Opposing microtubule motors control motility, morphology and cargo segregation during ER-to-Golgi transport

Abstract

We recently demonstrated that dynein and kinesin motors drive multiple aspects of endosomal function in mammalian cells. These functions include driving motility, maintaining morphology (notably through providing longitudinal tension to support vesicle fission), and driving cargo sorting. Microtubule motors drive bidirectional motility during traffic between the endoplasmic reticulum (ER) and Golgi. Here, we have examined the role of microtubule motors in transport carrier motility, morphology, and domain organization during ER-to-Golgi transport. We show that, consistent with our findings for endosomal dynamics, microtubule motor function during ER-to-Golgi transport of secretory cargo is required for motility, morphology, and cargo sorting within vesicular tubular carriers en route to the Golgi. Our data are consistent with previous findings that defined roles for dynein-1, kinesin-1 (KIF5B) and kinesin-2 in this trafficking step. Our high resolution tracking data identify some intriguing aspects. Depletion of kinesin-1 reduces the number of motile structures seen, which is in line with other findings relating to the role of kinesin-1 in ER export. However, those transport carriers that were produced had a much greater run length suggesting that this motor can act as a brake on anterograde motility. Kinesin-2 depletion did not significantly reduce the number of motile transport carriers but did cause a similar increase in run length. These data suggest that kinesins act as negative regulators of ER-to-Golgi transport. Depletion of dynein not only reduced the number of motile carriers formed but also caused tubulation of carriers similar to that seen for sorting nexin-coated early endosomes. Our data indicated that the previously observed anterograde-retrograde polarity of transport carriers in transit to the Golgi from the ER is maintained by microtubule motor function.

Keywords: Dynein; Endoplasmic reticulum; Golgi; Kinesin; Microtubule motor; Secretory cargo trafficking.

Fig. 1.. Projections of time lapse sequence from TIRF imaging of tsO45-G-GFP translocation.

Image sequences comprise 1000 frames acquired at 12.04 frames per second. (A,D,G,J,M) Maximum intensity projections generated from all 1000 frames. B,E,H,K,N (enlarged in panels C,F,I,L,O) Colour coded time projections of every fourth frame from these sequences (332 ms intervals, 250 frames total). (P) The colour coding for time projections. Scale bars: 10 µm (A–O).

Fig. 2.. Proportion of motile objects and track length.

(A) Proportion of moving objects as a percentage of the total puncta detected. Statistical significance was tested using an ANOVA with a Dunnett's post-hoc test. (B) Quantification of time-lapse imaging. Images were processed to generate maximum intensity projections as shown in Fig. 1. Track lengths were then automatically measured in Volocity and plotted. Each spot represents an individual particle. Measurements are derived from 3 independent experiments, 3–8 cells in each experiment, total of 75 objects. Statistical significance was tested using an ANOVA with a Dunnett's post-hoc test. Red asterisks indicate statistically detectable decrease in track length; green asterisks represent a statistically detectable increase in track length.

Fig. 3.. Increased tubulation of ER-to-Golgi carriers following suppression of DHC1 expression.

Examples are taken from the time lapse movie sequence associated with Fig. 1D. Boxed regions show 3× enlargements. Time from the start of the image sequence is shown (mins:secs). Scale bars: 10 µm.

Fig. 4.. Motor subunit depleted cells expressing tsO45-G-GFP were fixed and immunolabelled to detect COPI (β′-COP).

(A) An example of a control (GL2 siRNA transfected) cell. (B,C) Image quantification was performed manually to determine (B) the number of tsO45-G-GFP puncta with associated COPI labelling and (C) the polarity of these structures with respect to the Golgi apparatus. “Correctly oriented” structures are those with the COPI-positive domain facing away from the Golgi and the tsO45-G-GFP-positive domain facing towards the Golgi. “G” indicates the user-defined point of the Golgi apparatus. Asterisks in panels B and C dictate those cases for which a statistically detectable difference was found using ANOVA with Dunnett's post-hoc test. (D) RPE1 cells were transfected with siRNA against the various motor proteins for 72 hours. The cells were then lysed and after a protein assay equal amounts of protein analyzed by immunoblotting with an anti-KDEL or an anti-α-tubulin antibody. (E) Immunofluorescence labelling of control or KIF5B depleted cells with antibodies directed against giantin or KDEL bearing proteins. Scale bars: 10 µm (A,E).