Received: 26 December 2018 Revised: 4 March 2019 Accepted: 5 March 2019 Uncorrected manuscript published: 7 March 2019 Published on: 9 April 2019 DOI: 10.1111/tra.12640 ORIGINAL ARTICLE TRAPPC11 functions in autophagy by recruiting ATG2BWIPI4/WDR45 to preautophagosomal membranes Daniela Stanga1 | Qingchuan Zhao2 | Miroslav P. Milev1 | Djenann Saint-Dic1 | Cecilia Jimenez-Mallebrera3 | Michael Sacher1,4 1 Concordia University, Department of Biology, Montreal, Quebec, Canada TRAPPC11 has been implicated in membrane traffic and lipid-linked oligosaccharide synthesis, 2 and mutations in TRAPPC11 result in neuromuscular and developmental phenotypes. Here, we University of Montreal, Department of Medicine and Institute for Research in Immunology and Cancer, Montreal, Quebec, Canada 3 Neuromuscular Unit, Neuropaediatrics Department, Institut de Recerca Sant Joan de Déu, Hospital Sant Joan de Déu and CIBERER, Barcelona, Spain 4 McGill University, Department of Anatomy and Cell Biology, Quebec, Canada Correspondence Michael Sacher, Concordia University, Department of Biology, 7141 Sherbrooke Street West SP-457.01, Montreal, Quebec H4B1R6, Canada Email: michael.sacher@concordia.ca show that TRAPPC11 has a role upstream of autophagosome formation during macroautophagy. Upon TRAPPC11 depletion, LC3-positive membranes accumulate prior to, and fail to be cleared during, starvation. A proximity biotinylation assay identified ATG2B and its binding partner WIPI4/WDR45 as TRAPPC11 interactors. TRAPPC11 depletion phenocopies that of ATG2 and WIPI4 and recruitment of both proteins to membranes is defective upon reduction of TRAPPC11. We find that a portion of TRAPPC11 and other TRAPP III proteins localize to isolation membranes. Fibroblasts from a patient with TRAPPC11 mutations failed to recruit ATG2BWIPI4, suggesting that this interaction is physiologically relevant. Since ATG2B-WIPI4 is required for isolation membrane expansion, our study suggests that TRAPPC11 plays a role in this process. We propose a model whereby the TRAPP III complex participates in the formation and expansion of the isolation membrane at several steps. Funding information Institute of Genetics; Natural Sciences and Engineering Research Council of Canada; McGill University.; Concordia University; European Regional Development Fund; Instituto de Salud Carlos III; Canadian Institutes of Health Research; Christian-AlbrechtsUniversität zu Kiel KEYWORDS ATG2, autophagy, isolation membrane, LC3, TRAPP, TRAPPC8, TRAPPC11, TRAPPC12, WIPI4 1 | INTRODUCTION creating autolysosomes where the degradative enzymes of the lysosome hydrolyze the contents of the autophagosome to help re- Macroautophagy (autophagy) is a basal process in cells that removes establish a pool of cellular nutrients. damaged organelles and ensures a ready supply of amino acids for Proteins acting at various stages of autophagy (ATG) have been protein synthesis. This process is upregulated during starvation when identified by genetic screens in yeast and many of these proteins are nutrients are scarce. Autophagy starts with the nucleation of mem- conserved throughout evolution (reviewed in Reference 7). Much branes referred to as the phagophore assembly site in yeast or isola- effort has been put into defining the sequence of events in tion membranes in higher eukaryotes. The expansion of the isolation autophagy. Nucleation is initiated by the action of the ULK1 kinase membrane is due in part to the input of membranes from sources which activates a phosphatidylinositol-3 kinase-containing complex.8 including (ER), The resulting phosphatidylinositol-3-phosphate (PI3P)-rich mem- and branes of the ER subdomain called the omegasome serve as a plat- mitochondria, mitochondria-ER endosomes. 1–6 contact the sites, endoplasmic the plasma reticulum membrane Upon sealing of the isolation membrane into a fully- form to recruit ATG5-ATG12-ATG16L1, a complex required to enclosed, double-membrane-bound autophagosome, its cargo is lipidate LC3-I, converting it into the autophagy-specific LC3-II.9 Iso- delivered to the lysosome by autophagosome-lysosome fusion lation membrane expansion requires other ATG proteins including © 2019 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd Traffic. 2019;20:325–345. wileyonlinelibrary.com/journal/tra 325 326 STANGA ET AL. ATG2,10 a protein that interacts with a PI3P effector called WIPI4/ 2 | RE SU LT S WDR45 (hereafter WIPI4).11–14 Little is known about how the isolation membrane is sealed into a mature autophagosome, although ATG5 and members of the LC3/GABARAP family have been implicated in this process.10,15 Identifying novel regulators of these key autophagy proteins will be of importance for a more complete TRAPP is a family of multisubunit complexes implicated in membrane trafficking that were first identified in yeast and are also found in higher eukaryotes.16–20 Each complex is composed of an identical core of proteins, and additional, complex-specific proteins define their 21,22 Most of the subunits of the complexes are conserved throughout evolution but metazoa have several subunits not found in Saccharomyces cerevisiae including TRAPPC11 and TRAPPC12.23 These latter two subunits are components of TRAPP III that also contains TRAPPC8 and TRAPPC13.16 While a role for yeast TRAPP III in autophagy is established,17,24–28 the role of the human complex in this process is not well-understood. A proteomics study initially implicated human TRAPP III in autophagy,13 and several recent studies have further shown a role for TRAPP III in this process.29,30 To date, two of the human TRAPP III-specific proteins (TRAPPC11 and TRAPPC12) have been implicated in cellular processes distinct from membrane trafficking. These include lipid-linked oligosaccharide synthesis for TRAPPC11 and To begin to address the role of TRAPP III proteins in autophagy, we first examined the appearance of the autophagy-specific form of endoge- understanding of this process. identity, localization and function. 2.1 | Depletion of TRAPP III proteins affects autophagic flux chromosome congression for TRAPPC12.31,32 In addition, both proteins have been implicated in human disease that manifests as muscular and neurodevelopmental phenotypes.33–39 Given the overall size of the TRAPP complexes,40–42 it is not unreasonable to expect their constituent proteins to have unique interacting partners, thereby allowing the complexes to coordinate distinct events even within a common pathway. Thus, understanding the various roles that the TRAPP proteins play in cell physiology will shed light on the diseases caused by mutations in their genes. nous LC3 (LC3-II) in HeLa cells depleted of TRAPP III proteins by small interfering RNA (siRNA) (Figure 1A and Figure S1A). While control cells showed an increase in LC3-II levels after 2 hours of starvation that decreased by 4 hours, cells that were depleted of TRAPPC11 showed elevated levels of LC3-II prior to starvation and the increased levels persisted over the 4-hour time course of starvation. Similar results were seen for TRAPPC12-depleted cells. In contrast, and as previously reported,13,43 TRAPPC8 depletion blocked the formation of LC3-II, consistent with TRAPPC8 functioning upstream of both TRAPPC11 and TRAPPC12. A similar result was noted for TRAPPC2-depleted cells, the core TRAPP subunit that acts as an adaptor for TRAPPC8.24,44 The cells also contain mRFP-GFP-LC3 and we monitored them for the number of GFP-positive structures during starvation (Figure 1B and C). Consistent with the results in Figure 1A, both TRAPPC11- and TRAPPC12-depleted cells had a large number of GFP-positive punctae prior to starvation that did not significantly change over the 4 hours of starvation. TRAPPC8- and TRAPPC2-depleted cells did not show any starvation-dependent increase in the number of GFP-positive punctae compared to control cells. We then examined the formation of autolysosomes upon depletion of the TRAPP III proteins in the mRFP-GFP-LC3 cell line. Fusion of autophagosomes with lysosomes ultimately results in quenching of the GFP fluorescence in the more acidic environment, resulting in red punctae. If fusion is blocked, both a red and green (yellow) punctum is seen.45 As shown in Figure 1D and quantified in Figure 1E, prior to starvation, control cells have few punctae, some of which are red. Upon starvation, there is a large increase in these structures and a Here, we show that the TRAPP III-specific proteins TRAPPC11, large portion of them are red, indicating that autolysosome formation TRAPPC12 and TRAPPC8 affect autophagy. While depletion of has taken place. Formation of autolysosomes was blocked by the TRAPPC8 affected the formation of isolation membranes, TRAPPC11 inclusion of bafilomycin A1, resulting in yellow punctae. In contrast and TRAPPC12 functioned downstream of this step. We show that to the control, when either TRAPPC11 or TRAPPC12 was depleted, TRAPPC11 interacts with ATG2B, one of the ATG2 homologs, as well we observed a large number of yellow but few red punctae both prior as with its binding partner WIPI4. TRAPPC11 depletion phenocopies to and following starvation even in the absence of bafilomycin A1, that of both ATG2 and WIPI4 resulting in unsealed isolation mem- indicating that these structures have not fused with lysosomes. branes, suggesting that TRAPPC11 functions upstream of isolation Depletion of TRAPPC8 resulted in substantially fewer punctae, con- membrane after sistent with a defect in autophagosome formation. Defective auto- autophagosome formation. We demonstrate that the recruitment of lysosome formation in TRAPPC11- and TRAPPC12-depleted cells ATG2B and WIPI4 to isolation membranes is dependent upon was also seen by examining the lifetime of the GFP-LC3 fluorescence TRAPPC11 and implicate the carboxy-terminus of TRAPPC11 as (Figure S1B). Individual punctae were monitored over a 120-second important for this recruitment. Fibroblasts from an individual with time frame and the intensity of the GFP fluorescence, which is a mea- bi-allelic TRAPPC11 mutations including one at the carboxy- sure of autophagosome-lysosome fusion, was assessed. While the terminus are defective in ATG2B and WIPI4 recruitment. This majority of punctae in control cells showed a fluorescence intensity closure. In contrast, TRAPPC12 functions defect in the fibroblasts was suppressed by wild type TRAPPC11, of 20% of the initial (t = 0) intensity or lower, the majority of the suggesting that this newly-identified interaction is important in punctae in both TRAPPC11- and TRAPPC12-depleted cells were human health. We propose a model whereby different TRAPP III between 30% and 50% of the initial intensity, consistent with a proteins participate in various steps in autophagy including isola- defect in autophagic flux. In addition, the diameter of these struc- tion membrane formation and expansion, thus implicating TRAPP III tures in TRAPPC11-depleted cells was larger than those seen in con- in a key role in early autophagy events. trol cells (Figure S1C). Although the diameter of the structures in 327 STANGA ET AL. FIGURE 1 Legend on next page. TRAPPC12-depleted cells was similar to that of control cells, we We then assessed the ability of HeLa cells to clear noted a population of punctae that were larger in diameter. The rea- autophagosomes that were formed in the presence of the son for these larger structures is unknown. autophagosome-lysosome fusion inhibitor bafilomycin A1. Cells were 328 STANGA ET AL. starved for 2 hours in the presence of the inhibitor and then trans- function.47 When compared with control, the TRAPPC11, ATG2 and ferred to nutrient-rich medium for up to 40 minutes and the effect WIPI4 depletions showed elevated levels of LC3-II prior to starvation on LC3-II was monitored throughout (Figure 1F). In control cells, (Figure 2A and quantified in Figure S2). These levels increased slightly LC3-II was highest following treatment with bafilomycin A1, and during starvation. The phenotype was not further exaggerated by a was then rapidly cleared by the 40 minute time point. In contrast, double TRAPPC11/ATG2 or TRAPPC11/WIPI4 depletion, nor by a tri- both TRAPPC11- and TRAPPC12-depleted cells did not show any ple TRAPPC11/ATG2/WIPI4 depletion. A similar result was seen for significant clearance of the bafilomycin A1-induced LC3-II. In depletion of TRAPPC12, but depletion of TRAPPC8 did not result in TRAPPC8-depleted cells, we were unable to detect significant detectable levels of LC3-II consistent with our earlier result (see levels of LC3-II. Collectively, our data suggest that the TRAPP III Figure 1F). The latter phenotype was similar to that of depletion of proteins TRAPPC11, TRAPPC12 and TRAPPC8 are involved in ATG9A, a protein that acts early in autophagy and blocks formation autophagic flux, but the former two may be acting downstream of of isolation membranes.48 Using HeLa cells expressing mRFP-GFP- TRAPPC8. LC3, we found that, as previously reported,47 ATG2 depletion resulted in an increased number of GFP-LC3 punctae that were gen- 2.2 | TRAPPC11 interacts with ATG2B-WIPI4 erally larger in size compared to those in control cells and were seen even prior to starvation (Figure 2B and C). These structures persisted Given the involvement of TRAPPC11 in disease, we decided to focus on this protein to better understand its role in human health. We employed a biotin proximity ligation screen (BioID) to identify interacting partners that might help to elucidate the role of TRAPPC11 in autophagy.46 The screen was performed with both an amino- and carboxy-terminal fusion protein of TRAPPC11 with the BirA biotin ligase (BirA-C11 and C11-BirA, respectively), and it was done two times to verify the accuracy of the partners identified. As expected, a number of TRAPP III subunits were identified including TRAPPC8, TRAPPC12 and TRAPPC13 (Table 1). Interestingly, peptides for both over a 5-hour starvation period. A similar phenotype was noted upon depletion of either TRAPPC11, TRAPPC12 or WIPI4. In contrast, the phenotype resulting from depletion of TRAPPC8 was similar to that of ATG9A depletion where few if any GFP-positive punctae were seen. It was previously shown that many of the LC3-positive punctae that result from ATG2 depletion are also positive for autophagy proteins that are components of functional complexes including ULK1, WIPI1, ATG5, ATG14 and ATG9.47 Similarly, we found that depletion of TRAPPC11 and WIPI4 also result in LC3-positive TRAPPC8 and TRAPPC13 were identified more robustly with the structures that co-localize with these early-acting autophagy pro- C11-BirA construct compared with BirA-C11, suggesting that these teins (Figure 2D). Therefore, by several different criteria we find proteins may interact with the amino-terminus of TRAPPC11. Strik- that depletion of TRAPPC11 phenocopies that of either ATG2 or ingly, besides TRAPPC11 itself, peptides derived from the autophagy WIPI4. protein ATG2B were the most abundant identified. These peptides were nearly all detected using the BirA-C11 construct, suggesting that ATG2B interacts with the carboxy-terminus of TRAPPC11. We also identified peptides from WIPI4, a protein that forms a complex with 2.3 | Depletion of TRAPPC11 results in unsealed isolation membranes ATG2B. Like ATG2B, these peptides were preferentially identified by Because ATG2 proteins function prior to the sealing of isolation the BirA-C11 construct (Table 1). membranes into enclosed autophagosomes,10,47 we employed a pro- We next examined whether TRAPPC11 depletion in HeLa cells tease protection assay to determine whether TRAPPC11 depletion results in a similar phenotype to ATG2 and WIPI4 depletion. Note that also affected this step of autophagy. This assay faithfully reports the both ATG2 paralogues, ATG2A and ATG2B, were depleted because a state of the isolation membranes in both mammalian and lower previous study showed that these proteins may have a redundant eukaryotic systems.47,49,50 Following depletion, HeLa cells were Depletion of TRAPPC11 affects autophagic flux. (A) HeLa cells expressing mRFP-GFP-LC3 were subjected to a knockdown (KD) with either non-specific (NS) small interfering (si)RNA, or siRNA targeting TRAPPC2, TRAPPC8, TRAPPC11 or TRAPPC12. The cells were then placed in starvation (EBSS) medium for the indicated time (in hours). At each time point, lysates were prepared, fractionated on an SDS-polyacrylamide gel, transferred to PVDF and probed for endogenous LC3 and tubulin. The resulting bands were quantified using Image J and plotted as the ratio of LC3-II to tubulin. Results reflect data from three separate experiments with SEM indicated. Representative immunoblots are shown below the graphs. (B, C) HeLa cells stably expressing mRFP-GFP-LC3 were treated as in panel (A), transferred into starvation (EBSS) medium and imaged every 10 minutes for three hours. GFP-LC3-positive structures were counted using Imaris software and plotted in Graphpad Prism. A total of 10 cells were quantified at each time point in (C). The times above the panels in (B) indicate the time in starvation (EBSS) medium. (D) HeLa cells stably expressing mRFP-GFP-LC3 were left untreated or treated with siRNA as in (A). The cells were then either left in nutrient-rich (DMEM) medium or transferred to starvation (EBSS) medium for 3 hours either with or without 200 nM bafilomycin A1. Panels from the merged GFP and RFP channels are shown. (E) Quantification of the red (autolysosome) and yellow (isolation membrane or autophagosome) punctae from panel (D) was performed on a minimum of 10 cells. The error bars indicate SEM. (F) HeLa cells were treated as in (A). One sample also received 200 nM bafilomycin A1 for 2 hours (2B). The bafilomycin A1-treated cells were then transferred into fresh nutrient-rich (DMEM) medium for up to 40 minutes. Lysates were prepared and processed for western analysis and representative immunoblots are shown below each graph. The scale bars in panels (B) and (D) are 20 μm. Numbers to the right of immunoblots denote the migration of molecular size standards in kD. The extent of each of the knockdowns is shown in Figure S1 FIGURE 1 329 STANGA ET AL. Proteins identified and corresponding peptide count in a proximity biotinylation screen with TRAPPC11 TABLE 1 Total peptides Unique peptides BirA-C11 C11-BirA BirA-C11 C11-BirA Protein Trial#1 Trial#2 Trial#1 Trial#2 Trial#1 Trial#2 Trial#1 Trial#2 TRAPPC2L 19 17 3 8 5 5 2 4 TRAPPC8 4 6 15 23 4 6 12 19 TRAPPC11 384 432 68 164 58 58 33 45 TRAPPC12 12 12 14 19 9 8 11 12 TRAPPC13 1 0 11 17 1 0 7 11 ATG2B 110 128 1 2 59 58 1 2 WDR45/WPI4 21 23 0 1 11 10 0 1 Cell lines stably expressing BirA-TRAPPC11 (BirA-C11) or TRAPPC11-BirA (C11-BirA) were generated in Flp-In T-Rex 293 cells. The cell lines and controls (just expressing BirA) were treated with tetracycline to induce expression of the BirA constructs and supplemented with biotin. Lysates were prepared, the biotinylated proteins were captured on streptavidin-agarose beads and processed for mass spectrometric analysis. The experiment was performed on two separate occasions and the number of total peptides and unique peptides was quantified. Note that in all cases except for ATG2B in C11-BirA the background was 0 peptides. In the stated exception, the background exceeded the experimental case. either left in nutrient-rich medium or subjected to starvation for in an increase in the number of LC3-positive cellular structures, only 2 hours in the presence of bafilomycin A1, after which various mem- those formed in response to TRAPPC11 depletion are unsealed brane fractions were collected. The membranes were either membranes, similar to those seen upon ATG2 or WIPI4 depletion. untreated, treated with proteinase K, or solubilized in Triton X-100 before proteinase K treatment (Figure 3A). In control cells under nutrient-rich conditions, LC3 (in the form of LC3-I) was mostly in the 2.4 | Recruitment of ATG2B and WIPI4 to isolation membranes is dependent upon TRAPPC11 supernatant fraction (HS). Upon starvation, LC3-II was formed and largely found in the low-speed membrane fraction (LP). The protein in both the LP and the high-speed membrane fractions (HP) were resistant to proteinase K treatment but could be digested if the We next investigated whether TRAPPC11 was required to recruit ATG2B and WIPI4 to membranes during starvation. The recruitment was first examined biochemically using the cell fractionation assay membranes were first solubilized with detergent, suggesting that performed in Figure 3. In this case, we probed for the presence of autophagosomes were formed. In contrast, TRAPPC11-depleted ATG2B, WIPI4, TRAPPC11, TRAPPC12 and TRAPPC8. We also cells under nutrient-rich conditions showed LC3-II in both the LP included ATG9A and ATG12 as two autophagy proteins that function and HP membrane fractions, consistent with the presence of abun- at different steps compared with TRAPPC11. ATG9A also served as a dant LC3-positive structures prior to starvation. These pools of membrane protein control and in all cases where ATG9A was not the LC3-II were sensitive to proteinase K in the absence of detergent, subject of the knockdown, the protein appeared exclusively in the suggesting that the LC3-II on the membranes is completely exposed membrane fractions as expected. The recruitment assay was per- to the surroundings. The same result was observed upon starvation formed in HeLa cells that were unstarved or starved for 2 hours in the of the presence of bafilomycin A1. The cells were either untreated or TRAPPC11-depleted cells were similar to those seen for either depleted of either one of the TRAPP III proteins, ATG2, WIPI4 or ATG2- or WIPI4-depleted cells. In contrast, while LC3-II was ATG9A (Figure 4). Upon depletion of TRAPPC11, less ATG2B and detected in WIPI4 were found in the membrane fractions following starvation TRAPPC12-depleted cells, this form of LC3 was resistant to protein- when compared with control. In contrast, TRAPPC12 depletion did ase K treatment in the absence of detergent, suggesting that not affect the membrane association of either ATG2B or WIPI4 fol- autophagosome formation has occurred in the absence of lowing starvation compared with control. Neither TRAPPC11 nor TRAPPC12. In order to confirm that TRAPPC12 functions down- TRAPPC12 depletion affected the recruitment of ATG12 to mem- stream of both ATG2 and TRAPPC11, the assay was performed fol- branes. In contrast, while TRAPPC8 depletion also affected the lowing the TRAPPC11-depleted both prior to cells. and The following results in starvation double recruitment of ATG2B and WIPI4 to membranes, this depletion depletions. In both cases, LC3-II was sensitive to proteinase K treat- prevented recruitment of ATG12 to membranes (Figure 4), likely due ment in the absence of detergent (Figure 3B), supporting the notion to the absence of isolation membranes. Although depletion of ATG2 that both ATG2 and TRAPPC11 function upstream of TRAPPC12. prevented recruitment of WIPI4 to membranes, and WIPI4 depletion Consistent with the near complete absence of presumably LC3-II- prevented recruitment of ATG2B to membranes, neither depletion positive structures following TRAPPC8 depletion, only LC3-I was affected recruitment of any TRAPP III proteins nor of ATG12 to mem- detected in this assay and this form of the protein did not fractionate branes (Figure 4). As expected, and similar to the TRAPPC8-depleted in either of the membrane fractions (Figure 3A). This was phenotypi- cells, depletion of ATG9A prevented the recruitment of all proteins cally similar to what was seen for depletion of ATG9A (Figure 3A). examined to membranes, consistent with its role in isolation mem- Thus, although depletion of either TRAPPC11 or TRAPPC12 results brane formation. TRAPPC11/TRAPPC12 and ATG2/TRAPPC12 330 STANGA ET AL. FIGURE 2 Legend on next page. We then assessed recruitment of ATG2B and WIPI4 to endoge- nor LC3 localize to a substantial amount of punctae in control cells nous LC3 structures in HeLa cells depleted of either TRAPPC11, under nutrient-rich conditions. Upon starvation, these cells showed an TRAPPC12 or TRAPPC8 (Figure 5A and B). Neither ATG2B, WIPI4 increased number of LC3, ATG2B and WIPI4 structures, many of STANGA ET AL. 331 Depletion of TRAPPC11 results in unsealed isolation membranes. HeLa cells were treated with non-specific (NS) siRNA or siRNA targeting either TRAPPC11, TRAPPC12, TRAPPC8, ATG2, WIPI4 or ATG9A. The cells were left in nutrient-rich medium (DMEM) or transferred to starvation medium (EBSS) for 2 hours in the presence of bafilomycin A1. Lysates were prepared and processed for the membrane sealing assay as described in Materials and Methods. Prior to fractionation, some samples were treated with proteinase K (ProK) or with ProK in the presence of 1% Triton X-100 (TX). The fractions, composed of a postnuclear supernatant (PN), low-speed pellet (LP), high-speed pellet (HP) and high-speed supernatant (HS), were probed for LC3. Note that isolation membranes and autophagosomes appear in the LP and/or HP fractions. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD. (B) HeLa cells were depleted of both TRAPPC11 and TRAPPC12 or of ATG2 and TRAPPC12. The cells were starved for 2 hours in the presence of bafilomycin A1. Lysates were prepared and processed as described in (A) FIGURE 3 TRAPPC11 depletion phenocopies an ATG2B or WIPI4 depletion. (A) HeLa cells were treated with non-specific (NS) siRNA or siRNA targeting either TRAPPC11, TRAPPC12, TRAPPC8, ATG2A and ATG2B (ATG2), WIPI4, ATG9A, both TRAPPC11 and ATG2 or WIPI4, or siRNA targeting the three transcripts TRAPPC11, ATG2 and WIPI4. The cells were left in nutrient-rich (DMEM) medium or transferred into starvation (EBSS) medium for 2 hours with or without the protease inhibitors E64d and pepstatin A (PIs). Lysates were prepared and the levels of LC3 and tubulin were detected by western analysis. (B) HeLa cells were treated with non-specific (NS) siRNA or siRNA targeting the indicated transcripts. The cells were then transferred to starvation (EBSS) medium for the times indicated. The cells were then fixed and stained with anti-LC3 antibody. (C) The number of LC3-positive punctae in panel (B) were quantified using Imaris from deconvoluted images and plotted. A total of 25 cells at each time point were used in the quantification. Error bars indicate SEM. (D) HeLa cells or HeLa cells expressing GFP-tagged ULK1, WIPI1, ATG5 or ATG14 as indicated in the figure were treated with non-specific (NS) siRNA or siRNA targeting either TRAPPC11, ATG2 or WIPI4. The cells were then stained for LC3 or, in the case of HeLa cells, LC3 and ATG9A. Endogenous GFP fluorescence or ATG9A fluorescence was imaged and overlayed onto the LC3 fluorescent signal. The scale bars in (B) and (D) represent 20 μm. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD FIGURE 2 332 STANGA ET AL. ATG2B-WIPI4 recruitment to membranes is impaired by TRAPPC11 depletion. HeLa cells were treated with non-specific (NS) siRNA or siRNA targeting either TRAPPC11, TRAPPC12, TRAPPC8, ATG2, WIPI4 or ATG9A. Cells were grown in either nutrient-rich (DMEM) or starvation (EBSS) medium for 2 hours. Lysates were prepared and processed for membrane recruitment as described in the Materials and Methods section. The fractions, composed of a postnuclear supernatant (PN), low-speed pellet (LP), high-speed pellet (HP) and high-speed supernatant (HS), were probed for the indicated proteins. Note that half the amount of tubulin was loaded in the PN fraction in order to obtain an exposure that shows tubulin in the other fractions without grossly overexposing the PN signal. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD FIGURE 4 which overlap. As shown earlier in HeLa cells (Figure 1B and 2B), TRAPPC11 depletion resulted in a substantial number of LC3 punctae under nutrient-rich conditions and their numbers increased upon starvation. Importantly, we did not detect a concomitant increase in either ATG2B or WIPI4 punctae in either nutrient-rich or starved conditions following depletion of TRAPPC11 (Figure 5A and B). This was in contrast to the TRAPPC12-depleted cells where, although there were numerous LC3 structures both prior to and during starvation, there were also numerous ATG2B and WIPI4 punctae co-localizing with 2.5 | The TRAPPC11 interaction with ATG2B-WIPI4 is dependent upon ATG9 function Our results thus far suggest that TRAPPC11 interacts with ATG2B-WIPI4, a complex that functions upstream of the sealing of isolation membranes into autophagosomes. Consistent with this notion is the fact that the consequences of TRAPPC11 depletion are phenotypically identical to that of either ATG2 or WIPI4 depletion. Suspecting that we might see an enhanced interaction between TRAPPC11 and ATG2B during starvation, LC3. Depletion of TRAPPC8 did not result in LC3-positive structures, we immunoprecipitated TRAPPC11 at various time points following starva- nor in ATG2B or WIPI4 punctae, similar to what was seen upon tion and probed for TRAPP proteins and ATG2B. As expected and ATG9A depletion (Figure 5A and B). We conclude that the presence reported previously,16,23 TRAPPC11 co-immunoprecipitated with of ATG2B and WIPI4 on LC3-positive punctae in starved cells is TRAPPC12, TRAPPC2 and TRAPPC8, but not with the TRAPP II subunit dependent upon TRAPPC11 but not upon TRAPPC12. TRAPPC10 (Figure S3). Although an interaction with ATG2B was detected, 333 STANGA ET AL. we did not see a starvation-dependent increase in this interaction over time. Rather, there seemed to be a spike in the interaction at the 2-hour time point. The reason for this transient increase is not clear. In yeast, the TRAPP III complex was shown to interact with Atg9.25 Another study demonstrated an interaction between yeast Atg2-Atg18 (the ATG2 and WIPI4 homologs, respectively) with Atg9.51 These studies prompted us to examine whether the interaction between ATG2B-WIPI4 and mammalian TRAPP III also included ATG9A. As shown in Figure 6, immunoprecipitation of TRAPPC11 coprecipitated ATG2B and WIPI4, but not ATG9A. However, depletion of ATG9A abrogated the interaction between TRAPPC11 and both ATG2B and WIPI4. Furthermore, depletion of ATG2B prevented WIPI4 from co-immunoprecipitating with TRAPPC11. These results suggest that the TRAPPC11 interaction with ATG2B-WIPI4 is dependent upon ATG9A function and further show that TRAPPC11 does not interact with WIPI4 in the absence of ATG2B. 2.6 | A portion of TRAPP III localizes to isolation membranes Our results thus far suggest that a portion of TRAPPC11, and perhaps TRAPP III, localizes to isolation membranes where it interacts with ATG2B-WIPI4. Autophagosome formation follows a maturation process where proteins function sequentially to produce these organelles. Two such proteins that are thought to associate with isolation membranes are ATG5 and WIPI1.52 Thus, we used HeLa cells that express GFP-tagged versions of these proteins to assess their colocalization with TRAPP III proteins. Cells were either starved for 0.5 hours in the presence of bafilomycin A1 to prevent autolysosome formation, or depleted of either ATG2 or WIPI4, conditions that lead to a buildup of isolation membranes (see Figure 2B). The cells were then immunostained for either TRAPPC8, TRAPPC11 or TRAPPC12 and GFP. As shown in Figure 7A, ~10% to 15% of the ATG5 or WIPI1 punctae formed during brief starvation were also positive for the TRAPP III proteins. This percentage increased to ~20% to 25% following depletion of either ATG2 or WIPI4. A similar increase in TRAPP protein co-localization was seen using an LC3-GFP cell line (Figure S4). The co-localization is not likely incidental because (a) overlap was assessed using thin (0.4-1 μm) three-dimensional sections and not flattened images, and (b) TRAPP protein-positive structures decreased in the ATG2 and WIPI4 knockdowns compared with starvation conditions (not shown) yet the co-localization increased during the knockdowns. These co-localization results suggest that Recruitment of ATG2B-WIPI4 to LC3-positive membranes is impaired by TRAPPC11 depletion. (A) HeLa cells were treated with non-specific (NS) siRNA or siRNA targeting either TRAPPC11, TRAPPC12, TRAPPC8 or ATG9A. The cells were grown in nutrientrich (DMEM) or starvation (EBSS) medium for 2 hours at which time the cells were fixed and stained with anti-ATG2B or anti-WIPI4 IgG. (B) The LC3-positive and the ATG2B- or WIPI4-positive punctae in (A) were quantified using Imaris. A total of 25 cells were used for quantification in each sample. Error bars represent SEM. The scale bar in (A) represents 20 μm FIGURE 5 a portion of TRAPP III proteins, including TRAPPC11, transiently localize to early autophagic membranes. Knowing that TRAPPC11 localizes to isolation membranes and is thus in proximity to ATG2B-WIPI4, we decided to more formally test the notion that the carboxy-terminus of TRAPPC11 is important for this interaction (see Table 1). To this end, we depleted HeLa cells of TRAPPC11 and assessed whether the defective recruitment of ATG2B and WIPI4 could be suppressed by expression of BirA-C11 (a free carboxy-terminus) or C11-BirA (a blocked 334 STANGA ET AL. The interaction between TRAPPC11 and ATG2B-WIPI4 is regulated by ATG9A. HeLa cells were treated with non-specific (NS) siRNA or siRNA targeting either TRAPPC11, ATG9A or ATG2. The NS-treated samples were subjected to starvation to induce the formation of isolation membranes. Lysates were prepared and incubated with anti-TRAPPC11 IgG. The total lysates and the immunoprecipitates were probed by western analysis for the presence of TRAPPC11, ATG2B, WIPI4 and ATG9A. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD FIGURE 6 carboxy-terminus) using the same biochemical membrane recruit- to nutrient-rich medium and the appearance/disappearance of ment assay as above (see Figure 4). As shown in Figure 7B, deple- LC3-II was monitored. In control fibroblasts, the bafilomycin tion of TRAPPC11 prevented the starvation-induced increase in A1-dependent increase in LC3-II levels was reverted to near membrane-associated ATG2B and WIPI4 but not that of ATG12, unstarved levels following 40 minutes in nutrient-rich medium an ATG protein acting at a step upstream of TRAPPC11. (Figure 8A). In contrast, fibroblasts from the TRAPPC11 patient Overexpression of BirA-C11 resulted in an increase in membrane- (PatC11) showed elevated levels of LC3-II prior to starvation that associated while did not substantially increase after bafilomycin A1 treatment. overexpression of C11-BirA resulted in a much smaller increase. These elevated levels of LC3-II did not diminish upon transfer into The differences were not due to expression-level differences nutrient-rich medium. The fibroblasts from the patient with the (Figure 7C). These results are consistent with the notion that the TRAPPC12 mutation (PatC12) showed a starvation-induced carboxy-terminus of TRAPPC11 is important for the recruitment of increase in LC3-II levels that were slightly reduced upon transfer ATG2B-WIPI4 to isolation membranes. to nutrient-rich medium. These results suggest that both PatC11 ATG2B and WIPI4 following starvation, and PatC12 have a defect in autophagic flux. We then examined the recruitment of ATG2B-WIPI4 by the 2.7 | Fibroblasts from an individual with TRAPPC11 mutations fail to recruit ATG2B-WIPI4 to membranes biochemical A number of individuals with TRAPPC11 mutations have now been appearance of ATG2B and WIPI4 in the membrane fractions, reported.33–38 Because the carboxy-terminus of TRAPPC11 PatC11 cells did not (Figure 8B). In contrast, PatC12 cells showed appears to be important for ATG2B-WIPI4 recruitment to mem- ATG2B and WIPI4 in the membrane fraction even prior to starva- branes (see Figures 7C and Table 1), we examined the fibroblasts tion. Because reduced recruitment of ATG2B and WIPI4 to mem- from one individual harboring the compound heterozygous muta- branes could result in unsealed autophagosomes, we employed the tions p.Ala372_Ser429del and p.Asp1127Valfs*47 in TRAPPC11 protease protection assay to examine this scenario. While the (note that a more detailed characterization of this individual will be LC3-II-positive membranes in control and PatC12 fibroblasts were reported elsewhere) to determine if this newly-identified interac- largely resistant to proteinase K treatment, those in PatC11 were tion was physiologically relevant. This individual was chosen sensitive (Figure 8C). fractionation procedure described above (see Figure 3). While control fibroblasts showed a starvation-induced because the latter allele results in a frameshift at amino acid 1127, Finally, we examined the fibroblasts for recruitment of ATG2B only 6 residues from the carboxy-terminus. We first examined and WIPI4 to punctae during starvation. Control fibroblasts whether these cells displayed a defect in autophagic flux compared showed a starvation-dependent increase in ATG2B and WIPI4 with control fibroblasts. Included in this analysis were fibroblasts punctae (Figure 8D and Figure S5A). In contrast, PatC11 fibro- from a patient with the homozygous p.Glu49Argfs*14 mutation in blasts did not show such an increase, consistent with a role for TRAPPC12. 39 Fibroblasts were untreated or starved for 2 hours in TRAPPC11 in ATG2B-WIPI4 recruitment. In PatC12, a large num- the presence of bafilomycin A1, at which point they were returned ber of ATG2B- and WIPI4-positive punctae were present prior to STANGA ET AL. FIGURE 7 335 Legend on next page. starvation, suggesting that recruitment of ATG2B-WIPI4 was due to the TRAPPC11 mutations since this defect was suppressed already taking place in these cells. The defect in starvation- by wildtype TRAPPC11 fused at its amino-terminus, but not fused dependent ATG2B and WIPI4 punctae formation in PatC11 was at its carboxy-terminus, to BirA (Figure 8E and Figure S5B). This 336 STANGA ET AL. latter result is consistent with the earlier notion that the carboxy- that autolysosomes are indeed formed. As expected for unsealed terminus of TRAPPC11 is important for the recruitment of ATG2B- membranes that form in response to depletion of either ATG2 or WIPI4. Collectively, our studies on these fibroblasts from an indi- TRAPPC11, there was little overlap between the LC3-positive struc- vidual with TRAPPC11 mutations suggest that some TRAPPC11 tures and LAMP1 (Figure 9C and D). Similarly, depletion of TRAPPC12 mutations may result in defective recruitment of ATG2B-WIPI4 also resulted in low overlap between the LC3-positive structures and and dysregulated autophagy. LAMP1, suggesting that the autophagic flux defect following TRAPPC12 depletion is upstream of autolysosome formation. Thus, 2.8 | Depletion of TRAPPC11 and TRAPPC12 also affect lysosomal activity Both TRAPPC11 and TRAPPC12 function in membrane traffic23 and one study implicated TRAPPC11 in lysosomal function.33 While the autophagic flux defect resulting from TRAPPC11 depletion is due to unsealed autophagosomes, altered lysosomal proteolytic activity might explain the autophagic flux defect associated with TRAPPC12 depletion given that, in this case, sealed autophagosomes are forming although TRAPPC12 depletion results in lysosomes with reduced proteolytic activity, this does not account for the autophagic flux defect in these cells. Overall, our results indicate that during autophagy, ATG2 and TRAPPC11 function upstream of TRAPPC12. They further illustrate that TRAPPC11 and TRAPPC12 also affect lysosome activity most likely through a defect in biosynthetic trafficking to the lysosome, although this second function does not account for the autophagic flux defect seen upon TRAPPC11 or TRAPPC12 depletion. (see Figure 3A). We first assessed lysosomal proteolytic activity by examining the ability of HeLa cells depleted of either TRAPPC12, TRAPPC11 or ATG2 to process the lysosomal protease cathepsin 3 | DI SCU SSION B. Processing of this enzyme relies upon a proteolytically-active lysosome resulting in a cleaved, active, mature form.53 Depletion of either Here, we have shown that depletion of three different TRAPP III- TRAPPC12 or TRAPPC11 resulted in greatly reduced levels of mature associated proteins results in an autophagic flux defect suggesting a cathepsin B (Figure 9A). This may be due to a defect in trafficking of function in autophagy, but the specific steps in which these proteins certain proteases, including cathepsin B, to the lysosome along the function differ. Whereas TRAPPC8 affects formation of isolation 23,54 biosynthetic pathway. In contrast, depletion of ATG2 did not result in a cathepsin B processing defect. membranes, TRAPPC11 is required upstream of the isolation membrane enclosure that leads to the formation of autophagosomes. We then more directly examined lysosomal proteolytic activity TRAPPC12 acts downstream of TRAPPC11 after autophagosome for- using the fluorogenic substrate DQ-BSA. Upon delivery of DQ-BSA to mation. We also demonstrate that a fraction of each protein is associ- the lysosomes, fluorescence is dequenched as a result of proteolytic ated with LC3-positive membranes suggesting that a portion of activity.55 As shown in Figure 9B, depletion of either TRAPPC12 or TRAPP III localizes to isolation membranes/autophagosomes. We TRAPPC11 resulted in much less fluorescence of DQ-BSA in the lyso- examined the function of TRAPPC11 in more detail and found that it some compared with control. This was again in contrast to ATG2 is required for recruitment of ATG2B-WIPI4. depletion which showed strong DQ-BSA-dependent fluorescence. Combined with previous studies, we propose a model for the role The DQ-BSA defect in TRAPPC11- or TRAPPC12-depleted cells was of the TRAPPC11-containing TRAPP III complex in isolation mem- not because of an endocytic defect since uptake of a fluid-phase brane formation and expansion (Figure 10). Work in S. cerevisiae has marker in the absence of either protein was not affected (Figure S6). shown that yeast TRAPP III interacts with Atg9-containing vesicles.25 This result supports the notion that both TRAPPC11 and TRAPPC12 This interaction was mediated by Trs85, the yeast homolog of affect lysosomal activity, likely due to a membrane trafficking defect TRAPPC8. In the yeast system, initial recruitment of TRAPP III was in the biosynthetic pathway. Atg9-dependent. A functional link between mammalian ATG9A and If reduced proteolytic activity in lysosomes following TRAPPC12 TRAPPC8 was also demonstrated.29,43 Thus, it is not unexpected that depletion accounts for the autophagic flux defect (see Figures 1 and TRAPPC8 depletion would function upstream of isolation membrane 2), we would expect to observe an overlap between LC3 and the lyso- formation and resemble an ATG9A depletion as we report here. Our somal marker LAMP1 following TRAPPC12 depletion, an indication model also suggests that recruitment of ATG2B-WIPI4 by TRAPPC11 A portion of TRAPP III localizes to isolation membranes. (A) HeLa cells stably expressing either ATG5-GFP or WIPI1-GFP were either starved for 30 minutes in EBSS containing bafilomycin A1 or treated with siRNA targeting either ATG2 or WIPI4. The cells were then immunostained for one of the TRAPP III proteins (TRAPPC8, TRAPPC11 or TRAPPC12) and GFP. Representative images are shown below the graph, with examples of co-localizing punctae indicated by a white arrowhead. The co-localization between TRAPP protein- and ATG5- or WIPI1-positive structures was quantified using Imaris and is plotted as the percentage of GFP-positive punctae that were positive for the indicated TRAPP III protein. A minimum of 10 cells were counted and the SEM is indicated. The scale bar represents 5 μm. In non-starved cells there were no visible ATG5-GFP or WIPI1-GFP punctae. (B) HeLa cells were depleted of TRAPPC11 using siRNA. The cells were then left untransfected or transfected with either BirA-C11 or C11-BirA. The cells were subsequently starved for 2 hours in EBSS and the recruitment of ATG2B, WIPI4 or ATG12 to membranes was monitored as described in the Materials and Methods section. (C) Lysate from the non-specific knockdown (lane 1), TRAPPC11 knockdown (lanes 2-4) either untransfected (lane 2) or transfected with either BirA-C11 (lane 3) or C11-BirA (lane 4) were probed with anti-TRAPPC11 antibody to assess expression levels of the BirA-tagged TRAPPC11 constructs. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD FIGURE 7 STANGA ET AL. FIGURE 8 Legend on next page. 337 338 STANGA ET AL. serves several purposes. Previous work demonstrated that depletion TRAPPC11 mutations display some form of muscular pathology or an of ATG2 resulted in unsealed membranes that did not fuse with lyso- elevated creatine kinase level. Using fibroblasts derived from an indi- somes and it was suggested that ATG2 is required for isolation mem- vidual with a compound heterozygous TRAPPC11 mutation, we have brane elongation.10,47 More recently, it has been suggested by both shown that they display a defect in autophagic flux likely due to in vitro and in vivo studies that ATG2 can function as a tether during impairment in ATG2B-WIPI4 recruitment to isolation membranes. this process,51,56,57 prior to an ATG5-dependent sealing of the mem- With TRAPPC11 now linked to a role in autophagy, there are three branes into autophagosomes.10 It has also been implied that WIPI4 distinct cellular roles for this protein including lipid-linked oligosaccha- regulates the association of ATG9A with LC3-containing mem- ride synthesis and membrane traffic in the endomembrane sys- branes.58 with tem.23,31 It is unclear which dysregulated cellular process contributes LC3-membranes,48 it is tempting to speculate that another conse- to the clinical phenotype or whether it is a combination of all three. quence of the TRAPPC11-dependent recruitment of ATG2B-WIPI4 is Indeed, both defective autophagy and glycosylation have been impli- the facilitation of the release of ATG9A from nascent isolation cated in muscular dystrophies, particularly those caused by improper membranes/omegasomes. glycosylation of the protein dystroglycan.62,63 Interestingly, one indi- Because ATG9A only associates transiently It is unclear at which step TRAPPC12 functions. Here, we have vidual with a TRAPPC11 mutation was recently shown to have a shown that autophagosomes are sealed following depletion of dystroglycanopathy.36 Furthermore, the glycosyltransferases essential TRAPPC12 (Figure 3) and thus the protein functions downstream of TRAPPC11. The poor co-localization between LC3 and the lysosomal marker LAMP1 following TRAPPC12 depletion suggests that the role of TRAPPC12 in autophagy is prior to autolysosome formation. We can envision three possible functions for TRAPPC12 in this scenario. First, TRAPPC12 could recruit a late-acting factor necessary for the fusion event, such as the putative interacting partner TECPR1,13 a protein previously implicated in autophagosome-lysosome fusion.59 Second, it could be required to facilitate the activity of an as yet unidentified autophagosomal membrane protein necessary for formation of autolysosomes. Finally, it may be required as a second or supplemental tethering factor between autophagosomes and lysosomes. Although TRAPP complexes were first thought to act as tethers, a direct role in this process has yet to be shown. In addition, the HOPS complex has been suggested to mediate the tethering between autophagosomes and lysosomes,60 so it is unclear if another tethering factor would be required. Further studies on TRAPPC12 will be required to determine its role in this late step of autophagy. Mutations in TRAPPC11 have been linked to disease.61 The phenotypes are broad and include liver and eye involvement, achalasia, alacrima and intellectual deficit. However, virtually all individuals with for some of the modifications on dystroglycan are Golgi-resident enzymes that utilize the biosynthetic pathway to reach their destination.64–67 Further studies are required to tease apart the various functions of TRAPPC11 and their effects on muscle physiology. We also considered the possibility that the cellular phenotypes seen upon depletion of the TRAPP III proteins might be due to impaired guanine nucleotide exchange factor (GEF) activity for Rab1 because this GTPase has been implicated in autophagosome biogenesis.68 In yeast, it is known that TRAPP III acts as a GEF for the Rab1 homolog Ypt1,69 and that Ypt1 is important for recruitment of both Atg1, the functional homolog of mammalian ULK1, and Atg11.26,70 In higher eukaryotes, only TRAPP II, and not TRAPP III, was reported to act as a GEF for Rab1.71,72 Thus, it seemed unlikely that depletion of TRAPP III proteins would affect GEF activity for Rab1. Nevertheless, we examined whether a dominant-negative form of Rab1 could phenocopy a TRAPPC11 depletion, and whether a constitutively-active form of Rab1 could suppress the phenotype resulting from TRAPPC11 depletion. Neither was observed (not shown) suggesting that the phenotypes seen upon depletion of TRAPP III proteins are not because of an effect on the activity of Rab1. Fibroblasts from an individual with TRAPPC11 mutations are defective in ATG2B-WIPI4 recruitment. (A) Control fibroblasts and fibroblasts derived from patients with bi-allelic TRAPPC12 (PatC12) or TRAPPC11 (PatC11) mutations were transferred into starvation medium with or without bafilomycin A1 for 2 hours. The bafilomycin A1-treated cells were then transferred into fresh nutrient-rich (DMEM) medium for up to 40 minutes. Lysates were prepared and LC3 and tubulin were detected by western analysis. A schematic of the treatment is shown above the graph. The inset shows a western analysis for TRAPPC11 in all three cell lines. (B) The same fibroblasts as in (A) were grown in either nutrientrich (DMEM) or starvation (EBSS) medium for 2 hours. Lysates were prepared and processed for membrane recruitment detection by floatation as described in the Materials and Methods section. The indicated proteins were detected by western analysis. Note that half the amount of tubulin was loaded in the PN fraction in order to obtain an exposure that shows tubulin in the other fractions without grossly overexposing the PN signal. (C) The same fibroblasts as in (A) were grown in starvation (EBSS) medium for 2 hours in the presence of bafilomycin A1. Lysates were prepared and processed for the membrane sealing assay as described in the Materials and Methods section. Prior to fractionation, some samples were treated with proteinase K (ProK) or with ProK in the presence of 1% Triton X-100 (TX). The fractions, composed of a postnuclear supernatant (PN), low-speed pellet (LP), high-speed pellet (HP) and high-speed supernatant (HS), were probed for LC3. Note that the exposure for the PatC12 immunoblot is longer than for the others in order to reveal the LC3-II bands. (D) The same fibroblasts as in (A) were grown in nutrient-rich (DMEM) or starvation (EBSS) medium for 2 hours. The cells were then fixed and stained with anti-ATG2B or anti-WIPI4 antibody. The ATG2B (light gray bars) and WIPI4 (dark gray bars) punctae were quantified from a total of 25 cells for each condition. Representative images are shown in Figure S5A. (E) Control (Ctrl) and PatC11 fibroblasts were grown in nutrient-rich (DMEM) medium and left untransfected (Ctrl and PatC11) or PatC11 cells were transfected by electroporation with a construct expressing either BirA-C11 or C11-BirA as indicated. The cells were then fixed and stained with anti-ATG2B or anti-WIPI4 antibody. The ATG2B (light gray bars) and WIPI4 (dark gray bars) punctae were quantified from a total of 25 cells for each condition. Representative images are shown in Figure S5B. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD FIGURE 8 339 STANGA ET AL. Lysosomal proteolytic activity is reduced upon depletion of either TRAPPC11 or TRAPPC12. (A) HeLa cells were treated with nonspecific (NS) siRNA, or siRNA targeting either TRAPPC11, TRAPPC12 or ATG2. Lysates were prepared and fractionated by SDS-PAGE and then probed for cathepsin B. The mature (m) form of the enzyme is indicated. (B) The same cells in (A) were treated with DQ-BSA and lysotracker (LysoT), and processed for fluorescence microscopy as described in the Materials and Methods section. All images were obtained using the same exposure conditions. (C) The same cells as in (A) were fixed and stained for the lysosomal marker protein LAMP1 and for LC3. Co-localization between the two signals was determined using Imaris. Representative cells are displayed in (D). A minimum of 25 cells was used for the quantification and the percentage of LAMP1 punctae that were also positive for LC3 is displayed. Numbers to the right of the immunoblots denote the migration of molecular size standards in kD. The scale bars in (B) and (D) represent 20 μm FIGURE 9 Several pieces of evidence implicate the carboxy-terminus of specific affinity is in addition supported by the fact that C11-BirA TRAPPC11 in the recruitment of the ATG2B-WIPI4 complex. First, of showed preference for some other proteins compared to BirA-C11 the TRAPPC11 mutations in the compound heterozygous individual (eg, TRAPPC8 and TRAPPC13). Finally, we demonstrate better examined in this study, one allele (p.Asp1127Valfs*47) has a frame- recruitment of ATG2B and WIPI4 to membranes with BirA-C11 com- shift mutation starting six amino acids from the carboxy terminus of pared to C11-BirA in both HeLa and fibroblast cells. (p. Our study not only implicates TRAPP III proteins in autophagy, Ala372_Ser429del) results in an unstable protein.33 Because a but it provides evidence that TRAPPC11 and TRAPPC12 are also reduced level of TRAPPC11 is detected in the cell lysate from these required for proper proteolytic function of the lysosome. This is likely fibroblasts, it is likely to be due to the p.Asp1127Valfs*47 variant. due to their role in biosynthetic membrane trafficking.23,54 While it is Given that ATG2B-WIPI4 is not recruited to membranes in these cells, possible that altered lysosomal activity could affect formation of a role for the carboxy-terminus in ATG2B-WIPI4 recruitment could be autolysosomes,73 we do not favor this possibility for several reasons. inferred. The BioID results are also consistent with this notion. In this In the case of TRAPPC11 depletion, the autophagic flux defect is cau- case, biotinylation of both ATG2B and WIPI4 was more efficient using sed by the resulting unsealed isolation membranes because sealed BirA-C11 (free-carboxy-terminus) compared to C11-BirA. This was membranes are required for autolysosome formation.10 This notion is not due to an expression-level difference because expression levels of further supported by the ATG2 knockdown that showed normal lyso- the BirA constructs were essentially the same (not shown). The somal activity with respect to cathepsin B and DQ-BSA (Figures 9A this 1133 amino acid-long protein. The second allele 340 STANGA ET AL. 4 | M A T E R I A L S A N D M ET H O D S 4.1 | Cell culture HeLa cells, HeLa cells stably expressing the autophagosomal marker mRFPGFP-LC376 and primary fibroblasts were cultured in Dulbecco's modified eagle medium (DMEM) (Wisent, St. Bruno, Quebec, Canada) supplemented with 10% (vol/vol) fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, Massachusetts) at 37
C in a humidified incubator with 5% CO2. A model for the role of TRAPP proteins in autophagy. The model depicts the various stages of autophagy and which stage is affected by the TRAPP III proteins. The green arrows indicate that the TRAPP protein influences that particular step. The dark line connecting TRAPPC11 to ATG2B is based on the BioID result (Table 1). See text for details FIGURE 10 4.2 | Starvation treatments The cells were seeded in 6- or 10-cm diameter dishes, washed twice with phosphate-buffered saline (PBS) and incubated with EBSS (Wisent, St. Bruno, Quebec, Canada) for the times indicated in each figure. In some cases, 200 nM Bafilomycin A1 (Sigma-Aldrich, St. Louis, Missouri) was included during a 2-hour starvation to block the fusion of and B), yet these unsealed structures did not fuse with the lysosomes autophagosomes with lysosomes. Proteolytic activity of lysosomes was (Figure 9C and D). Thus, there appears to be no correlation between inhibited by including both 10 μM E64d (Sigma-Aldrich, St. Louis, Mis- lysosomal proteolytic activity and isolation membrane sealing. In the souri) and 10 μM pepstatin A (Sigma-Aldrich, St. Louis, Missouri) for case of TRAPPC12 depletion, there was poor co-localization between 3 hours. In some instances, cells starved and treated with Bafilomycin A1 the lysosomal marker LAMP1 and LC3, suggesting that its functions in biosynthetic membrane traffic and autophagy are probably unrelated, although this will require more experimentation to better clarify. The precise function of ATG2 in autophagosome formation and its recruitment to membranes is unclear, but recent studies suggest it may function in isolation membrane expansion. An ultrastructural study on mouse and human cells indicated that upon depletion of ATG2, isolation membranes failed to expand.10 Some of the structures in the ATG2-depleted cells appeared circular in shape. However, because circular structures were not seen in ATG2A/ATG2B double knockout cells,74 were returned to nutrient-rich medium by washing the cells twice with PBS and incubating in DMEM with 10% FBS for 20 or 40 minutes. The cells were either lysed by harvesting in lysis buffer (150 mM NaCl, 50 mM Tris pH 7.2, 1 mM DTT, 1% Triton X-100, 0.5 mM EDTA, Complete protease inhibitors [Roche, Basel, Switzerland]) and analyzed by western blotting, or plated on 18-mm glass coverslips pre-coated with poly-L-lysine (Thermo Fisher Scientific, Waltham, Massachusetts) for immunostaining, or plated on 35-mm glass-bottom four micro-well separation dishes (Greiner, Kremsmunster, Austria) for time-lapse microscopy. such structures in the depleted cells probably resulted from an incomplete 4.3 | Transfection and RNA interference inactivation of the protein. Indeed, only small membranes were seen in Transfection of siRNAs (see Table 2) was performed using JetPrime the absence of both ATG2 proteins, supporting the notion that ATG2 is (Polyplus, Illkirch, France) as per the manufacturer's protocol. Sixteen required for isolation membrane elongation. Recent studies demonstrated hours prior to transfection, the cells were seeded at 80% confluence. an in vitro tethering ability for ATG2A,56,57 consistent with its role in For TRAPPC12 knockdown, cells were incubated with 20 nM siRNA membrane elongation. While the extreme amino-terminus of ATG2 is for 24 hours. For TRAPPC11, TRAPPC8, TRAPPC2, ATG9A, WIPI4 and 74 required for isolation membrane binding in vivo, the interactions that regulate its membrane association are unclear. Because WIPI proteins are ATG2 knockdowns the cells were transfected with 20 nM siRNA on day 1, passaged 1:2 on day 2, and re-transfected with siRNA on day 3. and WIPI4 interacts with The rescue experiments in HeLa cells were performed using siRNA- ATG2, WIPI4 is thought to recruit ATG2 to such membranes that form resistant BirA-FLAG-C11 and C11-FLAG-BirA. HeLa cells at 80% con- after starvation. However, this complex still interacts in vitro with artificial fluence were transfected with 20 nM TRAPPC11 siRNA, passaged after 52 recruited to PI3P-containing membranes, 14,56,57 liposomes devoid of PI3P. Thus, the recruitment of ATG2 in vivo may in fact be regulated by a number of different factors, and we propose that TRAPPC11 is one such factor. Our present study places TRAPPC11 upstream of ATG2B, prior to the ATG5-dependent sealing of isolation membranes. Furthermore, our study implicates ATG2B-WIPI4 recruitment to isolation membranes as critical for human health. Indeed, the function of WIPI4 has also been implicated in human health.58,75 Future studies will be aimed at characterizing the membranes that form in the 24 hours to 50% confluence on a 10-cm dish and transfected with 20 nM TRAPPC11 siRNA and 10 μg of DNA. The cells were harvested after 48 hours and processed for membrane recruitment of ATG2B as described below. The constructs used in the rescue experiment in primary fibroblasts were introduced into the cells by electroporation using the Neon Transfection System (Thermo Fisher Scientific, Waltham, Massachusetts). In this case, 1 μg of DNA was used per transfection and the cells were subjected to 1600 V for 20 mseconds, single pulse. absence of TRAPPC11 in both HeLa cells and primary fibroblasts derived from individuals with TRAPPC11 mutations at the ultrastructural level, as 4.4 | Time-lapse microscopy well as detailing the distinct roles of each TRAPP III subunit in the forma- For live-cell imaging, HeLa cells stably expressing mRFP-GFP-LC3 were tion of isolation membranes. plated in 35-mm glass-bottom dishes. Time-lapse microscopy was 341 STANGA ET AL. TABLE 2 List of siRNAs used in this study Targeted gene Sense sequence (50 ! 30 ) Source TRAPPC2 UCCAUUUUAUGAACCCAAUTT Life Technologies TRAPPC8 CAGCUCUCCUAAUACGGUUTT Life Technologies TRAPPC11 GGAUUUAUAAACUACAAGGATT Life Technologies TRAPPC12 CGGACAAGCUGAACGAACATT Life Technologies ATG2A GCAUUCCCAGUUGUUGGAGUUCCUA Life Technologies ATG2B AGGUCUCUCUUGUCUGGCAUCUUUA Life Technologies WDR45/WIPI4 CCAGTGATAAGGGTACTGTCCATAT Life Technologies ATG9A CCAGAACUACAUGGUGGCACUGGUU Life Technologies TABLE 3 List of antibodies used in this study Antigen Type Host IF dilution WB dilution Size (kDa) Catalog number Source TRAPPC2 P r N/A 1:1000 16 N/A Sacher laboratory TRAPPC8 P r N/A 1:500 160 ab122692 Abcam TRAPPC10 M m N/A 1:1000 130 H00007109-M01 Abnova TRAPPC11 P r N/A 1:1000 129 N/A Sacher laboratory TRAPPC12 P r N/A 1:2500 78 N/A Sacher laboratory tubulin M m N/A 1:5000 50 T6199 Sigma-Aldrich LC3 P r 1:300 1:2500 15/17 ab51520 Abcam P m 1:100 1:1000 15/17 sc-398 822 Santa Cruz p62 (SQSTM1) P r N/A 1:2500 48 ab155686 Abcam ATG2B P r 1:500 1:1000 240 ab116215 Abcam ATG12 M m N/A 1:1000 55a sc-271 688 Santa Cruz WDR45/WIPI4 M m 1:100 1:1000 40 sc-398 272 Santa Cruz P r 1:100 1:1000 40 OAAB09622 Aviva ATG9A M r N/A 1:1000 95 ab108338 Abcam FLAG M m 1:300 1:5000 N/A F3165 Sigma-Aldrich cathepsin B M m N/A 1:1000 24-31 ab58802 Abcam LAMP1 P r 1:300 N/A N/A ab24170 Abcam GFP M m 1:200 N/A N/A 11 814 460 001 Sigma-Aldrich HCA g 1:500 N/A N/A A-11034 Life Technologies Secondary antibodies Alexa Fluor 488 anti–rabbit Alexa Fluor 488 anti–mouse HCA g 1:500 N/A N/A A-11029 Life Technologies Alexa Fluor 568 anti–rabbit HCA g 1:500 N/A N/A A-11036 Life Technologies Alexa Fluor 568 anti–mouse HCA g 1:500 N/A N/A A-11031 Life Technologies Alexa Fluor 647 anti–mouse HCA g 1:500 N/A N/A A-21235 Life Technologies HRP-labeled anti-mouse HCA g N/A 1:5000 N/A KP-474-1806 KPL HRP-labeled anti-rabbit HCA g N/A 1:5000 N/A KP-474-1506 KPL IF, immunofluorescence, g, goat; HCA, highly cross-adsorbed; M, monoclonal; m, mouse; N/A, not applicable; P, polyclonal; r, rabbit; WB, western blot. a molecular size of the ATG12-ATG5 conjugate. performed after transfection using a 60× (NA 1.4) or 100× (NA 1.46) oil Belfast, United Kingdom) and ImageJ (National Institutes of Health). For immersion objective, no binning, on an inverted confocal microscope measuring the fusion of autophagosomes a Zeiss LSM 880 Airyscan (LiveScan Swept Field; Nikon, Tokyo, Japan), Piezo Z stage (Nano- microscope with a 100× (NA 1.4) oil immersion objective was used. Z100N; Mad City Labs, Inc., Madison, Wisconsin), and an electron- Images were acquired every 20 seconds for 15 minutes at 0.3 μm incre- multiplying charge-coupled device camera (512 × 512; iXon X3; Andor ment sizes. The vesicles were tracked over time and the intensities of Technology, Belfast, United Kingdom). The microscope was equipped GFP and RFP were measured using Imaris. with an environmental chamber heated to 37
C with 5% CO2. Images were acquired with NIS-Elements Version 4.0 acquisition software every 10 minutes using a 0.2 second exposure at 0.3 μm increment sizes. 4.5 | Immunofluorescence microscopy Images were deconvolved with AutoQuant X3 software (Media Cyber- The cells were gently washed with PBS then fixed with 4% parafor- netics, Rockville, Maryland) and analyzed on Imaris version 7.6 (Bitplane, maldehyde (PFA) for 20 minutes at room temperature, quenched 342 STANGA ET AL. with 0.1 M glycine for 10 minutes and permeabilized with 0.1% Triton X-100 for 7 minutes. The cells were then blocked in 5% normal goat serum (Cell Signaling Technology, Danvers, Massachusetts) in PBS for 40 minutes at room temperature. Primary antibodies (see Table 3) were diluted in 5% normal goat serum and were added to coverslips and incubated for 16 hours at 4
C. Cells were then washed three times with PBS for 10 minutes per wash. DAPI was added during the first wash at a final concentration of 0.1 μg/mL. Cross-adsorbed secondary antibodies (Life Technologies, Carlsbad, California; see Table 3) were diluted in 5% normal goat serum in PBS and applied for 1 hour at room temperature. Coverslips were washed three times with PBS for 10 minutes per wash, mounted with Prolong Gold AntiFade reagent (Life Technologies, Carlsbad, California), and sealed with nail polish. 12-bit images (1024 × 1024-pixel resolution) were recorded on an Olympus FluoView confocal laser scanning microscope. The images were acquired with a 0.3 μm increment size. 4.8 | Subcellular fractionation and proteinase K treatment Subcellular fractionation and treatment of lysates with proteinase K was performed essentially as described in Velikkakath et al.47 Briefly, cells were seeded at full confluence on six-well dishes. HeLa cells were treated with siRNA as indicated in the figure and were cultured in either DMEM with 10% FBS (non-starved) or EBSS containing 200 nM bafilomycin A1 for 2 hours (starved and autophagosomelysosome fusion-impaired). Cells were then collected and homogenized in 250 μL homogenization buffer (10 mM 4-(2-hydroxyethyl)1-piperazineethanesulfonic acid [HEPES]-KOH pH 7.4, 0.22 M mannitol, 0.07 M sucrose, and protease inhibitors) by 10 cycles using a syringe with a 27 gauge needle. The post-nuclear supernatant (PN) was obtained by centrifugation at 300×g for 5 minutes in a refrigerated microfuge. A total of 150 μg of PN was further centrifuged at 7700×g for 5 minutes in a refrigerated microfuge to generate the low-speed pellet (LP). The resulting supernatant was centrifuged at 100000×g for 30 minutes in a table-top ultracentrifuge in a TLA100 rotor to generate 4.6 | Western blotting Samples (7 μg total protein for LC3 analysis or 30 μg total protein for all other proteins) were analyzed on either 8% or 15% SDSpolyacrylamide gels. The proteins were transferred to nitrocellulose membranes (BioRad, Hercules, California) for ATG2B or polyvinylidene difluoride (PVDF) membrane (Millipore, Burlington, Massachusetts) for all other proteins for 1 hour at 100 V. Membranes were blocked with 5% skimmed milk powder in PBS-T (PBS with 0.1% Tween 20 [vol/vol]) for 1 hour. The primary and secondary antibodies used, and their dilutions, are listed in Table 3. Primary antibodies were incubated in PBS-T overnight and secondary antibodies were incubated for 1 hour. Membranes were then incubated with ECL reagent (GE Healthcare, Chicago, a high-speed pellet (HP) and a high speed supernatant (HS). The LP and HP were resuspended in the same volume of homogenization buffer. To analyze the integrity of the membranes, the LP and HP (250 μL) were either left untreated or incubated with 100 μg/mL proteinase K (Invitrogen, Carlsbad, California) with or without 1% Triton X-100 on ice for 30 minutes. All samples (LP, HP and HS) were precipitated by incubating in 10% trichloroacetic acid (TCA) on ice for 30 minutes, washed with ice-cold homogenization buffer and resuspended in 1X Laemmli sample buffer. The resulting sample was then heated to 95
C for 5 minutes and analyzed by SDS-PAGE followed by western analysis for the indicated proteins. Primary fibroblasts were processed as described for the HeLa cells. Illinois) and detected using an Amersham Imager 600. 4.9 | ATG2 recruitment assay 4.7 | Immunoprecipitation HeLa cells (107 cells/10 cm dish) were washed twice with PBS and lysed in 1 mL of ice-cold RIPA buffer (10 mM Tris-Cl pH 8.0, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 140 mM NaCl, Complete protease inhibitors [Roche, Basel, Switzerland]). The lysate was clarified at 14000×g in a precooled microcentrifuge for 15 minutes. A total of 500 μg of lysate was incubated with 30 μL of protein A-agarose beads (50% slurry) (Bioshop Canada, Burlington, Ontario, Canada) at 4
C for 30 minutes on an orbital shaker. The agarose beads were removed by a brief centrifugation at 5000 rpm in a refrigerated microfuge for 30 seconds. The cleared lysate was incubated overnight at 4
C with 10 μg of TRAPPC11 antiserum. Immune Cells were seeded at full confluence on 10 cm dishes and mechanically lysed as described above. A PN was produced as described above and a total of 500 μg of PN was centrifugated at 7700×g for 30 minutes to generate the LP. The resulting supernatant fraction was centrifuged at 100000×g for 2 hours in a table-top ultracentrifuge in a TLA100 rotor to generate HP and HS fractions. The LP and HP were washed with ice-cold homogenization buffer to remove all traces of non-membrane-bound proteins and resuspended in 1X Laemmli sample buffer. The HS was precipitated with TCA as above, washed with homogenization buffer and resuspended in sample buffer. Samples were heated to 95
C for 5 minutes prior to western analysis. complexes were captured by adding 30 μL protein A-agarose bead 4.10 | Proximity biotinylation (BioID) slurry (blocked with 5% BSA in PBS-T) and gently rocking on an orbital The Flp-In/T-REx 293 cell line from Life Technologies was used to gener- shaker overnight at 4
C. The beads were collected by a brief centrifu- ate BioID cell lines with BirA fusion proteins.46 Parental cells were seeded gation at 5000 rpm in a refrigerated microfuge for 30 seconds and in a 10 cm dish with 15 ug/mL blasticidin 1 day before transfection to washed 3 times with 1 mL ice-cold PBS-T. The beads were obtain 70% confluency on the day of transfection. Transfections were resuspended in 60 μL 2× Laemmli sample buffer and heated at 95
C performed using Jetprime (Polyplus, Illkirch, France) containing 1.5 μg BirA for 5 minutes to dissociate the immunocomplexes. The beads were construct (C11-BirA or BirA-C11 in pDEST_pcDNA5_BirA-FLAG_Cterm pelleted by centrifugation and SDS-PAGE was performed with the or pDEST_pcDNA5_BirA-FLAG_Nterm, respectively, produced by supernatant fraction. Gateway cloning) and 13.5 μg pOG44. Transfected cells were 343 STANGA ET AL. incubated in a 37
C incubator for 24 hours before changing the medium with fresh antibiotic-free DMEM containing 10% FBS. The following day, cells were passaged at 1:4 dilution (~25% confluence) onto two 10 cm dishes in medium containing 100 μg/mL hygromycin and 15 μg/mL blasticidin. The medium was changed every 2 to 3 days during the 2-week selection until colonies were visible. The colonies were pooled to obtain a polyclonal cell line which was expanded in medium containing hygromycin and blasticidin. Two control cell lines were generated with pDEST_pcDNA5_BirA-FLAG_Nterm or pDEST_pcDNA5_BirA-FLAG_Cterm plasmids followed by selection as described above. To obtain purified TRAPPC11-interacting proteins, BirA-C11 and C11-BirA cell lines were grown in 15 cm dishes. At 60% confluency, they were treated with 50 ng/mL tetracycline for 16 hours at which Author contributions D.S. performed the experiments, analyzed and interpreted the data, prepared figures, and wrote and edited the manuscript. Q.Z. performed the BioID experiments, analyzed the data and edited the manuscript. M. M. performed the experiments, analyzed and interpreted the data, prepared figures and edited the manuscript. D. S.-D. performed the experiments and edited the manuscript. C. J.-M. identified the individual with a TRAPPC11 mutation (PatC11) and edited the manuscript. M. S. conceptualized the study, analyzed and interpreted the data, and wrote and edited the manuscript. ORCID Michael Sacher https://orcid.org/0000-0003-2926-5064 time the medium was changed and supplemented with 20% FBS, 50 ng/mL tetracycline and 50 μM biotin. After 8 hours, cells were RE FE RE NC ES pelleted and lysed in ice-cold RIPA buffer (50 mM Tris pH 7.4, 1. Axe EL, Walker SA, Manifava M, et al. Autophagosome formation from membrane compartments enriched in phosphatidylinositol 3-phosphate and dynamically connected to the endoplasmic reticulum. J Cell Biol. 2008;182(4):685-701. 2. Hailey DW, Rambold AS, Satpute-Krishnan P, et al. Mitochondria supply membranes for autophagosome biogenesis during starvation. Cell. 2010;141(4):656-667. 3. Hamasaki M, Furuta N, Matsuda A, et al. Autophagosomes form at ER-mitochondria contact sites. Nature. 2013;495(7441):389-393. 4. Hayashi-Nishino M, Fujita N, Noda T, Yamaguchi A, Yoshimori T, Yamamoto A. A subdomain of the endoplasmic reticulum forms a cradle for autophagosome formation. Nat Cell Biol. 2009;11(12):14331437. 5. Ravikumar B, Moreau K, Jahreiss L, Puri C, Rubinsztein DC. Plasma membrane contributes to the formation of pre-autophagosomal structures. Nat Cell Biol. 2010;12(8):747-757. 6. Young ARJ, Chan EY, Hu XW, et al. Starvation and ULK1-dependent cycling of mammalian Atg9 between the TGN and endosomes. J Cell Sci. 2006;119(18):3888-3900. 7. Levine B, Klionsky DJ. Autophagy wins the 2016 Nobel prize in physiology or medicine: breakthroughs in baker's yeast fuel advances in biomedical research. Proc Natl Acad Sci U S A. 2017;114(2):201-205. 8. Russell RC, Tian Y, Yuan H, et al. ULK1 induces autophagy by phosphorylating Beclin-1 and activating VPS34 lipid kinase. Nat Cell Biol. 2013;15(7):741-750. 9. Ichimura Y, Kirisako T, Takao T, et al. A ubiquitin-like system mediates protein lipidation. Nature. 2000;408(6811):488-492. 10. Kishi-Itakura C, Koyama-Honda I, Itakura E, Mizushima N. Ultrastructural analysis of autophagosome organization using mammalian autophagy-deficient cells. J Cell Sci. 2014;127(Pt 18):4089-4102. 11. Bakula D, Muller AJ, Zuleger T, et al. WIPI3 and WIPI4 beta-propellers are scaffolds for LKB1-AMPK-TSC signalling circuits in the control of autophagy. Nat Commun. 2017;8:15637. 12. Baskaran S, Ragusa MJ, Boura E, Hurley JH. Two-site recognition of phosphatidylinositol 3-phosphate by PROPPINs in autophagy. Mol Cell. 2012;47(3):339-348. 13. Behrends C, Sowa ME, Gygi SP, Harper JW. Network organization of the human autophagy system. Nature. 2010;466(7302):68-76. 14. Zheng JX, Li Y, Ding YH, et al. Architecture of the ATG2B-WDR45 complex and an aromatic Y/HF motif crucial for complex formation. Autophagy. 2017;13(11):1870-1883. 15. Weidberg H, Shvets E, Shpilka T, Shimron F, Shinder V, Elazar Z. LC3 and GATE-16/GABARAP subfamilies are both essential yet act differently in autophagosome biogenesis. EMBO J. 2010;29(11):1792-1802. 16. Bassik MC, Kampmann M, Lebbink RJ, et al. A systematic mammalian genetic interaction map reveals pathways underlying ricin susceptibility. Cell. 2013;152(4):909-922. 17. Lynch-Day MA, Bhandari D, Menon S, et al. Trs85 directs a Ypt1 GEF, TRAPPIII, to the phagophore to promote autophagy. Proc Natl Acad Sci U S A. 2010;107(17):7811-7816. 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA) supplemented with complete-mini protease inhibitor cocktail (Roche, Basel, Switzerland), 1 mM PMSF, 62.5 Units/ml benzonase, and incubated at 4
C for 1 hour. Samples were sonicated in a 4
C ice bath (4 bursts of 10 seconds on low power). After sonication, samples were centrifuged at 4
C for 30 minutes at 13200 rpm in a refrigerated microfuge. Supernatants were incubated with 14 μL streptavidin-Sepharose beads (Sigma) per dish at 4
C for 3 hours to capture the denatured biotinylated proteins. After the incubation, beads were washed five times with RIPA and four times with lowdetergent buffer (25 mM Tris pH 7.4, 100 mM NaCl, 0.025% SDS). Following the last wash, beads were pelleted and analyzed by mass spectrometry as previously described77 to identify proteins captured by the BioID purification. Data were analyzed using Scaffold 4 proteomic software. ACKNOWLEDGMENTS We are grateful to Dr. Markus Damme (Christian-AlbrechtsUniversität zu Kiel) for many helpful discussions and suggestions throughout the course of this study, and to Drs. Anne-Claude Gingras (Lunenfeld-Tanenbaum Research Institute), Noboru Mizushima (University of Tokyo) and David Rubinsztein (University of Cambridge) for providing the pDEST_pcDNA5_BirA plasmids for amino- and carboxy-terminal tagging, the GFP-tagged cell lines used in Figure 2, and the HeLa cell line stably expressing mRFP-GFP-LC3, respectively. We thank all members of the Sacher laboratory as well as Dr. Christopher Brett (Concordia University) for critical comments on this manuscript. This study was funded by research grants from the Canadian Institutes of Health Research and the Natural Sciences and Engineering Research Council of Canada (to M.S.), the Instituto de Salud Carlos III Subdirección General de Evaluación y Fomento de la Investigación Sanitaria (PI16/00579, CP09/00011 to C.J.-M.) and the European Regional Development Fund (FEDER a way to achieve Europe to C.J.-M.). All microscopy in this study was performed at the Centre for Microscopy and Cellular Imaging at Concordia University. Mass spectrometry was performed at the RI-MUHC at McGill University. 344 18. Riedel F, Galindo A, Muschalik N, Munro S. The two TRAPP complexes of metazoans have distinct roles and act on different Rab GTPases. J Cell Biol. 2017;217(2):601. 19. Sacher M, Barrowman J, Wang W, et al. TRAPP I implicated in the specificity of tethering in ER-to-Golgi transport. Mol Cell. 2001;7(2): 433-442. 20. Sacher M, Jiang Y, Barrowman J, et al. TRAPP, a highly conserved novel complex on the cis-Golgi that mediates vesicle docking and fusion. EMBO J. 1998;17(9):2494-2503. 21. Brunet S, Sacher M. In sickness and in health: the role of TRAPP and associated proteins in disease. Traffic. 2014;15(8):803-818. 22. Kim JJ, Lipatova Z, Segev N. TRAPP complexes in secretion and autophagy. Front Cell Dev Biol. 2016;4:20. 23. Scrivens PJ, Noueihed B, Shahrzad N, Hul S, Brunet S, Sacher M. C4orf41 and TTC-15 are mammalian TRAPP components with a role at an early stage in ER-to-Golgi trafficking. Mol Biol Cell. 2011;22(12): 2083-2093. 24. Brunet S, Shahrzad N, Saint-Dic D, Dutczak H, Sacher M. A trs20 mutation that mimics an SEDT-causing mutation blocks selective and non-selective autophagy: a model for TRAPP III organization. Traffic. 2013;14(10):1091-1104. 25. Kakuta S, Yamamoto H, Negishi L, Kondo-Kakuta C, Hayashi N, Ohsumi Y. Atg9 vesicles recruit vesicle-tethering proteins Trs85 and Ypt1 to the autophagosome formation site. J Biol Chem. 2012;287 (53):44261-44269. 26. Lipatova Z, Belogortseva N, Zhang XQ, Kim J, Taussig D, Segev N. Regulation of selective autophagy onset by a Ypt/Rab GTPase module. Proc Natl Acad Sci U S A. 2012;109(18):6981-6986. 27. Shirahama-Noda K, Kira S, Yoshimori T, Noda T. TRAPPIII is responsible for vesicular transport from early endosomes to Golgi, facilitating Atg9 cycling in autophagy. J Cell Sci. 2013;126(Pt 21): 4963-4973. 28. Taussig D, Lipatova Z, Segev N. Trs20 is required for TRAPP III complex assembly at the PAS and its function in autophagy. Traffic. 2014; 15(3):327-337. 29. Lamb CA, Nuhlen S, Judith D, et al. TBC1D14 regulates autophagy via the TRAPP complex and ATG9 traffic. EMBO J. 2016;35(3): 281-301. 30. Ramirez-Peinado S, Ignashkova TI, van Raam BJ, et al. TRAPPC13 modulates autophagy and the response to Golgi stress. J Cell Sci. 2017;130(14):2251-2265. 31. DeRossi C, Vacaru A, Rafiq R, et al. trappc11 is required for protein glycosylation in zebrafish and humans. Mol Biol Cell. 2016;27(8):12201234. 32. Milev MP, Hasaj B, Saint-Dic D, Snounou S, Zhao Q, Sacher M. TRAMM/TrappC12 plays a role in chromosome congression, kinetochore stability, and CENP-E recruitment. J Cell Biol. 2015;209(2): 221-234. 33. Bogershausen N, Shahrzad N, Chong JX, et al. Recessive TRAPPC11 mutations cause a disease spectrum of limb girdle muscular dystrophy and myopathy with movement disorder and intellectual disability. Am J Hum Genet. 2013;93(1):181-190. 34. Fee DB, Harmelink M, Monrad P, Pyzik E. Siblings with mutations in TRAPPC11 presenting with limb-girdle muscular dystrophy 2S. J Clin Neuromuscul Dis. 2017;19(1):27-30. 35. Koehler K, Milev MP, Prematilake K, et al. A novel TRAPPC11 mutation in two Turkish families associated with cerebral atrophy, global retardation, scoliosis, achalasia and alacrima. J Med Genet. 2017;54(3): 176-185. 36. Larson AA, Baker PR 2nd, Milev MP, et al. TRAPPC11 and GOSR2 mutations associate with hypoglycosylation of alpha-dystroglycan and muscular dystrophy. Skelet Muscle. 2018;8(1):17. 37. Liang WC, Zhu W, Mitsuhashi S, et al. Congenital muscular dystrophy with fatty liver and infantile-onset cataract caused by TRAPPC11 mutations: broadening of the phenotype. Skelet Muscle. 2015;5:29. 38. Matalonga L, Bravo M, Serra-Peinado C, et al. Mutations in TRAPPC11 are associated with a congenital disorder of glycosylation. Hum Mutat. 2017;38(2):148-151. 39. Milev MP, Grout ME, Saint-Dic D, et al. Mutations in TRAPPC12 manifest in progressive childhood encephalopathy and Golgi dysfunction. Am J Hum Genet. 2017;101(2):291-299. STANGA ET AL. 40. Kim YG, Raunser S, Munger C, et al. The architecture of the multisubunit TRAPP I complex suggests a model for vesicle tethering. Cell. 2006;127(4):817-830. 41. Tan D, Cai Y, Wang J, et al. The EM structure of the TRAPPIII complex leads to the identification of a requirement for COPII vesicles on the macroautophagy pathway. Proc Natl Acad Sci U S A. 2013;110(48): 19432-19437. 42. Yip CK, Berscheminski J, Walz T. Molecular architecture of the TRAPPII complex and implications for vesicle tethering. Nat Struct Mol Biol. 2010;17(11):1298-1304. 43. Imai K, Hao F, Fujita N, et al. Atg9A trafficking through the recycling endosomes is required for autophagosome formation. J Cell Sci. 2016; 129(20):3781-3791. 44. Zong M, Wu XG, Chan CW, et al. The adaptor function of TRAPPC2 in mammalian TRAPPs explains TRAPPC2-associated SEDT and TRAPPC9-associated congenital intellectual disability. PLoS One. 2011;6(8):e23350. 45. Kimura S, Noda T, Yoshimori T. Dissection of the autophagosome maturation process by a novel reporter protein, tandem fluorescenttagged LC3. Autophagy. 2007;3(5):452-460. 46. Roux KJ, Kim DI, Raida M, Burke B. A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells. J Cell Biol. 2012;196(6):801-810. 47. Velikkakath AKG, Nishimura T, Oita E, Ishihara N, Mizushima N. Mammalian Atg2 proteins are essential for autophagosome formation and important for regulation of size and distribution of lipid droplets. Mol Biol Cell. 2012;23(5):896-909. 48. Orsi A, Razi M, Dooley HC, et al. Dynamic and transient interactions of Atg9 with autophagosomes, but not membrane integration, are required for autophagy. Mol Biol Cell. 2012;23(10):1860-1873. 49. Klionsky DJ, Abdelmohsen K, Abe A, et al. Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy. 2016;12(1):1-222. 50. Zhou F, Zou S, Chen Y, et al. A Rab5 GTPase module is important for autophagosome closure. PLoS Genet. 2017;13(9):e1007020. 51. Gomez-Sanchez R, Rose J, Guimaraes R, et al. Atg9 establishes Atg2-dependent contact sites between the endoplasmic reticulum and phagophores. J Cell Biol. 2018;217(8):2743-2763. 52. Grimmel M, Backhaus C, Proikas-Cezanne T. WIPI-mediated Autophagy and longevity. Cell. 2015;4(2):202-217. 53. Hanewinkel H, Glossl J, Kresse H. Biosynthesis of cathepsin B in cultured normal and I-cell fibroblasts. J Biol Chem. 1987;262(25):1235112355. 54. Wendler F, Gillingham AK, Sinka R, et al. A genome-wide RNA interference screen identifies two novel components of the metazoan secretory pathway. EMBO J. 2010;29(2):304-314. 55. Mora R, Dokic I, Kees T, et al. Sphingolipid rheostat alterations related to transformation can be exploited for specific induction of lysosomal cell death in murine and human glioma. Glia. 2010;58(11): 1364-1383. 56. Chowdhury S, Otomo C, Leitner A, et al. Insights into autophagosome biogenesis from structural and biochemical analyses of the ATG2AWIPI4 complex. Proc Natl Acad Sci U S A. 2018;115:E9792-E9801. 57. Kotani T, Kirisako H, Koizumi M, Ohsumi Y, Nakatogawa H. The Atg2-Atg18 complex tethers pre-autophagosomal membranes to the endoplasmic reticulum for autophagosome formation. Proc Natl Acad Sci U S A. 2018;115(41):10363-10368. 58. Saitsu H, Nishimura T, Muramatsu K, et al. De novo mutations in the autophagy gene WDR45 cause static encephalopathy of childhood with neurodegeneration in adulthood. Nat Genet. 2013;45(4):445449. 449e441. 59. Chen D, Fan W, Lu Y, Ding X, Chen S, Zhong Q. A mammalian autophagosome maturation mechanism mediated by TECPR1 and the Atg12-Atg5 conjugate. Mol Cell. 2012;45(5):629-641. 60. Jiang P, Nishimura T, Sakamaki Y, et al. The HOPS complex mediates autophagosome-lysosome fusion through interaction with syntaxin 17. Mol Biol Cell. 2014;25(8):1327-1337. 61. Sacher M, Shahrzad N, Kamel H, Milev MP. TRAPPopathies, an emerging set of disorders linked to variations in the genes encoding transport protein particle (TRAPP)-associated proteins. Traffic. 2019; 20(1):5-26. 345 STANGA ET AL. 62. Michele DE, Barresi R, Kanagawa M, et al. Post-translational disruption of dystroglycan-ligand interactions in congenital muscular dystrophies. Nature. 2002;418(6896):417-422. 63. Sandri M, Coletto L, Grumati P, Bonaldo P. Misregulation of autophagy and protein degradation systems in myopathies and muscular dystrophies. J Cell Sci. 2013;126(Pt 23):5325-5333. 64. Grewal PK, McLaughlan JM, Moore CJ, Browning CA, Hewitt JE. Characterization of the LARGE family of putative glycosyltransferases associated with dystroglycanopathies. Glycobiology. 2005;15(10):912-923. 65. Keramaris-Vrantsis E, Lu PJ, Doran T, et al. Fukutin-related protein localizes to the Golgi apparatus and mutations lead to mislocalization in muscle in vivo. Muscle Nerve. 2007;36(4):455-465. 66. Matsumoto H, Noguchi S, Sugie K, et al. Subcellular localization of fukutin and fukutin-related protein in muscle cells. J Biochem. 2004; 135(6):709-712. 67. Yoshida A, Kobayashi K, Manya H, et al. Muscular dystrophy and neuronal migration disorder caused by mutations in a glycosyltransferase, POMGnT1. Dev Cell. 2001;1(5):717-724.  68. Zoppino FCM, Militello RD, Slavin I, Alvarez C, Colombo MI. 69. 70. 71. 72. Autophagosome formation depends on the small GTPase Rab1 and functional ER exit sites. Traffic. 2010;11(9):1246-1261. Thomas LL, Joiner AMN, Fromme JC. The TRAPPIII complex activates the GTPase Ypt1 (Rab1) in the secretory pathway. J Cell Biol. 2018; 217(1):283-298. Wang J, Menon S, Yamasaki A, et al. Ypt1 recruits the Atg1 kinase to the preautophagosomal structure. Proc Natl Acad Sci U S A. 2013;110 (24):9800-9805. Yamasaki A, Menon S, Yu S, et al. mTrs130 is a component of a mammalian TRAPPII complex, a Rab1 GEF that binds to COPI-coated vesicles. Mol Biol Cell. 2009;20(19):4205-4215. Li C, Luo X, Zhao S, et al. COPI-TRAPPII activates Rab18 and regulates its lipid droplet association. EMBO J. 2017;36(4):441-457. 73. Seranova E, Connolly KJ, Zatyka M, et al. Dysregulation of autophagy as a common mechanism in lysosomal storage diseases. Essays Biochem. 2017;61(6):733-749. 74. Tamura N, Nishimura T, Sakamaki Y, Koyama-Honda I, Yamamoto H, Mizushima N. Differential requirement for ATG2A domains for localization to autophagic membranes and lipid droplets. FEBS Lett. 2017; 591(23):3819-3830. 75. Haack TB, Hogarth P, Kruer MC, et al. Exome sequencing reveals de novo WDR45 mutations causing a phenotypically distinct, Xlinked dominant form of NBIA. Am J Hum Genet. 2012;91(6): 1144-1149. 76. Sarkar S, Korolchuk VI, Renna M, Winslow AR, Rubinsztein DC. Methodological considerations for assessing autophagy modulators: a study with calcium phosphate precipitates. Autophagy. 2009;5(3):307-313. 77. Brunet S, Saint-Dic D, Milev MP, Nilsson T, Sacher M. The TRAPP subunit Trs130p interacts with the GAP Gyp6p to mediate Ypt6p dynamics at the late Golgi. Front Cell Dev Biol. 2016;4:48. SUPPOR TI NG I NFORMATION Additional supporting information may be found online in the Supporting Information section at the end of this article. How to cite this article: Stanga D, Zhao Q, Milev MP, SaintDic D, Jimenez-Mallebrera C, Sacher M. TRAPPC11 functions in autophagy by recruiting ATG2B-WIPI4/WDR45 to preautophagosomal membranes. Traffic. 2019;20:325–345. https://doi.org/10.1111/tra.12640