Critical Review. Trafficking and Signaling in Mammalian Autophagy

Transcription

1 IUBMB Life, 62(7): , July 2010 Critical Review Trafficking and Signaling in Mammalian Autophagy Sharon A. Tooze*, Harold B. J. Jefferies, Eyal Kalie, Andrea Longatti, Fiona E. Mcalpine, Nicole C. Mcknight, Andrea Orsi, Hannah E. J. Polson, Minoo Razi, Deborah J. Robinson, and Jemma L. Webber Secretory Pathways Laboratory, London Research Institute, Cancer Research UK, London, UK Summary Macroautophagy, here called autophagy, is literally a selfeating catabolic process, which is evolutionarily conserved. Autophagy is initiated by cellular stress pathways, resulting in the sequestration or engulfment of cytosolic proteins, membranes, and organelles in a double membrane structure that fuses with endosomes and lysosomes, thus delivering the sequestered material for degradation. Autophagy is implicated in a number of human diseases, many of which can either be characterized by an imbalance in protein, organelle, or cellular homeostasis, ultimately resulting in an alteration of the autophagic response. Here, we will review the recent progress made in understanding the induction of autophagy, with emphasis on the contributions from our laboratory. Ó 2010 IUBMB IUBMB Life, 62(7): , 2010 Keywords autophagy; autophagosome; phagophore; matg9; ULK1; WIPI. INTRODUCTION Autophagy was identified using electron microscopy in the 1950s, recognized as a degradative pathway and correctly grouped with the lysosomal pathways [for review of the original publications see (1)]. The recent increased awareness of what autophagy is and the development of tools and assays for autophagy have catalyzed a significant boost in autophagy research. Following on from this are a number of key findings implicating, and in some cases elucidating the pivotal role played by autophagy in tissue homeostasis and human disease, such as cancer, neurodegeneration, infection, immunity, and aging (2). Autophagy is also tightly linked to cellular metabolism, as shown in genetically altered mice which lack key autophagy genes. A cellular lack of amino acids and decrease in ATP levels cause neonatal mortality in these newborn mice (3). Autophagy is a vesicular process that begins at a structure called the PAS (preautophagosome structure or phagophore assembly site). This structure has been identified in the yeast Saccharomyces cerevisiae as the single location where proteins required for autophagy, called the Atg (Autophagy related) proteins, are localized. The PAS or its equivalent has not been firmly confirmed to exist in mammalian cells, although membranes containing early acting Atg proteins may be the equivalent. In yeast, PAS expands to form the phagophore or isolation membrane (Fig. 1), which is the structure that recruits and harbors all the Atg proteins required for induction of autophagy (4, 5). The phagophore expands and sequesters cytosolic components, eventually closing to become a double membrane vesicle, known as the autophagosome. Autophagosomes have been purified from rat hepatocytes and characterized, and the fact that they lack of markers from the ER, Golgi, or endosomal system supports the idea that the autophagosome arises from a unique membrane source, the phagophore (6 9). In mammalian cells, the autophagosome is easily identified by morphology, as the internal sequestered material looks exactly like the cytosolic material. This immature autophagosome then fuses with endosomes and lysosomes to become an autolysosome (10). After fusion of the immature autophagosome with the endosomes and lysosomes, the immature autophagosome is then referred to as an amphisome or degradative autophagosome. In these structures, the sequestered content loses its identity, eventually becoming a homogeneous autolysosome (Fig. 1). In yeast, the autophagosome fuses directly with the vacuole, and no intermediates have been detected. Received 2 March 2010; accepted 19 March 2010 Address correspondence to: Sharon A. Tooze, Secretory Pathways Laboratory, London Research Institute, Cancer Research UK, London WC2A 3PX, UK. Tel.: (44) Fax: (44) Sharon.tooze@cancer.org.uk THE MOLECULAR MACHINERY The core molecular machinery driving autophagy are the Atg proteins, originally identified in yeast S. cerevisiae (11, 12), [and more up-to-date information can be found in recent ISSN print/issn online DOI: /iub.334

2 504 TOOZE ET AL. Figure 1. The autophagic pathway in mammalian cells after induction of autophagy, the phagophore membrane expands, enveloping cytosolic proteins, organelles (such a mitochondria picture inside), and finally closes forming an autophagosome. The autophagosome fuses with early endosomes, multivesicular bodies, late endosomes, or lysosomes. An amphisome is formed from the former set of fusion steps (early endosome and the late endosome), whereas an autolysosome is formed either by the latter fusion with the latter, the lysosome, or as a product of the step-wise fusion events beginning with the early endosome, followed by the late endosome/mvb and finally the lysosome. reviews (13 16)]. The best-characterized autophagy pathway in mammalian cells is the one downstream of the mammalian Target of Rapamycin complex 1 (mtorc1). mtorc1 is an important regulator of cell growth, and a potent negative regulator of autophagy. In nutrient-rich conditions, mtorc1 represses autophagy by a mechanism that involves regulation of the activity of the ULK1 kinase complex, consisting of the serine-threonine kinase ULK1, Atg13, and FIP200 (17 21). This complex is the equivalent of the yeast Atg1 complex, in which ULK1, Atg13, and FIP200 are the orthologues of yeast Atg1, Atg13, and Atg17, respectively. A new component, Atg101, which binds to Atg13 has also been shown to exist in this complex (22). The association of mtorc1 with the ULK1 complex in nutrient-rich conditions results in increased phosphorylation of ULK1 and Atg13, an apparent inhibition of the ULK1 kinase activity and autophagy. While there is a single Atg1 kinase in yeast, in mammalian cells, there are five ULK kinase family members, of which ULK1 and 2 are most closely related (17). A second key complex is the class III phosphatidylinositol 3- kinase (PtdIns 3-kinase): inhibition of its activity or loss of the lipid kinase components inhibits autophagy. The kinase complex consists of the mammalian orthologues of Vps34, Vps15, Atg6/ Vps30, and Atg14, which are called Vps34, p150, Beclin1, and Atg14. Importantly, Beclin1 interacts with a number of proteins, forming multiple functionally distinct complexes, and this allows Beclin1 to respond to different signals regulating autophagy (23, 24). As the activity of the ULK1 and Vps34/Beclin1 complexes are modulated by external stimuli, understanding how they are regulated and associate with the phagophore is essential in developing our knowledge of the initial stages of autophagy. Additional core complexes include two ubiquitin-like conjugation systems. The first is the Atg complex, in which the Atg12 molecule acts as the ubiquitin-like protein. Atg5 is covalently linked to Atg12, whereas Atg16 associates with the Atg12-5 conjugate to form a larger complex. The second ubiquitin-like conjugation system involves Atg8, which is covalently modified by the lipid phosphatidylethanolamine (PE). The function of these two conjugation systems are intimately linked in the formation of the phagophore (25) (for further information see (26) and references therein). Mammalian cells have several Atg8-like molecules, called LC3, GABARAP, and GATE-16, of which the first two have multiple family members. The most prevalent and best-studied Atg8 protein in mammalian cells is LC3, which associates with autophagosomal membranes once modified by PE, and this lapidated form is called LC3-II. A GFP-tagged version of LC3 is widely used as a marker for these membranes. The other core proteins are mammalian Atg9 (matg9), the only multispanning membrane protein so-far identified as being required for autophagy in both yeast and mammalian cells (27, 28), and the WIPI proteins, the putative orthologues of Atg18 and Atg21 (29, 30). The localization of these proteins will be described in detail below. THE FORMATION OF THE AUTOPHAGOSOME: INSIGHTS FROM ULK1 and matg9 In yeast, the Atg1 complex, and in particular Atg17, has been assigned an apical position in the hierarchy of autophagy proteins required for assembly of the PAS (4). In mammalian cells, recent evidence supports the primacy of ULK1 kinase complex in the autophagy hierarchy. ULK1 and the closely related ULK2 are cytoplasmic kinases with no apparent membrane association domains. Nevertheless, it has been shown that

3 TRAFFICKING AND SIGNALING IN MAMMALIAN AUTOPHAGY 505 Figure 2. A model for the formation of the phagophore expansion. Steps for the initiation of autophagy: (1) Activation of protein kinase ULK1 complex and the class III PtdIns 3-kinase complex, called here the Beclin1 complex. The subunits in the two complexes are shown on the bottom left (ULK1 complex:1, ULK1/2; 13, Atg13; FIP, FIP200; 101, Atg101. Beclin1 complex: 150, p150 (hvps15); 34, Vps34; 14, Atg14; Bec, Beclin1). p150 is myristoylated allowing its association to membranes. Atg14 is shown associating alone to the phagophore membrane although there is no evidence to support this, it simply indicates a PtdIns3P-independent recruitment. The omegasome is shown here to form a cradle for the membrane (either the PAS or phagophore) in the first step of induction. DFCP-1 marks the location of the omegasome structure, which also contains PtdIns 3-P (not shown here). While recent data supports this model, it remains to be shown that this is the exclusive site of phagophore expansion. (2) matg9 is transported to the autophagosome, and WIPI is recruited to the phagophore membrane. We have illustrated this as a process that continues during expansion of the phagophore. The membrane which donates matg9 to the phagophore is not yet established. Our results suggest it is likely to be the TGN or an endosomal membrane, or a vesicular intermediate. It is also possible that delivery of matg9 occurs from multiple matg9-positive compartments. (3) Immediately after or simultaneously with matg9 delivery, the Atg5-Atg12-Atg16 complex is recruited. (4) Lastly, LC3 in it lipidated form (LC3-II) is transferred to the phagophore. The association of the phagophore with the omegasome is presumed to be lost in Step 4, however, there is no data yet to confirm this. (5) After dissociation of the Atg complex and the loss of LC3-II on the outer surface, the phagophore is primed and starts to close forming an autophagosome (not shown). [Color figure can be viewed in the online issue, which is available at both proteins are tightly associated with membranes and are present in a number of punctate cytoplasmic structures, some of which colocalize with the early autophagy complex, Atg , and LC3 (19, 31, 32). My lab is interested in how and where the ULK1 complex exerts its function as a regulator of autophagy during the nucleation phase. Our current studies are based on our identification of the C-terminal domain (CTD) of ULK1 as a potent inhibitor of autophagy (31). The CTD contains a region responsible for membrane association and a region required for interaction with Atg13. In addition, FIP200 has also been shown to associate with ULK1 via the CTD (19). Finally, we have shown that the region of the CTD that inhibits autophagy is a short sequence at the extreme C-terminus of the CTD. On the basis of these results, we have proposed that the CTD regulates the catalytic activity of the ULK1 kinase and speculate that this is due to an alteration in the conformation of ULKl; from a closed conformation when associated with mtorc1 to an open conformation after mtorc1 inactivation. Importantly, the colocalization of ULK1 with Atg16 and LC3 demonstrates that ULK1 is found on autophagosome membranes. Therefore, we speculate that the ULK1 complex is either recruited to the phagophore (or prephagophore sites which would be the so-far uncharacterized precursor of the phagophore membranes, the PAS) after dissociation from mtorc1 in starvation, or that it is already localized to these membranes but in an inactive conformation. Dissociation of ULK1 from mtorc1 would be accompanied by an alteration in its conformation, to an open state, along with an activation of ULK1 kinase activity allowing its interaction with other pro-

4 506 TOOZE ET AL. teins required for autophagy including Atg13, FIP200, and other so-far unidentified proteins which are recruited to the extreme C-terminal ULK1 motif. The ULK1 complex may then catalyze the further recruitment of the subsequent Atg protein complexes. The order for this recruitment has not yet been definitively established, and in Fig. 2, we propose a tentative model. The class III PtdIns 3-kinase complex, called the Beclin1 complex in Fig. 2, is also important at early stages. Mammalian Atg14, also known as Barkor, is a subunit of the Vps34 complex, and Atg14 is recruited to Atg positive membranes and is required for autophagy (33 35). Interestingly, Atg14 is recruited to membranes, where it forms discrete puncta even after treatment with wortmannin, which inhibits the Vps34 kinase activity and therefore inhibits autophagy. In addition, Atg14 can be recruited to Atg16-positive membranes independently of its association with Vps34 or Beclin1 (34). This suggests that Atg14 recruitment does not require Vps34/Beclin1 activity (Fig. 2). Vps34 and Beclin1 may either be targeted to Atg14-positive membranes or be recruited independently of Atg14 to the phagophore membrane then assemble with Atg14, and subsequently, PtdIns-3P would be produced. The ability of wortmannin to potently inhibit autophagy has led to the conclusion that production of PtdIns-3P is critical for autophagy. With this in mind, the identification of DFCP-1 (double FYVE domain containing protein-1) was significant for several reasons: the FYVE domain of DFCP-1 binds PtdIns-3P; DFCP-1 translocates from the Golgi to the ER during starvation; DFCP-1 has been shown to delineate subdomains of the ER which are Atg5 and LC3- positive. These data have led to the hypothesis that DFCP-1 positive ER membranes form a cradle for the phagophore. Such structure has been called the omegasome (36), and see Fig. 2. Indeed, recent 3D tomography data supports this hypothesis, with putative phagophore membranes found to be sandwiched between two ER cisternae, and ER have been shown to be positive for DFCP-1 (37, 38). It remains to be definitively shown that the membranes traced in the tomograms represent canonical phagophore membranes, but nonetheless the existing data is compelling and suggests that the ER may, at least in some circumstances, be the source of the phagophore. However, it is not clear how and when the PtdIns-3P that recruits DFCP-1 is produced. A conundrum then arises when considering the data suggesting that the ER is the source of the phagophore with the localization of matg9, which in yeast has been proposed to deliver lipids to the PAS (39). In fact, mammalian Atg9 is found both on the TGN and endosomes in nutrient-rich cells and on LC3- positive autophagosomes in nutrient-starved cells (27). It is not known from which compartment matg9-traffics from to reach the forming autophagosome, but it requires ULK1 and Atg13: In the absence of either protein, matg9 remains in a juxta-nuclear region, colocalized with TGN46. This is significant when one considers the data in yeast showing that the Atg1 complex, with Atg29, is sufficient to assemble a PAS (40). We propose, therefore, that concurrent with the recruitment of ULK1 and Beclin1 complexes, and activation of the Vps34, that matg9 is delivered to the phagophore (Fig. 2). We further propose that matg9 is delivered continuously to the phagophore while it undergoes expansion, and that the matg9 donor membrane could be the TGN or endosomes (Fig. 2). It is not known if matg9 is delivered to the omegasome to allow the expansion of the omegasome. We have shown that the targeting of matg9 requires the ULK1 and Belcin1 complex activities, and therefore, matg9 may be transported to the omegasome maybe dependent upon the products of these protein and lipid kinases. As matg9 is a multispanning membrane protein, it is almost certainly delivered in membrane vesicles, and the content and composition of these vesicles would contribute to the formation of the PAS. The laboratory is striving to now understand matg9 trafficking during starvation, and in particular, the identification of the membranes from which matg9 originates; the composition of the vesicles targeted to the phagophore: the function of matg9. Importantly, we have also shown that trafficking of matg9 is controlled by p38mapk signaling through p38ip (p38 interacting protein) (41). Like ULK1, p38ip is required for trafficking of matg9 in nutrient-rich medium between the TGN and endosome. p38ip is also required for matg9 trafficking during starvation, and we propose that p38ip is required for delivery of matg9 to the phagophore. Furthermore, we propose that the control of matg9 localization by p38ip would to be coupled to the activation state of p38mapk and provide a regulatory loop, as p38mapk inhibits autophagy, and reactivation of p38mapk during nutrient replenishment could be influences by dissociation of p38ip from matg9 (42). THE ROLE OF PtdIns 3-PHOSPHATE AND WIPI2 Another consequence of the production of PtdIns-3P on the phagophore is the recruitment of the PtdIns 3P-binding protein WIPI, a family of proteins called WIPI1 to WIPI4. So far, the best-characterized member of this family is WIPI-1a or WIPI- 49 (29, 30), which has been shown to colocalize with LC3 on autophagosomes and dissociate from these membranes when PtdIns-3P is hydrolyzed by the PtdIns 3-phosphatase Jumpy/ MTMR14 (43). Interestingly, Jumpy colocalizes with matg9 and is also required for the distribution of matg9 to autophagosomes. It is not known whether this is due to a direct interaction of Jumpy with matg9, or because Jumpy hydrolyses PtdIns-3P causing the dissociation of WIPI from the membranes. Indeed, in yeast, Atg9 cycling from the PAS requires Atg18, the yeast orthologue of WIPI. We are currently investigating the WIPI proteins to determine how their association with membranes via PtdIns-3P is controlled, and how they function. We have data demonstrating that WIPI2 is also essential for autophagy, and in particular autophagy involving DFCP-1, and we are investigating the role of WIPI2 recruitment and the delivery of matg9 so see if there are parallels to the yeast pathway. It is also of interest to consider the function of the PtdIns- 3P, produced by the Beclin1 complex on the expanding phago-

5 TRAFFICKING AND SIGNALING IN MAMMALIAN AUTOPHAGY 507 phore. There are at least two PtdIns-3P binding proteins, DFCP- 1 and WIPI, whose localization is altered after starvation by the autophagy-specific pool of PtdIns-3P produced by the Beclin1 complex (Fig. 2). Our data shows suggest that a percentage of WIPI2 and DFCP-1 colocalize, which suggests they maybe present on the same compartments or closely opposed membranes. However, it is not known how the recruitment of DFCP-1 and WIPI2 to the autophagosomal pool of PtdIns-3P is coordinated, or if DFCP-1 and WIPI2 compete for the same pool of PtdIns-3P. It is, however, very clear, and significant that the pool of PtdIns 3-phosphate produced is solely involved in autophagy. CONCLUSIONS Autophagosome formation is a complex process likely to require the action of many proteins, most of which are probably not yet identified. High and low-throughput screens in mammalian cells will be essential in identifying new players in the process. We have completed two screens for novel proteins required for autophagy: one being a sigenome screen, whereas the other is a screen for Rab proteins and their regulators. Our knowledge regarding key players in autophagy has increased immensely in the last years. The combined wealth of new data emerging in the literature, from our screens, and other screens yet to be reported, provides a further expansion of the molecules involved and the understanding of the process of autophagy. REFERENCES 1. Tooze, S. A. The role of membrane proteins in mammalian autophagy. Semin. Cell. Biol., doi: /j.semcdb Mizushima, N., Levine, B., Cuervo, A. M., and Klionsky, D. J. (2008) Autophagy fights disease through cellular self-digestion. Nature 451, Mizushima, N. and Klionsky, D. J. (2007) Protein turnover via autophagy: implications for metabolism. Annu. Rev. Nutr. 27, Suzuki, K. and Ohsumi, Y. Current knowledge of the pre-autophagosomal structure (PAS). FEBS Lett. 7, Geng, J., Baba, M., Nair, U., and Klionsky, D. J. (2008) Quantitative analysis of autophagy-related protein stoichiometry by fluorescence microscopy. J. Cell. Biol. 182, Yamamoto, A., Masaki, R., Fukui, Y., and Tashiro, Y. (1990) Absence of cytochrome P-450 and presence of autolysosomal membrane antigens on the isolation membranes and autophagosomal membranes in rat hepatocytes. J Histochem Cytochem 38, Yamamoto, A., Masaki, R., and Tashiro, Y. (1990) Characterization of the isolation membranes and the limiting membranes of autophagosomes in rat hepatocytes by lectin cytochemistry. J Histochem Cytochem 38, Seglen, P. O., Gordon, P. B., and Holen, I. (1990) Non-selective autophagy. Semin. Cell. Biol. 1, Stromhaug, P. E., Berg, T. O., Fengsrud, M., and Seglen, P. O. (1998) Purification and characterization of autophagosomes from rat hepatocytes. Biochem. J. 335, Eskelinen, E. L. (2005) Maturation of autophagic vacuoles in mammalian cells. Autophagy 1, Tsukada, M. and Ohsumi, Y. (1993) Isolation and characterization of autophagy-defective mutants of Saccharomyces cerevisiae. FEBS Lett. 333, Thumm, M., Egner, R., Koch, B., Schlumpberger, M., Straub, M., Veenhuis, M., and Wolf, D. H. (1994) Isolation of autophagocytosis mutants of Saccharomyces cerevisiae. FEBS Lett. 349, He, C. and Klionsky, D. J. (2009) Regulation mechanisms and signaling pathways of autophagy. Annu. Rev. Genet. 43, Longatti, A. and Tooze, S. A. (2009) Vesicular trafficking and autophagosome formation. Cell. Death Differ. 16, Noda, T., Fujita, N., and Yoshimori, T. (2009) The late stages of autophagy: how does the end begin[quest]. Cell. Death Differ. 16, Orsi, A., Polson, H. E., and Tooze, S. A. Membrane trafficking events that partake in autophagy. Curr. Opin. Cell. Biol. 22, Chan, E. Y. and Tooze, S. A. (2009) Evolution of Atg1 function and regulation. Autophagy 5, Ganley, I. G., Lam Du, H., Wang, J., Ding, X., Chen, S., and Jiang, X. (2009) ULK1.ATG13.FIP200 complex mediates mtor signaling and is essential for autophagy. J Biol Chem 284, Hara, T., Takamura, A., Kishi, C., Iemura, S., Natsume, T., Guan, J. L., and Mizushima, N. (2008) FIP200, a ULK-interacting protein, is required for autophagosome formation in mammalian cells. J. Cell. Biol. 181, Hosokawa, N., Hara, T., Kaizuka, T., Kishi, C., Takamura, A., Miura, Y., Iemura, S.-I., Natsume, T., Takehana, K., Yamada, N., Guan, J.-L., Oshiro, N., and Mizushima, N. (2009) Nutrient-dependent mtorc1 association with the ULK1-Atg13-FIP200 complex required for autophagy. Mol. Biol. Cell 20, Jung, C. H., Jun, C. B., Ro, S. H., Kim, Y. M., Otto, N. M., Cao, J., Kundu, M., and Kim, D. H. (2009) ULK-Atg13-FIP200 complexes mediate mtor signaling to the autophagy machinery. Mol. Biol. Cell 20, Mercer, C. A., Kaliappan, A., and Dennis, P. B. (2009) A novel, human Atg13 binding protein, Atg101, interacts with ULK1 and is essential for macroautophagy. Autophagy 5, He, C. and Levine, B. (2010) The Beclin 1 interactome. Curr. Opion Cell. Biol. 22, Maiuri, M. C., Criollo, A., and Kroemer, G. (2010) Crosstalk between apoptosis and autophagy within the Beclin 1 interactome. EMBO J. 29, Sou, Y. S., Waguri, S., Iwata, J.-I., Ueno, T., Fujimura, T., Hara, T., Sawada, N., Yamada, A., Mizushima, N., Uchiyama, Y., Kominami, E., Tanaka, K., and Komatsu, M. (2008) The Atg8 conjugation system is indispensable for proper development of autophagic isolation membranes in mice. Mol. Biol. Cell 19, Noda, T., Fujita, N., and Yoshimori, T. (2008) The Ubi brothers united. Autophagy 4, Young, A. R., Chan, E. Y. W., Hu, X. M., Köchl, R., Crawshaw, S. G., High, S., Hailey, D. W., Lippincott-Schwartz, J., and Tooze, S. A. (2006) Starvation and ULK1-dependent cycling of mammalian Atg9 between the TGN and endosomes. J. Cell. Sci. 119, Saitoh, T., Fujitac, N., Hayashic, T., Takaharaa, K., Takashi Satoh, T., Lee, H., Matsunagac, K., Kageyamac, S., Omoric, Hs., Nodac, T., Yamamotoe, N., Kawaia, T., Ishiia, K., Takeuchia, O., Tamotsu Yoshimoric, T., and Akira, S. Atg9a controls dsdna-driven dynamic translocation of STING and the innate immune response. Proc. Natl. Acad. Sci. USA 106, Jeffries, T. R., Dove, S. K., Michell, R. H., and Parker, P. J. (2004) PtdIns-specific MPR pathway association of a novel WD40 repeat protein, WIPI49. Mol. Biol. Cell 15, Proikas-Cezanne, T., Waddell, S., Gaugel, A., Frickey, T., Lupas, A., and Nordheim, A. (2004) WIPI-1alpha (WIPI49), a member of the novel 7- bladed WIPI protein family, is aberrantly expressed in human cancer and is linked to starvation-induced autophagy. Oncogene 23,

6 508 TOOZE ET AL. 31. Chan, E. Y., Longatti, A., Mcknight, N. C., and Tooze, S. A. (2009) Kinase-inactivated ULK proteins inhibit autophagy via their conserved C- terminal domain using an Atg13-independent mechanism. Mol. Cell Biol. 29, Chan, E. Y. W., Kir, S., and Tooze, S. A. (2007) sirna screening of the kinome identifies ULK1 as a multi-domain modulator of autophagy. J. Biol. Chem. 282, Zhong, Y., Wang, Q. J., Li, X., Yan, Y., Backer, J. M., Chait, B. T., Heintz, N., and Yue, Z.. (2009) Distinct regulation of autophagic activity by Atg14L and Rubicon associated with Beclin 1-phosphatidylinositol-3-kinase complex. Nat. Cell Biol. 11, Itakura, E., Kishi, C., Inoue, K., and Mizushima, N. (2008) Beclin 1 forms two distinct phosphatidylinositol 3-kinase complexes with mammalian Atg14 and UVRAG. Mol. Biol. Cell 19, Sun, Q., Fan, W., Chen, K., Ding, X., Chen, S., and Zhong, Q. (2008) Identification of Barkor as a mammalian autophagy-specific factor for Beclin 1 and class III phosphatidylinositol 3-kinase. Proc. Natl. Acad. Sci. USA 105, Axe, E. L., Walker, S. A., Manifava, M., Chandra, P., Roderick, H. L., Habermann, A., Griffiths, G., and Ktistakis, N. T. (2008) Autophagosome formation from membrane compartments enriched in phosphatidylinositol 3-phosphate and dynamically connected to the endoplasmic reticulum. J. Cell. Biol. 182, Yla-Anttila, P., Vihinen, H., Jokitalo, E., and Eskelinen, E. L. (2009) 3D tomography reveals connections between the phagophore and endoplasmic reticulum. Autophagy 5, Hayashi-Nishino, M., Fujita, N., Noda, T., Yamaguchi, A., Yoshimori, T., and Yamamoto, A. (2009) A subdomain of the endoplasmic reticulum forms a cradle for autophagosome formation. Nat. Cell Biol. 11, Reggiori, F., Tucker, K. A., Stromhaug, P. E., and Klionsky, D. J. (2004) The Atg1-Atg13 complex regulates Atg9 and Atg23 retrieval transport from the pre-autophagosomal structure. Dev. Cell 6, Cao, Y., Cheong, H., Song, H., and Klionsky, D. J. (2008) In vivo reconstitution of autophagy in Saccharomyces cerevisiae. J. Cell. Biol. 182, Webber, J. L. and Tooze, S. A. (2010) Coordinated regulation of autophagy by p38alpha MAPK through matg9 and p38ip. EMBO J. 29, Webber, J. L. and Tooze, S. A. (2010) New insights into the function of Atg9. FEBS Lett. 7, Vergne, I., Roberts, E., Elmaoued, R. A., Tosch, V., Delgado, M. A., Proikas-Cezanne, T., Laporte, J., and Deretic, V. (2009) Control of autophagy initiation by phosphoinositide 3-phosphatase jumpy. EMBO J. 28,