Early endosomes and endosomal coatomer are required for autophagy

Abstract

Autophagy, an intracellular degradative pathway, maintains cell homeostasis under normal and stress conditions. Nascent double-membrane autophagosomes sequester and enclose cytosolic components and organelles, and subsequently fuse with the endosomal pathway allowing content degradation. Autophagy requires fusion of autophagosomes with late endosomes, but it is not known if fusion with early endosomes is essential. We show that fusion of AVs with functional early endosomes is required for autophagy. Inhibition of early endosome function by loss of COPI subunits (beta', beta, or alpha) results in accumulation of autophagosomes, but not an increased autophagic flux. COPI is required for ER-Golgi transport and early endosome maturation. Although loss of COPI results in the fragmentation of the Golgi, this does not induce the formation of autophagosomes. Loss of COPI causes defects in early endosome function, as both transferrin recycling and EGF internalization and degradation are impaired, and this loss of function causes an inhibition of autophagy, an accumulation of p62/SQSTM-1, and ubiquitinated proteins in autophagosomes.

Figure 1.

siRNA depletion of COPI increases autophagy. (A) siRNA depletion of β′-, β-, and α-COPI (top) and Sec23A and B (bottom) in 293/GFP-LC3 cells for 48 h. Decreased protein levels of the corresponding subunits is shown by immunoblots. (B) Loss of β′-, α-, and β-COP but not subunits of COPII (Sec 23A and B) increase GFP-LC3-II. In A and B, tubulin was the loading control. (C) GFP-LC3–positive vesicles were detectable in 293/GFP-LC3 cells after siRNA depletion of β′-, α-, and β-COP but only basal levels of GFP-LC3–positive vesicles are observed in control siRNA-depleted cells.

Figure 2.

Time course of COPI depletion shows that Golgi disperses before AVs are formed, but Golgi dispersal does not cause AV formation. (A) COPI subunits were depleted in 293/GFP-LC3 cells and analyzed 24, 36, and 48 h after addition of siRNA. Immunoblotting for β′- and α-COP confirms loss of COP subunits. (B) GFP-LC3-I and -II were monitored by immunoblots in parallel lysates using anti-GFP antibodies at the indicated times. (C) After 24 h siRNA treatment, depletion of β′-COP caused morphological changes and a reduction in perinuclear population of GM130. Box indicates enlarged area. (D) Rab1a/b was depleted using siRNAs for 48 h. Lysates were probed with anti-Rab1 to confirm depletion. (E) β′-COP or Rab1a/b were depleted as in D, and analyzed by indirect immunofluorescence using anti-GM130 and GFP fluorescence. In β′-COP panel, the asterisk indicates cells that did not show an accumulation of GFP-LC3–positive AVs.

Figure 3.

AV formation after siRNA depletion of COPI requires Atg5 and Atg7. 293/GFP-LC3 cells were incubated with siRNA for β′-, α-COP, or control siRNA alone or combined with siRNA against Atg5 or Atg7 for 48 h, then (A) lysed and analyzed by immunoblot for Atg5, Atg7, GFP-LC3, and tubulin (loading control), or (B) fixed for immunofluorescence analysis using anti-ERGIC-53 antibodies, or GFP fluorescence. In A, the asterisk indicates nonspecific band; in B and C, asterisk indicates cells that did not show a fragmented ERGIC-53.

Figure 4.

Correlative light-electron microscopy analysis of GFP-LC3 puncta demonstrates they are AVs. (A) Phase, (B) fluorescent, and (C) TEM of the same field of 293/GFP-LC3 cells 48 h after siRNA depletion of α-COP. In B the image was inverted to allow better visualization of the GFP-LC3–positive structures. (D) Higher magnification of cell shown with asterisk in A and C. Top region of cell of interest is enlarged in E to show the two AVs structures indicated by arrows in B. (F and G) Higher magnification of AVs shown by arrows in E.

Figure 5.

Ub and p62 accumulate in α- and β′-COP–depleted 293/GFP-LC3 cells. (A) 293/GFP-LC3 cells were treated with control or β′-COP siRNA for 48 h, then fixed and labeled with anti-p62 and anti-ubiquitin antibodies for colocalization with GFP-LC3. (B) HEK293 cells treated as in A and labeled with anti-ubiquitin and anti-p62 antibodies. (C) Cryosections from β′-COP–depleted cells were labeled with anti-Ub antibodies followed by 10-nm protein A–gold. AV, autophagosomes.

Figure 6.

AVs in COPI-depleted cells are not degradative. (A) Long-lived protein degradation was assessed 72 h after transfection with control or β′-COP siRNA in 293/GFP-LC3 cells. Cells were either starved of amino acids for 2 h (St) or incubated in fresh growth medium (Fed) for 2 h. The amount of protein degradation is presented as mean ± SEM. (B) siControl-treated cells and β′- and α-COP were depleted for 48 h, then incubated with Lysotracker red (50 nM) for 30 min before fixation. Merged image shows very little colocalization of GFP-LC3 AV-positive structures. Asterisk shows cell enlarged in inset. (C) After β′- and α-COP siRNA depletion, Magic Red-RR₂ (MR) was used to detect active cathepsin B in cells which were fed or starved (St) for 2 h. The intensity of MR labeling was measured in live cells by flow cytometry.

Figure 7.

GFP-LC3 AVs are LAMP2 positive. (A) 293/GFP-LC3 cells were incubated with control or β′- or α-COP siRNA for 48 h, fixed and labeled with anti-LAMP2 antibodies, followed by Alexa 555 anti–mouse antibodies. After COPI depletion many GFP-LC3–positive AVs are positive for LAMP2. Asterisk indicates cell enlarged in inset. (B) Cryosections of β′-COP-depleted cells labeled with anti-GFP (left) or anti-GFP and anti-LAMP2 (right) followed by protein A–gold as indicated. AV, autophagosomes, M, mitochondria.

Figure 8.

GFP-LC3–positive AVs colocalize TGN46 and early endocytotic markers after COPI depletion. (A) 293/GFP-LC3 cells were incubated with siRNA for β′-COP or control siRNA (Cont) for 48 h, processed for immunofluorescence, and labeled with antibodies for TGN46. In (B) siRNA-treated cells were fixed and labeled, or starved for 2 h (St), and labeled with antibodies to EEA1. (C) Control-starved HEK293 cells were labeled for endogenous LC3 and EEA1 by double labeling after methanol fixation. (D) After siRNA treatment, rhodamine-labeled dextran was internalized for 20 min, followed by a 40-min chase after which the 293/GFP-LC3 cells were fixed and analyzed by confocal microscopy.

Figure 9.

Transferrin and EGF trafficking is inhibited by COPI depletion. 293/GFP-LC3 cells were treated with β′-COP siRNA or control siRNA. After 48 h cells were incubated with Alexa 555 transferrin (A) for 2 min, (B) 30 min, or (C) 30 min followed by a 30-min chase, then washed and fixed and analyzed by confocal microscopy. (D and E) ¹²⁵I-EGF was added to HeLa cells after 48 h of siRNA depletion of β′-COP, or control siRNA for 10 min. The cells were then washed as described in the Materials and methods, and chased for 30, 60, 120, and 180 min (including 10-min incubation). (D) The amount of ¹²⁵I-EGF internalized was shown as a percentage of the total ¹²⁵I-EGF taken up in 10 min. (E) The amount of degraded ¹²⁵I-EGF was determined as a percentage of the total radioactivity present after TCA precipitation over the total ¹²⁵I-EGF internalized. The experiment was performed three times in duplicate and the data shown is the mean ± SEM. The significance was determined by Student's t test. *, P ≤ 0.05.