ABSTRACT
The macroautophagy (hereafter autophagy) process involves de novo formation of double-membrane autophagosomes; after sequestering cytoplasm these transient organelles fuse with the vacuole/lysosome. Genetic studies in yeasts have characterized more than 40 autophagy-related (Atg) proteins required for autophagy, and the majority of these proteins play roles in autophagosome formation. The fusion of autophagosomes with the vacuole is mediated by the Rab GTPase Ypt7, its guanine nucleotide exchange factor Mon1-Ccz1, and soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins. However, these factors are not autophagosome-vacuole fusion specific. We recently showed that 2 autophagy scaffold proteins, the Atg17-Atg31-Atg29 complex and Atg11, regulate autophagosome-vacuole fusion by recruiting the vacuolar SNARE Vam7 to the phagophore assembly site (PAS), where an autophagosome forms in yeast.
The Atg17-Atg31-Atg29 complex translocates to the PAS at a very early stage of autophagy, recruiting other Atg proteins such as Atg1-Atg13 to facilitate autophagy induction. We found that the ternary complex also interacts with the SNARE Vam7 at the PAS even before a complete autophagosome forms. This observation raises a very interesting question, that is, how is premature fusion of the phagophore membrane with the vacuole inhibited? To answer this question, we carefully examined the interaction of the ternary complex with Vam7 because the latter is involved in the terminal fusion step.
First, we demonstrated that Atg17 is the component in the complex that directly interacts with Vam7. We then mapped the interaction site and found that the Vam7 SNARE domain mediates the Atg17-Vam7 interaction, which is the same region that is responsible for the formation of a SNARE complex with other SNAREs during membrane fusion events. Thus, we maintain that Atg17 recruits Vam7 to the PAS at an early stage of autophagy, but inhibits its function by occupying the SNARE domain to prevent premature fusion of uncompleted autophagosomes (i.e., phagophores) with the vacuole.
The next question we sought to address is whether the recruitment of Vam7 to the PAS by Atg17 is necessary for later fusion of mature autophagosomes with the vacuole. To explore this issue we wanted to identify mutants that specifically block the Atg17-Vam7 interaction; however, we could not achieve this goal by deleting the Vam7 SNARE domain without affecting other fusion events at the vacuole. Thus, we decided to map critical sites in Atg17 that mediated the interaction. We showed that some hydrophobic residues in Atg17 helix 1 and helix 4, including L105, F317 and I325, are required for binding Vam7. The Atg17F317D mutant in particular showed partially diminished interaction with Vam7 without affecting its binding affinity toward Atg13 and Atg31, which was critical to eliminate indirect effects. However, the Atg17F317D mutant unexpectedly did not show autophagy activity defects. Previous studies indicated a partial overlap in function between Atg17 and another scaffold protein, Atg11. Indeed, we found that Atg11 plays a partially redundant role in recruiting Vam7 to the PAS; Atg11 interacts with Vam7 independent of the Atg17-Atg31-Atg29 complex. Accordingly, in the atg11Δ atg17Δ background the Atg17F317D mutant displayed an approximately 30% decrease in autophagy activity compared to the wild-type protein. Moreover, using a protease protection assay that monitored accessibility to the precursor form of the vacuolar hydrolase aminopeptidase I, and correlative light and electron microscopy, we showed that a pool of complete autophagosomes accumulate in the cytosol of the Atg17F317D mutant cells. These data suggest that the recruitment of Vam7 to the PAS by Atg17 and Atg11 is required for autophagosome-vacuole fusion.
Based on our results, we propose a model where the Atg17-Atg31-Atg29 complex coordinates with Atg11 to recruit Vam7 to the PAS, mediating efficient autophagosome-vacuole fusion. However, some questions still remain to be explored. First, how does Vam7 get released from Atg17 after autophagosome formation is completed? It is possible that other SNARE proteins such as Vam3 and Vti1 are recruited to the PAS after autophagosome maturation, and that they have stronger binding affinity toward Vam7 than that of Atg17. Thus they may outcompete Atg17 to form a SNARE complex with Vam7, facilitating fusion of autophagosomes with the vacuole. Another possibility is that Atg17 dissociates from mature autophagosomes after clearance of PtdIns3P by Ymr1, a PtdIns3P phosphatase, in the process releasing Vam7. Second, it has been shown that Atg11, but not Atg17, is required for the cytoplasm-to-vacuole targeting (Cvt) pathway that delivers resident hydrolases to the vacuole under growing conditions. Along these lines, does Atg11 play the major role in recruiting Vam7 in the Cvt pathway? Further analysis of the Atg11-Vam7 interaction will be needed to answer this question. Third, does the mammalian homolog of Atg17, RB1CC1/FIP200, also regulate autophagosome-lysosome fusion? In mammals, ATG14 interacts with STX17, the autophagosomal SNARE protein, and SNAP29 on mature autophagosomes to promote membrane tethering and the fusion of autophagosomes with lysosomes. Does RB1CC1 coordinate with ATG14 to regulate autophagosome-lysosome fusion? If this is the case, it may suggest that a conserved molecular machinery is employed to regulate the fusion of autophagosomes with the vacuole/lysosome.
Funding
This work was supported by NIH grant GM053396 to DJK.
Disclosure of potential conflicts of interest
No potential conflicts of interest were disclosed.