The ESCRT-III subunit hVps24 is required for degradation but not silencing of the epidermal growth factor receptor

Abstract

The endosomal sorting complexes required for transport, ESCRT-I, -II, and -III, are thought to mediate the biogenesis of multivesicular endosomes (MVEs) and endosomal sorting of ubiquitinated membrane proteins. Here, we have compared the importance of the ESCRT-I subunit tumor susceptibility gene 101 (Tsg101) and the ESCRT-III subunit hVps24/CHMP3 for endosomal functions and receptor signaling. Like Tsg101, endogenous hVps24 localized mainly to late endosomes. Depletion of hVps24 by siRNA showed that this ESCRT subunit, like Tsg101, is important for degradation of the epidermal growth factor (EGF) receptor (EGFR) and for transport of the receptor from early endosomes to lysosomes. Surprisingly, however, whereas depletion of Tsg101 caused sustained EGF activation of the mitogen-activated protein kinase pathway, depletion of hVps24 had no such effect. Moreover, depletion of Tsg101 but not of hVps24 caused a major fraction of internalized EGF to accumulate in nonacidified endosomes. Electron microscopy of hVps24-depleted cells showed an accumulation of EGFRs in MVEs that were significantly smaller than those in control cells, probably because of an impaired fusion with lyso-bisphosphatidic acid-positive late endosomes/lysosomes. Together, our results reveal functional differences between ESCRT-I and ESCRT-III in degradative protein trafficking and indicate that degradation of the EGFR is not required for termination of its signaling.

Figure 1.

hVps24 is localized to late endosomes. Hep2 cells were permeabilized before fixation. The cells were stained with antibodies against hVps24 (red) and EEA1 (green; A–C), CD63 (green; D–F), LAMP2 (green; G–I), or TGN46 (green; J–L). Colocalization between hVps24 and the endosome/TGN markers is indicated in yellow.

Figure 2.

Localization of hVps24 and EGFR in EGF-stimulated cells. HEp2 cells were starved overnight before stimulation with 50 ng/ml EGF (5 min) followed by 15-min chase in 10 μg/ml cycloheximide-containing medium. The cells were permeabilized before fixation and stained for EGFR (red) and hVps24 (green). Colocalization between EGFR and hVps24 is indicated in yellow in the merged image.

Figure 3.

Depletion of hVps24 retards EGFR down-regulation. HeLa cells treated with scrambled RNA (−) or with siRNA against hVps24 (+) were stimulated for 0, 2, and 4 h with EGF. The cells were lysed and analyzed by SDS-PAGE and sequential blotting with antibodies against EGFR and hVps24. The same membrane was then reblotted with anti-tubulin to verify equal loadings.

Figure 4.

EGFR is retained in early endosomes in the absence of hVps24. HeLa cells treated with scrambled RNA (A) or with siRNA against hVps24 (B) were given a 30-min pulse of EGF and chased for 3 h. The cells were permeabilized before fixation and stained with antibodies against the EGFR (green), EEA1 (red), and LBPA (blue). Colocalization between EGFR and EEA1 is shown in yellow.

Figure 5.

Fluid phase transport from early to late endosomes is unaffected in hVps24-depleted cells. HeLa cells treated with siRNA against hVps24 were left to internalize prebound dextran for 30 min at 37°C. The cells were washed and chased for 0 h (A) and 3 h (B) before they were permeabilized, fixed, and stained for immunofluorescence microscopy. Dextran is shown in red, EEA1 in green, and LBPA in blue. Colocalization between dextran and EEA1 is shown in yellow, and colocalization between dextran and LBPA is in purple.

Figure 6.

Different signaling downstream of the EGFR in cells depleted of Tsg101 or hVps24. (A) HeLa cells treated with scrambled RNA, siRNA against Tsg101, or siRNA against hVps24 were plated out in one six-well dish each and starved for 3 h in the presence of cycloheximide. All wells except one in each dish were given a 5-min pulse of EGF and then chased in normal medium supplied with cycloheximide for the indicated times. The cells were then lysed and analyzed by SDS-PAGE and antibodies against phospho-MEK, total MEK, phospho-ERK, and total ERK as described in Materials and Methods. The intensities of the phospho-MEK (B) and phospho-ERK (C) bands were quantified by using the software provided by the ChemiGenius imaging system and plotted as percentage of the respective phosphorylation intensities after 5 min of EGF stimulation and 0 min of chase. Intensities are adjusted for different loadings. Note that the values in C are obtained from a separate set of experiments with shorter chase times, an example of which is provided in Supplemental Figure S1. Control cells, square symbols; Tsg101 siRNA-treated cells, triangle symbols; and hVps24 siRNA-treated cells, round symbols. The error bars represent SEM of four experiments.

Figure 7.

Effect of hVps24 and TSG101 ablation on the vesicular pH of EGF-containing endocytic vesicles. Ratiometric fluorescence video image analysis of internalized EGF-FITC (and transferrin-FITC; our unpublished data) containing vesicles was performed in live HeLa cells transfected with the indicated siRNA as described in Materials and Methods. The mean pH of vesicle population was calculated by the Origin software, measured in five independent experiments (±SEM). In each experiment, the luminal pH of 100–300 endocytic vesicles was determined.

Figure 8.

Depletion of hVps24 causes accumulation of small MVEs. EM of EGFR-labeled endosomes/MVEs. (A) EGFR labeling in control cell within a typical endosome with multivesicular appearance. Note that the MVE contains numerous intraluminal vesicles and that most of the labeling is found on these vesicles. (C) Overview of a group of endosomes illustrating the homogenous morphology of EGFR-positive endosomes. (B) In RNAi-treated cells, the EGFR-positive vesicles are significantly smaller than control MVEs and contain seemingly fewer intraluminal vesicles. EGFR labeling is found on these vesicles, which do not seem so homogeneous in size compared with control cells. (D) Overview of a group of endosomes and lysosomes in an RNAi-treated cell. Arrowheads in C and D indicate EGFR labeling. Bars, 200 nm. (E) Estimation of the mean vesicle diameter of EGFR-positive endosomes in control and RNAi-treated cells (control cells, n = 76; RNAi-treated cells, n = 68; 3 separate experiments). The difference in diameter between the two groups is ∼30%. Significance level by Student's t test (p ≤ 0.005).