Autophagy is a catabolic process that delivers cytoplasmic material to the lysosome for degradation. The mechanisms regulating autophagosome formation and size remain unclear. Here, we show that autophagosome formation was triggered by the overexpression of a dominant-negative inactive mutant of Myotubularin-related phosphatase 3 (MTMR3). Mutant MTMR3 partially localized to autophagosomes, and PtdIns3P and two autophagy-related PtdIns3P-binding proteins, GFP-DFCP1 and GFP-WIPI-1α (WIPI49/Atg18), accumulated at sites of autophagosome formation. Knock-down of MTMR3 increased autophagosome formation, and overexpression of wild-type MTMR3 led to significantly smaller nascent autophagosomes and a net reduction in autophagic activity. These results indicate that autophagy initiation depends on the balance between PI 3-kinase and PI 3-phosphatase activity. Local levels of PtdIns3P at the site of autophagosome formation determine autophagy initiation and the size of the autophagosome membrane structure.
Macroautophagy (hereafter referred to as autophagy) is a highly conserved intracellular degradation pathway that mediates the turnover of long-lived proteins and organelles (1,2). Autophagy is induced by cellular starvation, and increased autophagic activity leads to recycling of cellular macromolecules and promotes cell survival. During autophagy, a unique double membrane-bound structure, the autophagosome, is formed that ultimately fuses with lysosomes, leading to the degradation and sequestration of the degraded substrates (3).
Phosphoinositides (PIs), the phosphorylated derivatives of phosphatidylinositol (PtdIns), are involved in the regulation of diverse cellular events including cytoskeletal dynamics, signaling, nuclear processes and membrane trafficking in eukaryotic cells, and a role for PI in autophagy induction and regulation has recently been demonstrated. Among the PIs, PtdIns3P is generated by the action of PI 3-kinases, which phosphorylate PtdIns or PIs at the 3-position of the inositol ring (4). The class III PI 3-kinase, hVps34, and its regulatory subunit hVps15, along with Atg14L, and Beclin-1 are intimately involved in autophagy (5–8). Additionally, several lines of evidence have suggested a role for PI signaling in autophagy. Treatment with wortmannin, a potent PI 3-kinase inhibitor, suppresses autophagy (9). Additionally, PtdIns3P localizes to the inner membrane of the yeast autophagosome, and WIPI-1α (WIPI49/Atg18) binds PtdIns3P and localizes to the autophagosome (10–12). Finally, DFCP1, a PtdIns3P binding protein, accumulates at the omegasome, the structure thought to be the site of autophagosome formation (13). However, despite numerous studies, the precise role of PtdIns3P in autophagy remains unclear.
The cellular levels of PIs are reciprocally regulated by the action of PI phosphatases and PI kinases (14). The myotubularin (MTM) family are PI 3-phosphatases with specificity for PtdIns3P and PtdIns(3,5)P2(15), and there are 16 family members characterized by a conserved protein tyrosine phosphatase (PTP) domain (16). MTMR3 (KIAA0371) is ubiquitously expressed, but its specific role is not known. It contains the conserved MTM phosphatase domain, but it also contains an FYVE domain at its C-terminus and an N-terminal PH-GRAM (PH-G) domain (15,17). However, unlike conventional FYVE domains, the FYVE domain of MTMR3 does not confer endosomal localization or bind PtdIns3P (18). In contrast, the PH-G domain binds PIs (18). Additionally, yeast two-hybrid and coimmunoprecipitation experiments showed that MTMR3 binds MTMR4 (17), and overexpression of a catalytically inactive MTMR3 mutant (C413S) in mammalian cells induces the formation of unique vacuolar compartments characterized by cup-shaped membranous structures (19). These structures resembled autophagosomes, but their association with the autophagosomal markers LC3 or Atg5 was not performed (1,2,20).
In this study, we demonstrate that MTMR3 modulates the local PtdIns3P levels and negatively regulates autophagy. These results suggest that autophagy initiation and the regulation of autophagosome size are controlled by local PtdIns3P levels.
To examine whether MTMR3 functions in autophagy, A549 cells stably expressing GFP-LC3 or GFP-Atg5, both specific markers of autophagic structures, were transfected with wild-type mStrawberry-MTMR3 or the inactivated dominant-negative C413S mutant using an adenoviral expression system. Cells were then analyzed by confocal laser fluorescence microscopy. The overexpression of mStrawberry-MTMR3C413S increased the number of GFP-LC3 and GFP-Atg5 puncta even under nutrient-rich conditions (Figure 1). This was not observed in cells overexpressing wild-type MTMR3, or transfected with empty vector or mutant MTM1 and MTMR2 (Figures 1 and S1). Furthermore, immunoelectron microscopic analysis confirmed that GFP-LC3 labeled the double-membrane structures generated in mStrawberry-MTMR3C413S overexpressing cells, demonstrating that these structures are morphologically similar to autophagosomes (see below). PI 3-kinase activity is required for autophagy, and treatment with wortmannin, a potent PI 3-kinase inhibitor, suppresses autophagy (9). In MTMR3C413S overexpressing cells, the observed increase in the number of GFP-Atg5 puncta was PtdIns3P level dependent because wortmannin treatment impaired puncta formation (Figure S3). Additionally, mStrawberry-MTMR3C413S colocalized with both GFP-LC3 (Figure 1A) and GFP-Atg5 puncta (Figure 1B) as well as demonstrating an endoplasmic reticulum (ER) localized pattern. Thus, these data indicate that there is a direct relationship between MTMR3 function and autophagy.
MTMR3 is a PtdIns3P phosphatase, and to assess the effect of MTMR3 on the cellular distribution of PtdIns3P, we used recombinant GST-2xFYVE to monitor the localization of PtdIns3P within cells. In MTMR3C413S-overexpressing cells, the number of GST-2xFYVE puncta was increased under nutrient-rich conditions compared to control cells (Figure 2A,B). Moreover, the distribution of MTMR3C413S substantially overlapped with both GST-2xFYVE and GFP-LC3 (Figure 2C), demonstrating that the amount of PtdIns3P is focally increased in areas of MTMR3C413S accumulation on autophagosomes. In control cells, the number of GST-2xFYVE puncta was increased under starvation conditions (Figure 2A,B), but overexpression of wild-type MTMR3 abrogated these increases (Figure 2A,B).
WIPI-1α (WIPI49/Atg18), a homolog of the yeast autophagy protein Atg18, directly binds PtdIns3P and forms puncta in mammalian cells in a PI 3-kinase-dependent manner (11,12). Under starvation conditions, WIPI-1α accumulates at LC3-positive membrane structures, and WIPI-1α puncta formation correlates with autophagosome formation (11,12). Thus, we next examined the localization of WIPI-1α in cells overexpressing wild-type and C413S MTMR3. WIPI-1α puncta formation was extremely limited in control and MTMR3wt-overexpressing cells under nutrient-rich conditions (Figure 3A), but, under the same conditions, there was an increased number of WIPI-1α puncta (Figure 3A). Additionally, these WIPI-1α-containing puncta colocalized with MTMR3C413S (Figure 3A).
The PtdIns3P-binding protein DFCP1 is associated with autophagosome formation, and DFCP1 localizes to the ER and omegasomes, the structure where the autophagosome formation is thought to take place inside it (13). The localization of DFCP1 is dependent upon its PtdIns3P binding and ER localization motif (13). Given the relationship of PtdIns3P, DFCP1 and autophagosome formation, we wished to further characterize the possible relationship between DFCP1 and MTMR3. Under nutrient-rich conditions, few DFCP1-positive omegasomes were present within cells, but, upon induction of starvation, omegasomes and DFCP1-positive puncta appeared. In control cells and MTMR3wt-overexpressing cells, few GFP-DFCP1 puncta localized to omegasomes under nutrient-rich conditions (Figure 3B,C). In contrast, the number of GFP-DFCP1 spots was significantly increased in cells expressing MTMR3C413S under nutrient-rich conditions (Figure 3B,C). Furthermore, DFCP1 colocalized with MTMR3 (Figure 3B). Thus, overexpression of inactive MTMR3 leads to increased PtdIns3P level at pre-autophagosomal structures. These results suggest that MTMR3 localizes at autophagosomes and/or omegasomes, and may regulate the local levels of PtdIns3P.
To determine if there is a direct role for MTMR3 in autophagy, we used RNAi to knock down MTMR3 expression. From three siRNA sequences tried, we selected two different sequences that efficiently knocked down MTMR3 expression (Figure S4). Compared to cells transfected with control siRNA, when MTMR3 was knocked down, the numbers of GFP-LC3, -Atg5 and -WIPI-1α puncta were significantly increased (Figure 4). These data demonstrate that MTMR3 directly suppresses the initiation of autophagy under nutrient-rich conditions.
The data we have presented argue that inactivation of MTMR3 initiates autophagosome formation, but whether the formation and subsequent fusion to endo/lysosomes is completed is not determined. To ask whether the autophagic flow is complete, we used a tandemly-tagged fluorescent protein LC3 construct (mRFP-GFP-LC3), tfLC3. Previous studies have shown that mRFP(+)GFP(+)-LC3 spots represent autophagosomes or their precursors and mRFP(+)GFP(-)-LC3 spots indicate autolysosomes, the fusion products of autophagosomes and lysosomes, because only the GFP signal is susceptible to lysosomal acidic and proteolytic degradation (21). In MTMR3C413S-overexpressing A549 cells, the proportion of mRFP(+)GFP(-)-LC3 spots increased, indicating that autophagy proceeds normally and to completion (Figures 5A,B and S5). In contrast, the percentage of mRFP(+)GFP(-)-LC3 puncta was significantly reduced in MTMR3wt-overexpressing cells compared to control transfected cells under nutrient-rich conditions, suggesting that the basal autophagy was frustrated by MTMR3 overexperession (Figures 5A,B and S5).
We next measured the effect to MTMR3 overexpression on the degree of protein degradation seen during autophagy. In cells overexpressing MTMR3C413S, autophagy-dependent protein degradation was significantly increased under nutrient-rich conditions, but it was not changed by cell starvation (Figures 5C and S6). In contrast, degradation of long-lived proteins was significantly decreased under both nutrient-rich and starvation conditions in MTMR3wt-overexpressing cells to a similar extent as that seen following Atg4BC74 A overexpression, an Atg4B mutant known to completely inhibit autophagy (Figures 5C and S6) (22). Overexpression of wild-type MTMR3 also impaired the DFCP1-positive puncta formation under starvation conditions (Figure S2). This effect is scarcely seen by overexpression of wild-type MTM1 or MTMR2, suggesting specific role of MTMR3 (Figure S2). These results show that MTMR3 negatively regulates autophagic degradation.
Intriguingly, the number of GFP-LC3 and GFP-Atg5 puncta in MTMR3wt-expressing cells was approximately equal to that of control cells (Figure 1) despite the suppression of autophagy (Figure 5). However, the pixel sizes of the GFP-LC3 signals in MTMR3wt-overexpressing cells were significantly smaller than in control cells (Figures 5A and 6A). We hypothesized that autophagy was initiated to some extent in MTMR3wt-expressing cells, but, because either only small autophagosomes were formed or the formation of autophagosomes was interrupted, autophagy could not be driven to the autophagosome/lysosome fusion stage. We, therefore, observed cells by electron microscopy (EM) under starvation conditions and counted the number of both autophagosomes and autolysosomes over 500 nm in diameter (Figure 6B,C). The size and number of autophagosomes or autolysosomes did not differ between MTMR3C413S-overexpressing cells and control cells (Figure 6B,C,d,e,h,i). In contrast, the number of autophagosomes or autolysosomes over 500 nm in diameter was significantly lower in cells overexpressing MTMR3wt (Figure 6C). However, autophagosome- or autolysosome-like structures less than 500 nm in diameter were frequently observed in MTMR3wt-overexpressing cells (Figure 6B,f,g). To investigate whether these structures are related to autophagy, we examined A549 cells stably expressing GFP-LC3 by immunoelectron microscopy with anti-GFP antibodies following transfection with mStrawberry-empty vector, -MTMR3wt or -MTMR3C413S. Nearly all of the membrane structures less than 500 nm in diameter observed in MTMR3wt-expressing cells were labeled with gold particles (Figure 6D), indicating that these small membrane structures were autophagosomes. These results suggest that the local levels of PtdIns3P affect both autophagy initiation and autophagosome size.
In this study, we identified several novel aspects of autophagy regulation including a role for MTMR3 in autophagosome formation and maintenance. Our data clearly demonstrate that autophagy is induced by increased local levels of PtdIns3P in MTMR3C413S-overexpressing cells under nutrient-rich conditions, and this suggests that autophagy effector molecules are ‘primed’ and able to rapidly respond to the changes in PtdIns3P levels. Additionally, local PtdIns3P level is a key factor determining autophagy induction, and increased PtdIns3P appears both necessary and sufficient for the induction of autophagy. Autophagosome formation in MTMR3C413S-expressing cells was less than that seen under starvation conditions, and this is likely because of poor recruitment of the autophagy-specific class III PI 3-kinase complex under nutrient-rich conditions (5). Consistent with this, overexpression of MTMR3C413S did not affect autophagic degradation under starvation conditions, and we hypothesized that the MTMR3-mediated suppression of autophagy was most relevant under nutrient-rich conditions and these effects became negligible under starvation conditions. For increased autophagic activity to occur under starvation conditions, MTMR3 function may be inhibited either by its degradation or modification. In addition, the autophagy-specific PI 3-kinase complex is recruited to the autophagosome, leading to elevated local PtdIns3P levels during cell starvation (5). mTOR, a key regulator of autophagy, is a prime candidate protein to regulate this balance.
We showed that excess MTMR3wt suppresses autophagy (Figures 5, 6, S2, S5 and S6), and this is the first evidence demonstrating that PtdIns3P itself, not just PI 3-kinase activity, is needed for autophagy in mammalian cells. Our EM analyses showed that normal autophagosomes are not formed in MTMR3wt-overexpressing cells, but small, LC3-positive autophagosomes were seen in these cells. Interestingly, treatment of cells with wortmannin completely inhibits LC3-puncta formation (20), but LC3-positive puncta were present in cells overexpressing MTMR3wt (Figures 5A and 6). Thus, the high levels of wild-type MTMR3 in these cells did not completely eliminate the local generation of PtdIns3P, and the residual PtdIns3P allowed the formation of the small autophagosomes seen. These results further suggest that autophagosome size is also regulated by PtdIns3P. During yeast autophagosome formation, the LC3 homolog Atg8 is thought to be important for determination of autophagosome size (23), and our study is the first suggestion that PtdIns3P plays a role in this process. Under nutrient-rich conditions, the autophagy is suppressed to basal levels, and both rates of autophagosome formation and the size of formed autophagosomes may contribute for the overall levels of autophagy.
Finally, our data clearly demonstrate the direct involvement of MTMR3 in the regulation of autophagy. MTMR3 is one of 16 members of the human MTM family of PI phosphatases (16), and there is a high degree of redundancy among the cellular roles of PI phosphatases (14,24). Recently, Vergne et al. reported that the MTM family protein jumpy (MTMR14) also plays an important role in autophagy (25). Jumpy activity was seen under both starvation and nutrient-rich conditions, while our data show that MTMR3 is important only under nutrient-rich conditions. Additionally, they showed that MTMR3 knock-down does not affect to the LC3 lipidation level (25). However, we observed that autophagy was induced by MTMR3 depletion, but this was to a lesser extent than seen following MTMR3C413S overexpression (Figures 1, 3, 4 and S4). Moreover, MTMR3C413S specifically localized to the autophagosome, in addition to the ER, strongly suggesting a direct role for MTMR3 in autophagy, possibly in redundant way with other PI phosphatase(s) including jumpy. An interesting problem to pursue is how the localization of MTMR3 is determined.
Based on the data presented here, we propose that autophagy is regulated by local PtdIns3P levels on nascent autophagosomes, and this, in turn, is tightly controlled by the nutrient status of the cell. Such a fine degree of control requires the reciprocal regulation on PI phosphorylation by kinases and phosphatases, especially under nutrient-rich conditions in which low, basal levels of autophagy are observed. Further studies are needed to further illuminate the regulatory mechanisms controlling autophagy induction and completion.
Materials and Methods
The following antibodies were used: anti-c-myc (clone 9E10; Gentaur Molecular Products); anti-GST (sc-459; Santa Cruz); anti-GFP [MP 06455, Molecular Probes]; AlexaFluor® 405 Goat Anti-mouse IgG (Molecular Probes); and AlexaFluor® 568 Goat Anti-rabbit IgG (Molecular Probes).
Cell culture, plasmid transfections, adenovirus infections and generation of stable cell lines
HEK-293 cells stably expressing GFP-DFCP1 (clone 201) were previously described (13) and Plat-E cells were kindly provided by Dr. T. Kitamura [The University of Tokyo; (26)]. A549, HEK-293 and Plat-E cells were grown in DMEM (Sigma) containing 10% FBS (GIBCO) and 4 mml-glutamine (Invitrogen) in a 5% CO2 incubator at 37°C. For nutrient starvation, cells were washed once with Eisen's balanced salt solution (EBSS) (Sigma) before incubation in EBSS, and cells were cultured for 1–3 h. Transient transfections were carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol or the protocol associated with the adenovirus expression system. Adenovirus infections were carried out as follows: on the day before infection, ∼2 × 105 cells were plated into 6-well plates and incubated at 37°C overnight in a CO2 incubator. The adenovirus-containing medium was added to the culture medium, and cells were incubated for 12 h. The medium was then replaced with 2-mL growth medium and incubated for 24 h prior to experiments. Stable transformants were established by retrovirus infection and selection in growth medium with 2–5 µg/mL puromycin.
Plasmids, recombinant protein, recombinant adenoviruses and recombinant retroviruses
The plasmid encoding monomeric red fluorescent protein (pmStrawberry) was a generous gift from Dr. Roger Y. Tsien [University of California, San Diego, CA; (27)]. The plasmid encoding GST-2xFYVE was a kind gift from Dr. H. Stenmark (the Norwegian Radium Hospital, Oslo, Norway) and recombinant protein was purified as described (28). The open reading frame (ORF) for human MTMR3 was amplified by PCR from pBluescript-KIAA0371 (Kazusa DNA Research Institute in Chiba), and subcloned into pmStrawberry-C3 using engineered HindIII and SalI sites. C413S mutants were generated by site-directed mutagenesis (QuickChange, Stratagene) and subcloned into appropriate expression vectors. To produce recombinant adenoviruses, the cDNAs corresponding to mStrawberry, EGFP or myc, mStrawberry-, EGFP or myc-tagged-MTMR3wt or -MTMR3C413S were subcloned into the pENTR 1A plasmid (Invitrogen). The cDNA inserts in pENTR-1A were transferred to the pAd/CMV/V5-DEST vector (Invitrogen) with the Gateway system using LR clonase (Invitrogen). Recombinant adenoviruses were prepared with the ViraPower Adenovirus Expression System (Invitrogen) according to the manufacturer's instructions. pMRX-IRES-puro was donated by Dr. S. Yamaoka [Tokyo Medical and Dental University, Japan; (29)]. For production of recombinant retroviruses, the cDNA corresponding to enhanced green fluorescent protein (EGFP)-WIPI-1α was transferred to pMRX-IRES-puro. Recombinant retroviruses were prepared as described previously (29).
Cells cultured on glass coverslips were fixed with 3% formaldehyde (FA) in PBS for 10 min. Samples were analyzed by a laser scanning confocal microscope, FV1000 (Olympus). The number or size of GFP-LC3 or -Atg5 puncta was analyzed by G-Count (G-Angstrom).
Assessment of the amount of PtdIns3P by GST-FYVE
A549 cells were transfected with control vector, MTMR3wt and MTMR3C413S by adenovirus expression system. After infection for 24 h, the cells were subcultured on glass coverslips and incubated for 24 h. Cells were then washed with EBSS medium and cultured in EBSS for 4 h. Twenty minutes prior to fixation with 3% FA in PBS, the cells were prepermeabilized with 0.05% saponin in EBSS for 5 min. After blocking for 30 min with 1% BSA in PBS, cells were incubated with 20 µg/mL GST-2xFYVE for 1 h and stained with anti-GST and/or anti-myc(9E10) followed by AlexaFluor® 568 and/or AlexaFluor® 405 secondary antibodies. These samples were analyzed by a FV1000 (Olympus) laser scanning confocal microscope.
The siRNAs for MTMR3 knock-down were designed using siDESIGN Center (http://www.dharmacon.com/DesignCenter/DesignCenterPage.aspx), and three different sequences were selected. A549 cells were transfected twice with siRNAs using Lipofectamine RNAiMAX reagent (Invitrogen) according to the manufacturer's protocol. The second transfection was done 24 h after the first transfection, and the cells were analyzed 24 h after second transfection.
Evaluation of autophagy by tfLC3
A549 cells were transfected with empty virus, myc-MTMR3wt or -MTMR3C413 S encoding adenovirus. After overnight culture, the cells were infected with adenoviruses encoding tfLC3 for 4 h. Cells were subcultured on glass coverslips and incubated for 24 h. Cells were fixed with 3% FA in PBS for 10 min, washed three times with PBS, and permeabilized with 50 µg/mL digitonin in PBS for 10 min. After washing three times with PBS, immunostaining was carried out using anti-myc and cells were analyzed using a FV1000 (Olympus) laser scanning confocal microscope.
Long-lived protein degradation assay
A549 cells expressing mStrawberry-empty vector, -MTMR3wt or -MTMR3C413S with or without adenoviral transfected FLAG-tagged Atg4BC74A were plated on 12-well plates and cultured in DMEM for 8–12 h. Cells were then incubated for ∼18 h in DMEM including 0.6 µCi/mL l-[14C] valine and 1/100 vol. penicillin-streptomycin (GIBCO). Cells were washed three times with 500 µL DMEM and incubated for 4 h in DMEM containing 10 mm cold valine. After three washes with DMEM, the cells were incubated in either DMEM or EBSS supplemented with 0.1% BSA and 10 mm cold valine for 4 h. The medium was subsequently precipitated with 10% trichloroacetic acid (TCA), and TCA-soluble radioactivity was measured. Cells were lysed with RIPA buffer [150 mm NaCl, 50 mm Tris–HCl (pH 7.5), 5 mm ethylenediaminetetraacetic acid (EDTA), 0.1% SDS, 1% TritonX-100, 1× protease inhibitor cocktail and 1 mm phenylmethylsulphonyl fluoride (PMSF)] and precipitated in 10% TCA. The precipitates were washed once with acetone. Total cell fraction was measured after solubilization with 6 m urea. l-[14C] valine release was estimated as a percentage of the radioactivity in the TCA-soluble material relative to the total cell radioactivity.
A549 cells stably expressing GFP-LC3 were transfected with adenovirally encoded mStrawberry-empty vector, -MTMR3wt or -MTMR3C413S. The cells, cultured on glass-bottom dishes (MatTek), were washed once with EBSS before incubation in EBSS. After 3-h incubation, the cells were fixed with 4% FA in 0.1 m sodium phosphate (pH 7.4) for 1 h at room temperature, and examined using a FV1000 (Olympus) confocal laser scanning microscope. The same specimens were further incubated with 2.5% glutaraldehyde and 2% FA in 0.1 m sodium phosphate (pH 7.4) at 4°C overnight. After three washings with 0.1 m sodium phosphate (pH 7.4) containing 7% sucrose, the samples were postfixed with 1% osmium tetroxide and 0.5% potassium ferrocyanide for 1 h, washed three times with distilled water, dehydrated in ethanol, and embedded in Epon812 (TAAB Laboratories Equipment, Ltd.). Ultrathin sections (70-nm thick) were stained with saturated uranyl acetate and Reynolds lead citrate solution. The electron micrographs were taken with a JEOL JEM-1011 transmission electron microscope (JEOL, Ltd.). The ratio of the profile area of the object in the sections relative to total area of the sections (VV) were obtained as described (30).
The mStrawberry-empty vector, -MTMR3wt or -MTMR3C413S were transfected into A549 cells stably expressing GFP-LC3 using the adenoviral expression system. After ∼36 h in culture, cells were washed with PBS and incubated in EBSS for 3 h. Adherent cells were fixed by adding an equal volume of 2× concentrated fixative (8% FA in H2O [EMS], 0.2% glutaraldehyde [GA in H2O, EMS] in PHEM [containing 120 mm piperazine-1,4-bis(2-ethanesulfonic acid)/PIPES, 100 mm 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid/HEPES, pH 6.9, 40 mm ethylene glycol tetraacetic acid/EGTA, 4 mm MgCl2]) to the cell culture medium. After 10 min this mixture was removed and replaced with 1× fixative (4% FA, 0.1% GA in HEM, pH 6.9) for 90 min at room temperature. Cells were scraped and centrifuged. Cell pellets were washed by resuspending with PBS-glycine four times and incubated for 10 min in PBS-glycine at 37°C to quench aldehyde groups. Cells were then resuspended and incubated in 10% gelatin/PBS for 10 min at 37°C, centrifuged and immediately placed on ice until the gelatin solidified. The pellet was cut into 1-mm3 pieces. Gelatin-embedded samples were cut into small pieces, infiltrated with 2.3 m sucrose (as a cryoprotectant) at 4°C over night on a rotating wheel, mounted onto sample pins and frozen in liquid nitrogen. Subsequently, 60–70-nm feed cryo sections were cut with a FC6/UC6-cryo ultramicrotome (Leica) and a 45° diamond knife (Diatome), picked up with a 1:1 mixture of 2% methyl cellulose (25 centipoises, Sigma-Aldrich) and 2.3 m sucrose (USB Corporation) (31). On gelatin plates, thawed sections were incubated in blocking solution [1.5% BSA (w/v), 0.1% (w/v)] fish skin gelatin in PBS] for 15 min at room temperature. Sections were incubated with primary antibody [anti-GFP (MP 06455, Molecular Probes)] diluted in blocking solution for 30 min at room temperature. Sections were washed with PBS and incubated with 10-nm protein-A gold diluted in blocking buffer for 30 min. Sections were washed with PBS, and extensively washed with water. Sections were then stained/embedded in 4% UA/2% methyl cellulose mixture (ratio 1:9) (32). Images were recorded with JEM-1011 JEOL [with a top-mounted 2k × 2k CCD Camera (Veleta, Olympus)].
All values shown in the figures are represented with SD. Statistical significance (p value) was determined by two-tailed Student's t test.
The authors thank Dr. Harald Stenmark (the Norwegian Radium Hospital, Oslo, Norway) for the gift of GST-2xFYVE, Dr. Roger Y. Tsien (University of California, San Diego, CA) for the gift of mStrawberry cDNA, Dr. Shoji Yamaoka for the gift of pMRX-IRES-puro, Dr. Toshio Kitamura (The University of Tokyo, Japan) for the gift of Plat-E cells, Dr. Shunsuke Kimura (Yoshimori lab) and Kentaro Ikegami (Yoshimori lab) for the gift of adenoviruses encoding tfLC3, and Christopher K.E. Bleck (Yoshimori lab) for technical assistance. The work was supported in part by the Special Coordination Funds for Promoting Science, Technology of the Ministry of Education, Culture, Sports, Science and Technology (MEXT). N. T is supported by the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for JSPS Fellow.