10 μm. A anterior, P posterior. (B) Statistics of the number of OSM-3-G444E puncta at different developmental stages. YA young adult. N = 15 animals for each stage. Green asterisks represent comparisons between L1-stage and the other four-stage animals; orange asterisks represent comparisons between L2–L4-stage and young-adult animals. n.s. not significant; *P < 0.05, ***P < 0.001 by one-way ANOVA using BH method to adjust P values. Data are mean ± SD. (C) Representative fluorescence images of OSM-3-G444E localization relative to the ciliated sensory neurons. The histone marker HIS-54::BFP and membrane marker Myristoylation::mScarlet are expressed under the control of Pdyf-1 (a ciliated neuron-specific promoter) in OSM-3-G444E::GFP KI animals. Ph pharynx. Scale bar, 10 μm. A anterior, P posterior. (D) Statistics of the distribution of OSM-3-G444E puncta at the young adult stage. N ≥ 15 animals. Data are mean ± SD. (E) Representative fluorescence image of OSM-3-G444E puncta at the tips of axons or ectopic neurites. The histone marker HIS-54::BFP and the membrane marker Myristoylation::mScarlet are expressed under the control of the ciliated sensory neuron-specific promoter Pdyf-1 in OSM-3-G444E::GFP KI animals. The dashed box is enlarged on the right. Orange arrowheads, an ectopic neurite; arrows, the tip of the neurite. Scale bars, 10 μm. (F) Schematic diagram of the fate of hyperactive OSM-3 in sensory neurons. (G) Fluorescence time-lapse images of endogenous OSM-3-G444E in the sensory neuron. #1 and #2 indicate the two sequentially released OSM-3-G444E puncta. Scale bar, 5 μm. Source data are available online for this figure.

Figure 3. Dynamics of OSM-3CA within ejecta and neurites.

(A) Representative FRAP images of released OSM-3CA ejecta after partial photobleaching. Endogenous GFP-tagged OSM-3CA ejecta are partially bleached and images are taken every 100 ms. Yellow dashed line marks the bleached region. Scale bar, 1 μm. (B) Statistics of fluorescence recovery of bleached regions in (A). Gray dots with lines are mean values of each time point while dotted lines represent SD. The dashed line is the recovery curve fitted with an exponential equation as described in ”Methods”. Colored full lines show 95% CI (confidence interval) of the fitted curve. t_1/2, time used to achieve half of the maximum recovered intensity. I_R maximum recovered intensity, n total number of events. (C) upper panel, schematics of the photobleaching experiment. Unreleased OSM-3CA ejecta are marked with yellow dashed lines while abnormal neurites are marked with green dashed lines. Yellow regions are photobleached; Lower panel, representative FRAP images after photobleaching of the ejecta regions. Images are taken every second. Scale bar, 2 μm. (D) Statistics of fluorescence recovery of yellow region marked ejecta that are photobleached in (B). (E) Statistics of fluorescence intensity in abnormal neurites in (B) after photobleaching. Curve is fitted with an exponential equation as described in ”Methods”. T_1/2 fluorescence half-life, I_P plateau value indicates the minimum intensity in the region after it is balanced. (F) Representative FRAP images of released OSM-3CA ejecta after full-region photobleaching. Lower panel shows the enhanced images of the upper panel. Images are taken every 10 s. Scale bar, 1 μm. (G) Statistics of fluorescence recovery of yellow region marked ejecta that are photobleached in (F). Source data are available online for this figure.

Figure 4. The disposed OSM-3CA were engulfed by neighboring glial cells.

(A) Schematic depiction of the relative position of postembryonic amphid neurons and CEPsh glia. Axons of bilateral amphid entering the nerve ring around the pharynx are arranged in bundles and enveloped by CEPsh glia. A anterior, P posterior, D dorsal, V ventral. (B) Fluorescence images of endogenous OSM-3-G444E and CEPsh glia. mScarlet is expressed under the control of a CEPsh glia-specific promoter Phlh-17 in OSM-3-G444E::GFP KI animals. Ph pharynx. Scale bar, 10 μm. A anterior, P posterior. (C) Representative fluorescence images (left) and quantification scheme (right) of OSM-3-G444E puncta in the animals of the indicated genotypes. Count, the number of puncta. Scale bar, 10 μm. (D) Quantification of the number, size and fluorescence intensity of OSM-3-G444E puncta in the animals of different genotypes. N ≥ 50 animals for all genotypes (left and middle). Each dot represents the average measurements of 5–35 individual puncta in one worm (middle). N = 10 animals, 168–244 puncta for per strain (right). Green asterisks represent comparisons between OSM-3-G444E::GFP KI animals (Control) and other strains; pink asterisks represent comparisons between OSM-3-G444E::GFP; ced-1(e1735) double mutants and the rescue strains. n.s. not significant; *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA using BH method to adjust P values. Data are mean ± SEM. Source data are available online for this figure.

Figure 5. Intragenic mutations of osm-3 restore the ciliary localization of OSM-3-G444E and rescue ciliary defects in osm-3(G444E).

(A) Schematic of the mutations on OSM-3 that suppress the phenotypes of osm-3(G444E). (B) The intragenic mutations on OSM-3. Structure of dimerized OSM-3 was predicted by Alphafold2 using LocalCoLabFold. Motor domain is labeled with green in one copy of OSM-3 and the stalk and tail regions are labeled with wheat color. (C) Amphid and phasmid cilia in the WT, H207Q mutant osm-3(H207Q), R238W mutant osm-3(R238W) animals, and the two suppressors osm-3(H207Q; G444E), osm-3(R238W; G444E). Cilia are visualized with GFP-tagged endogenous WT OSM-3 or the corresponding mutant OSM-3. Dyf, dye-filling defective; N ≥ 100. Arrowheads indicate the ciliary base. Arrows indicate the junctions between the middle and distal segments. Scale bar, 5 μm. (D) Quantification of cilium length in WT and osm-3 mutant animals. Statistical significance compared to the control with matching color codes. N = 27–53, number of worms analyzed. ***P < 0.001 by one-way ANOVA using BH method to adjust P values. Data are mean ± SD. (E) Histogram of anterograde IFT speeds of WT and mutant OSM-3 motor at the middle (top) and distal (bottom) ciliary segments. Velocities (V) and numbers of IFT particles (N) are indicated. The plots were fit by a Gaussian distribution. Comparisons were performed between the WT and the mutants. ***P < 0.001 by unpaired Student’s t tests. Data are mean ± SD. (F) Relative ATPase activity of WT and mutant OSM-3. The assays were performed as described in methods following the established protocols. The mean activity of WT OSM-3 was set to 100%. The results were obtained from three independent experiments. Statistical significance compared to the control with matching color codes. *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA using BH method to adjust P values. Data are mean ± SD. (G) Motility properties of WT and mutant OSM-3. N.D. not detected. Data are mean ± SD. Source data are available online for this figure.

Figure 6. Fluorescence lifetime imaging microscopy (FLIM) images of the WT and mutant OSM-3 proteins in sensory neuron soma.

(A) Schematic diagram designed for FLIM imaging. (B) Representative FLIM images of WT OSM-3 with different fluorescence labeling. OSM-3::GFP, OSM-3::GFP::mScarlet and mScarlet::OSM-3::GFP are expressed under the control of Pdyf-1 in osm-3(p802) mutant animals. Fluorescence lifetimes of GFP are shown with a scale of pseudo color ranging from 2.40 to 3.00 nanoseconds (ns). Scale bar, 5 μm. (C) Frequency distribution of GFP fluorescence lifetimes measured from the images as shown in (B). Over 10⁶ events from 30 areas in 10 worms of each strain were measured. the plots were fit with a Gaussian distribution. Comparisons were performed between OSM-3::GFP and OSM-3::GFP::mScarlet or mScarlet::OSM-3::GFP. ***P < 0.001 by one-way ANOVA using BH method to adjust P values. (D) Representative FLIM images of WT (image from (B)) or mutant OSM-3 whose N-terminus was tagged with mScarlet tag and C-terminus was tagged with GFP in sensory neurons. Fluorescent protein-tagged OSM-3 are expressed under the control of Pdyf-1 in osm-3(p802) mutant animals. Fluorescence lifetimes of GFP are shown with a scale of pseudo color ranging from 2.40 to 3.00 nanoseconds (ns). Scale bar, 5 μm. (E) Frequency distribution of GFP fluorescence lifetimes measured from the images as shown in (D). The statistical analysis was performed the same as in (C). Blue asterisks represent comparisons between WT OSM-3 and the other three mutant OSM-3. Green asterisks represent comparisons between OSM-3-G444E and OSM-3-H207Q-G444E or OSM-3-R238W-G444E. n.s. not significant, represents the comparison between OSM-3-H207Q-G444E and OSM-3-R238W-G444E. ***P < 0.001 by one-way ANOVA using BH method to adjust P values. (F) Representative FLIM images in dendrites of sensory neurons in the strains as labeled. Scale bar, 5 μm. (G) Frequency distribution of GFP fluorescence lifetimes measured from (F). Over 5 × 10⁵ events from 30 areas in 10 worms of each strain were measured. The statistical analysis was performed the same as in (C). Source data are available online for this figure.

Figure 7. Neurons dispose of hyperactive kinesin-1 resulting from an ALS-associated mutation.

(A) Amphid and phasmid cilia in the WT, G235A mutant osm-3(G235A) (OSM-3-G235A::GFP KI animals) and osm-3(G235A); klp-11(tm324) double-mutant animals. Dyf, dye-filling defective; N ≥ 100. Arrowheads indicate the ciliary base. Arrows indicate the junctions between the middle and distal segments. The WT panel is from Fig. 5C. Scale in image, 5 μm. Scales in kymograph, vertical, 5 s; horizontal, 5 μm. (B) The absence of ciliary localization of the endogenous GFP-tagged OSM-3-G235A-G444E in KI animals. The mScarlet-tagged endogenous CHE-3 is shown in magenta. Arrowheads indicate the ciliary base. Arrows indicate the junctions between the middle and distal segments. Scale bar, 5 μm. (C) Representative fluorescence images of the endogenous OSM-3-G444E or OSM-3-G235A-G444E localization relative to the ciliated sensory neurons. HIS-54::BFP and Myristoylation::mScarlet are expressed under the control of Pdyf-1 in OSM-3-G444E::GFP or OSM-3-G235A-G444E::GFP KI animals. Dashed boxes are enlarged on the right. The OSM-3-G444E panel is from Fig. 2C. Scale bars, 10 μm (left) and 5 μm (right). A anterior, P posterior. (D) Representative fluorescence images of the localization of KIF5A or KIF5AΔExon27 in sensory neurons. GFP-tagged WT KIF5A or the 27^th exon deleted mutant KIF5A (KIF5AΔExon27) and the membrane marker Myristoylation::mScarlet are expressed under the control of Pdyf-1. The dashed boxes (yellow: soma; white: axon) are enlarged on the right. Orange arrowheads, an ectopic neurite; orange arrows, the KIF5AΔExon27 granules colocalize with axon; white arrows, the KIF5AΔExon27 puncta that does not colocalize with axon. Scale bars, 5 μm. (E) Representative fluorescence images of the localization of KIF5A-G235A or KIF5A-ΔExon27-G235A in sensory neurons co-expressed with the membrane marker Myristoylation::mScarlet under the control of Pdyf-1. Scale bar, 10 μm. (F) Statistics of puncta numbers of KIF5A mutants showed in (D, E). The center of the box is the mean value; the upper and lower bounds of the box are the first and third quartile, respectively; the whiskers show the minima and maxima. For each strain, n > 15 worms were analyzed. (G) A proposed model for the regulation of hyperactive OSM-3 in vivo. Source data are available online for this figure.

Figure EV1. The OSM-3-G444E mutation is recessive and loss-of-function, related to Fig. 1.

(A, B) Frequency distribution of single-molecule velocity of WT and G444E mutant OSM-3. The velocity curve was fitted with a Gaussian distribution. n number of single-molecule events measured, v velocity. Data are mean ± SD. (C) Representative image of non-fluorescently-tagged OSM-3-G444E in osm-3(sa125), yellow dashed lines showed the pharynx, A, anterior; P, posterior. Scale bar, 10 μm. (D) Heterozygotic GFP-KI animals expressing one copy of WT osm-3 and the other copy of osm-3(G444E). Cilia that are visualized with mScarlet-tagged endogenous DYF-11 are indistinguishable from WT animals. Arrowheads indicate the ciliary base. Arrows indicate the junctions between the middle and distal segments. Scale bar, 5 μm. (D) shows GFP fluorescence in cilia whereas (E) shows GFP puncta around axons or ejecta outside of sensory neurons. Scale bar, 10 μm. (F) Expression of the WT osm-3 gene under the control of a ciliated neuron-specific promoter Pdyf-1 did not rescue the ejecta formation in osm-3 (G444E) mutants. Scale bar, 10 μm. (G) Expression profiles at osm-3 locus in WT and osm-3(G444E) strains. Gene body schematic is shown at the bottom. (H) Representative RNA editing analysis at osm-3 locus in WT and osm-3CA strains. No editing events were found. The bar showing editing level and gene body schematic are shown above.

Figure EV2. Localization and disposal of hyperactive OSM-3 at the tips of ectopic neurites, related to Figs. 2 and 3.

(A, B) Representative fluorescence images of OSM-3-G444E granules at the tips of axons or ectopic neurites. HIS-54::BFP and Myristoylation::mScarlet are expressed under the control of Pdyf-1 in OSM-3-G444E::GFP KI animals. Dashed box is enlarged on the right (A) or the bottom (B). Arrows, the tip of the neurites where the hyperactive OSM-3 localized. Scale bars, 10 μm. A anterior, P posterior. (C, D) Fluorescence time-lapse images of endogenous OSM-3-G444E in the sensory neurons. White arrows, the punctum that is about to be released; yellow arrows, the released punctum. Scale bars, 5 μm.

Figure EV3. Colocalization of OSM-3-G444E ejecta with actin, ESCRT I component TSG-101 and microtubules.

(A) Fluorescence time-lapse images of OSM-3-G444E and its colocalization with actin in the sensory neurons. GFP-tagged MoesinABD is expressed under the control of Pdyf-1 in OSM-3-G444E-mScarlet KI worms. Scale bar, 5 μm. (B) Representative fluorescence images of OSM-3-G444E puncta that colocalize with actin. Dashed box is enlarged in (C). Scale bar, 5 μm. (C) Line scan shows colocalization of OSM-3-G444E-mScarlet and MoesinABD labeled actin. Relative intensities are plotted using normalized gray values along the scanning line. Yellow arrow shows the scanning direction on the scanning line. Scale bar, 5 μm. (D) Enrichment fold of MoesinABD labeled actin. Enrichment fold = [(Mean Gray Value of Colocalized Area)-(Background Gray Value)]/ [(Mean Gray Value of Non-Colocalized Neurites)-(Background Gray Value)]. n number of events analyzed. (E) Representative fluorescence images of OSM-3-G444E puncta that colocalize with GFP-tagged TSG-101. Dashed box is enlarged in (F). Scale bar, 5 μm. (F) Line scan shows colocalization of OSM-3-G444E-mScarlet and GFP::TSG-101. Relative intensities are plotted using normalized gray values along the scanning line. Yellow arrow shows the scanning direction on the scanning line. Scale bar, 5 μm. (G) Enrichment fold of GFP::TSG-101. n number of events analyzed. (H) Representative fluorescence images of OSM-3-G444E puncta that colocalize with TBB-2, TBB-2 forms ring-like structures at the edges of the puncta. Scale bar, 5 μm.

Figure EV4. ced-2 and ced-5 are not involved in the clearance of OSM-3-G444E ejecta.

(A) Representative fluorescence image of OSM-3-G444E ejecta in ced-1(Q375Stop) mutants with the overexpression of wild-type ced-1 under the control of PY37A1.B. Count, the number of ejecta. Scale bar, 10 μm. (B) Quantification of the number of OSM-3-G444E ejecta in the strains shown in (A), n > 50 worms were analyzed for each strain. Data are mean ± SD. (C) Representative fluorescence image of OSM-3-G444E ejecta in ced-2(R102Stop) or ced-5(E28Stop) strains. (D) Quantification of the number of OSM-3-G444E ejecta in the strains shown in (C), n > 50 worms were analyzed for each strain. Data are mean ± SD. (E) Representative fluorescence image of OSM-3-G444E ejecta in ced-1(Q375Stop); ced-2(R102Stop) double mutant. (F) Quantification of the number of OSM-3-G444E ejecta in the strains shown in (E), n > 50 worms were analyzed for each strain. Data are mean ± SD. ***P < 0.001, *P < 0.05, n.s. not significant, by one-way ANOVA using BH method to adjust P values.

Figure EV5. Intragenic mutations that rescue the ciliary phenotypes in osm-3(G444E) mutant and restore the localization of endogenous OSM-3-G444E.

(A) Flowchart of the suppressor screening. OSM-3-G444E::GFP KI animals at the late L4 stage were treated with ethyl methanesulfonate (EMS). By filling with the fluorescent dye DiI, dye-positive F2 progenies were considered putative suppressors. About 10⁵ haploid genomes were screened in 5 rounds of screenings. The localization of OSM-3-G444E were examined via confocal microscopy. Mutant genes were cloned by the whole-genome sequencing. (B) Molecular lesions of the osm-3 and dyf-5 suppressor alleles. (C) Amphid and phasmid cilia in WT and the suppressors harbor the mutations shown in Fig. 5A, B. GFP-tagged endogenous WT OSM-3 or the corresponding mutant OSM-3 are shown. Dyf, dye-filling defective; N ≥ 100. Arrowheads indicate the ciliary base. Arrows indicate the junctions between the middle and distal segments. The WT image is from Fig. 5C. Scale bar, 5 μm. (D) Distribution of endogenous OSM-3-H207Q-G444E (left), OSM-3-R238W-G444E (middle) and OSM-3-H207Q-G445E (right) in sensory cilia. For the strain shown on the left and in the middle, H207Q or R238W mutation was generated in OSM-3-G444E::GFP KI animals by CRISPR/Cas9-triggered genome editing. The OSM-3-H207Q-G445E strain shown on the right was obtained from a suppressor screen restoring ciliary defects caused by H207Q mutation in OSM-3-H207Q::GFP KI worms. Dyf, dye-filling defective; N ≥ 100. Arrowheads indicate the ciliary base. Arrows indicate the junctions between the middle and distal segments. Scale bar, 5 μm. (E) Frequency distribution of MT gliding velocities showed in Fig. 5G. The velocity distribution curves were fitted with a Gaussian distribution. n number of MTs measured, v velocity. Data are mean ± SD. (F) Distribution of the endogenous OSM-3-H207Q-G444E (top left), OSM-3-H207Q (top right) and OSM-3-R238W-G444E (bottom left) in the soma of sensory neurons. Ph, pharynx. Scale bar, 10 μm. A anterior, P posterior. (G) Left, overlays of the elution profiles of WT (line in black), G444E (red) and H207Q or R238W mutant (blue) OSM-3. The molar mass of OSM-3-H207Q or R238W determined from the MALS fit is shown on the left; right, SDS–PAGE analyses to identify the elution peak *1 shown on the left. Protein constituents were determined by subsequent liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.