If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
[S] The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1–S5. * This work was supported in part by grants-in-aid for Scientific Research on Priority Areas and Targeted Proteins Research Program from Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan.
The autophagy-related protein 8 (Atg8) conjugation system is essential for the formation of double-membrane vesicles called autophagosomes during autophagy, a bulk degradation process conserved among most eukaryotes. It is also important in yeast for recognizing target vacuolar enzymes through the receptor protein Atg19 during the cytoplasm-to-vacuole targeting (Cvt) pathway, a selective type of autophagy. Atg3 is an E2-like enzyme that conjugates Atg8 with phosphatidylethanolamine. Here, we show that Atg3 directly interacts with Atg8 through the WEDL sequence, which is distinct from canonical interaction between E2 and ubiquitin-like modifiers. Moreover, NMR experiments suggest that the mode of interaction between Atg8 and Atg3 is quite similar to that between Atg8/LC3 and the Atg8 family interacting motif (AIM) conserved in autophagic receptors, such as Atg19 and p62. Thus, the WEDL sequence in Atg3 is a canonical AIM. In vitro analyses showed that Atg3 AIM is crucial for the transfer of Atg8 from the Atg8∼Atg3 thioester intermediate to phosphatidylethanolamine but not for the formation of the intermediate. Intriguingly, in vivo experiments showed that it is necessary for the Cvt pathway but not for starvation-induced autophagy. Atg3 AIM attenuated the inhibitory effect of Atg19 on Atg8 lipidation in vitro, suggesting that Atg3 AIM may be important for the lipidation of Atg19-bound Atg8 during the Cvt pathway.
Autophagy is an intracellular degradation system conserved among eukaryotes from yeasts to mammals. In higher eukaryotes, autophagy plays a crucial role in fundamental biological processes such as intracellular clearance, differentiation, development, programmed cell death, and antigen representation, and autophagy dysfunction is associated with severe diseases such as neurodegenerative disorders and cancers (
). Although autophagy is in principle a non-selective degradation process, it also mediates selective degradation of various targets that include aggregated proteins, surplus or damaged organelles, and even invasive bacterial cells (
). In yeast, aminopeptidase 1 (Ape1) is selectively transported into the vacuole via an autophagy-like process called the cytoplasm-to-vacuole targeting (Cvt) pathway under vegetative conditions and via autophagy under starvation conditions that strongly induce autophagy (
). Because of its mechanistic similarity to autophagy, the biosynthetic Cvt pathway has been regarded as a model of selective autophagy. During autophagy and the Cvt pathway, double-membrane structures called autophagosomes and Cvt vesicles sequester a portion of the cytoplasm and the Ape1 complex, respectively, and fuse with the vacuole (lysosome in the case of mammalian autophagy) to deliver their inner contents into the organelle lumen (
). Atg3 is composed of three characteristic regions: the E2 core region, the flexible region, and the handle region (HR). The latter two regions are necessary for the interaction with Atg7 and Atg8, but the detailed interaction mode is unknown (
). Recently, we have shown that Atg8 and its mammalian ortholog LC3 recognize the WXXL sequence conserved in Atg19 and p62, autophagic receptors for Ape1 and ubiquitinated protein aggregates, respectively, in a similar manner (
). Because Atg3 HR, which is responsible for the interaction with Atg8, also has a WXXL sequence (Trp270-Glu271-Asp272-Leu273), we postulated that Atg3 also interacts with Atg8 through the sequence.
Here, we show by NMR spectroscopy that Atg3 directly interacts with Atg8 through the WEDL sequence in HR and that the interaction is quite similar to Atg8-Atg19AIM and LC3-p62AIM interactions. Thus, the WEDL sequence of Atg3 is a canonical AIM. In vitro analyses showed that Atg3AIM contributes to the efficient formation of Atg8-PE conjugates particularly under the existence of Atg19, and in vivo analyses showed that Atg3AIM is necessary for the Cvt pathway. These results suggest that AIM identified in autophagic receptors is also utilized in a non-receptor protein, Atg3, and plays a crucial role in autophagic processes.
Protein Expression and Purification
Plasmid construction, expression, and purification of Atg7, Atg3, Atg8, and Atg12-Atg5 conjugate bound to the N-terminal region (residues 1–46) of Atg16 (Atg16N) were performed as described previously (
). The Atg3 HR (256–280) and Atg19 genes were amplified by polymerase chain reaction and cloned into pGEX6P-1 (GE Healthcare). Mutations leading to the specific amino acid substitutions were introduced by PCR-mediated site-directed mutagenesis. All constructs were sequenced to confirm their identities and were expressed in Escherichia coli BL21 (DE3). The cells were lysed, and glutathione S-transferase (GST)-fused Atg proteins were purified by affinity chromatography using a glutathione-Sepharose 4B column (GE Healthcare). GST-fused Atg3 and GST-fused Atg19 proteins for pulldown assay were further purified using a HiTrap Desalting column (GE Healthcare). For NMR spectrometry, GST was excised from Atg8 K26P (a fully active mutant of Atg8) (
) and Atg3HR with a PreScission protease (GE Healthcare). Atg8 K26P was further purified using a CM cation-exchange column (GE Healthcare) followed by a Superdex75 gel filtration column (GE Healthcare). Atg3 HR was further purified using a Superdex30 gel filtration column (GE Healthcare). For in vitro conjugation assays, GST was excised from Atg3 and Atg19 with a PreScission protease. Atg3 was further purified using a DEAE anion-exchange column (GE Healthcare) followed by a Superdex75 gel filtration column. Atg19 was further purified with a glutathione-Sepharose 4B column to remove GST followed by a Superdex200 gel filtration column. 15N-Labeled and 13C,15N-double labeled proteins were prepared by growing E. coli in M9 media using 15NH4Cl and [13C]glucose as the sole nitrogen and carbon sources, respectively. 2H,15N-Double- labeled proteins for the transferred cross-saturation experiment were prepared by growing E. coli in 99.8% D2O M9 media using 15NH4Cl and 97% [H6]glucose as the sole nitrogen and carbon sources.
NMR experiments were carried out at 298 K on a Varian UNITY INOVA 600 spectrometer. 0.3 mm13C,15N-labeled Atg3 HR was prepared, and 13C,1H,15N resonance assignments of Atg3 HR were performed using the following sets of spectra: [1H,15N] HSQC, HNCO, HN(CO)CA, HNCA, HN(CA)HA, HBHA(CO)NH, [13C,1H] HSQC, C(CO)NH, CCH-TOCSY, HCCH-TOCSY, HbCbCgCdHd, and HbCbCgCdCeHe. Spectra were processed by NMRpipe (
). The sample solution of the 15N-labeled Atg3 HR complexed with a 1.5 molar eq of non-labeled Atg8 K26P dissolved in 20 mm phosphate buffer (pH 6.8), 100 mm NaCl, and 5 mm dithiothreitol was prepared for chemical shift perturbation and steady state NOE measurements. Likewise, the sample solution of 15N-labeled Atg8 K26P complexed with a 1.5 m equivalent of non-labeled Atg3 HR dissolved in 20 mm phosphate buffer (pH 6.8) and 100 mm NaCl was prepared for chemical shift perturbation studies. Chemical shift perturbations (Δppm) were calculated using the equation Δppm = [(ΔδHN)2 + (ΔδN/5)2]1/2, where ΔδHN and ΔδN are the differences in chemical shift between the free and complex states along the 1H and 15N axes, respectively. The transferred cross-saturation experiment was carried out using the pulse scheme (
) at 298 K on a Varian UNITY INOVA 800 spectrometer. A 0.87 mm2H,15N-labeled Atg8 K26P in complex with 2 mol % of the nonlabeled Atg3 was dissolved in 50 mm phosphate buffer (pH 6.8), 100 mm NaCl, and 5 mm2H-labeled dithiothreitol containing 10% H2O, 90% D2O. Saturation of the aliphatic protons of Atg3 was made using the WURST-2 decoupling scheme (
). The saturation frequency was set at 0 ppm, and the maximum radiofrequency amplitude was 0.26 kHz for WURST-2. The measurement time was 14 h, with a relaxation delay of 1.0 s and a saturation time of 2.0 s. To evaluate the effect of the residual aliphatic protons within Atg8 K26P, a transferred cross-saturation experiment without Atg3 was also performed under the same conditions. Spectra were processed by the NMRpipe program (
). Intensity ratio was calculated as sat (+)/sat (−).
In Vitro Pulldown Assay
For the interaction study between Atg3 and Atg8, the purified GST-Atg3s (wild type and mutants) were incubated with glutathione-Sepharose 4B beads for 15 min at 277 K, then Atg8s (wild-type and mutants) were loaded to the beads, and they were further incubated for 15 min at 277 K. After washing the beads with phosphate-buffered saline (PBS), proteins were eluted with 10 mm glutathione in 50 mm Tris-HCl buffer (pH 8.0). The eluates were subjected to SDS-PAGE, and the proteins were detected by Coomassie Brilliant Blue (CBB) staining. For the competition study the purified GST-Atg19 was incubated with glutathione-Sepharose 4B beads for 15 min at 277 K, then Atg8 and the wild-type or W270A mutant of Atg3 was loaded to the beads, and they were further incubated for 15 min at 277 K. After washing the beads with PBS, proteins were eluted with 10 mm glutathione in 50 mm Tris-HCl buffer (pH 8.0). The eluates were subjected to SDS-PAGE, and the proteins were detected by CBB staining. Quantification of the results was performed using ImageJ software (
). For detecting Atg8∼Atg3 thioester intermediates, Atg7 (1 μm), Atg3s (5 μm), and Atg8 (10 μm) were incubated in reaction buffer (20 mm Tris-HCl (pH 8.0), 150 mm NaCl, 1 mm ATP, 1 mm MgCl2, and 1 mm dithiothreitol) at 303 K for 15 min. The reaction was stopped by mixing with SDS-PAGE sample buffer and boiling. The samples were subjected to NuPAGE 4–12% Bis-Tris gels (Invitrogen) (
), and proteins were detected by CBB staining (Fig. 3A) or immunoblotting using anti-Atg8 antibody (Fig. 6B). Liposomes were prepared as follows. All lipids were purchased from Avanti Polar Lipids. Dioleoylphosphatidylethanolamine and 1-palmitoyl-2-oleoylphosphatidylcholine were mixed in a glass tube at a molar ratio of 2:8 in chloroform. The chloroform solvent was removed by rotary evaporation. The mixture was further dried in a desiccator under vacuum for 12 h. The resulting lipid film was hydrated in 20 mm Tris-HCl buffer (pH 8.0) at a final concentration of 1 mm phospholipid for 30 min and suspended by sonication for 5 min to obtain small unilamellar liposomes. For detecting Atg8-PE, Atg7 (1 μm), Atg3 variants (1 μm), Atg8 (5 μm), and liposomes (400 μm lipids) were incubated in reaction buffer (20 mm Tris-HCl (pH 8.0), 150 mm NaCl, 1 mm ATP, 1 mm MgCl2, and 1 mm dithiothreitol) at 303 K for 1 h. The reaction was stopped by boiling the sample with SDS-PAGE sample buffer. The samples were subjected to urea-SDS-PAGE (
). This strain was transformed with centromeric plasmids, pRS315-ATG3, pRS315-ATG3 W270A, and the empty vector pRS315. These cells were grown in synthetic dextrose media supplemented with appropriate nutrients to mid-log phase and incubated in SD-N media (2% glucose, 0.17% yeast nitrogen base without amino acids and ammonium sulfate) for 4 h to induce autophagy. The preparation of samples for SDS-PAGE and immunoblotting analyses was performed as described previously (
) with minor modifications. Cells were grown to early log phase, and then 550∼800 A600 units were harvested and spheroplasted. Spheroplasts then suspended in 2.5 ml of 15% Ficoll solution were disrupted with a polycarbonate filter (3.0 μm pore, 47 mm), and the lysed cell solution was loaded in the bottom of an ultraclear SW41 tube (Beckman Coulter, Fullerton CA) and overlaid with 8, 4, and 0% Ficoll solutions. Vacuoles were collected from the 0%/4% float interface with a Pasteur pipette after spinning for 1.5 h at 110,000 × g at 4 °C. Vacuoles were checked under a microscope.
Activity Assay of α-Mannosidase (Ams1)
Cell lysates and harvested vacuoles were assayed for Ams1. Ams1 activity was determined based on the established protocol (
) with minor modifications. Samples were treated with Triton X-100 (2.5% final concentration), and the volume was then brought up to 400 μl with distilled H2O. 100 μl of 5× substrate mix (200 mm sodium acetate (pH 6.5), 2 mm 4-metylumbelliferyl-α-d-mannopyranoside) was added to start the reaction, and the samples were then incubated for 1 h at 37 °C. The reaction was stopped with 200 μl of 10% trichloroacetic acid, and any particulates were spun down in a microcentrifuge for 5 min. An equal volume of 1 m glycine (pH 10.4) was added to neutralize the reaction before fluorescence measurement at 450 nm. Results from assays for each strain were tabulated from three independent vacuole preparations.
Atg3 Directly Interacts with Atg8 through the WEDL Sequence in HR
). Thus, we postulated that the WEDL sequence in HR might be responsible for the direct interaction with Atg8. To confirm the hypothesis, we first performed an in vitro pulldown assay using wild-type and mutant Atg3s fused to GST and wild-type Atg8. GST-Atg3 can bind to Atg8 without a thioester bond (
) (Fig. 1B, first lane). However, Atg3 W270A (Trp270 was substituted with alanine), D272A, and L273A mutants, particularly W270A, lost their ability to bind to Atg8 (Fig. 1B, second through fifth lanes; supplement Fig. S4A). These data suggest that the WEDL sequence in HR is critical for Atg3-Atg8 interaction. To confirm this interaction, we performed an NMR titration experiment using 15N-labeled Atg3 HR and a non-labeled K26P mutant of Atg8 (a stable mutant with full autophagic activity) (
). A number of HN cross-peaks originated from the main chains of Asp272 and Leu273, and the side chain of Trp270 shifted significantly (Fig. 1C; supplement Fig. S1). Moreover, we measured steady state NOE of the Atg3 HR complexed with Atg8 K26P, which suggested that residues 266–275, including the WEDL sequence, were mainly involved in the interaction with Atg8 K26P (Fig. 1D). Next, NMR titration experiments were performed using 15N-labeled Atg8 K26P and non-labeled Atg3 HR to show the interaction mode between Atg3 HR and Atg8. In the case of the interaction between Atg8 and an Atg19 peptide (Trp412-Glu413-Glu414-Leu415), the side chains of Trp412 and Leu415 interact with the two hydrophobic pockets of Atg8; one is composed of Glu17, Ile21, Pro30, Ile32, Lys48, and Leu50, and the other is composed of Tyr49, Val51, Pro52, Leu55, Phe60, and Val63 (
). When non-labeled Atg3 HR was titrated into the solution containing 15N-labeled Atg8 K26P, a number of HN cross-peaks in the HSQC spectrum of Atg8 K26P, including Val31, Lys46, Lys48, Tyr49, Leu50, Val51, Val63, and Ile64 shifted significantly (Fig. 2A; supplement Fig. S2). Residues with large chemical shift perturbations are plotted on the crystal structure of the Atg8-Atg19AIM complex (Fig. 2A) (
). The residues that showed large chemical shift perturbations were similar to those involved in the interaction with Atg19AIM, and all cross-peaks were perturbed in a dose-dependent manner. These NMR data suggest that the Atg8-Atg3 HR interaction is quite similar to the Atg8-Atg19AIM interaction (
) was performed using 2H,15N-labeled Atg8 K26P and non-labeled Atg3. The signal intensity ratio was calculated based on the spectra with and without irradiation. The result showed that the intensities of the residues in β1, β2, and α3, including Val31, Lys48, Tyr49, Val51, Phe60, and Val63, were significantly decreased (Fig. 2B; supplement Fig. S3). No residues were affected on the back of the molecule. These attenuated residues coincided with the residues that showed large chemical shift perturbations in the titration experiments using 15N-labeled Atg8 K26P and non-labeled Atg3 HR (Fig. 2, A and B), clearly showing that the direct interaction between Atg3 HR and Atg8 is also maintained in the full-length Atg3-Atg8 interaction. We further performed in vitro pulldown assays to determine which residue of Atg8 is important for the interaction. As expected, Atg8 R28A, L50A, P52A, and R67A mutants, which were defective in the interaction with Atg19AIM (
were severely impaired in binding to Atg3 (Fig. 2C; supplement Fig. S4B). These results clearly indicate that the WEDL sequence of Atg3 HR is a functional AIM and that Atg3AIM directly interacts with Atg8 in a manner similar to Atg8-Atg19AIM and LC3-p62AIM interactions.
Atg3AIM Is Responsible for the Lipidation of Atg8 but Not for the Formation of the Atg8∼Atg3 Thioester Intermediate
After Atg3 receives Atg8 from the Atg7∼Atg8 thioester to form Atg3∼Atg8 thioester, it transfers Atg8 to the amino group of PE. Although Atg3AIM was shown to be essential for the direct interaction between Atg3 and Atg8, its significance in Atg3 functions is elusive. To elucidate the significance of Atg3AIM in the Atg8 system, in vitro conjugation assays were performed using wild-type and W270A mutant Atg3s. Purified Atg7, Atg8, and wild-type or mutant Atg3 were incubated with ATP at 303 K for 15 min. Atg3 W270A formed Atg8∼Atg3 thioester intermediates, which reached a plateau within 15 min as in the case of wild-type Atg3 (Fig. 3A; supplement Fig. S4C). These results show that the interaction between Atg3 and Atg8 through Atg3AIM is not required for Atg3 to receive Atg8 from Atg7. Next, purified Atg proteins were incubated with ATP and 20 molar percentage PE-containing liposomes at 303 K for 1 h. The lipidation level of Atg8 using Atg3 W270A was about half compared with that using wild-type Atg3 (Fig. 3B). These results show that the interaction of Atg8 with Atg3AIM is crucial for the efficient transfer of Atg8 from Atg3 to PE.
Atg3AIM Is Dispensable for Atg8 Lipidation in the Presence of Atg12-Atg5 Conjugates
To investigate the in vivo significance of Atg3AIM in Atg8 lipidation, we next examined the accumulation of Atg8-PE conjugates in atg3Δ cells expressing Atg3 W270A. Unexpectedly, the accumulation of Atg8-PE conjugates in atg3Δ cells expressing Atg3 W270A was comparable with that expressing wild-type Atg3 under both nutrient-rich and starvation conditions (Fig. 4A, supplement Fig. S4D). This suggests that Atg3AIM is dispensable for Atg8 lipidation in vivo. Because in vivo conditions contain Atg12-Atg5 conjugates that promote Atg8 lipidation, we next performed in vitro conjugation assays in the presence of Atg12-Atg5 conjugates, which showed that the lipidation level of Atg8 using Atg3 W270A was comparable with that using wild-type Atg3 (Fig. 4B, supplement Fig. S4E). These data suggest that Atg3AIM is dispensable for Atg8 lipidation in the presence of Atg12-Atg5 conjugates.
Atg3AIM Is Important for the Cvt Pathway but Not for Starvation-induced Autophagy
We next examined whether Atg3AIM has a role in autophagic processes. The activity of starvation-induced autophagy was estimated by an alkaline phosphatase assay (
). The alkaline phosphatase assay monitors the transport of alkaline phosphatase from the cytoplasm into the vacuole via autophagy by measuring the activity of alkaline phosphatase, which is artificially expressed in the cytosol as a proform and transported via autophagy into the vacuole, where it is processed into the mature form. atg3Δ cells expressing Atg3 W270A showed autophagic activity comparable with that expressing wild-type Atg3 under starvation (Fig. 5A), suggesting that Atg3AIM is dispensable for starvation-induced autophagy. Similar results were obtained when autophagy was induced by rapamycin treatment (supplement Fig. S5A). We then observed the autophagic bodies accumulated in vacuoles under starvation using an optical microscope, which also suggested that Atg3AIM is dispensable for starvation-induced autophagy (supplement Fig. S5B). Next, we monitored the maturation of the proform of Ape1, which is selectively transported into the vacuole via the Cvt pathway under nutrition-rich conditions. The protein is then processed into a mature form within the vacuole. Ape1 maturation in atg3Δ cells expressing Atg3 W270A was significantly lower than in those expressing wild-type Atg3 (Fig. 5B). Moreover, we measured the activity of Ams1 recovered in the vacuole, which is another selective cargo protein transported to the vacuole via the Cvt pathway. The activity of Ams1 recovered in the vacuole of atg3Δ cells expressing Atg3 W270A was much lower than that expressing wild-type Atg3 (Fig. 5C). These results suggest that Atg3AIM is important for the Cvt pathway.
Atg3AIM Is Important for Atg8 Lipidation in the Presence of Atg19
In vivo experiments showed that Atg3AIM is dispensable for Atg8 lipidation and starvation-induced autophagy. Nevertheless, Atg3AIM was shown to be important for the Cvt pathway. In the Cvt pathway, Atg19, the receptor protein for Ape1, directly interacts with Atg8 through the C-terminal AIM (Trp412-Glu413-Glu414-Leu415), and this interaction is crucial for the Cvt pathway. Because Atg19AIM may compete with Atg3AIM for Atg8 binding, we then examined the effects of Atg19 on Atg3-Atg8 interaction and on Atg8 lipidation in vitro. In vitro pulldown assays showed that Atg8 binding to Atg19 was attenuated by Atg3 in a dose-dependent manner (Fig. 6A, left; supplement Fig. S4F). This attenuation is less observed when Atg3 W270A instead of wild-type Atg3 was used (Fig. 6A, right; supplement Fig. S4F). These results indicate that Atg19AIM competes with Atg3AIM for Atg8 binding. Next, in vitro conjugation reactions were performed using wild-type or the W270A mutant of Atg3 in the presence of Atg19. Formation of the Atg8∼Atg3 thioester intermediate was not affected by Atg19 using both wild-type and W270A Atg3 (Fig. 6B). In contrast, lipidation of Atg8 using wild-type Atg3 was slightly inhibited by Atg19, and that using Atg3 W270A was significantly inhibited by Atg19 in a dose-dependent manner (Fig. 6, C and D). These results suggest that Atg3AIM is important for Atg8 lipidation in the presence of Atg19.
In this study we examined the direct interaction between Atg8 and Atg3 by NMR analyses. NMR data together with in vitro pulldown assays showed that the direct Atg8-Atg3 interaction is mediated by AIM in Atg3 HR and that the interaction between Atg8 and Atg3AIM is quite similar to the Atg8-Atg19AIM and LC3-p62AIM interactions. Furthermore, in vitro conjugation assays showed that this interaction is important for the formation of Atg8-PE conjugates but not for the formation of Atg8∼Atg3 thioester intermediates. Because the Atg8∼Atg3 thioester intermediate is formed on Atg7 and Atg7 should directly interact with both Atg8 and Atg3, the Atg8-Atg3AIM interaction may be not required for Atg8∼Atg3 thioester formation. In contrast, the Atg8-Atg3AIM interaction contributed to the efficient formation of Atg8-PE conjugates, suggesting that the interaction promotes the transfer of Atg8 from the Atg8∼Atg3 thioester intermediate to PE. In the case of SUMO, a ubiquitin-like modifier, Nup358/RanBP2 acts as an E3 by binding both SUMO and Ubc9, the E2 enzyme for SUMO, to position the SUMO∼Ubc9 thioester at optimal orientation to enhance the conjugation of SUMO to its target protein (
). The SUMO interacting motif in the N-terminal region of Nup358/RanBP2 forms an antiparallel β-sheet with SUMO β2, resulting in reduced conformational flexibility of the SUMO∼Ubc9 complex, thus preventing non-productive SUMO∼Ubc9 conformation and aligning the complex and thioester for SUMO transfer. Similarly, Atg3AIM forms a parallel β-sheet with Atg8 β2, indicating the possibility that Atg8-Atg3AIM interaction may reduce the conformational flexibility of the Atg8∼Atg3 thioester intermediate and regulate the orientation of each protein for the efficient lipidation of Atg8.
In contrast to the reduced conjugating activity of Atg3 W270A mutant in vitro, the same mutant retained activity comparable with that of wild-type Atg3 in vivo. This discrepancy can be attributed to the existence of Atg12-Atg5 conjugates in vivo. Atg12-Atg5 conjugates are constitutively expressed and facilitate the lipidation of Atg8 in vivo. An in vitro conjugation assay in the presence of Atg12-Atg5 conjugates showed that promotion of Atg8 lipidation by Atg3AIM was superseded by Atg12-Atg5 conjugates (Fig. 4B). This observation suggests that the E3-like functions of Atg12-Atg5 conjugates may include the E3-like activity of Atg3AIM, reducing the conformational flexibility of Atg8∼Atg3 thioester and regulating the orientation of each protein for the efficient lipidation of Atg8.
Although the importance of Atg3AIM in Atg8 lipidation was not validated in vivo, it was unexpected to find that Atg3AIM is important for the Cvt pathway. This observation is consistent with the fact that Atg3AIM is conserved among yeasts (Fig. 1A) but not among higher eukaryotes and that the Cvt pathway has been identified only in yeasts. In vitro conjugation assays in the presence of Atg19 suggest that Atg3AIM attenuates the inhibitory effect of Atg19 on Atg8 lipidation (Fig. 6, C and D). This is attributed to the competition between Atg3AIM and Atg19AIM for Atg8 binding (Fig. 6A; supplement Fig. S4F). Then, what is the role of Atg3AIM in the Cvt pathway? During the Cvt pathway, the cargo proteins (mainly Ape1) facilitate receptor recruitment and vesicle formation rather than the forming vesicle concentrates the cargo proteins (
). This is in contrast to the case of starvation-induced autophagy, in which the formation of autophagosomes is independent of cargo proteins. Therefore, it is likely that Cvt vesicle formation proceeds along the surface of the Ape1-Atg19 complex. During this process, Atg8 plays two pivotal roles; one is recognizing Atg19, and the other is expanding the membranes of the Cvt vesicle. It is not clear how Atg8 exerts these two functions simultaneously when the active sites overlap (
). Although current knowledge is restricted, the following model can be proposed to clarify these observations and the crucial role of Atg3AIM in the Cvt pathway (Fig. 7). First, a PE-free form of Atg8 binds to the Atg19-Ape1 complex. Next, Atg3 and unknown membranes are recruited to the complex and Atg3 conjugates Atg8 to PE in these membranes, during which Atg3AIM plays a crucial role by releasing Atg8 from Atg19AIM. Finally, Atg8-PE functions for membrane expansion. A low concentration of Atg19 around forming autophagosomes may be the reason why Atg3AIM is dispensable for starvation-induced autophagy. Whatever the case, further studies are required to elucidate the molecular events during the Cvt pathway and autophagy. The most critical issues are when and where Atg8-PE conjugates form.
AIMs were initially identified in the autophagic receptors Atg19 and p62 (
). Therefore, AIMs appear to be widely utilized in autophagic receptors for interaction with Atg8/LC3. We have recently shown that LC3 specifically recognizes the YEDL sequence in the N-terminal tail of Atg4B in a similar manner to Atg8-Atg19AIM and LC3-p62AIM interactions, showing that the YEDL sequence in Atg4B is an AIM (
). Here, we showed that Atg3 also has an AIM and utilizes the motif for interaction with Atg8. These observations suggest that AIMs are utilized not only in autophagic receptors but also in enzymes modifying Atg8/LC3 and exert crucial roles in autophagosome formation.