NEKL-3 phosphorylates OSM-3 at its “elbow”

(A) Schematic of the full-length OSM-3. Motor domain (blue), neck (gray) and coiled-coils (green) are indicated. CC, coiled-coil. (B) Mass spectrum of an OSM-3 peptide that was phosphorylated by NEKL-3. Phosphorylated gel bands were subjected to MS analysis searching for phosphorylation modifications. Residues 487 to 490 of OSM-3 were phosphorylated and were marked by green color. (C) Phosphorylated 488-490 residues are at the “elbow” of OSM-3. It shows the overall structure of the homodimeric OSM-3 predicted by alphafold2. The dashed square marks the “elbow” region and is zoomed-in in (D). (E) Genome editing constructs of the elbow, showing the PD (phosphor-dead) and PM (phosphor-mimic) sequences comparing to wild type. Residue 487-490 of OSM-3 was edited to “FAAA” for PD strain or edited to “DDEE” for PM strain. Abbreviations: Y, Tyr; S, Ser; T, Thr; F, Phe; A, Ala; D, Asp; E, Glu.

Phosphorylation at the elbow of OSM-3 is inhibitive in vivo

(A) Representative images of the phospho-dead (PD) and phospho-mimic (PM) knock-in worms showing their OSM-3 signal at amphid cilia, amphid neuronal soma and phasmid cilia, respectively. The contours of the worms are marked by white dashed lines. Scale bar, 10μm. (B) Representative images of the cilia from PD and PM worms marked by the ciliary marker DYF-1::mScarlet. White arrows indicate the junction between middle and distal segments while the asterisks indicate the ciliary base. M.S., middle segment; D.S., distal segment. Scale bar, 5μm. (C) Statistics of the cilium length of the strains showed in (B). The lengths of DYF-1::mScarlet signals were measured and analyzed. (D) IFT velocities of PD and PM worms. ****, p < 0.0001, analyzed by one way ANOVA, p values were adjusted by BH method.

Structural models of the OSM-3 kinesin and its mutants

(A) Relaxed structure models of OSM-3 and mutants. Black arrow heads indicate the elbow while red dashed circles mark the C-terminus of the protein. Black dashed lines showed the extending direction from the elbow towards the C-terminus. WT and PM showed close interaction between the tail and motor domain while G444E and PD showed that the tails are far away from the motor. (B) Heatmaps of the energy states of the pre-relaxed structure models from amino acid 481st to 500th, as labeled on the left; the amino acids between the white lines are the elbow region; each row represents an energy item as labeled on the bottom. (C) Heatmaps comparing the energy states by direct subtraction between the mutants and WT. The PD mutant has lower “fa_dun” energy while has higher “ref” energy than that of the PM mutant. Energy terms are explained in Fig. S4.

Phospho-dead OSM-3 behaves constitutively active in vitro while Phospho-mimic OSM-3 stays autoinhibited.

(A) Microtubule stimulated ATPase activity of WT OSM-3 and mutants. G444E, the hyperactive positive control; KHC, kinesin heavy chain. Average activity of KHC was set to 100% and others was normalized to KHC. (B) Summary of the single-molecular assay and the microtubule gliding assay. R.D., rarely detected. N.A., not available. Data are [mean ±SD (number of events)]. (C) Velocity distributions of microtubule gliding assays of the indicated OSM-3 constructs. n, total evens measured. v, μm·s−1, average velocity with standard deviation. (D) Statistics of microtubule gliding velocities shown in (C). (E) Representative kymographs of the single-molecular movements of WT OSM-3 and mutants as indicated. Scale bars, vertical, 10 s; horizontal, 5 μm. (F) Velocity distributions of the single-molecular assays. n, total evens measured. v, μm·s−1, average velocity with standard deviation. The distribution of G444E was fitted with a Gaussian distribution curve while the distribution of PD was fitted with a one-phase decay curve. (G) Run length distributions of the single-molecular assays. n, total evens measured. l, average run length. The curves were fitted with the one-phase decay distribution. *, p < 0.05, **, p < 0.01, ****, p< 0.0001, analyzed by one way ANOVA, p values were adjusted by BH method.

Genetic screening identified T489 as the key regulatory residue in the elbow of OSM-3.

(A) Schematics of the forward genetic screen. 100% Dyf osm-3(PD)::GFP KI worms were mutated by EMS and F2 progenies were screened for dye filling positive mutants. (B) Two independent suppressor mutants cloned from the genetic screening. (C) Amino acid sequences of the suppressor mutants at the elbow. (D) Representative images of the cilia of the suppressors and the kymographs showing the velocity of OSM-3. The rightest panel shows the same OSM-3 version (487-489: “FATA”) with the suppressors but over-expressed under the ciliary Pdyf-1 promoter in osm-3(p802) worm. The arrows indicate the junction between middle and distal segments while the asterisks indicate the ciliary base. M.S., middle segment; D.S., distal segment. Scale bars, vertical, 10 s; horizontal, 5 μm. (E) Summary of the OSM-3 velocity. M.S., middle segment. D.S. distal segment. Data are [mean ±SD (number of events)].

Proposed model for elbow regulation of OSM-3 by NEKL-3 phosphorylation

OSM-3 is phosphorylated at the elbow by NEKL-3 and behaved autoinhibited while after dephosphorylation, OSM-3 turns from a compact state to an extended state due to the elbow conformational changes.

Kinesin-2 family members have conserved elbow motif

Sequence alignment of OSM-3 and other members in kinesin-2 family (KIF3A, KIF3B, KIF3C and KIF17) from different model organisms as indicated. This figure shows the alignment results around elbow region. Green arrow heads indicate the four phosphorylated residues. The position of amino acids of OSM-3 is labeled at the top of the figure. Sequences were aligned using CLUSTAL O (1.2.4) and presented by ESPript 3.0.

PAE and pLDDT of OSM-3 dimer models

(A) PAE (Predicted Aligned Error) of OSM-3 dimer presented by LocalColabFold. A total of five predicted model were presented and the “rank_1” model was selected for following analysis. x axis, residue number; y axis, A and B represent for two identical OSM-3 peptide chains. (B) pLDDT (predicted Local Distance Difference Test) for all five predicted models shown in (A).

Statistical analysis of the IFT velocities in osm-3pd and osm-3pm worms, corresponding to Fig. 2D

(A-C) IFT velocity marked by DYF-1-mScarlet. Genotypes are shown on the left. Left panel, frequency distribution of IFT particles in the middle segment. Middle panel, frequency distribution of IFT particles in the distal segment. Data were fitted with a Gaussian distribution. Right panel, overlay of the fitted curves of middle and distal segments. (D) OSM-3::GFP velocity in osm-3pm worms. N.A., not available.

Summary of energy terms in Rosetta Energy Function, corresponding to Fig.3

(B-C) This table explains the terms used in the analysis in Fig.3 (B-C). Modified from Alford et al., J Chem. Theory. Comput., 2017 (Alford et al., 2017).

SDS-PAGE of purified recombinant OSM-3 mutants

SDS-PAGE of the purified recombinant OSM-3::eGFP and mutants, Coomassie Blue stained. The elution peaks of each recombinant protein were shown, labeled on the top.

Statistical analysis of the IFT velocities in osm-3pd suppressors, corresponding to Fig. 5E

(A-C) OSM-3 velocity of WT and mutant proteins marked by GFP. Genotypes are shown on the left. Left panel, frequency distribution of IFT particles in the middle segment. Middle panel, frequency distribution of IFT particles in the distal segment. Data were fitted with the Gaussian distribution. Right panel, overlay of the fitted curves of middle and distal segments. (D) Velocity and frequency distribution of overexpressed OSM-3 (487-489: “FATA”) under the ciliary Pdyf-1 promoter in osm-3(p802) worm.

T489E and T489A of OSM-3 cause aggregated signals.

(A-B) Representative images of OSM-3(T489E) and OSM-3(T489A) overexpressed in osm-3(p802) strain under the ciliary Pdyf-1 promoter, respectively. The images showed the areas around nerve rings. Scale bar, 10 μm. (C) Statistics of strains in (A-B), showing the percentage of worms with bright puncta around nerve rings for each strain. n, number of worms checked.