I began to proceed with the section 2.1.2.1.1 of the project (Domains involved in Ymr1 association to the PAS), which involved the co-localization of Ymr1 with Ape1 protein. Due to the low protein expression of Ymr1, the fluorescence signal of genomically GFP-tagged protein was very weak. Therefore, with the advice of Prof. Fulvio Reggiori, I decided to continue studying the regulation of the autophagosome biogenesis, but in a slightly different context.
I first set out to characterize the dynamics of the PAS association to the ER and vacuole. The observations suggest that the dynamic direction of the autophagosome formation could occur, at least in part, from the ER network towards the vacuole. These observations are in agreement with previous reports that indicate the formation of new autophagosomes occurs in the proximity of the ER in yeast.
Once I showed that the PAS localize between the ER and the vacuolar membrane over the entire autophagosome formation cycle, I analyzed the influence of the core Atg machinery in this unique positioning. I concluded that the PI3K complex I is involved in the PAS positioning in the proximity of the ER.
In order to understand better the role of the PI3K complex I, I characterized the positioning of the PAS in cells knocked out for the other components of this complex. That is, the PI3K complex I has a dual role in the positioning the PAS between the ER and vacuole, being the Atg6-Atg14 subcomplex required for its association with the ER and the Vps15-Vps34 subcomplex responsible of its localization adjacently to the vacuole.
In parallel, I further characterized the PI3K complex I subunit mutants to know whether the PAS proximity to the ERES was affected as well. I concluded that although all the subunits have a role in the association of Atg8 with the autophagosomal membranes, only Atg6 and Atg14 are needed for the PAS positioning in the vicinity of the ERES.
Once I showed that PI3K complex I are essential for the proper localization of the PAS the ER and the vacuole, I asked which subunit was responsible of this specific positioning. Altogether, the results indicated that the N-terminal region of Atg14 plays a role in the PAS association with the ER, while the BATS domain is required for the PAS recruitment in proximity of the vacuole.
Next, I analyzed the distribution of both mutants to determine if these proteins were still recruited to the PAS. I concluded that the BATS domain of Atg14 is required for the PI3K complex I localization at the PAS and for its recruitment at the vicinity of the vacuole.
Because of the differences in Atg14 mutant distribution, I asked about the composition of the PI3K complex I. To further substantiate these data, I checked PI3K complex I activity in cells carrying each mutant by exploring PI3P generation at the PAS. Collectively, these data indicate that the first 24 N-terminal amino acids of Atg14 are required for the stability of the PI3K complex I, and therefore both arms (Atg6-Atg14 and Vps15-Vps34) are recruited separately to the PAS, where they probably assemble together. Additionally, the amphipathic a-helix of Atg14 is responsible of the PI3K complex I recruitment to the PAS, but also to the membranes, where PI3P synthesis takes place.
To investigate whether these domains affect the assembly of the Atg machinery at the PAS, I analyzed the localization of the different Atg complexes on autophagosomal membranes. I concluded that N-terminal region of Atg14 plays an important role in assuring the PI3P synthesis where the Atg machinery assembles, while BATS domain is required for the PI3K complex I localization together with the 2 complexes (Atg1 complex and Atg9 system).
For all the experimental details and results, please refer to the Technical Report.