EGF induces KRas translocation to the early endosomal compartment. COS-1 cells were transiently transfected with GFP-KRas, and its intracellular localization was analyzed by confocal microscopy. (A) Colocalization of GFP-KRas and EEA1 on intracellular structures. (B) Cells treated with 100 ng/ml EGF-TRITC for 20 min show increased colocalization between GFP-KRas, EEA1, and EGF-TRITC. (C) Quantification of cells in B. Statistical significances of differences between control and EGF treatment were determined using the Student's t test. Data are means ± SEM; **, P < 0.01. (D) A more detailed confocal image of early endosomal structures positive for GFP-KRas, EEA1, and EGF-TRITC is shown. Red dotted circles indicate the perimeter of magnified endosomes. (E) Likewise, in EGF-stimulated cells, endocytosed EGFR detected by immunofluorescence colocalizes with GFP-KRas and EGF-TRITC. (F) NIH3T3-Wt8 cells treated with 100 ng/ml EGF-TRITC for 20 min also confirmed colocalization of GFP-KRas, EEA1, and EGF-TRITC. (G) In NIH3T3 cells, a significant increase of GFP-KRas together with EGFR is detected in endosomal fractions isolated by sucrose density gradients after EGF treatment (Western blot), and the results presented are representative of three independent experiments (graph). (H) Similarly, after EGF stimulation, subcellular fractionation by sucrose gradients using nontransfected NIH3T3-Wt8 cells shows enrichment of endogenous KRas and EGFR in endosome fractions (Western blot). Colocalization is indicated with white (merge) or red (separate channels) arrows. (A, B, E, and F) Dotted boxes define the areas from which the corresponding insets were generated. Bars: (A, B, E, and F) 20 µm; (D) 500 nm.

Translocation of KRas to the early endosomal compartment is a CaM- and PKC-independent process. COS-1 cells expressing GFP-KRas or the indicated mutants were treated with EGF-TRITC for 20 min, fixed, and immunolabeled with anti-EEA1. (A) Colocalization of GFP-KRas (wild type [wt]), EEA1, and EGF-TRITC on intracellular vesicles was observed after treatment with 10 µg/ml of the CaM inhibitor W13 for 1 h. (B) The same as in A but using 10 µM of the general PKC inhibitor BIM for 1 h. (C) YFP-KRas (Ser181D [S181D]) colocalizes with EEA1 and EGF-TRITC. (D) Similarly, YFP-KRas (Ser181A [S181A]) colocalizes with EEA1 and EGF-TRITC. (E) The presence of YFP-KRas (Ser181A) on EEs was analyzed in cells treated with W13. Quantification of colocalization is shown underneath each panel. Statistical significances of differences between control and EGF treatment were determined using the Student's t test. Data are means ± SEM; **, P < 0.01; ***, P < 0.001. Colocalization is indicated with white (merge) or red (separate channels) arrows. Dotted boxes define the areas from which the corresponding insets were generated. Bars, 20 µm.

KRas follows the clathrin-dependent endocytic pathway upon EGF stimulation. (A–G) COS-1 cells were transiently transfected with GFP-KRas, and its intracellular localization was analyzed by confocal microscopy. (A) Starved cells were fixed and immunolabeled with an anticlathrin antibody. (B) Clathrin detection on starved cells after 1 h at 4°C. (C) Cells treated with EGF-TRITC for 2.5 min were fixed and immunolabeled with an anticlathrin antibody. The red dotted lines point to colocalization in the image and in the z-axis profile. (D) The same as in C, but cells were incubated with EGF-TRITC for 1 h at 4°C. Reconstruction of the z-axis profile was performed to confirm the visualization on basal PM (C and D, insets). (E) Cells treated with EGF-TRITC for 2.5 min were fixed and immunolabeled with anti-EGFR. (F) Cells coexpressing GFP-KRas and the HA-tagged dynamin mutant K44A (DynK44A) were treated with EGF-TRITC for 20 min and immunolabeled with anti-HA. Red dotted contours indicate the perimeter of cells cotransfected with GFP-KRas and HA-DynK44A. (G) Quantification of EEA1 and GFP-KRas colocalization in control and dynamin K44A–expressing cells ± EGF is shown. Statistical significances of differences between control and EGF treatment were determined using the Student's t test. Data are means ± SEM; **, P < 0.01; ***, P < 0.001. (A–E) Colocalization is indicated with white (merge) or red (separate channels) arrows. Dotted boxes define the areas from which the corresponding insets were generated. Bars, 20 µm.

U18666A and bafilomycin induce accumulation of KRas in a Rab7/LBPA late endosomal compartment. (A) HA-KRas and GFP-Rab7–expressing cells were treated with EGF for 20 min and immunolabeled with an anti-HA (red channel) and anti-EEA1 (blue channel). Arrows indicate the colocalization of KRas and EEA1, and arrowheads point to Rab7-positive structures. (B) Localization of GFP-KRas in cells treated with U18666A. Arrows point to PM KRas labeling, and arrowheads show intracellular ringlike structures. (C) HA-KRas was coexpressed with GFP-Rab7 and treated with U18666A plus EGF for 45 min. Arrows show colocalization of HA-KRas with GFP-Rab7. (D–F) Localization of KRas in cells treated with 20 nM B-A1 overnight and with 100 ng/ml EGF for 45 min. Arrows indicate colocalization of HA-KRas and Rab7 (E) or GFP-KRas and LBPA (F) on enlarged endosomes. The white dotted contour in D indicates the cell perimeter, and the arrowheads point to ringlike endosomal structures. Bars: (A–D) 8 µm; (E and F) 2 µm.

Lysosomal targeting of KRas. (A) COS-1 cells expressing GFP-KRas were treated with 100 ng/ml EGF-TRITC for 2 h. Triple labeling of GFP-KRas, LAMP1 (blue channel), and EGF-TRITC (red channel) is shown, and arrows indicate colocalization. (B and C) GFP-KRas–expressing cells were treated either with 20 nM B-A1 (B) or 25 µM leupeptin (C) and EGF. Immunolabeling with anti-LAMP1 antibody shows colocalization of KRas (green channel) with LAMP1 (red channel) on perinuclear ringlike structures (see arrows in insets of B and insets 1–3 of C); arrowheads (B, insets) point to KRas-positive vesicles not labeled with LAMP1. TO-PRO 3 is a nuclear marker (blue channel). (D) Quantification of LAMP1/GFP-KRas colocalization from experiments shown in A–C. Statistical significances of differences between different treatments were determined using the Student's t test. Data are means ± SEM; *, P < 0.05; ***, P < 0.001. (E and F) GFP-HRas–expressing cells treated with B-A1 (E) or leupeptin (F) and EGF show colocalization of HRas (green channel) with GMAP210 (Golgi marker; blue channel) but not with LAMP1 (red channel). (G) In vivo imaging of GFP-KRas–expressing cells after treatment with B-A1, leupeptin, and EGF reveals MVB-like structures containing intralumenal GFP-KRas–positive membranes (z1–z12). (H and H′) 3D tomographic reconstruction of confocal sections z1–z11 (H) and z1–z7 (H′). (I) GFP-Rab7–expressing cells treated with B-A1 and EGF and labeled with anti–pan-Ras exhibit considerable colocalization (arrows). Arrowheads point to Rab7-positive vesicles not labeled with the pan-Ras antibody. (J) HA-KRas stability was measured in control and B-A1 plus leupeptin–treated cells. Before stimulation with 100 ng/ml EGF, starved cells were preincubated for 6 h with 100 µg/ml CHX (control) or CHX in combination with B-A1 plus leupeptin (B-A1 + Leu). Equal amounts (15 µg) of cell lysates from 0, 3, 6, 15, and 21 h after EGF addition were electrophoresed, and the levels of HA-KRas detected with an anti-HA antibody in both control and B-A1 plus leupeptin–treated cells are shown (Western blot; n = 5). Actin was used as a loading control. Quantification analysis by densitometry was performed. Data are the mean ± SEM; *, P < 0.05; **, P < 0.01 (Student's t test). Bars: (A–C, E, F, and I), 20 µm; (G, H, and insets) 2 µm.

Time-lapse recording of KRas activation on the late endosomal compartment. (A and B) COS-1 cells were transfected with Raichu-KRas (A) or Raichu-Pak-Rho (B). Starved cells were treated with 20 nM B-A1 overnight and then with 100 ng/ml EGF for 50 min. In the top panels, representative ratio images of YFP/CFP at the indicated time points after EGF stimulation are shown. In the bottom panels, representative images of CFP–Raichu-KRas (A) and CFP–Raichu-Pak-Rho (B) indicate localization of the probes at each corresponding FRET image frame. All images in A and B were obtained every 5 min with a time-lapse epifluorescent microscope. High FRET efficiency can be observed at the PM (arrowheads) and on endosomal structures (arrows) between 25 and 45 min in Raichu-KRas–expressing cells only. (C) Quantification of FRET (FRET ratio of YFP/CFP) efficiency on the endocytic structures from at least six independent experiments was plotted for each time point by measuring the increase over the basal activity, which was averaged over 10 min before EGF addition. (D) Time-lapse video microscopy of COS-1 cells expressing GFP-KRas treated with B-A1 and with 100 ng/ml EGF-TRITC for 1.5 h. Confocal section images were collected every 15 s. Endosomal fusion of KRas-positive vesicles (green channel) harboring internalized EGF (red channel) is observed (arrows). Arrowheads point to PM-localized EGF-TRITC. Endosomal diameters d1 and d2 are 2.6 µm and 3.4 µm, respectively. Data are means ± SEM. Bars: (A and B) 20 µm; (D) 2 µm.

Grb2 and Raf1 are recruited to KRas-positive LEs. (A and B) COS-1 cells were transfected with Grb2-YFP and treated with B-A1 for 6 h. (A) Starved cells show diffuse cytosolic distribution of Grb2-YFP. (B) After 45 min of EGF-TRITC stimulation, Grb2 is redistributed onto perinuclear vesicles. Colocalization of EGF-TRITC (red channel), immunolabeled EGFR (blue channel), and Grb2 (green channel) is observed (inset). (C and D) Grb2-YFP and CFP-KRas were coexpressed in COS-1 cells, and then cells were treated with B-A1. (C) Control cells show diffuse cytoplasmic distribution of Grb2-YFP. Insets show colocalization of immunolabeled EGFR and CFP-KRas but not Grb2-YFP on both endosomes (arrows) and the PM (arrowheads). (D) After 45 min of EGF-TRITC stimulation, Grb2 is redistributed onto vesicular structures. Colocalization of EGF (red channel), immunolabeled EGFR (gray channel), Grb2 (green channel), and KRas (blue channel) is observed (arrows). (E and F) COS-1 cells were transfected with YFP-Raf1 and treated with B-A1 for 6 h. (E) Starved cells show diffuse cytoplasmic distribution of YFP-Raf1. (F) After 45 min of EGF-TRITC stimulation, Raf1 was recruited, to some extent, to vesicular structures (inset). Colocalization of internalized EGF (red channel) and Raf1 (green channel) could be observed (inset). (G and H) YFP-Raf1 and CFP-KRas were coexpressed in COS-1 cells, and then cells were treated with B-A1. (G) Control cells show a diffuse cytosolic pattern of YFP-Raf1. Some degree of colocalization of YFP-Raf1 (green channel) and CFP-KRas (blue channel) can be observed (white arrows in the merge inset and red arrows in the color channel insets). (H) After 45 min of EGF-TRITC stimulation, Raf1 is redistributed onto vesicular structures. Clear colocalization of endocytosed EGF (red channel), Raf1 (green channel), and KRas (blue channel) is observed (white arrows in the merge inset and red arrows in the color channel insets). (B–D and F–H) Dotted boxes define the areas from which the corresponding insets were generated. Bars, 20 µm.

KRas localizes on LEs containing the p14–MP1 scaffolding complex. (A and B) COS-1 cells transiently coexpressing GFP-KRas and Xpress-tagged p14. (A) Colocalization of KRas (green channel) with p14 (blue channel) and endocytosed EGF-TRITC (100 ng/ml for 45 min) on perinuclear endosomes (insets). (B) Late endosomal colocalization of GFP-KRas and p14 after B-A1 treatment (20 nM for 6 h) in nonstimulated cells (insets). TO-PRO 3 is a nuclear marker (blue channel). (C and D) COS-1 cells were cotransfected with GFP-KRas and myc-tagged MP1. (C) Colocalization of KRas (green channel) with MP1 (blue channel) and endocytosed EGF-TRITC (red channel) on LEs (insets). (D) Triple labeling of GFP-KRas, MP1 (red channel), and cathepsin D (CatD; blue channel) is shown in cells treated with leupeptin for 8 h and EGF (100 ng/ml for 45 min; insets). Colocalization is indicated with white (merge) or red (separate channels) arrows. Dotted boxes define the areas from which the corresponding insets were generated. Bars, 20 µm.

KRas-mediated MAPK signaling on LEs. (A–E) Cells coexpressing GFP-KRas with Xpress-tagged p14 were treated with 20 nM B-A1plus 25 µM leupeptin for 6 h and then were left starved or pulse-chase stimulated with 100 ng/ml EGF for various times before fixation. Cells were then subjected to indirect immunofluorescence analysis using anti-Xpress and anti-pERK1/2 antibodies. (A) The pERK1/2 antibody showed no staining in starved cells. KRas (green channel) colocalizes with p14 (blue channel)-positive endosomes (insets 1 and 2). (B) After 5 min of EGF stimulation, pERK1/2 labeling (red channel) can be readily detected on the PM (inset 3) and nuclei. Endosomal colocalization of KRas (green channel) and p14 (blue channel) is also shown (inset 4). (C–E) In cells that were stimulated with EGF for 10 (C), 30 (D), or 60 min (E), colocalization of pERK1/2 (red channel) with GFP-KRas and p14 (blue channel) was detected on endosomal structures (insets 5–10). After 10 min of EGF stimulation, pERK1/2 labeling can also be observed on the PM (inset 5). Insets 1–10 show magnifications of boxed areas. (F and G) COS-1 cells (F) and NIH3T3-Wt8 (G) cells were transiently transfected with HA-KRas and serum starved for 20 h. Control cells or cells treated with the lysosomal inhibitors B-A1 plus leupeptin (B-A1 + Leu) were then stimulated with EGF by pulse and chase for the indicated times. Cell lysates were separated by SDS-PAGE and probed with the indicated antibodies (Western blot; n = 5). The graphs show a densitometric analysis of ERK1/2 relative activation. The pERK1/2 signals were normalized against the ERK1/2 signals. Data are the mean ± SEM; *, P < 0.05 (Student's t test). (H) B-A1–treated cells expressing GFP-KRas were stimulated with 100 ng/ml EGF-TRITC for 45 min before fixation. Colocalization of KRas (green channel) with immunolabeled pERK1/2 (blue channel) and EGF (red channel) is observed on ringlike structures (insets). (I) Leupeptin-treated cells expressing GFP-KRas were loaded with LysoTracker red DND-99 and incubated in the presence of 100 ng/ml EGF at 37°C for 45 min before fixation. Colocalization of KRas (green channel) with immunolabeled pERK1/2 (blue channel) and LysoTracker (red channel) is observed on ringlike structures (insets). (H and I) Dotted boxes define the areas from which the corresponding insets were generated. Bars, 20 µm.

Model of the endosomal KRas traffic and signaling. EGF binding triggers endocytosis of the EGFR and downstream signaling molecules to the early endosomal compartment. There, Ras signaling modules settle to elicit signal outputs. HRas is ubiquitinated and is recycled back to the PM, whereas KRas is sorted en route to LEs/MVBs. Once there, KRas can be GTP loaded, thus being able to recruit effectors such as Raf1, which initiates the MAPK signaling cascade on this compartment. Finally, KRas down-regulation/degradation may occur in the late endosomal/lysosomal compartment. CCV, clathrin-coated vesicle.