Journal: eLife

Article Title: Kinesin-4 KIF21B is a potent microtubule pausing factor

doi: 10.7554/eLife.24746

Figure Lengend Snippet: ( A ) Histograms of fluorescence intensities at the initial moment of observation of single molecules of the indicated proteins immobilized on coverslips (symbols) and the corresponding fits with lognormal distributions (lines). n = 3107, 5802 and 4674 molecules and fluorophore density was 0.15, 0.28 and 0.23 µm −2 for GFP, GFP-EB3 and KIF21B-MD-CC1-GFP proteins. ( B ) Representative photobleaching time traces of GFP, GFP-EB3 and KIF21B-MD-CC1-GFP individual molecules (background subtracted). ( C ) Kymographs illustrating the dynamics of MTs grown in vitro in the presence of 20 nM mCherry-EB3 alone, with 10 nM purified GFP or with 2 and 10 nM KIF21B-MD-CC1-GFP. Zooms of the boxed areas are shown on the right. Kymographs were generated from movies acquired using a Photometrics Evolve 512 EMCCD camera (Roper Scientific) (stream acquisition, exposure time 500 ms). ( D ) Histograms of fluorescence intensities of single GFP molecules immobilized on coverslips and KIF21B-MD-CC1-GFP moving on MTs in a separate chamber on the same coverslip (symbols) and the corresponding fits with lognormal distributions (lines). n = 4815 and 1381 molecules; fluorophore density was 0.16 and 0.09 µm −2 for GFP and KIF21B-MD-CC1-GFP proteins (for the latter, MT-containing regions were manually selected for analysis). Dashed lines show corresponding relative median values. ( E ) Histogram of KIF21B-MD-CC1-GFP velocities in vitro is shown with black bars. Red line shows fitting with a normal distribution. n = 675 in two independent experiments. ( F ) Histogram of KIF21B-MD-CC1-GFP run lengths in vitro is shown with black bars. Red line shows fitting with an exponential distribution. n = 675 in two independent experiments. ( G ) Upper panel - quantification of the MT growth rate illustrated in C. n = 71 for control, n = 65 for purified GFP, n = 71, 67 and 54 for 2, 5 and 10 nM KIF21B-MD-CC1-GFP, respectively. Lower panel shows quantification of the MT growth rate with 15 µM tubulin alone or with 10 nM purified GFP or with 2, 5 and 10 nM KIF21B-MD-CC1-GFP as illustrated in . n = 67 for control, n = 57 for purified GFP, n = 71, 66 and 80 for 2, 5 and 10 nM KIF21B-MD-CC1-GFP, respectively, two independent experiments, p<0.0001, Mann-Whitney U test (indicated by asterisks). ( H ) Quantification of the MT growth rate with different concentrations of tubulin along with 20 nM EB3 in the absence and presence of 2 nM KIF21B-MD-CC1-GFP as illustrated in . ND; not determined, n = 71 for all conditions. two independent experiments, p<0.0001, Mann-Whitney U test (indicated by asterisks). ( I ) Kymographs illustrating the dynamics of MTs grown in vitro in the presence of 20 nM EB3 and 1 nM KIF21B-MD-CC1-GFP. Zoom of the boxed area is shown on the right. Kymographs were generated from a movie acquired using Photometrics Evolve 512 EMCCD camera (Roper Scientific) (stream acquisition, exposure time 100 ms). ( J ) Distribution of EB3 fluorescence intensity fluctuations over time (normalized to its maximum value during a course of a growth event) at MT tip in the presence of 1 nM GFP or 1 nM KIF21B-MD-CC1-GFP (solid line) with Gaussian fit (dotted line). n = 25 in both cases. Thick dotted lines show the peak of the Gaussian fitting. MT dynamics assay was performed in the presence of 15 µM tubulin, 20 nM EB3 and 1 nM GFP or 1 nM KIF21B-MD-CC1-GFP in the separate chambers of the same coverslip. ( I ) Plot of the mean and SD values of Gaussian fits shown in . DOI: http://dx.doi.org/10.7554/eLife.24746.006 10.7554/eLife.24746.007 Figure 2—source data 1. An excel sheet with numerical data on the quantification of KIF21B-MD-CC1-GFP dimer analysis, photobleaching-step analysis, velocities, run length, effects on MT growth rate and distribution of EB3 fluorescence intensity represented as plots in . DOI: http://dx.doi.org/10.7554/eLife.24746.007

Article Snippet: In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific).

Techniques: Fluorescence, In Vitro, Purification, Generated, Control, MANN-WHITNEY

Journal: eLife

Article Title: Kinesin-4 KIF21B is a potent microtubule pausing factor

doi: 10.7554/eLife.24746

Figure Lengend Snippet: Kymographs were generated from the movies of 600 frames (stream acquisition, exposure time 500 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific). DOI: http://dx.doi.org/10.7554/eLife.24746.009

Article Snippet: In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific).

Techniques: Generated

Journal: eLife

Article Title: Kinesin-4 KIF21B is a potent microtubule pausing factor

doi: 10.7554/eLife.24746

Figure Lengend Snippet: ( A ) Kymographs illustrating the in vitro dynamics of MTs grown in the presence of different concentrations of tubulin along with 20 nM EB3 in the absence and presence of 2 nM KIF21B-MD-CC1-GFP. Kymographs were generated from the movies of 600 frames (stream acquisition, exposure time 500 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific). ( B ) Quantification of the velocity of MT minus end growth in the presence of 15 µM tubulin and 20 nM mCherry-EB3 without (control) or with 10 nM KIF21B-MD-CC1-GFP. n = 50 in both cases. DOI: http://dx.doi.org/10.7554/eLife.24746.010 10.7554/eLife.24746.011 Figure 2—Figure Supplement 3—Source Data 1. An excel sheet with numerical data on the quantification of the MT minus end growth rates represented as plot in . DOI: http://dx.doi.org/10.7554/eLife.24746.011

Article Snippet: In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific).

Techniques: In Vitro, Generated, Control

Journal: eLife

Article Title: Kinesin-4 KIF21B is a potent microtubule pausing factor

doi: 10.7554/eLife.24746

Figure Lengend Snippet: ( A ) Kymographs showing the behavior of KIF21B in in vitro reconstitution assays on dynamic MTs grown from Rhodamine-tubulin-labeled seeds in the presence of 15 µM tubulin, 100 nM mCherry-EB3 (red) and 3 nM KIF21B-FL-GFP (green). Kymographs were generated from movies acquired using a Photometrics Evolve 512 EMCCD camera (Roper Scientific) (stream acquisition, exposure time 500 ms). Pausing and catastrophe events are indicated by arrows. ( B ) Quantification of MT seed length-dependent blocking of MT growth by 0.5 nM KIF21B- FL-GFP in the presence of 20 nM mCherry-EB3. 188 MT seeds of different lengths were analyzed in four independent experiments. ( C ) Kymographs illustrating pausing events induced by KIF21B-FL-GFP (0.5 nM) on dynamic MTs in vitro in the presence of 15 µM tubulin, 20 nM mCherry-EB3, 3% Rhodamine-tubulin. MTs were grown from GMPCPP-stabilized seeds labeled with Rhodamine-tubulin. Kymographs were generated from the movies acquired using a CoolSNAP HQ2 CCD camera (Roper Scientific) with a 1.2-s interval between frames and an exposure time of 100 ms. The rightmost panels show tracked positions of the kinesins and MT tips together with the fluorescence intensity of the kinesins over time for the corresponding kymographs. ( D ) Kymographs illustrating various events induced by KIF21B-FL-GFP (0.5 nM) on dynamic MTs in vitro in the presence of 15 µM tubulin, 20 nM mCherry-EB3 and 3% Rhodamine-tubulin. MTs were grown from GMPCPP-stabilized seeds labeled with Rhodamine-tubulin. Different events are indicated by arrows. Kymographs were generated from movies acquired as described for . ( E ) Quantification of different events observed after KIF21B-FL-GFP (0.5 nM) reaches a growing MT plus end, as illustrated in C and D. n = 132 events, four independent experiments were analyzed. ( F ) Kymograph illustrating a long pause event induced by multiple KIF21B-FL-GFP molecules on dynamic MTs in vitro in the presence of 15 µM tubulin, 20 nM mCherry-EB3 and 3% Rhodamine-tubulin in solution. Kymographs are generated from a movie acquired as described for . ( G ) Kymographs illustrating the effects of KIF21B-FL-GFP (0.5 nM) on dynamic MTs in vitro in the presence of 30 µM tubulin with 3% Rhodamine-tubulin and 20 nM mCherry-EB3. MTs were grown from GMPCPP-stabilized seeds labeled with Rhodamine-tubulin. The arrows show the position of KIF21B at the site of MT pause and the asterisk indicates the growing MT tip beyond the position of KIF21B binding; note that the slope of the kymograph after KIF21B attachment is less steep than before, indicating that the growth rate is reduced. Kymographs are generated from movies acquired as described for . ( H, I ) Quantification of MT growth rate and catastrophe frequency in vitro in the presence of 15 or 30 µM tubulin with 20 nM mCherry-EB3 alone or together with 0.5 nM KIF21B-FL-GFP. MTs were grown in the presence of 3% Rhodamine-tubulin. For 15 µM tubulin, n = 71 for control and n = 100 for KIF21B-FL-GFP, three independent experiments. For 30 µM tubulin, n = 71 for control and n = 80 for KIF21B-FL-GFP, three independent experiments, p<0.0001 Mann-Whitney U test (indicated by asterisks). DOI: http://dx.doi.org/10.7554/eLife.24746.012 10.7554/eLife.24746.013 Figure 3—source data 1. An excel sheet with numerical data on the quantification of KIF21B-FL seed blocking activity, pause induction, effects on MT growth rate and catastrophe frequency and outcomes of KIF21B-FL-GFP arrival at MT plus ends represented as plots in . DOI: http://dx.doi.org/10.7554/eLife.24746.013

Article Snippet: In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific).

Techniques: In Vitro, Labeling, Generated, Blocking Assay, Fluorescence, Binding Assay, Control, MANN-WHITNEY, Activity Assay

Journal: eLife

Article Title: Kinesin-4 KIF21B is a potent microtubule pausing factor

doi: 10.7554/eLife.24746

Figure Lengend Snippet: ( A ) GFP intensity analysis of kinesins during MT pausing events. Values are normalized to the GFP intensity of proteins immobilized on the same coverslip in areas devoid of MTs. Data are from two independent experiments. ( B ) Kymographs showing the behavior of KIF5B-560-GFP in an in vitro reconstitution assay on dynamic MTs grown from Rhodamine-tubulin-labeled seeds in the presence of 15 µM tubulin, 20 nM mCherry-EB3 (red) and 5 nM KIF5B-560-GFP (green). ( C ) Histograms of fluorescence intensities of single GFP molecules immobilized on coverslips (initial moment of observation of single molecules) or KIF5B-560-GFP moving on MTs in a separate chamber on the same coverslip (symbols) and the corresponding fits with lognormal distributions (lines). n = 452 and 2040 molecules; fluorophore density was 0.05 and 0.09 µm −2 for GFP and KIF5B-560-GFP proteins (for the latter, MT-containing regions were manually selected for analysis). Dashed lines show the corresponding relative median values. ( D ) Kymographs showing the behavior of 5 nM KIF5B-560-GFP (moving on dynamic MTs, green) and 0.5 nM KIF21B-FL-GFP (moving on seeds, green) in an in vitro reconstitution assay with MTs grown from Rhodamine-tubulin-labeled seeds in the presence of 15 µM tubulin and 20 nM mCherry-EB3 (red). Histograms illustrate fluorescence intensities of KIF5B-560-GFP moving on MTs and KIF21B-FL-GFP moving on seeds in separate chambers on the same coverslip (symbols) and the corresponding fits with lognormal distributions (lines). n = 5123 and 8728 molecules; fluorophore density was 0.15 and 0.18 µm −2 for KIF5B-560-GFP and KIF21B-FL-GFP proteins (MT-containing regions were manually selected for analysis). Ratio of the corresponding median values is also indicated. ( E ) Kymographs illustrating KIF21B-FL-GFP (0.5 nM) movement on seeds and dynamic MTs, histograms of the corresponding fluorescence intensities measured within the same sample (symbols) and their fits with lognormal distributions (lines). Median values are also indicated. ( F–H ) Kymographs illustrating motility of KIF5B-560-GFP (5 nM) on dynamic MTs and KIF21B-FL-GFP (0.5 nM) moving from an MT seed to the dynamic MT lattice and inducing a MT pause in an in vitro reconstitution assay with MTs grown from Rhodamine-tubulin-labeled seeds in the presence of 15 µM tubulin and 20 nM mCherry-EB3 (red). Histograms show fluorescence intensities of motile KIF5B-560-GFP molecules and KIF21B-FL-GFP inducing a MT pause in separate chambers on the same coverslip (symbols) and the corresponding fits with lognormal distributions (lines). Median values are also indicated. In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific). Positions of seeds in each kymograph are indicated. ( I ) Fitted peak intensity time trace for the trajectory of a moving KIF21B-FL-GFP molecule from the event shown in . Dashed lines correspond to the scaled values of median fluorescence fitted peak intensity of KIF5B-560-GFP molecules moving on dynamic MT in a parallel chamber on the same coverslip. DOI: http://dx.doi.org/10.7554/eLife.24746.023 10.7554/eLife.24746.024 Figure 5—source data 1. An excel sheet with numerical data on the quantification of KIF21B-FL intensity during MT pausing events, KIF5B-560 dimer analysis and comparison of fluorescence intensities of KIF5B-560 with KIF21B-FL represented as plots in . DOI: http://dx.doi.org/10.7554/eLife.24746.024 10.7554/eLife.24746.025 Figure 5—figure Supplement 2—Source data 1. An excel sheet with numerical data on the quantification of photobleaching traces of KIF21B-FL-GFP represented as plots in . DOI: http://dx.doi.org/10.7554/eLife.24746.025

Article Snippet: In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific).

Techniques: In Vitro, Reconstitution Assay, Labeling, Fluorescence, Generated, Comparison

Journal: eLife

Article Title: Kinesin-4 KIF21B is a potent microtubule pausing factor

doi: 10.7554/eLife.24746

Figure Lengend Snippet: ( A ) Histograms of fluorescence intensities at the initial moment of observation of single molecules of the indicated proteins immobilized on coverslips (symbols) and the corresponding fits with lognormal distributions (lines). n = 3907, 5002, 6725 and 6943 molecules; fluorophore density was 0.19, 0.24, 0.30 and 0.33 µm −2 for GFP, GFP-EB3, KIF21B-FL-ΔrCC-GFP and KIF21B-MD-CCΔrCC-GFP proteins. Insets show representative photobleaching traces of individual molecules (background subtracted). ( B ) Kymographs illustrating the behavior of the indicated deletion mutants of KIF21B at 3 nM concentration on dynamic MTs in the presence of 100 nM mCherry-EB3. GMPCPP-stabilized MT seeds were labeled with Rhodamine-tubulin (lines below kymographs). Kymographs were generated from the movies acquired using Photometrics Evolve 512 EMCCD (Roper Scientific) camera (stream acquisition with an exposure time of 500 ms). ( C ) Quantification of the velocity of KIF21B-FL-GFP and the deletion mutants on seeds and freshly polymerized MT lattices, shown in and . Seed: n = 295 for KIF21B-FL-GFP, n = 195 for KIF21B-FL-ΔrCC-GFP, n = 434 for KIF21B-MD-CCΔrCC-GFP; lattice: n = 131 for KIF21B-FL, n = 133 for KIF21B-FL-ΔrCC-GFP, n = 434 for KIF21B-MD-CCΔrCC-GFP. Data are from two or three independent experiments. Values significantly different from each other are indicated by asterisks, p<0.0001, Mann-Whitney U test. ( D ) Kymographs illustrating the interaction of purified GFP-L-WD40 (100 nM) with dynamic MTs grown from Rhodamine-tubulin labeled GMPCPP- or taxol-stabilized seeds (as indicated) in the presence of 20 nM mCherry-EB3. Kymographs were generated from the movies acquired in stream acquisition mode with an exposure time of 500 ms using Photometrics Evolve 512 EMCCD camera (Roper Scientific). ( E ) Kymographs illustrating the behavior of KIF21B deletion mutants on dynamic MTs in the presence of 20 nM mCherry-EB3 and 3% Rhodamine-tubulin. Pauses and KIF21B detachment from a depolymerizing MT end are indicated by arrows and asterisks, respectively. Arrowheads indicate kinesin detachment from the growing MT tip. Kymographs were generated from the movies acquired using CoolSNAP HQ2 CCD camera (Roper Scientific) with a 1.2-s interval between frames and an exposure time of 100 ms. ( F ) Quantification of MT growth rate in vitro in the presence of 15 µM tubulin with 20 nM mCherry-EB3 alone (n = 71) or together with 3 nM KIF21B-FL-ΔrCC-GFP (n = 79) or KIF21B-MD-CCΔrCC-GFP (n = 79). MTs were grown in the presence of 3% Rhodamine-tubulin. two independent experiments. ( G ) Quantification of different events observed after KIF21B-FL or its mutants reach a growing MT plus end. Data shown in are included here for comparison. n = 501 for KIF21B-FL-ΔrCC-GFP, n = 647 for KIF21B-MD-CCΔrCC-GFP. Data are from at least two independent experiments. ( H ) Percentage of pausing events induced by a single event of kinesin arrival from all detected pauses. Total number of pausing events: n = 51 for KIF21B-FL-GFP, n = 46 for KIF21B-FL-ΔrCC-GFP, n = 22 for KIF21B-MD-CCΔrCC-GFP. Data are from at least two independent experiments. ( I ) Quantification of the duration of MT pausing induced by a single kinesin arrival event at the growing MT plus end. n = 36 for KIF21B-FL-GFP, n = 8 for KIF21B-FL-ΔrCC-GFP, n = 3 for KIF21B-MD-CCΔrCC-GFP. Data are from at least two independent experiments.**p<0.0001, *p<0.0004 Mann-Whitney U test. ( J ) Model for the regulation of KIF21B motility and pause induction by the tail domain. In solution, KIF21B motor domains are inhibited by the regulatory region, while the WD40 domains are available for the interaction with MTs; WD40 domains show preference for the GMPCPP-stabilized seeds (red). After binding to seeds, KIF21B becomes activated and can walk to the plus end; it is likely that both the WD40 and the CC2 region contribute to MT binding. The kinesin can transfer from the seed to the freshly polymerized MT lattice; the interaction of the CC2 but not of the WD40 with the lattice promotes motor processivity. At the tip, the conversion to the autoinhibited conformation and the WD40 domain can prevent KIF21B from stepping off the MT plus end. This allows the motor to prevent both elongation and shortening of a small number of protofilaments with which it interacts. The remaining protofilaments might undergo short excursions of growth and shrinkage (upper panel); alternatively, they might elongate for some time and such an incomplete MT will be prone to bending and catastrophe (lower panel). DOI: http://dx.doi.org/10.7554/eLife.24746.035 10.7554/eLife.24746.036 Figure 7—source data 1. An excel sheet with numerical data on the quantification of KIF21B mutants dimer analysis, photobleaching step analysis, velocities on seeds and MT lattices, MT growth rate in vitro and outcomes of the arrival of KIF21B mutants at MT plus ends, represented as plots in . DOI: http://dx.doi.org/10.7554/eLife.24746.036

Article Snippet: In all the conditions, kymographs were generated from movies of 1500 frames (stream acquisition, exposure time 100 ms) using Photometrics Evolve 512 EMCCD camera (Roper Scientific).

Techniques: Fluorescence, Concentration Assay, Labeling, Generated, MANN-WHITNEY, Purification, In Vitro, Comparison, Binding Assay