Measuring NDC80 binding reveals the molecular basis of tension-dependent kinetochore-microtubule attachments
Figures

FLIM-FRET measurement of NDC80-kMT binding in human tissue culture cells.
(A) Engineered U2OS cell expressing mTurquoise2-NDC80 and β-tubulin-TC-FlAsH. NDC80 (gray), mTurquoise2 (blue) and TC-FlAsH (green). (B) Two-photon microscopy images of the engineered U2OS cells not exposed to FlAsH (top) and exposed to FlAsH (bottom). 3 μm scale bar. mTurquoise2 (blue) and FlAsH (green). (C) Example fluorescence decay curves of mTurquoise2-NDC80 in the engineered U2OS cells not exposed to FlAsH (left, green circle) and exposed to FlAsH (right, orange triangle), plotted with the best-fit single-exponential decay models (black and blue dotted lines). Corresponding weighted residuals (the deviation of data from model, divided by the square root of the number of photons) are plotted below after being smoothened to display systematic deviations. (D) The fluorescence decay curve of mTurquoise2-NDC80 in the presence of FlAsH labeling (orange triangle, same as (C)), plotted with the best-fit two-exponential model (blue solid line). The single-exponential model fit to the fluorescence decay curve in the absence of FlAsH labeling (black dotted line) plotted together for comparison. Corresponding smoothened weighted residual (described above) for the two-exponential model is plotted below. Long- and short-lifetime exponentials correspond to the mTurquoise2-NDC80 populations in non-FRET state and FRET state, respectively, and their relative amplitudes give the fraction of each population. To facilitate the comparison, the fluorescence decay curves in the absence and presence of FlAsH labeling were normalized such that they asymptotically overlap. Data points and source FLIM data are available in Figure 1-Data (Yoo et al., 2018).

β-tubulin labeling fraction measurement.
(A) Two-photon fluorescent microscopy images of 9 mitotic cells with β-tubulin-TC-FlAsH. 5 μm scale bar. (B) Example 3D segmentation using active contour algorithm. (C) (top left) The number of voxels in segmented 3D cell images, (top right) the total number of photons collected from entire cells, (bottom left) the total number of photons divided by the number of voxels and the measurement time, and (bottom right) the number of photons collected per second at a voxel in the cytoplasmic region. Gray lines are the average over the 9 different cells. (D) and (E) FCS measurements on Alexa Fluor 488 and TC-FlAsH in solution, respectively. Black circles are averaged over 5 or 6 autocorrelation functions, and red lines are a single-component FCS model fit to the averages. Corresponding weighted residuals (the difference between data and model, divided by the standard deviation of data) are plotted below to show any systematic deviations of the fitted models from the data.

Kinetochore FLIM-FRET measurement.
(A) Illustration of fluorescence decay acquisition in a TCSPC (time-correlated single photon counting) FLIM system. A Ti:Sapphire pulsed laser is used for excitation and a photomultiplier tube (PMT) for detection. The photon arrival time, the difference in timing between the emitted photon and excitation laser pulse, is measured by TCSPC, which is accumulated to a fluorescence decay curve over many laser repetition periods. (B) The fluorescence lifetime is the average time the fluorophore stays in the excited state. (left) When the donor fluorophore (D) is not engaged in FRET with an acceptor (A) fluorophore, the donor fluorophore has a single-exponential fluorescence decay with fluorescence lifetime τD=1/(kr + knr), where kr and knr are the radiative and non-radiative decay rates, respectively. (middle) FRET provides an additional relaxation pathway to the excited donor, reducing the fluorescence lifetime of the donor to τFRET = 1/(kr + knr + kFRET), where kFRET is the FRET rate. (right) The fluorescence decay of a mixture of donors engaged in FRET and not engaged in FRET is a sum of two exponentials with two different lifetimes, τD and τFRET, which corresponds to the non-FRET and FRET populations, respectively. The relative amplitude of the short-lifetime exponential decay provides the fraction of the FRET population, and the lifetime ratio provides the intrinsic FRET efficiency. (C) TCSPC FLIM provides fluorescence decay curve at each pixel, and the total photon counts in each pixel provides a two-photon fluorescence intensity image. To quantify the FRET fraction at each kinetochore, kinetochores were identified based on the intensity image, then the fluorescence decay curves in the pixels within each kinetochore were summed. Then we performed a Bayesian analysis to obtain the posterior distribution of the FRET fraction at each kinetochore. The posterior distributions of the kinetochores in a group of kinetochores were combined by multiplication to compute the mean and SEM of the FRET fraction for the group.

Negative control data for NDC80-kMT FLIM-FRET measurements.
(A) to (D) Schematic descriptions, example cell images, and example mTurquoise2 fluorescence decay curves from three different FRET-negative control experiments and a nocodazole treatment experiment. mTurquoise2 fluorescence decay curves (blue circles) are plotted with best-fit single- (black dotted line) or two-exponential decay model (black solid line), and the associated weighted residuals are plotted below (blue curve). 3 μm scale bar. (A) Negative control 1. Nuf2 N-terminally labeled with mTurquoise2, and no FlAsH labeling. (B) Negative control 2. Nuf2 C-terminally labeled with mTurquoise2 (far from kMT), and no FlAsH labeling. (C) Negative control 3. Nuf2 C-terminally labeled with mTurquoise2 (far from kMT), and β-tubulin C-terminally labeled with FlAsH. (D) Nocodazole treatment experiment. Nuf2 N-terminally labeled with mTurquoise2, and β-tubulin C-terminally labeled with FlAsH. Cell was incubated with 5 µM nocodazole for >10 min to depolymerize microtubules. (E) Boxplot of fluorescence lifetimes estimated from single-exponential models fit to the negative control fluorescence decays. n = 32, 11, and 6 cells for Neg Ctrl 1, 2, and 3, respectively.

Förster radius estimation by FLIM-FRET measurements and Monte Carlo protein simulations.
(A) Fluorescence decay curves of cells expressing mTurquoise2-TC in the absence (green circle) and the presence (orange triangle) of FlAsH. A single-exponential model (black solid line) was fit to the fluorescence decay curve in the absence of FlAsH. For easier comparison, the fluorescence decay curves were normalized such that they asymptotically overlap. (B) The conformational ensemble of the flexible tether between mTurquiose2 and TC were modeled by Monte Carlo protein simulations, and the distance, r, between mTurquiose2 (blue cartoon) and TC-FlAsH (green ball) was estimated. This distribution is denoted by p(r). (C) The measured fluorescence decay of mTurquoise2-TC-FlAsH (orange triangles, same as (A) but not normalized) plotted with the best-fit decay model (black dotted line, model described in the box and derived in Materials and methods). Associated weighted residual (deviation of model from data, divided by the square root of the number of photons) plotted below. Fitting the decay model to the data estimated the Förster radius to be 5.90 ± 0.10 nm (SE).

Characterization and calibration of NDC80-kMT FLIM-FRET measurement.
(A) The conformational ensemble of the flexible tether between mTurquoise2 and Nuf2 (red) and the disordered C-terminal tails of beta-tubulins around the NDC80 (green) were modeled by large-scale Monte Carlo protein simulations, which were then used to calculate the distances, ri, between the mTurquoise2 and the TC motifs. FlAsH labeling was assigned to the TC motifs with 26.1% probability (which is the measured labeling fraction of beta-tubulin). Fluorescence lifetimes of the mTurquoise2 were calculated for randomly sampled sets of distances, , based on which fluorescence decay curves were simulated. (B) (top) Fluorescence decay curves for various distances between NDC80 and MT were simulated and then were fit using single- and two-exponential decay models. Difference in Bayesian information criteria (BIC) between single- and double-exponential models is plotted against the NDC80-MT distance. Data points are mean and SD. (bottom) Example simulated fluorescence decay curves (green dots) for 0, 4, and 9 nm NDC80-MT distances are plotted with the best-fit single- (blue line) and two-exponential (red line) models. Corresponding smoothened weighted residuals plotted below. (C) Fluorescence decay curves for various NDC80 binding fractions (fb) were simulated and fit by using two-exponential decay model to estimate FRET fraction (fFRET). (left) Three example simulated fluorescence decay curves (green dots) for 0, 30, and 60% binding fractions with the best-fit two-exponential models (red line), and the corresponding smoothened weighted residuals plotted below. (right) NDC80 FRET fractions (fFRET) plotted against NDC80 binding fractions (fb) (blue dots), and the linear fit (black line). Gray-shaded area represents the uncertainty in the slope, which was determined from the uncertainties in the measured beta-tubulin labeling fraction and Förster radius (see Supplemental experiments in Materials and methods).

NDC80-kMT binding is regulated in a chromosome-autonomous fashion.
(A) Example cell images and time course of NDC80 FRET fraction from prometaphase to metaphase to anaphase (n = 11 cells). Black squares are the mean, y-error bars are the SEM, and x-error bars are the SD of the data points (green circles) in equally spaced time intervals. 5 μm scale bar. (B) Kinetochores at each time point in prometaphase cells are divided into two groups, centered and off-centered, based on their distances from the metaphase plate. Kinetochores less than 1 μm away from the metaphase plate were classified as centered, and kinetochores more than 2.5 μm away were classified as off-centered. (C) Time course of the fraction of centered (green) and off-centered (orange) kinetochores in prometaphase. (D) Time course of NDC80 FRET fraction of centered (green circles) and off-centered (orange squares) kinetochores in prometaphase (n = 11 cells, 2886 centered and 572 off-centered kinetochores). Data points are the mean, y-error bars the SEM, and the x-error bars the SD in equally spaced time intervals. Gray areas are the 95% confidence intervals for the linear fits. Data points and source FLIM data are available in Figure 2-Data (Yoo et al., 2018).

NDC80-kMT binding is correlated with kMT dynamics and centromere tension.
(A) (left) kMTs predominantly depolymerize at leading kinetochores and polymerize at trailing kinetochores. (right) K-K distance is a proxy for centromere tension. Measuring NDC80-kMT binding along with the kinetochore movement and K-K distance therefore reveals how NDC80-kMT binding is related to the kMT dynamics and centromere tension. (B) Image of a metaphase cell with mTurquoise2-NDC80 (blue) and β-tubulin-TC-FlAsH (green), and kinetochore tracking (yellow circles) and pairing (red lines) results. 3 μm scale bar. (C) NDC80 FRET fraction vs. kinetochore speed for leading (green circle) and trailing (orange triangle) kinetochores (n = 17 cells, 681 kinetochores/data point). Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges within groups of kinetochores with similar velocities. (D) NDC80 FRET fraction vs. K-K distance for untreated cells (green circle, n = 17 cells, 984 kinetochores/data point), cells treated with 10 μM taxol (orange triangle, n = 7 cells, 525 kinetochores/data point), and cells treated with 5 μM STLC (purple square, n = 16 cells, 493 kinetochores/data point). For STLC data, only poleward-facing kinetochores are plotted (see Figure 3—figure supplement 2 for comparison between poleward and anti-poleward kinetochores). Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges within groups of kinetochores with similar K-K distances. Gray area is the 95% confidence interval for the linear fit to the combined data. (E) Histograms of K-K distances for the untreated (top, green), taxol-treated (middle, orange), and STLC-treated (bottom, purple) cells. 3 μm scale bar in the cell images of mTurquoise2-NDC80 (blue) and beta-tublin-TC-FlAsH (green). ***p-value (Welch’s t-test) less than 10−30. Data points and source FLIM data are available in Figure 3-Data (Yoo et al., 2018).

K-K distance and kinetochore velocity are not correlated.
(A) Each data point represents the fraction of leading kinetochores within a group of kinetochores with similar K-K distances. Gray region is the 95% confidence interval of the linear fit. (B) Histogram of K-K distances of leading (green) and trailing (orange) kinetochores.

NDC80 FRET fraction of poleward- and anti-poleward-facing kinetochores in STLC-treated cells with monopolar spindles.
NDC80 FRET fraction vs. K-K distance for poleward-facing kinetochores (purple square, same as Figure 3D) and anti-poleward-facing kinetochores (pink triangles) in cells treated with 5 μM STLC (n = 16 cells, 493 kinetochores/data point). Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges within groups of kinetochores with similar K-K distances. 3 µm scale bar. Gray area is the 95% confidence interval for the linear fit.

Aurora B kinase regulates NDC80-kMT binding in a graded fashion in vivo.
(A) (top) Cell images showing mTurquoise2-NDC80 (blue) and beta-tubulin-TC-FlAsH (green). (bottom) Time course of NDC80 FRET fraction in response to Aurora B inhibition by 3 μM ZM447439 (n = 15 cells). (B) Images of cells with mTurquoise2-NDC80 (blue) and beta-tubulin-TC-FlAsH (green) after depleting endogenous Hec1 by siRNA and expressing siRNA-insensitive WT or three different phosphomimetic mutants of Hec1: 9A-, 2D-, and 9D-Hec1 (see Materials and methods). (C) NDC80 FRET fraction of cells whose endogenous Hec1 are replaced with WT or phosphomimetic Hec1 (see Materials and methods). Black dots are from individual cells and red error bars are mean ± SEM. n = 19, 22, 12, and 17 cells for WT-, 9A-, 2D-, and 9D-Hec1. *p<0.1; **p<0.01; ***p<0.001; ****p<0.0001. (D) Time course of NDC80 FRET fraction of 2D-Hec1-expressing cells in response to Aurora B inhibition by 3 μM ZM447439 (n = 12 cells). (E) (top) Cell images color-coded with Aurora B sensor non-FRET fraction. (bottom) Time course of the non-FRET fraction of the cytoplasmic Aurora B FRET sensor in response to 3 μM ZM447439 (n = 10 cells). In (A), (D) and (E), black squares and error bars are the weighted mean and SEM of the data points (green circles) in equally spaced time intervals of 1 min. Red solid and dashed lines are the best-fit exponential decay models and their 95% confidence intervals, respectively. 5 μm scale bar for all images. (F)NDC80 FRET fraction (from (A)) and NDC80 binding fraction (converted from the FRET fraction) plotted against the fraction of phosphorylated Aurora B phosphorylation sites in NDC80 (converted from Aurora B FRET sensor non-FRET fraction in (E)). Red solid and dashed lines are the best-fit NDC80-kMT binding model (derived in Mathematical modeling in Materials and methods) and its 95% confidence interval. Data points and source FLIM data are available in Figure 4-Data (Yoo et al., 2018).

Supplemental data for Aurora B inhibition experiments.
(A) Time course of NDC80 FRET fraction in response to 0.03% DMSO (n = 5 cells, negative control for Figure 4A). (B) The design of Aurora B FRET biosensor. The FRET sensor contains a kinesin-13 family Aurora B substrate (gray) whose phosphorylation results in its binding to the forkhead-associated domain (FHA2, green) in the sensor, which constrains the sensor in the open non-FRET state. Therefore, measuring the non-FRET fraction of the FRET sensor allows the quantification of Aurora B activity. (C) Time course of the non-FRET fraction of cytoplasmic Aurora B FRET sensor in response to 0.03% DMSO (n = 3 cells, negative control for Figure 4E). (D) Time course of the non-FRET fraction of Nuf2-targeted Aurora B FRET sensor in response to 3 μM ZM447439 (n = 9 cells). Black squares and error bars are the weighted mean and SEM of the data points (green circles) in equally spaced time intervals of 1 min. Red solid and dashed lines are the best-fit exponential model and its 95% confidence interval, respectively. 5 μm scale bars.

Haspin-dependent centromere-localized Aurora B is responsible for the tension dependency of NDC80-kMT binding.
(A) NDC80 FRET fraction vs. K-K distance for 9A-Hec1-expressing cells with no drug treatment (green circle, n = 12 cells, 803 kinetochores/data point), with 10 μM taxol treatment (orange triangle, n = 9 cells, 1113 kinetochores/data point), or with 5 μM STLC treatment (purple square, n = 10 cells, 855 kinetochores/data point). For STLC data, only poleward-facing kinetochores are included (see Figure 5—figure supplement 1A for comparison between poleward-facing and anti-poleward-facing kinetochores). Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges within groups of kinetochores with similar K-K distances. Gray area is the 95% confidence interval for the linear fit to the combined data. (B) Histograms of K-K distances for the untreated 9A-Hec1 cells (top, green), untreated cells with endogenous Hec1 (top, black line), 9A-Hec1 cells treated with taxol (middle, orange), and 9A-Hec1 cells treated with STLC (bottom, purple). 3 μm scale bar in the cell images of mTurquoise2-NDC80 (blue) and beta-tubulin-TC-FlAsH (green). ***p<10−6 (Welch’s t-test). (C) Haspin kinase phosphorylates histone H3 at Thr3 (H3T3), which recruits the chromosome passenger complex (CPC, red) to centromeres. 5-Iodotubercidin (5-ITu) inhibits haspin kinase, thereby displacing Aurora B from centromeres. (D) Spinning-disk confocal microscopy images of cells expressing mNeonGreen-Nuf2 (green) and INCENP-mCherry (red) before (top) and after (bottom) haspin inhibition by 10 μM 5-ITu treatment. 3 μm scale bar. (E) NDC80 FRET fraction vs. K-K distance for cells treated with 10 μM 5-ITu (green circle, n = 15 cells, 1170 kinetochores/data point), for cells treated with both 10 μM 5-ITu and 10 μM taxol (orange triangle, n = 3 cells, 359 kinetochores/data point), and for cells treated with 10 μM 5-ITu and 5 μM STLC (purple square, n = 12 cells, 564 kinetochores/data point). For 5-ITu + STLC data, only poleward-facing kinetochores are included (see Figure 5—figure supplement 1B for comparison between poleward-facing and anti-poleward-facing kinetochores). Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges within groups of kinetochores with similar K-K distances. Gray area is the 95% confidence interval for the linear fit to the combined data. (F) Histograms of K-K distances for the 5-ITu-treated (top, green), untreated (top, black line), 5-ITu + taxol treated (middle, orange), and 5-ITu + STLC treated cells (bottom, purple). 3 μm scale bar in the cell images of mTurquoise2-NDC80 (blue) and beta-tubulin-TC-FlAsH (green). ***p<10−6 (Welch’s t-test). (G) NDC80 FRET fraction and (H) the non-FRET fraction of Nuf2-targeted Aurora B FRET sensor (proxy for Aurora B activity at NDC80) for different drug treatments. Each data point (gray circle) corresponds to an individual cell, and the error bar (red) shows the mean and SEM. P-values from two-sided Welch’s t-test. Data points and source FLIM data are available in Figure 5-Data (Yoo et al., 2018).

NDC80 FRET fraction of poleward and anti-poleward kinetochores in STLC-induced monopolar spindles of 9A-Hec1-expressing cells and haspin inhibited cells.
NDC80 FRET fraction vs. K-K distance for poleward-facing kinetochores (purple squares, same as Figure 5) and anti-poleward-facing kinetochores (pink triangles) in (A) 9A-Hec1-expressing cells treated with 5 μM STLC (n = 10 cells, 855 kinetochores/data point) and (B) endogenous-Hec1-expressing cells treated both with 10 μM 5-ITu and 5 μM STLC (n = 12 cells, 564 kinetochores/data point). Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges within groups of kinetochores with similar K-K distances. 3 µm scale bar.

The concentration of Aurora B at the location of NDC80 decreases with centromere tension.
(A) Spinning-disk confocal microscopy image of mNeonGreen-Nuf2 (green) and INCENP-mCherry (red). 3 µm scale bar. The location of NDC80 was determined to sub-pixel accuracy, using the mNeonGreen-Nuf2 image. For each pair of sister kinetochores, the intensity of INCENP-mCherry at the location of NDC80 was measured and normalized on a cell-by-cell basis. (B) Normalized INCENP-mCherry intensity at the location of NDC80 were averaged within groups of kinetochores with similar K-K distances, and plotted against the K-K distances for untreated (green circles), taxol-treated (orange triangles), and 5-ITu-treated (purple squares) cells. Data points are the mean, y-error bars the SEM, and the x-error bars the interquartile ranges. Black solid and dotted lines are the linear fits to DMSO+taxol combined data and 5-ITu data, respectively. 906 kinetochore pairs in 9 cells, 599 pairs in 8 cells, and 680 pairs in 6 cells were analyzed for DMSO control, taxol treatment, and 5-ITu treatment data, respectively. Data points and source FLIM data are available in Figure 6-Data (Yoo et al., 2018).

A biophysical model of tension dependent NDC80-kMT binding.
(A) Plot of NDC80 binding fraction, fbound, (converted from NDC80 FRET fraction in Figures 3D and 5E) vs. Aurora B concentration at NDC80, [A] (converted from INCENP-mCherry intensity in Figure 6B, see Materials and methods). Data points (black circles) are the mean and SEM. We constructed a mathematical model that predicts NDC80 binding fraction from Aurora B concentration at NDC80 through three steps: intermolecular Aurora B auto-activation, NDC80 phosphorylation, and NDC80-kMT binding. Red line shows the mathematical model fit to the data. (B) NDC80 binding fraction vs. K-K distance before (left) and after (right) haspin inhibition by 5-ITu. The data points (black circles) are adapted from Figures 3D and 5E. Red lines are the predictions from the mathematical model. Data points are available in Figure 7-Data (Yoo et al., 2018).
Tables
Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
---|---|---|---|---|
Cell line (Homo sapiens) | U2OS | ATCC | HTB-96 | |
Transfected construct (Homo sapiens) | pBABE-puro mTurquoise2-Nuf2 | this paper | Nuf2 N-terminally labeled with mTurquoise2; in retroviral vector with puromycin selection marker | |
Transfected construct (Homo sapiens) | pBABE-hygro mTurquoise2-Nuf2 | this paper | Same as above, but with hygromycin selection marker | |
Transfected construct (Homo sapiens) | pBABE-blast mTurquoise2-Nuf2 | this paper | Same as above, but with blasticidin marker | |
Transfected construct (Homo sapiens) | pBABE-blast Aurora B FRET sensor (mTurquoise2/YPet) | this paper | modified from Addgene #45215; Fuller et al. (2008) | |
Transfected construct (Homo sapiens) | Nuf2-targeted Aurora B FRET sensor (mTurquoise2/Ypet) | this paper | modified from Addgene #45215; Fuller et al. (2008) | |
Transfected construct (Homo sapiens) | mTurquoise2-TC | this paper | mTurquoise2 with tetracysteine motif at the C-terminus | |
Transfected construct (Homo sapiens) | WT-Hec1-LSSmOrange | this paper | modified from WT-Hec1-GFP from Jennifer DeLuca | |
Transfected construct (Homo sapiens) | 9A-Hec1-LSSmOrange | this paper | modified from 9A-Hec1-GFP from Jennifer DeLuca | |
Transfected construct (Homo sapiens) | 2D(S44,55D)-Hec1 -LSSmOrange | this paper | modified from 2D-Hec1-GFP from Jennifer DeLuca | |
Transfected construct (Homo sapiens) | 9D-Hec1-LSSmOrange | this paper | modified from 9D-Hec1-GFP from Jennifer DeLuca | |
Transfected construct (Homo sapiens) | INCENP-mCherry | other | Gift from Michael Lampson | |
Recombinant DNA reagent | pSpCas9(BB)−2A-GFP (pX458) | Ran et al. (2013) | Addgene: #48138 | |
Sequence-based reagent | Donor single-stranded DNA for TC tag insertion at the C-terminus of TUBB | IDT | ssDNA: cgtctctgagtatcagcagtacca ggatgccaccgcagaagaggaggaggattt cggtgaggaggccgaagaggaggcctGCT GTCCCGGCTGTTGctaaggcagagcccc catcacctcaggcttctcagttcccttagccgtc ttactcaactgcccctttcctctccctcaga; sgRNA target sequence: GAGGCCGAA GAGGAGGCCTA | |
Sequence-based reagent | Hec1 siRNA | Qiagen | Cat#: SI02653567 | |
Peptide, recombinant protein | TC-peptide | Genscript | Custom designed | Synthesized, Ac-AEEEACCPGCC-NH2 |
Commercial assay or kit | Amaxa Cell Line Nucleofector Kit V | Lonza | Cat#:VCA-1003 | |
Commercial assay or kit | Ingenio Electroporation Kit | Mirus | Cat#: MIR 50118 | |
Commercial assay or kit | Lipofectamine RNAiMax | Thermo Fisher | Cat#:13778075 | |
Chemical compound, drug | FlAsH-EDT2 | Thermo Fisher | Cat#:T34561 | |
Chemical compound, drug | 1,2-Ethanedithiol (EDT) | Alfa Aesar | Cat#:540-63-6 | |
Chemical compound, drug | ZM447439 | Enzo Life Sciences | Cat#:BML-EI373 | |
Chemical compound, drug | Paclitaxel (Taxol) | Enzo Life Sciences | Cat#:BML-T104 | |
Chemical compound, drug | 5-iodotubercidin (5-ITu) | Enzo Life Sciences | Cat#:BML-EI29 | |
Chemical compound, drug | S-Trityl-L-cysteine | Sigma Aldrich | Cat#:164739–5G | |
Chemical compound, drug | Alexa Fluor 488 | Thermo Fisher | Cat#:A20000 | |
Chemical compound, drug | Sodium 2- mercaptoethanesulfonate | Sigma Aldrich | Cat#:M1511 | |
Software, algorithm | Interactive kinetochore FLIM-FRET analysis GUI (MATLAB 2016) | This paper | http://doi.org/10.5281/zenodo.1198705; copy archived at https://github.com/elifesciences-publications/FLIM-Interactive-Data-Analysis | |
Software, algorithm | Aurora B concentration at NDC80 analysis (Python 3) | This paper | http://doi.org/10.5281/zenodo.1198702;copy archived at https://github.com/elifesciences-publications/AuroraConcentrationAnalysis | |
Software, algorithm | CAMPARI (v2) | Pappu Lab | http://campari.sourceforge.net/V2/index.html | |
Software, algorithm | Rosetta 3.8 | RosettaCommons | RRID:SCR_015701 | |
Other | 25 mm #1.5 poly-D-lysine coated round coverglass | neuVitro | Cat#:GG-25–1.5-pdl | |
Other | FluoroBrite DMEM | Thermo Fisher | Cat#:A1896701 | |
Other | Microtubule structure | Zhang et al. (2015) | PDB 3JAS | |
Other | Human NDC80 bonsai decorated tubulin dimer | Alushin et al. (2010) | PDB 3IZ0 | |
Other | mTurquoise structure | Stetten et al. (unpublished) | PDB 4B5Y |
Figure | Parameter | Mean | 95% CI |
---|---|---|---|
4A | A | 0.088 | (0.069,0.106) |
(min) | 3.26 | (1.31,5.21) | |
c | 0.089 | (0.080,0.099) | |
4D | A | 0.024 | (0.011,0.038) |
(min) | 0.50 | (−0.70,1.71) | |
c | 0.059 | (0.048,0.071) | |
4E | A | 0.17 | (0.16,0.18) |
(min) | 1.95 | (1.46,2.45) | |
c | 0.37 | (0.36,0.38) | |
4-S1D | A | 0.076 | (0.061,0.090) |
(min) | 1.12 | (0.23,2.00) | |
c | 0.56 | (0.55,0.57) |
Additional files
-
Transparent reporting form
- https://doi.org/10.7554/eLife.36392.020