Acetylated Microtubules Are Preferentially Bundled Leading to Enhanced Kinesin-1 Motility

Abstract

The motor proteins kinesin and dynein transport organelles, mRNA, proteins, and signaling molecules along the microtubule cytoskeleton. In addition to serving as tracks for transport, the microtubule cytoskeleton directs intracellular trafficking by regulating the activity of motor proteins through the organization of the filament network, microtubule-associated proteins, and tubulin posttranslational modifications. However, it is not well understood how these factors influence motor motility, and in vitro assays and live cell observations often produce disparate results. To systematically examine the factors that contribute to cytoskeleton-based regulation of motor protein motility, we extracted intact microtubule networks from cells and tracked the motility of single fluorescently labeled motor proteins on these cytoskeletons. We find that tubulin acetylation alone does not directly affect kinesin-1 motility. However, acetylated microtubules are predominantly bundled, and bundling enhances kinesin run lengths and provides a greater number of available kinesin binding sites. The neuronal MAP tau is also not sensitive to tubulin acetylation, but enriches preferentially on highly curved regions of microtubules where it strongly inhibits kinesin motility. Taken together, these results suggest that the organization of the microtubule network is a key contributor to the regulation of motor-based transport.

Figure 1

The motility of single kinesin motors is enhanced by MT bundling. (A) To isolate endogenous MT networks, cells were treated with detergent to remove the membrane whereas the MT network was stabilized with Taxol. Purified fluorescently tagged motor proteins were added with ATP onto the isolated cytoskeleton. (B) (Left) Time-lapse movies of GFP-conjugated kinesin-1 motility are captured with TIRF microscopy (Movie S1). (Middle) After the motility assay, MTs were fixed and stained with tubulin antibodies (acetylated tubulin shown in red and α-tubulin in green). Note that the 6-11B-1 antibody recognizes both acetylated and deacetylated MTs, but not filaments that have never been acetylated (59). (Right) Trajectories of kinesin-1 were tracked with subpixel resolution (∼25 nm) using TrackMate. (C) The MT cytoskeleton is shown in a living cell (pre-extraction) and after the extraction treatment, using SiR-tubulin (Spirochrome). (D) Zoomed images from yellow box in (B). The SD map of kinesin-1 was generated from the time-lapse movie. Single-molecule tracking produced the trajectories of kinesin (shown in different colors). Yellow lines on the SD map represent the regions of interest for drawing kymographs shown in (E). (E) Kymographs show a higher number of binding events for bundled MTs that are either acetylated or nonacetylated. (F) Kinesin-1 motors show increased run lengths, (G) binding times, and (H) number of runs on acetylated and nonacetylated bundles compared to the single MTs. For each cell, the values are normalized to the motility on nonacetylated single MTs (baseline). F–H) (Top plots) Colors represent different cells for a total of eight cells over seven experiments. Large black circles and error bars indicate the mean ± SE for all cells. The raw means for each cell for all plots are shown in Fig. S4A. (F and G) (Bottom) The normalized distribution of kinesin run lengths (F) and binding times (G) for all cells (mean in red, SEM in yellow). The grayscale represents the density of points with black as the highest density. (I) Off-axis displacement of runs of kinesin-1 does not differ on the single versus bundled MTs (p = 0.08) (SD). Acetylated single: n = 362, acetylated bundled: n = 4648, nonacetylated single: n = 2070, nonacetylated bundled: n = 1976 total runs (all cells, runs > 50 nm). Full-factorial two-way ANOVA was used to determine the effect of acetylation (p = 0.09) and bundling (p < 0.001) on kinesin-1 log-transformed run lengths (interaction term: p = 0.46), as well as on binding times. Posthoc Tukey test then compared the differences among the four MT populations as shown on the bottom plots: ^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001; N.S., no significance. To see this figure in color, go online.

Figure 2

Acetylated MTs are often bundled in COS-7 fibroblast cells. (A) Acetylated MTs (left) colocalize with MT bundles as indicated by intense tubulin staining (right). Single MTs show low signal intensity (see Fig. S2 for detailed description of characterizing single and bundled MTs). The majority of acetylated MTs are bundled (red arrows indicate bundles). The superresolution image generated by kinesin binding on MTs shows that bundles are composed of two or more MT filaments (Fig. S2). (B) Lines were drawn on acetylated and nonacetylated MTs using the images in (A), which were then categorized into single or bundled MTs. (C) The fraction of acetylated and nonacetylated MTs in bundles (by length) were quantified as described in (B) for nonextracted cells (n = 3) and extracted cells (n = 3). An average of 95 ± 2 (mean ± SE)% of all acetylated MTs were found to be bundled in cells, compared to 43 ± 3% of nonacetylated MTs. The fraction of bundled MTs is similar in nonextracted and extracted cells (ANOVA, acetylated: p = 0.96, nonacetylated: p = 0.06). To see this figure in color, go online.

Figure 3

Tau inhibits kinesin-1 motility and induces more frequent pausing. (A) At a concentration of 1 nM, individual tau-3RS patches (left) are seen binding and diffusing along MTs as shown on the SD map (middle) (Movie S5). Kymographs (right) show that tau exists in stable (kymograph 1 generated from yellow line 1) and dynamic populations (kymographs 1 and 2). (B) Kinesin-1 motility assays were performed on extracted MT networks in the absence (− tau) and presence of 1 nM tau (+ tau). Kinesin-1 kymographs show decreased displacement on both single and bundled tau-decorated MT populations. Kymographs also show a higher occurrence of pausing (yellow arrows) and immobile kinesin motors (red arrows) in the presence of tau. (C) Kinesin run lengths are decreased in the presence of tau for all four MT populations (p < 0.01), whereas (D) binding times are increased (p < 0.01). Colors represent different cells (four cells over three experiments, which were also used for Fig. 1 represented with the same colors). The mean of all experiments and the SE of the mean is shown in black symbols. For each cell, experimental means (C–E, left side) and distributions (C and D, right side) are normalized to nonacetylated single MTs without tau (baseline). Raw means are shown in Fig. S4E. (E) The number of kinesin binding events was unaffected by tau (p > 0.40). (F) Kinesin average velocities decreased in the presence of tau (p < 0.01). (G) Off-axis displacement of kinesin-1 on tau-decorated MTs does not differ on single versus bundled MTs (p = 0.12) (SD). No tau, acetyl single: n = 148; acetyl bundled: n = 1359; nonacetyl single: n = 748; nonacetyl bundled: n = 1158 total runs (all cells, runs > 50 nm). With tau, acetyl single: n = 89; acetyl bundled: n = 1493; nonacetyl single: n = 1214; nonacetyl bundled: n = 1932 total runs (all cells, runs > 50 nm). Three-way ANOVA tested the effect of tau, acetylation, and bundling as well as their interactions (p > 0.25 for all interactions) on log-transformed normalized kinesin run length and binding time and posthoc Tukey test assessed the differences between the MT populations. To see this figure in color, go online.

Figure 4

Tau accumulates on curved regions of MTs. (A) Tau enriches itself on MT segments with high curvature. (B) Time-lapse frame of 1 nM tau and SD map of time-lapse movie. (Left) Given here is zoom of area in yellow square and its kymograph. At low concentrations, tau exhibits longer residence times on curved MT segments of MTs (yellow arrows). (C) Given here are kymographs of kinesin-1 (green) and 1 nM tau (red) representing the area of MT filaments left and right of the curvature’s peak at the center (shown by yellow arrow). Kinesin-1 runs (shown with red arrow) end near the apex of the curve, corresponding with stably bound tau patches. To see this figure in color, go online.