Changes in microtubule overlap length regulate kinesin-14-driven microtubule sliding

Transcription

1 Supplementary information Changes in microtubule overlap length regulate kinesin-14-driven microtubule sliding Marcus Braun* 1,2,3, Zdenek Lansky* 1,2,3, Agata Szuba 1,2, Friedrich W. Schwarz 1,2, Aniruddha Mitra 1,2, Mengfei Gao 1,2, Annemarie Lüdecke 1,2, Pieter Rein ten Wolde 4,#, Stefan Diez 1,2,# 1 B CUBE - Center for Molecular Bioengineering, Technische Universität Dresden, Arnoldstraße 18, Dresden, Germany; 2 Max Planck Institute of Molecular Cell Biology and Genetics, Pfotenhauerstraße 108, Dresden, Germany; 3 Institute of Biotechnology CAS, BIOCEV, Prumyslova 595, Vestec 25250, Czech Republic. 4 AMOLF, Science Park 104, 1098 XG Amsterdam, the Netherlands * equal contribution; # correspondence: tenwolde@amolf.nl, stefan.diez@tu-dresden.de 1

2 Supplementary results Supplementary Figure 1. HSET constructs used in the study. (a) SDS-PAGE of the HSET constructs used in our experiments: [1] HSET (78 kd), [2] GFP-HSET (104 kd), [3] GFP-HSET-tail (66 kd) and [4] GFP-HSET-motor (90 kd). (b) Profile from size exclusion chromatography demonstrating that the purified GFP-HSET is monodisperse. (c) HSET slides antiparallel microtubules. Upper panel shows a kymograph representing HSET-driven sliding of a transport microtubule along a surface-immobilized template microtubule (HSET here is not GFP-labeled). The transport microtubule slowed down when sliding off the template microtubule and the overlap was retained. The lower panel shows the movement of single GFP-labeled kinesin-1 molecules along the microtubules shown in the upper panel. Kinesin-1-GFP was imaged immediately after being added to the assay, subsequent to the end of the time lapse acquisition presented in the upper panel. The vertical dotted line represents the end of the transport microtubule. The movement of plus-end directed kinesin-1-gfp molecules indicates that the HSET-stabilized microtubule overlap is in antiparallel orientation. (d) 2

3 Multichannel kymograph showing a transport microtubule (magenta) sliding on a surface-immobilized template microtubule driven by GFP-HSET (green) with proteolytically removed hexa-histidine tag (Methods) showing qualitatively identical results to those presented in Fig. 1 and Supplementary Fig. 1c (the wavy position of the template-microtubule end is due to thermal drift of the microscope stage in x/y, while the z-position is stabilized by a Zeiss autofocus focus-maintaining device). Supplementary Figure 2. HSET-driven microtubule movement slows down with increasing motor density. (a) The amount of GFP-HSET molecules in the shortening overlaps remained approximately constant as the microtubules slid apart (time t = 0 corresponds to the moment, when microtubules started to slide apart), leading to progressive increase in GFP-HSET densities in the shortening overlaps. The data represent binned and averaged values (± SD; n = 33). (b) The density of GFP-HSET increased when two microtubules slid apart. When the overlap was halved, on average about 84% of the initial amount of GFP-HSET 3

4 remained in the overlap. To minimize bleaching effects, only the initial 5 minutes of sliding (of the data presented in Fig. 2a) are shown and analyzed. Individual colors represent individual events of microtubules sliding apart. (c) Displacement-weighted velocities 1 of gliding microtubules driven by HSET molecules, immobilized on a casein-coated surface, as function of microtubule length (same data as in Fig. 2d). Gliding assays were performed at three different concentrations of HSET used for the coating of the surface: 25 nm (n = 88, red), 85 nm (n = 75, yellow) and 850 nm (n = 82, blue), which set different surface densities of the GFP-HSET molecules. Each data point represents a single gliding microtubule. At a given GFP-HSET concentration, the longer microtubules have more GFP-HSET molecules attached than the shorter ones. For microtubules longer than approximately 1.5 µm, the velocity of gliding was independent of microtubule length, suggesting that above a certain motor number, the gliding velocity was independent of the motor number but dependent on the motor density. (d) Displacement weighted velocity of gliding microtubules driven by surface-immobilized HSET-motor-domains as function of microtubule length. Gliding assays were performed at three different concentrations of GFPantibodies used for the HSET-motor-domain-coating of the coverslip surface (see Methods): 10 nm (n = 124, red), 100 nm (n = 195, yellow), 300 nm (n = 30, blue), which set different surface densities of the motor-domains coating the coverslip surface. For microtubules longer than approximately 1.5 µm, the gliding velocity was independent of microtubule length, suggesting that above a certain motor number the gliding velocity was independent of the motor number but dependent on the motor density. (e) Median microtubule gliding velocities of the data shown in d as function of the HSET-motor-domain concentration represented in box-and whisker plots. 4

5 Supplementary Figure 3. Interaction of full-length and truncated HSET constructs with microtubules. (a) Mean square displacement (MSD) of fulllength GFP-HSET in presence of ATP on a single microtubule (n = 22) (red) and in a microtubule overlap (n = 14) (blue). Weighted linear fits provide diffusion constants of 0.39 ± 0.01 μm 2 s 1 (mean ± SEM) and ± μm 2 s 1, respectively (95% confidence band in grey). (b) MSD of full-length GFP-HSET in microtubule overlaps for two different motor densities in the presence of ATP. Circles (n = 14) represent data where the average distance between neighboring molecules (calculated as the length of the overlap divided by the average number of molecules in the overlap) was L > 500 nm. On the two second timescale we observe free diffusion (indicated by the linear fit). A weighted linear fit (red) provides a diffusion constant of ± μm 2 s 1. Triangles (n = 22) represent data for 250 < L < 430 nm. A weighted fit (blue) of the model of diffusion confined in boundaries (equation 11a from Kusumi at al. 2 ) provides a 5

6 diffusion constant of ± μm 2 s 1 s and a distance between the boundaries of L B = 354 ± 25 nm (95% confidence band in grey). The distance between the boundaries corresponds to the average distance between neighboring molecules suggesting that the molecules of HSET cannot pass each other in the microtubule overlap. (c) MSD of GFP-HSET-tail on single microtubules (n= 22). A weighted linear fit (red) provides a diffusion constant of 0.41 ± 0.02 (95% confidence band in grey). (d) Distribution of dwell times of the GFP-HSET-motor with microtubules (estimated from single molecule trajectories as in Fig. 3e, pooled data from single microtubules and overlaps) in the presence of ATP. A single-exponential fit provides an unbinding rate of 4.1 ± 0.9 s -1 (n = 473 events, 95 % confidence interval). Supplementary Figure 4. The number of GFP-HSET molecules in the overlap is constant during ATP-independent overlap expansion. The amount of GFP-HSET in microtubule overlaps during sliding was estimated from the integrated GFP intensity in the overlap in experiments as presented in Figure 4. Data points represent binned and averaged values (± SD; n = 8 microtubule overlaps). 6

7 Supplementary Figure 5. Simulations of HSET-driven microtubule sliding. (a) Simulation of the amount of HSET molecules in microtubule overlaps stays approximately constant during the time when the microtubules slide apart, leading to an increase in the HSET density in the shortening overlap. Time t = 0 corresponds to the moment, when microtubules started to slide apart (n = 15 microtubule pairs, mean ± SD). (b) Simulation of two microtubules sliding apart. Shown are the time traces of the overlap length and number of HSET molecules in the overlap. Time t = 0 corresponds to the moment, when microtubules started to slide apart. All simulation parameters where the same as employed in Fig. 5 and 6 and as described in the Methods (summarized in Supplementary Table 1). HSET motors are retained in the shortening overlap, leading to an increase in HSET density. Consequently, microtubule sliding slows down as the overlap length decreases and resulting in the retainment of the overlap; as 7

8 experimentally observed in Fig. 1. (c and d) Simulation of two microtubules sliding apart. All simulation parameters where the same as employed in Fig. 5 and 6, except that the binding rate of the tethered motor-domain k motor(tethered) on was reduced from 60 s -1 to 0.5 s -1 (c) or the unbinding rate of the tail-domain was increased from 0.01 s -1 to 10 s -1 (d). Under both conditions, HSET motors were not retained in the overlap. Consequently, the microtubules fully slid apart within less then 30 s. Supplementary Figure 6. Increasing the ionic strength of the assay buffer abolishes the retention of HSET in microtubule overlaps and the slowdown of sliding. (a) Kymograph showing single full-length GFP-HSET molecules at concentration of 0.15 nm diffusing along single microtubules in the presence of 8

9 ATP at increasing ionic strength. The time of GFP-HSET interaction with microtubules is markedly decreased (compare to Fig. 3a). (b) Multichannel kymograph showing GFP-HSET (green) -driven microtubule (magenta) sliding at increased ionic strength (at 150 mm KCl). Microtubules slide at approximately constant velocity until they fully separate. (c) The density of GFP-HSET does not increase when two microtubules slide apart at increased ionic strength (at 150 mm KCl). The GFP-HSET density and the overlap length were normalized to unity at the moment when the microtubules started to slide apart. Blue data points indicate combined measurements of 33 microtubules sliding apart (events such as presented in panel b). Red crosses indicate binned and averaged values (± SD). Compare to Fig. 2a. Supplementary Table 1. Diffusion constant of tail domain 0.4 µm 2 /s measured Maximum motor force 0.1 pn Furuta et al. 3 * Maximum motor velocity 0.1 µm/s Furuta et al. 3 Spring stiffness 5x10 4 k B Tµm -2 Lansky et al. 4 ** k motor(tethered) on 60 s -1 fit*** k motor(solution) off 4 s -1 measured k tail(solution) off 0.01 s -1 measured Free diffusion of transport MT 0.01 µm 2 /s Hunt et al. 5 **** Lattice spacing 0.01 µm length of tubulin dimer * For simulation of the experiment in the presence of ADP (absence of ATP), the maximum motor force was set to zero. ** The spring stiffness was set to a value estimated earlier for a different microtubule crosslinker. Our results are qualitatively independent of the spring stiffness in the range between ~ k B Tµm -2. For spring stiffnesses higher than ~ 10 5 k B Tµm -2 we did not observe any sliding. *** k motor(tethered) on was chosen such that the diffusion constant of a single molecular motor bound between two filaments was as observed experimentally. **** Free diffusion of an unbound transport microtubule in solution. 9

10 Supplementary references. 1. Alper, J. D., Tovar, M. & Howard, J. Displacement-Weighted Velocity Analysis of Gliding Assays Reveals that Chlamydomonas Axonemal Dynein Preferentially Moves Conspecific Microtubules. Biophys J 104, (2013). 2. Kusumi, A., Sako, Y. & Yamamoto, M. Confined lateral diffusion of membrane receptors as studied by single particle tracking (nanovid microscopy). Effects of calcium-induced differentiation in cultured epithelial cells. Biophys J 65, (1993). 3. Furuta, K. et al. Measuring collective transport by defined numbers of processive and nonprocessive kinesin motors. Proc Natl Acad Sci USA 110, (2013). 4. Lansky, Z. et al. Diffusible crosslinkers generate directed forces in microtubule networks. Cell 160, (2015). 5. Hunt, A. J., Gittes, F. & Howard, J. The force exerted by a single kinesin molecule against a viscous load. Biophys J 67, (1994). 10