Local cytoskeletal and organelle interactions impact molecular-motor- driven early endosomal trafficking

Abstract

Background: In the intracellular environment, motor-driven cargo must navigate a dense cytoskeletal network among abundant organelles.

Results: We investigated the effects of the crowded intracellular environment on early endosomal trafficking. Live-cell imaging of an endosomal cargo (endocytosed epidermal growth factor-conjugated quantum dots) combined with high-resolution tracking was used to analyze the heterogeneous motion of individual endosomes. The motile population of endosomes moved toward the perinuclear region in directed bursts of microtubule-based, dynein-dependent transport interrupted by longer periods of diffusive motion. Actin network density did not affect motile endosomes during directed runs or diffusive interruptions. Simultaneous two-color imaging was used to correlate changes in endosomal movement with potential obstacles to directed runs. Termination of directed runs spatially correlated with microtubule-dense regions, encounters with other endosomes, and interactions with the endoplasmic reticulum. During a subset of run terminations, we also observed merging and splitting of endosomes, deformation of the endoplasmic reticulum, and directional reversals at speeds up to 10-fold greater than characteristic in vitro motor velocities. These observations suggest that endosomal membrane tension is high during directed run termination.

Conclusions: Our results indicate that the crowded cellular environment significantly impacts the motor-driven motility of organelles. Rather than simply acting as impediments to movement, interactions of trafficking cargos with intracellular obstacles may facilitate communication between membrane-bound compartments or contribute to the generation of membrane tension necessary for fusion and fission of endosomal membranes or remodeling of the endoplasmic reticulum.

Figure 1

Motility in the early endosomal population A. Early endosomes move in rapid bursts towards the perinuclear region interrupted by periods of little net movement. Kymographs of early endosomes were contrast-inverted and background-subtracted. See also Movie S1. Scale bars: 2 s, 5 μm. B. Tracking of GFP-Rab5 movements for analysis. GFP-Rab5-positive early endosomes (single frame in top panel). Bottom panel: maximum intensity projection of GFP-Rab5 movie (45 s) overlaid with trajectories (red) and cell outline (blue). Scale bars: 5 μm. C. Early endosomes exhibit rapid movements towards the perinuclear region while late endosomes/lysosomes exhibit bidirectional rapid movements. Representative Rab5 and Rab7 trajectories from the cell in B, color-coded for time (first frame = blue, last frame = red). Rab5 trajectory position in B (blue box). Scale bar: 2 μm. See also Figure S1E,F, Table S1.

Figure 2

The effect of cytoskeletal disruption on early endosomal motion A. Potential roles for actin filaments in early endosomal movement. Early endosomes (green), actin filaments (blue), myosin motors (red). B. The effect of cytoskeletal perturbations on endosomal motion. All EGF-Qdot trajectories were analyzed for the indicated parameters using a sliding time window of 2.25 s and the % change in the median value per cell was calculated (n= 10 cells). DMSO: dimethyl sulfoxide control; LatB: 1 μM latrunculin B; Nocod: 10 μg/ml nocodazole; Jasp: 100 nM jasplakinolide. For all boxplots: medians (red lines), 25^th and 75^th percentiles (box edges), data extremes (whiskers), outliers (red crosses), 0% change (green lines). Significant changes (*p < 0.05) were determined using a binomial test of the % change in the median. See also Table 1, Figure S2. C. The effect of cytoskeletal perturbations on the percentage of time EGF-Qdots spend confined, diffusive or directed. D. Early endosomal trajectories contain two types of non-directed movement. An immotile set of trajectories lacks directed movement for the entire period of observation. In contrast, motile trajectories contain periods of directed movement but also pauses (arrows). Scale bar: 2 μm. E. The effect of cytoskeletal perturbations on the motion of the motile subset of trajectories. F. The effect of cytoskeletal perturbations on the percentage of time EGF-Qdots spend confined, diffusive or directed in the motile subset.

Figure 3

Parsing of trajectories into directed runs and pauses A. Early endosomal trajectories exhibited directed motion punctuated by pauses. A representative EGF-Qdot trajectory (black) with directed motion (green) and 2.25 s immediately before (blue) and after (magenta) the run highlighted. For α values (means ± SEMs), n=71 before run, n=29 directed run, n=54 after run. See also Figure S3. B. Track parsing criteria and motion analysis. The directed movement criteria (SCI and instantaneous speed (position smoothed over 1.2 s; unfiltered data in grey)) and the MSD α scaling exponent using a 2.25 s sliding window (Figure S2B, Table S2) are shown for the trajectory in A (5% false-positive thresholds in cyan (Figure S3C–E, Table S3)). Circles highlight the α values immediately before and after the run and the run center. The α values before the red asterisk were unreliable due to poor MSD fitting and are not plotted. C. MSD motion analysis of intact trajectories, directed runs, and pauses using a 2.25 s sliding time window. Whole tracks: 7% confined, 65% diffusive, 28% directed. Directed runs: 0% confined, 1% diffusive, 99% directed. Pauses: 7% confined, 82% diffusive, 11% directed. D. Characteristics of early endosomal motion after a run, during a pause, and before a run begins. Approximately scaled drawing of an early endosome (~200 nm diameter), its linkage to the MT track via dynein (~70 nm, adapted from [30]), and its predicted diffusion (root mean squared displacement (RMSD), blue circles) after different time intervals.

Figure 4

The motion of individual EGF-Qdot-containing early endosomes during encounters with other Rab5-positive early endosomes A. EGF-Qdots merge with and split away from other early endosomes during their trafficking. A kymograph showing an EGF-Qdot (magenta, yellow arrow) relative to several other GFP-Rab5 punctae (green) that merge (white arrows) and split away (green arrow). Scale bars: 5 μm, 5 s. Inset: x,y trajectory color-coded for time (blue = first frame, red = last frame). Scale bar: 5 μm. See also Movie S2. B. Pauses in early endosomal motility that occur during an encounter with another early endosome are mainly diffusive. The trajectory (black) of an EGF-Qdot that interacts with another endosome during a pause (3 time regions with distinct α are color-coded on the plot during the pause- red, cyan, green). The inset images show the EGF-Qdot-containing endosome (magenta, white arrow) moving towards a large GFP-Rab5-positive endosome and pausing (red arrows). Scale bar: 2 μm. C. An extreme example of a GFP-Rab5/EGF-Qdot particle snapping backwards after several periods of directed motion. (i) The trajectory is color-coded for time (blue = first frame, red = last frame). Scale bar: 5 μm. (ii) The instantaneous velocity of the trajectory in (i). The α values for three pause locations are indicated with arrows. (iii) The kymographs show one of the EGF-Qdot reversals (white arrow, red asterisk marks time-point of the reversals in (i) and (ii)) during which a puncta of GFP-Rab5 separates from the EGF-Qdot/GFP-Rab5 (yellow arrows) as the EGF-Qdot snaps backwards. Scale bar 2 μm. See also Movie S3. D. Transferrin-positive tubules extend from motile, EGF-positive endosomes. (i) A kymograph showing a pause in EGF-Alexa532/transferrin-Alexa488 trafficking and a fission event (white arrow). Scale bars: 2 μm, 1 s. (ii) A montage from the movie in (i) of EGF (magenta, white arrow) and transferrin (green, yellow arrow).

Figure 5

Early endosomal movement is affected by interactions with the ER and multiple MTs. A. Early endosomes remain attached to the ER for long periods during which repeated deformation of the ER occurs. The white arrow follows the same early endosome. Scale bar: 2 μm. See also Movie S4A. B. An early endosome appears to pull an ER tubule behind it during a directed run. The yellow-dotted line in the first panel indicates the trajectory used to generate the kymograph in C. Images are background-subtracted. Scale bar 2 μm. See also Movie S4B. C. The kymograph highlights the rapid (~1.6 μm/s) movement of the early endosome (green, white arrow) towards the perinuclear region followed by a slower reversal (~0.7 m/s, red arrow). Scale bars: 4 s, 2 μm. See also Movie S4B, Figure S4A. D. Pausing and rotation of two elongated early endosomes at a region of dense MTs. The white and yellow arrows indicate the two ends of an elongated early endosome (green) to highlight its rotation during the pause near dense MTs (magenta). The two rows show different early endosomes navigating the same region. Images are background-subtracted. Scale bar: 2 μm. See also Movie S5A. E. Rapid switch between 2 MT tracks by an early endosome. The dotted white line indicates the trajectory taken by the early endosome (green, arrow). Images are background-subtracted. Scale bar: 2 μm. See also Movie S5B, Figure S4B.