Roles of dynein and dynactin in early endosome dynamics revealed using automated tracking and global analysis

Abstract

Microtubule-dependent movement is crucial for the spatial organization of endosomes in most eukaryotes, but as yet there has been no systematic analysis of how a particular microtubule motor contributes to early endosome dynamics. Here we tracked early endosomes labeled with GFP-Rab5 on the nanometer scale, and combined this with global, first passage probability (FPP) analysis to provide an unbiased description of how the minus-end microtubule motor, cytoplasmic dynein, supports endosome motility. Dynein contributes to short-range endosome movement, but in particular drives 85-98% of long, inward translocations. For these, it requires an intact dynactin complex to allow membrane-bound p150(Glued) to activate dynein, since p50 over-expression, which disrupts the dynactin complex, inhibits inward movement even though dynein and p150(Glued) remain membrane-bound. Long dynein-dependent movements occur via bursts at up to ∼8 µms(-1) that are linked by changes in rate or pauses. These peak speeds during rapid inward endosome movement are still seen when cellular dynein levels are 50-fold reduced by RNAi knock-down of dynein heavy chain, while the number of movements is reduced 5-fold. Altogether, these findings identify how dynein helps define the dynamics of early endosomes.

Figure 1. PolyParticleTracker tracks early endosome movement accurately.

Movies of GFP-Rab5 were recorded for 1000 frames at 28 frames s⁻¹ in HeLaM cells: (A) control cell; (B) DHC1 depleted cell; (C) p50 expressing cell; (D) CC1 expressing cell; (E) nocodazole-treated cell. PolyParticleTracker was applied to each movie, and the resulting tracks overlaid on the first movie frame (bar = 10 µm).

Figure 2. First Passage Probability measurements allow a global description of early endosome movement.

Tracking data from 5 movies in HeLaM cells for each condition were used for FPP analysis of GFP-Rab5 in: (A) control cells; (B) DHC1 depleted cells; (C) p50 expressing cells; (D) CC1 expressing cells; (E) nocodazole-treated cells. For each length of passage, L (see key), the probability of that passage occurring at a given average speed is plotted.

Figure 3. FPP analysis identifies distinct short and long-range endosome dynamics.

FPP measurements of endosome movement for each condition (see key) in HeLaM cells (A), or RPE cells (B), were plotted to show the mean passage speed for given passage lengths.

Figure 4. Early endosomes move in fast bursts, irrespective of cellular dynein levels.

Representative tracks >2 µm in control RPE cells (A–B), DHC1-depleted RPE cells (C–D), control HeLaM cells (E–F), and DHC1-depleted HeLaM cells (G–H) were smoothed and plotted as displacement versus time (red lines). Constant speed segments were identified manually, colored according to the speed chart, and overlaid over the raw tracks (insets). The beginning (spots) and end (crosses) of each track is highlighted (inset), and the speed and duration of each segment is shown.

Figure 5. Effects of dynactin inhibition and DHC1 depletion on long-range endosome motility.

(A) Histogram showing the % of total tracks that are >2 µm under each condition, in either direction. (B) Membrane association of dynein and dynactin components in extracts from transfected cells. TfR (transferrin receptor) is a membrane control and Alix is largely cytosolic. Note that membrane: cytosol loading is ∼4∶1. (C) Histogram showing average speeds of >2 µm tracks.