Coupling fission and exit of RAB6 vesicles at Golgi hotspots through kinesin-myosin interactions

Abstract

The actin and microtubule cytoskeletons play important roles in Golgi structure and function, but how they are connected remain poorly known. In this study, we investigated whether RAB6 GTPase, a Golgi-associated RAB involved in the regulation of several transport steps at the Golgi level, and two of its effectors, Myosin IIA and KIF20A participate in the coupling between actin and microtubule cytoskeleton. We have previously shown that RAB6-Myosin IIA interaction is critical for the fission of RAB6-positive transport carriers from Golgi/TGN membranes. Here we show that KIF20A is also involved in the fission process and serves to anchor RAB6 on Golgi/TGN membranes near microtubule nucleating sites. We provide evidence that the fission events occur at a limited number of hotspots sites. Our results suggest that coupling between actin and microtubule cytoskeletons driven by Myosin II and KIF20A ensures the spatial coordination between RAB6-positive vesicles fission from Golgi/TGN membranes and their exit along microtubules.

Fig. 1

GFP-RAB6-positive vesicles exit the Golgi at fission hotspots. a HeLa cells stably expressing GFP-RAB6 were imaged every second for 1 min (Supplementary Movie 1). Orange circles show the location of five fission hotspots. Details of each fission event for the five regions of hotspots are displayed. In each case, sequences of images are displayed at the top and a scheme of the events at the bottom. The Golgi region is highlighted by a white line. Vesicles exiting the Golgi are drawn and painted either yellow-green or orange-red. A sum of the events is shown at the end of each sequence. b Cells were treated for 45 min with para-nitro-blebbistatin. Then, para-nitro-blebbistatin was washed out. After recovery of normal GFP-RAB6 vesicles trafficking, the cells were imaged every second for 60 s (Supplementary Movie 3). A scheme of the images is displayed at the bottom. The Golgi regions are highlighted by a white line. Tubes are drawn in yellow. Regions where the GFP-RAB6-positive fission hotspots are found are drawn in red. Bars: 10 µm

Fig. 2

Inhibition of KIF20A function inhibits the fission of Rab6 transport carriers from the Golgi. GFP-RAB6, Myosin II, and KIF20A are co-localized on dotted structures that correspond to the sites of fission. a HeLa cells stably expressing GFP-RAB6 were imaged by time-lapse microscopy, and images of the same cell before (t0) or 40 min (t40) after paprotrain (25 µM) treatment are presented (Supplementary Movie 4). Paprotrain was used as an inhibitor of KIF20A function. b Quantification of the number of transport carriers and of Golgi-connected tubules in cells treated as indicated in (a) (mean ± SEM, n = 38 cells). ***P < 10⁻¹⁰ (Student’s t test). c Staining of endogenous GM130 (green), KIF20A (red), and TGN46 (blue) in HeLa cells indicates a higher co-localization of KIF20A with the trans-Golgi marker TGN46 as compared to the cis-Golgi marker GM130 (see higher magnifications (boxes) on the bottom). Line profiles of the GM130 (green), the KIF20A (red), and the TGN46 (blue) fluorescence intensities (arbitrary units) along the white-dashed arrow. Bar, 10 µm. d Staining of GFP-RAB6 (green), MLC-mCherry (red), and endogenous KIF20A (blue) in HeLa cells indicates a partial co-localization of the three proteins on dotted structures at the Golgi complex (see higher magnifications (boxes) on the bottom). Line profiles of the GFP-RAB6 (green), the MLC-mCherry (red), and the KIF20A (blue) fluorescence intensities (arbitrary units) along the white-dashed arrow. Bar, 10 µm. e Staining of GFP-RAB6 (green), MLC-mCherry (red), and endogenous KIF20A (blue) in HeLa cells treated for 45 min with paprotrain indicates a co-localization (arrow) of the three proteins at the base of a GFP-RAB6-positive tube. Bar, 10 µm

Fig. 3

KIF20A interacts with Myosin II and stabilizes Myosin II at the Golgi. a Yeast two-hybrid interactions between the tail domain of Myosin II (1147–1653 fragment) and different domains of KIF20A. The Saccharomyces cerevisiae reporter strain L40 was co-transformed with a plasmid encoding fusion proteins to detect interactions between amino acids 1147 and 1653 of human myosin IIA heavy chain and the motor domain (1–530), the tail domain (529–887), and the RAB6 binding domain (RBD) (529–665) of KIF20A. Growth on medium lacking histidine (-W -L -H) indicates an interaction between the encoded proteins. b Endogenous Myosin II is pulled-down by GFP-KIF20A. GFP-KIF20A or GFP-transfected HeLa cell extracts were immunoprecipitated using the GFP-trap system. Myosin II bound to GFP-KIF20A was revealed by western blot analysis using anti-Myosin II antibody. GFP-KIF20A and KIF20A were revealed by anti-KIF20A antibody. Note that endogenous KIF20A is found in the IP fraction because it likely can form dimers with GFP-KIF20A. Input represents a 5% load of the total cell extracts used in all conditions. c Staining for endogenous Giantin (green) and endogenous Myosin II using the AD7 antibody (red) in Rat Clone 9 cells 3 days after transfection with specific KIF20A siRNAs. Bar, 10 µm. Of note, KIF20A is required for cytokinesis and KIF20A depletion leads to binucleated cells^,. Quantification of Golgi-associated Myosin II fluorescence intensity in cells treated as described above (mean ± SEM, n = 23–66 cells). *P = 0.02 (Student’s t test). MW molecular weight in kDa

Fig. 4

Definition of the RAB6-BD of KIF20A. RAB6 partially recruits KIF20A on the Golgi complex. RAB6 is required for KIF20A-Myosin II interaction. Relative binding affinities between RAB6, KIF20A, and Myosin II. a–c KIF20A-RAB6-binding domain (KIF20A-RBD) is a dimer composed of parallel helices that form a right-handed coiled-coil stabilized by an inter-helical cysteine bridge and two RAB6 molecules bind on opposite sides of the dimer b. The RAB6:KIF20A interface makes hydrophobic and polar contacts between the molecules b, c, in particular K629 and S631, are buried in the interface and involved in direct interactions at the center of the interface. d Yeast two-hybrid interactions between the KIF20A-RBD-529-665 fragment or KIF20A-RBD-529-665-K626W-S631W fragment with RAB6, Myosin II-RBD, and KIF20A-RBD. The Saccharomyces cerevisiae reporter strain L40 was co-transformed with a plasmid encoding fusion proteins to detect interactions between the KIF20A-RBD-529-665 fragment or KIF20A-RBD-529-665-K626W-S631W fragment and RAB6-Q72L, KIF20A-RBD, Myosin II-RBD, RAB1Q70L, and Lamin A. Growth on medium lacking histidine (-W -L -H) indicates an interaction between the encoded proteins. e Golgi association of the KIF20A-RBD-529-665 fragment or KIF20A-RBD-529-665-K626W-S631W fragments. Staining for endogenous Giantin (red) and overexpressed myc-tagged KIF20A-RBD-529-665 fragment or KIF20A-RBD-529-665-K626W-S631W fragment. f Staining of endogenous Giantin (green) and KIF20A (red) in HeLa cells 3 days after transfection with specific RAB6 siRNAs. Quantification of Golgi-associated KIF20A fluorescence intensity in cells treated as described above (mean ± SEM, n = 34-37 cells). **P < 10⁻³ (Student’s t test). Bar, 10 µm. g RAB6 is required for KIF20A/Myosin II interaction. GFP-KIF20A or GFP expressing HeLa cell extracts treated for 3 days with control or RAB6-specific siRNAs were immunoprecipitated using the GFP-trap system. Myosin II bound to GFP-KIF20A was revealed by western blot analysis using anti-Myosin II antibody. Input represents a 5% load of the total cell extracts used in all conditions. h 96-well plates were coated with recombinant GST-RAB6 or GST-Myosin II-1148-1652 and incubated with increasing amounts of recombinant 6 × His-KIF20A-RBD-529-665 or 6 × His-RAB6, as indicated (solid-phase assay). After washes, KIF20A-RBD-529-665 or RAB6 proteins bound to RAB6 or Myosin II-1148-1652 were detected (arbitrary units) using anti-6 × His antibodies and a chromogenic substrate (mean ± SEM, n = 3 experiments). No binding to GST-GFP alone was detected. MW molecular weight in kDa

Fig. 5

GFP-RAB6 diffuses on Golgi membranes and KIF20A inhibition affects the diffusion of GFP-RAB6. a Images of GFP-RAB6 expressing cells treated with DMSO or with paprotrain for 30 min. HeLa cells expressing GFP-RAB6 were bleached in a 30 pixels circular region of the Golgi apparatus (white circle) and then imaged for 2 min by spinning disk microscopy. Sequence of images of GFP-RAB6 in Golgi of a control cell (b) and a paprotrain-treated cell (c) before and after photobleaching. The round region delimited by a dotted line was photobleached immediately after the first image (−1). In both cases, images are displayed at the top and schematics of the images at the bottom. The orange line represents the recovery of the GFP-RAB6 fluorescence. In b, the photobleached area rapidly recovered fluorescence. The fluorescence flows from the center of the Golgi to the extremity of the bleached area. In c, the photobleached area recovered fluorescence twice faster than in b (compare t5). d The fluorescence recovery after photobleaching was measured in three independent experiments and plotted for GFP-RAB6 treated with DMSO (black) or with paprotrain (orange) (n = 17–19 Golgi). e Half-life recovery of GFP-RAB6 fluorescence in control cells and cells treated with paprotrain (mean ± SEM, n = 17–19 cells). *P = 0.01 (Student’s t test)

Fig. 6

KIF20A co-localizes with growing microtubules on Golgi membranes. a HeLa cells were incubated for 45 min at 4 °C to promote microtubule depolymerization (without affecting the Golgi morphology) and then incubated for 3 min at RT to allow microtubule repolymerization. Staining of endogenous KIF20A (green) and β-tubulin (red) in HeLa cells indicates a partial co-localization of KIF20A with growing microtubules from Golgi membranes (see higher magnifications (boxes) on the bottom). Line profiles of the KIF20A (green) and the β-tubulin (red) fluorescence intensities (arbitrary units) along the white dashed line. Bar, 10 µm. b Staining of GFP-RAB6 (green), MLC-mCherry (red), and endogenous GCC185 (blue) in HeLa cells indicates a partial co-localization of the three proteins on dotted structures at the Golgi complex (see higher magnifications (boxes) on the bottom). Line profiles of the GFP-RAB6 (green), the MLC-mCherry (red), and the GCC185 (blue) fluorescence intensities (arbitrary units) along the white dashed line. Bar, 10 µm. c Schematic of the sequence of events that can be envisioned for the generation of fission hotspots at the TGN. (1) RAB6 diffuses on Golgi/TGN membranes. (2) RAB6 participates in the recruitment and stabilization of KIF20A. When bound to KIF20A, RAB6 diffusion is decreased allowing the localization and anchoring of RAB6 molecules to sites of growing Golgi-associated microtubules. (3) Myosin II is recruited by KIF20A and RAB6. Either a complex between RAB6 and KIF20A is required for Myosin II recruitment, or KIF20A acts alone. Myosin II can be recruited by the KIF20A-RBD-529-665 domain in complex or not with RAB6 or through the 796–887 domain. (4) KIF20A, RAB6, and Myosin II, in association with actin filaments and microtubules, define a Golgi fission hotspot. There are around six hotspots per Golgi. (5) Myosin II and actin then drive the fission of RAB6-positive transport carriers from Golgi/TGN membranes. RAB6-positive vesicles are then transported along microtubules to the plasma membrane thanks to KIF5B