Single molecule motility assays

Microtubules were assembled from 5 μl of 18 mg/ml unlabelled pig tubulin, 0.2 μl of 1 mg/ml HiLyte Fluor 647 or X-rhodamine-labelled tubulin (#TL670M, #TL6290M, Cytoskeleton) and 0.5 μl of 0.5 mg/ml biotin tubulin (#T333P, Cytoskeleton) in 15 μl MRB80 (80 mM PIPES pH 6.8, 4 mM MgCl2, 1 mM EGTA, 1 mM DTT) supplemented with 4 mM GTP. The mixture was incubated at 37°C for 90 min before diluting in 80 μl MRB80 supplemented with 20 μM Taxol. Unincorporated tubulin was removed by pelleting microtubules through a glycerol cushion (30% glycerol in MRB80) at 20,238 × g for 12 min at room temperature. The microtubule pellet was resuspended in 80 μl of MRB80 with 20 μM Taxol and stored at 28°C covered from light for use on the same day.

Coverslips (22 × 22) were cleaned by incubating in 2.3 M hydrochloric acid overnight at 60 °C. The next day, coverslips were washed with Millipore water and sonicated at 60 °C for 5 min. The wash cycle was repeated five times. The coverslips were dried using a Spin Clean (Technical video) and plasma cleaned using Henniker plasma clean (HPT-200) for 3 min. Flow chambers were made using clean glass slides (Menzel Gläser Superfrost Plus, Thermo Scientific) and double- sided sticky tape (Scotch 3 M) by placing the cleaned coverslip on the sticky tape creating a 100 μm deep flow chamber. The surface was coated with (0.2 mg/ml) PLL(20)-g[3.5]-PEG(2)/PEG(3.4)-Biotin (50%) (#PLL(20)-g[3.5]-PEG(2)/PEGbi, Susos AG). Biotin-647-microtubules were attached to this surface with streptavidin (0.625 mg/ml) (#S4762 Sigma) and the surface was blocked with κ-casein (1 mg/ml) (#C0406 Sigma).

KIF1C-GFP, KIF1C_P176L-GFP and KIF1C_R169W-GFP were diluted in TIRF Assay Buffer (25 mM HEPES-KOH pH 7.2, 5 mM MgSO4, 1 mM EGTA, 1 mM DTT, 10 μM Taxol) supplemented with 0.05% Tween-20, 25 mM KCl and 200ng/ml κ-casein, spun at 100,000 × g for 10 min in an Airfuge (Beckman Coulter) and fluorescence at 507 nm measured using a NanoDrop™ 3300 Fluorospectrometer to verify concentrations used in the assay and determine landing rates. Proteins were then added to the motility mix (TIRF Assay Buffer supplemented with 5 mM ATP, 5 mM phosphocreatine (#P7936, Sigma), 7 U/ml creatine phosphokinase (#C3755 Sigma), 0.2 mg/ml catalase, 0.4 mg/ml glucose oxidase, 4 mM DTT, 50 mM glucose (#G8270, Sigma), 25 mM KCl, 10 μM taxol, 0.2 mg/ml κ-casein) and flown into the chamber.

Chambers were observed on an Olympus TIRF system using a ×100 NA 1.49 objective, 488 and 640 nm laser lines, an ImageEM emCCD camera (Hamamatsu Photonics) under the control of xCellence software (Olympus), an environmental chamber maintained at 25 °C (Okolab, Ottaviano, Italy). Images were acquired at 5 fps for 180 seconds using 2x2 binning, thus the resulting images have 162 nm pixels and 200 ms temporal resolution.

The fraction of microtubules that contain one or more kinesin motors at the plus end was determined by counting all microtubules for which both ends were visible in the first frame of the second movie taken from each chamber, i.e. about 3 minutes after start of the experiment, and scoring for the presence / absence of a GFP signal at one of the ends.

Motility was analysed by tracing microtubules and generating maximum intensity kymographs from 7 pixel wide lines using the ImageJ Kymograph plugin from Arne Seitz (EPFL Lausanne). Motor tracks from kymographs were traced by hand and recorded as ImageJ ROIs. These paths were analysed in a custom-built python analysis script. Individual phases of tracks were segmented from the ROI, and motors were said to be translocating towards the plus-end or minus-end of the microtubule in each phase if their speed towards that end of the microtubule was greater than 25 nm/s. When the motor’s speed was less than 25 nm/s, it was annotated as paused. Motors with an average speed of less than 25 nm/s or a total run length of less than 500 nm were considered as static. Non-static motors that underwent bidirectional motion of at least 324 nm or just moved towards the minus end where classed as diffusing. Only the remaining, plus end directed motors were analysed for run parameters. Dwell times were calculated as the total time the motor spent on the microtubule until it reaches the plus end or the end of the recording. Run length is the total distance covered by the motor. The pause-corrected average speed was calculated by dividing the run length by the dwell time minus any time when the motor was paused. Superplots show the mean speed, run-length, or dwell time for a given experimental day, with smaller colour-coded spots showing the individual measured values of motors. On each of the three experiment days, one to four chambers were prepared for each protein. Statistics were calculated at the per-motor level between experimental groups. The data were tested for normality using D’Agostino and Pearson’s test, and as they were not normally distributed, a Kruskal-Wallis test was used to determine if experimental groups differed significantly. Where a statistically significant difference was indicated, pairwise interactions were tested using Conover’s post-hoc test and p-values were corrected for multiple comparisons using the Holm-Bonferroni method.

To test homogeneity of the motors in the motility assay, the intensity of motors was determined using the ImageJ plugin GaussFit_OnSpot for each line traced in the kymograph. The normalised intensity value output by the function for the first 10 time points of each track were averaged to obtain the fluorescent intensity for the motor. To pool intensity measurements from different experiments, data were divided by the average intensity of all wildtype KIF1C detections and then grouped into highly processive motors with a run length of more than 1 μm and motors that moved less than 1 μm. While a small fraction of motor clusters were observed in the assays, the clusters did not primarily contribute to the highly processive runs (Figure S2). Data were plotted in MATLAB using the plotSpread function from Jonas Dorn.