Myo1c regulates lipid raft recycling to control cell spreading, migration and Salmonella invasion

Abstract

A balance between endocytosis and membrane recycling regulates the composition and dynamics of the plasma membrane. Internalization and recycling of cholesterol- and sphingolipid-enriched lipid rafts is an actin-dependent process that is mediated by a specialized Arf6-dependent recycling pathway. Here, we identify myosin1c (Myo1c) as the first motor protein that drives the formation of recycling tubules emanating from the perinuclear recycling compartment. We demonstrate that the single-headed Myo1c is a lipid-raft-associated motor protein that is specifically involved in recycling of lipid-raft-associated glycosylphosphatidylinositol (GPI)-linked cargo proteins and their delivery to the cell surface. Whereas Myo1c overexpression increases the levels of these raft proteins at the cell surface, in cells depleted of Myo1c function through RNA interference or overexpression of a dominant-negative mutant, these tubular transport carriers of the recycling pathway are lost and GPI-linked raft markers are trapped in the perinuclear recycling compartment. Intriguingly, Myo1c only selectively promotes delivery of lipid raft membranes back to the cell surface and is not required for recycling of cargo, such as the transferrin receptor, which is mediated by parallel pathways. The profound defect in lipid raft trafficking in Myo1c-knockdown cells has a dramatic impact on cell spreading, cell migration and cholesterol-dependent Salmonella invasion; processes that require lipid raft transport to the cell surface to deliver signaling components and the extra membrane essential for cell surface expansion and remodeling. Thus, Myo1c plays a crucial role in the recycling of lipid raft membrane and proteins that regulate plasma membrane plasticity, cell motility and pathogen entry.

Fig. 1.

Myo1c partitions into lipid rafts. Cell surface lipid rafts on HeLa cells transiently transfected with GFP–Myo1c (A) and non-transfected HeLa cells (B) were labeled with CTB–Alexa-Fluor-555 (CTB-Alexa555) and co-stained with antibodies against GFP (A) and endogenous Myo1c (B) for confocal microscopy. Cell nuclei are shown in blue in the merged image. Arrowheads highlight examples of colocalization of Myo1c with lipid rafts. (C) HeLa cells were co-transfected with GFP–GPI and mCherry–Myo1c, seeded onto fibronectin-coated coverslips and imaged using live-cell microscopy. The picture presented is a still image from the start point of a representative movie (supplementary material Movie 1). Arrowheads highlight the localization of Myo1c on lipid-raft-enriched tubules. Scale bars, 10 μm. The inserts are enlarged representations of the boxed regions. (D) HeLa cells stably expressing GFP–Myo1c were incubated with CTB–HRP on ice, lysed with cold 1% Triton X-100 and subjected to a sucrose gradient ultracentrifugation. The fractions were separated on SDS-PAGE and Coomassie stained. (E) Floating fractions (3–5) and detergent-soluble fractions (8–10) were identified based on the distribution of CTB–HRP by dot-blot analysis (top panel) and localization of marker proteins using western blotting with indicated antibodies (bottom panels).

Fig. 2.

Depletion of Myo1c causes accumulation of lipid rafts in the perinuclear region. (A) HeLa cells stably expressing GFP–GPI were either mock transfected or transfected with siRNA specific to MYO1C and labeled with antibodies against caveolin-1 and GFP for immunofluorescence microscopy. (B) Mock- or Myo1c-depleted HeLa cells were labeled with antibodies against the raft markers caveolin-1 and CD55. (C) HeLa cells transiently transfected with the dominant-negative GFP–Myo1c^K111R rigor mutant were labeled with antibodies against GFP and caveolin-1. Scale bars, 10 μm.

Fig. 3.

Overexpression of Myo1c increases surface raft levels. (A) Cell surface lipid rafts of HeLa cells transiently transfected with GFP–Myo1c were visualized with CTB–Alexa-Fluor-555. 82% of non-transfected control cells exhibited a modest level of surface CTB staining (indicated by * in the example pictures). Scale bar, 10 μm. Overexpression of GFP–Myo1c increases the percentage of CTB positive cells (arrowheads) from 18% in control cells to 49%. A total number of 2544 cells from three independent experiments were analyzed. Values are means±s.e.m. (B) A stable HeLa cell line expressing homogenous levels of GFP–Myo1c was created by FACS sorting. Cell lysates of control HeLa cells and HeLa cells stably expressing GFP–Myo1c were analyzed by SDS-PAGE and immunoblotted with antibodies against Myo1c and α-tubulin, used as a loading control, demonstrating that endogenous and exogenous Myo1c are expressed at similar levels. (C) For labeling of cell surface lipid rafts, control HeLa cells and HeLa cells stably expressing GFP–Myo1c were incubated with or without CTB–Alexa-Fluor-647 (CTB-Alexa647), while in suspension. The amount of cell-surface-bound CTB-Alexa647 was determined by FACS analysis.

Fig. 4.

Myo1c colocalizes with RalA on lipid raft enriched tubules. (A) HeLa cells were co-transfected with HA–RalA and GFP–Myo1c, or (B) transfected with HA–RalA and stained for HA–RalA and endogenous Myo1c for confocal microscopy. White arrowheads highlight colocalization of Myo1c and RalA on tubules, black arrowheads exemplify colocalization on membrane ruffles of the plasma membrane. Cell nuclei are shown in blue in the merged images. (C) HeLa cells expressing HA–RalA were labeled with antibodies against HA and the exocyst component ExoC2. (D–F) HeLa cells were co-transfected with GFP–GPI and HA–RalA or transfected with GFP–GPI and stained with antibodies against GFP and HA (D), CD59 (E) or CD55 (F). The inserts are enlarged representations of the boxed regions. Scale bars, 10 μm.

Fig. 5.

Myo1c depletion reduces formation of lipid raft enriched tubules, but does not affect recycling of transferrin receptor. (A) HeLa cells co-transfected with GFP–GPI and HA–RalA were treated with siRNA targeting MYO1C and labeled with antibodies against HA and GFP. (B) HeLa cells stably expressing GFP–Rab11 were mock- or Myo1c-depleted and stained with antibodies against GFP and caveolin-1. Scale bars, 10 μm. (C) Control and Myo1c-depleted cells were pulsed with Tfn–Alexa-Fluor-647 (Tfn-Alexa647) at 37°C for 30 minutes, then washed and incubated at 37°C in the presence of excess unlabeled Tfn for indicated times. The amount of Tfn-Alexa647 remaining in the cell was determined by FACS analysis and expressed as a percentage of total endocytosed Tfn-Alexa647 at time zero. Graphs represent the means±s.d. of three independent experiments. ns, not significant. (D) Cell lysates from mock- and Myo1c-depleted HeLa cells were blotted and probed with antibodies against transferrin receptor (TfnR), Myo1c and α-tubulin as a loading control to confirm that mock and knockdown cells express similar levels of TfR.

Fig. 6.

Myo1c facilitates lipid raft recycling, but is dispensable for lipid raft internalization. (A) Control and Myo1c-depleted cells were loaded with antibodies against CD55 at 37°C for 2 hours. After acid stripping the cells were incubated at 37°C for indicated times to allow CD55 recycling. The remaining anti-CD55 antibodies were detected with secondary antibodies conjugated to Alexa Fluor 488, quantified by automated microscopy and are expressed as a percentage of total endocytosed CD55 antibodies at time zero. Graphs represent the means±s.e.m. of three independent experiments. A total number of >76,000 cells were quantified, with a minimum of 3000 cells per time point. (B) To quantify steady-state levels of intracellular CD55, mock- or Myo1c-depleted cells were incubated for 2 hours with anti-CD55 antibodies. After acid stripping intracellular anti-CD55 antibodies were detected with Alexa-Fluor-488-conjugated secondary antibodies and quantified using automated microscopy. Graphs represent the means±s.e.m. for three independent experiments (>18,000 cells). (C) Cell lysates from mock- and Myo1c-depleted HeLa cells were blotted and probed with antibodies against Myo1c, CD55 and α-tubulin, as a loading control, to confirm that mock and knockdown cells express similar levels of CD55. (D) Antibodies against CD55 or (E) CD59 were taken up into mock- or Myo1c-depleted HeLa cells, before fixation and labeling with antibodies against caveolin-1 for immunofluorescence microscopy. (F) Control and Myo1c-knockdown cells were prelabeled with antibodies against CD55 on ice, then incubated at 37°C for indicated times to allow internalization in the presence of primaquine. Intracellular anti-CD55 antibodies were stained with fluorescently labeled secondary antibodies. Internalized CD55 antibodies are expressed as a percentage of the total amount of anti-CD55 antibodies bound to the cell surface before uptake. Graphs represent the means±s.e.m. of three independent experiments. A total number of >90,000 cells were quantified, with a minimum of 3000 cells per time point. ns, not significant.

Fig. 7.

Myo1c is required for exocytosis of lipid rafts during cell spreading and depletion of Myo1c impairs cell spreading and rearranges focal adhesions. (A) HeLa cells, either mock- or Myo1c-depleted, were detached from the tissue culture dish, held in suspension to encourage raft internalization before seeding onto fibronectin-coated coverslips. Their ability to spread after 2 hours was assessed using Rhodamine–phalloidin as a cell label, which stains F-actin. Scale bars, 20 μm. (B) For quantification of cell spreading, images of randomly selected fields were taken and the mean area covered by cells was measured using Volocity Imaging software. A total of 5781 cells from three independent experiments were analyzed. Values are means±s.e.m. (C) HeLa cells, either mock- or Myo1c-depleted, or transiently transfected with the GFP–Myo1c^K111R rigor mutant, were labeled with antibodies against the focal adhesion markers vinculin or paxillin. Scale bars, 20 μm. (D,E) Mock- and Myo1c-depleted HeLa cells were detached, held in suspension, reseeded onto fibronectin-coated coverslips and allowed to spread for 3 hours. Cell surface lipid rafts were labeled with antibodies against CD59 (D) or CD55 (E) and secondary antibodies conjugated to Alexa Fluor 555. For quantification of surface rafts, randomly selected fields were captured and the total fluorescence intensity per cell was measured using Volocity Imaging software. A total number of 1213 cells from three independent experiments (D) and 848 cells from three independent experiments (E) were analyzed. Values are means±s.e.m.

Fig. 8.

Random migration requires Myo1c and RalA. (A) Mock, Myo1c, RalA and both Myo1c and RalA were depleted through siRNA treatment in RPE cells, which were then lysed, blotted and probed with antibodies against Myo1c, RalA and α-tubulin, as a loading control, to confirm the successful protein knockdown. Single knockdowns of either Myo1c or RalA led to a near complete protein depletion with undetectable levels of Myo1c or RalA by western blotting. However, in the double-knockdown protein depletion was less efficient as shown in the representative immunoblots. (B) Time-lapse video microscopy was used to capture the movements of individual control, Myo1c and RalA single-knockdown and Myo1c and RalA double-knockdown cells on fibronectin-coated coverslips over 3 hours. Migration tracks of 249 cells from at least three independent experiments per knockdown were analyzed using Volocity Imaging software. Individual cell trajectories of one representative experiment are depicted. (C) Quantification revealed that loss of Myo1c, RalA or Myo1c in combination with RalA significantly reduces migration speed and track length. Values are means±s.e.m. ns, not significant.

Fig. 9.

Myo1c depletion impairs membrane ruffle formation crucial for macropinocytosis and Salmonella invasion. (A) To quantify membrane ruffling, mock- and MYO1C-siRNA-treated A549 cells were loaded with 70 kDa TMR-coupled dextran and the number of cells containing dextran-positive structures larger than 0.5 μm in size were quantified. A representative picture of one cell containing three dextran macropinosomes (arrowheads) is shown. A total number of 2611 cells from three independent experiments were analyzed. Values presented are means±s.e.m. (B) HeLa cells transiently transfected with GFP–Myo1c, HA–RalA or GFP–GPI were infected with wild-type Salmonella Typhimurium labeled with Alexa555 conjugated succinimidyl-esters and stained with antibodies against GFP, HA and endogenous ExoC2. The merged images (bottom panel) show bacteria in red. The inserts are enlarged, single colour representations of the white boxes. Bars, 10 μm (C) Mock, Myo1c and RalA siRNA treated HeLa cells were infected with Salmonella enterica serovar Typhimurium for 1 hour. After a gentamicin protection assay to kill extracellular bacteria, invasion was quantified by spreading cell lysates onto LB agar for counting of bacterial colony forming units. Invasion efficiency is presented as a percentage normalized to mock depleted cells. Values are means±s.e.m. for four independent experiments, each performed in triplicate. ns, non-significant.