Structure-based Analyses Reveal Distinct Binding Sites forAtg2 and Phosphoinositides in Atg18*□S

Received for publication, July 3, 2012, and in revised form, July 24, 2012 Published, JBC Papers in Press, July 31, 2012, DOI 10.1074/jbc.M112.397570

Yasunori Watanabe‡§, Takafumi Kobayashi¶, Hayashi Yamamoto¶, Hisashi Hoshida�, Rinji Akada�,Fuyuhiko Inagaki**, Yoshinori Ohsumi¶, and Nobuo N. Noda‡1

From the ‡Institute of Microbial Chemistry, Tokyo, Tokyo 141-0021, the §Graduate School of Life Science, Hokkaido University,Sapporo 060-0812, the ¶Frontier Research Center, Tokyo Institute of Technology, Yokohama 226-8503, the �Department of AppliedMolecular Bioscience, Yamaguchi University Graduate School of Medicine, Ube 755-8611, and the **Faculty of Advanced LifeScience, Hokkaido University, Sapporo 001-0021, Japan

Background: Atg18 plays a critical role in autophagy as a complex with Atg2 and phosphatidylinositol 3-phosphate.Results: The structure of the Atg18 paralog was determined, and important residues in Atg18 were identified.Conclusion: Atg18 recognizes Atg2 and membranes using distinct regions.Significance: This study will be a basis for elucidating the function of Atg18 in autophagy.

Autophagy is an intracellular degradation system by whichcytoplasmic materials are enclosed by an autophagosome anddelivered to a lysosome/vacuole. Atg18 plays a critical role inautophagosome formation as a complex with Atg2 and phos-phatidylinositol 3-phosphate (PtdIns(3)P). However, little isknown about the structure of Atg18 and its recognitionmode ofAtg2 or PtdIns(3)P. Here, we report the crystal structure ofKluyveromyces marxianusHsv2, an Atg18 paralog, at 2.6 Å res-olution. The structure reveals a seven-bladed �-propeller with-out circular permutation. Mutational analyses of Atg18 basedon the K. marxianus Hsv2 structure suggested that Atg18 hastwo phosphoinositide-binding sites at blades 5 and 6, whereastheAtg2-binding region is located at blade 2. Pointmutations inthe loops of blade 2 specifically abrogated autophagy withoutaffecting another Atg18 function, the regulation of vacuolarmorphology at the vacuolar membrane. This architectureenables Atg18 to form a complex with Atg2 and PtdIns(3)P inparallel, thereby functioning in the formation of autophago-somes at autophagic membranes.

Macroautophagy (hereafter referred to as autophagy) is anintracellular degradation system conserved among eukaryotesfrom yeast to mammals. During autophagy, a double-mem-brane structure called an autophagosome sequesters a portionof the cytoplasm and fuses with a vacuole (or lysosome in thecase of mammalian autophagy) to deliver its inner contents to

the lumenof the organelle (1). Autophagy is important in awiderange of physiological processes, such as adaptation to starva-tion, quality control of intracellular proteins and organelles,embryonic development, elimination of intracellular microbes,and prevention of neurodegeneration and tumor formation(2–4).Currently, �30 genes involved in autophagy have been iso-

lated in yeast and termed autophagy-related (ATG) genes.Among these genes, ATG1–10, ATG12–14, ATG16–18,ATG29, and ATG31 are essential for autophagosome forma-tion during starvation-induced autophagy, and the 18 Atgproteins they encode are classified into six functional groups(1, 5): (i) starvation-responsive Atg1 kinase complex, (ii)class III phosphatidylinositol (PtdIns)2 3-kinase complex I,(iii) proteins involved in the ubiquitin-like conjugation ofAtg12 with Atg5, (iv) proteins involved in the ubiquitin-likeconjugation of Atg8 with phosphatidylethanolamine, (v)multimembrane-spanning protein Atg9, and (vi) the Atg2-Atg18 complex. These Atg proteins localize, at least in part,to the pre-autophagosomal structure (PAS), which is proxi-mal to the vacuole and plays a central role in autophagosomeformation (6). The characterization of each of these proteinsis ongoing, and the interrelationships among these func-tional groups have also been studied systematically. How-ever, except for the proteins involved in ubiquitin-like con-jugation (7), structural studies on the Atg proteins involvedin autophagosome formation have been limited.Atg18 plays a critical role in autophagy (8, 9) by forming a

protein complex with Atg2 (10). Besides Atg2, Atg18 also bindsto PtdIns(3)P and PtdIns(3,5)P2 and is therefore considered tobe one of the effectors of these molecules (11, 12). Atg18 andAtg2 localize to the PAS in an interdependent manner, forwhich the ability of Atg18 to bind PtdIns(3)P is required (10,13). Furthermore, the production of PtdIns(3)P at the PAS byPtdIns 3-kinase complex I is required for autophagosome for-

* This work was supported in part by Japan Society for the Promotion ofScience (JSPS) Grants-in-aid for Scientific Research (KAKENHI) 23687012(to N. N. N.), 10J01988 (to Y. W.), and 23000015 (to Y. O.); Ministry of Educa-tion, Culture, Sports, Science, and Technology (MEXT) Grant-in-aid for Sci-entific Research 24113725 (to N. N. N.) and Targeted Proteins Research Pro-gram (to F. I. and Y. O.); and a Leave a Nest Grant BioGARAGE award(to T. K.).

□S This article contains supplemental Figs. S1 and S2.The atomic coordinates and structure factors (code 3VU4) have been deposited in

the Protein Data Bank, Research Collaboratory for Structural Bioinformatics,Rutgers University, New Brunswick, NJ (http://www.rcsb.org/).

1 To whom correspondence should be addressed: Inst. of Microbial Chemis-try, Kamiosaki 3-14-23, Shinagawa-ku, Tokyo 141-0021, Japan. Tel.: 81-3-3441-4173; Fax: 81-3-3441-7589; E-mail: nn@bikaken.or.jp.

2 The abbreviations used are: PtdIns, phosphatidylinositol; PAS, pre-autophago-somal structure; KmHsv2, K. marxianus Hsv2; EGFP, enhanced GFP; mRFP,monomeric red fluorescent protein; TAP, tandem affinity purification.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 287, NO. 38, pp. 31681–31690, September 14, 2012© 2012 by The American Society for Biochemistry and Molecular Biology, Inc. Published in the U.S.A.

SEPTEMBER 14, 2012 • VOLUME 287 • NUMBER 38 JOURNAL OF BIOLOGICAL CHEMISTRY 31681

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

mation (14, 15). These observations suggest that the interactionofAtg18withAtg2 andPtdIns(3)P at the PAS is essential for theformation of autophagosomes. In addition to autophagy, Atg18also has a role in regulating the vacuolar morphology of yeast,forwhichAtg18 localizes to the vacuolarmembrane through itsinteractionwith PtdIns(3,5)P2, but not with PtdIns(3)P (11, 16).Previous studies predicted the structure of Atg18 and itshomologs as a seven-bladed �-propeller fold and identified aputative phosphoinositide-binding motif (FRRG) within thepredicted �-propeller structure (11, 17, 18). However, little isknown about themolecularmechanisms underlying howAtg18recognizes Atg2 and PtdIns(3)P via its �-propeller structureand how theAtg2-Atg18 complex functions in the formation ofautophagosomes.In yeast, two Atg18 paralogs have been identified: Atg21 and

Hsv2 (11, 12, 19). AlthoughAtg21 andHsv2 are not essential forautophagy, they contain an FRRGmotif and bind to PtdIns(3)Pand PtdIns(3,5)P2 (12, 17, 20). Moreover, they were also pre-dicted to have a seven-bladed �-propeller similar to Atg18. Toreveal the architecture of Atg18, we herein report the crystalstructure of Hsv2 from a thermotolerant yeast, Kluyveromycesmarxianus (KmHsv2), at a resolution of 2.6 Å. The structurereveals a seven-bladed �-propeller fold. Mutational analyses ofAtg18 based on the structure of KmHsv2 showed that Atg18possesses two binding sites for phosphoinositides at blades 5and 6, whereas the loop regions in blade 2 are specificallyrequired for recognizing Atg2 and thus for autophagy. Theseresults suggest that Atg18 tethers Atg2 to the PAS andautophagic membranes through its simultaneous interactionwith Atg2 and PtdIns(3)P, thus playing a critical role in theformation of autophagosomes.

EXPERIMENTAL PROCEDURES

Protein Expression and Purification—KmHsv2was amplifiedby PCR and cloned into the pGEX-6P-1 vector (GEHealthcare)to produce GST fusion proteins. The construct was sequencedto confirm its identity and expressed in Escherichia coliBL21(DE3) cells that were cultured in 2� YTmedium contain-ing 10 g/liter yeast extract, 16 g/liter Tryptone, and 5 g/litersodium chloride. After cell lysis by sonication, GST-fused pro-teins were purified by affinity chromatography using a glutathi-one-Sepharose 4B column (GE Healthcare). The GST tag wasthen cleaved with PreScission protease (GE Healthcare) andremoved by affinity chromatography using a glutathione-Sep-harose 4B column. This process left a Gly-Pro-Leu-Gly-Sersequence at the N terminus of Hsv2. Further purification wasperformed using a Superdex 200 gel filtration column (GEHealthcare) and elution with 20 mM Tris-HCl (pH 8.0) and 150mM NaCl.X-ray Crystallography—Crystallization of Hsv2 was per-

formed using the sitting drop vapor diffusion method at 20 °C.Drops of 10 mg/ml Hsv2 in 20 mM Tris-HCl (pH 8.0), 150 mM

NaCl, and 2 mM dithiothreitol were mixed with equal amountsof reservoir solution (1.2 M (NH4)2SO4 and 100 mM acetatebuffer at pH 5.5) and equilibrated against 100 �l of the samereservoir solution by vapor diffusion. Crystals, typically withdimensions of 0.30 � 0.25 � 0.25 mm, were obtained withinone week. Diffraction data of the native and selenomethionine-labeled crystals were collected on anADSCQuantum210 char-ge-coupled device detector using beamlineAR-NW12A at KEK(Ibaraki, Japan). The diffraction data were indexed, integrated,and scaled using the HKL2000 program suite (21). The initialphasing was performed by themultiwavelength anomalous dis-

TABLE 1Data collection, phasing, and refinement statisticsValues in parentheses refer to the highest resolution shell. SeMet, selenomethionine; r.m.s.d., root mean square deviation.

Hsv2SeMet-labeled Native

Data collection statisticsX-ray source NW12A NW12ASpace group P6522 P6522Cell parameters a � b � 140.27, c � 251.52 Å;

� � � � 90°, � � 120°a � b � 140.27, c � 251.27 Å;

� � � � 90°, � � 120°Data set Peak Edge RemoteWavelength (Å) 0.97918 0.97932 0.96409 1.00000Resolution range (Å) 50.0–3.10 50.0–3.10 50.0–3.10 50.0–2.60Outer shell (Å) 3.15–3.10 3.15–3.10 3.15–3.10 2.64–2.60Observed reflections 1,177,881 1,183,346 1,182,747 821,218Unique reflections 27,356 27,480 27,462 45,734Completeness (%) 100.0 (100.0) 100.0 (100.0) 100.0 (100.0) 99.9 (99.6)Rsym 0.106 (0.618) 0.110 (0.684) 0.114 (0.748) 0.062 (0.542)

Phasing statisticsResolution range (Å) 50–3.1No. of Se sites 28Figure of merit (initial) 0.32Figure of merit (after SHELXE) 0.70

Refinement statisticsResolution range (Å) 50–2.6No. of protein atoms 5031No. of sulfate ions 13No. of water molecules 138R/Rfree 0.224/0.252r.m.s.d. from idealityBond length (Å) 0.007Angle 1.4°

Crystal Structure of KmHsv2

31682 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 38 • SEPTEMBER 14, 2012

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

persionmethod using the peak, edge, and remote data from theselenomethionine-labeled crystals. After the 28 selenium siteswere identified using the SHELXD program, the initial phaseand density modification were calculated using the SHELXEprogram (22). Automatedmodel building was performed usingthe Buccaneer program (23) in the CCP4 program suite (24).Further model building was performed manually using theCoot program (25), and crystallographic refinement was per-formed using the crystallography and nuclear magnetic reso-nance system software (26). Data collection, phasing, andrefinement statistics are summarized in Table 1.

Yeast Strains andMedia—We utilized standard methods foryeast manipulation (27). The Saccharomyces cerevisiae strainsused in this study are listed in Table 2. Yeast cultures wereincubated in rich YPD medium (1% Bacto-yeast extract, 2%Bacto-peptone, and 2% D-glucose) or SDCA medium (0.17%yeast nitrogen base (without amino acids and ammonium sul-fate), 0.5% ammonium sulfate, 0.5% casamino acid, and 2%D-glucose) containing appropriate amino acids. Gene disrup-tion or epitope tagging was carried out as reported previously(28, 29). To induce autophagy, the cells were grown to mid-logphase in YPD or SDCA medium and then incubated in 0.17%

FIGURE 1. Crystal structure of Hsv2. A, ribbon diagram of the Hsv2 structure. The seven blades and �-strands are labeled. B, phosphoinositide-binding sitesof Hsv2A (left) and Hsv2B (right). The side chains of the site 1 and 2 residues and the bound sulfate are shown as a stick model. The crystallographically adjacentHsv2 molecule bound to site 2 is shown in yellow. Sites 1 and 2 are shown in dashed circles.

TABLE 2Cell strains used in this study

Strain Genotype Source/Ref.

SEY6210 MATa leu2-3,112 ura3-52 his3�200 trp1�901 lys2-801 suc2-�9 Ref. 40BJ2168 MATa prc1-407 prb1-1122 pep4-3 leu2 trp1 ura3-52 Yeast Genetic Stock CenterKOY192 SEY6210 pho8�::PGPD-pho8�60:kanMX Ref. 39TKY1001 SEY6210 atg18�::kanMX Ref. 30TKY1051 KOY192 atg18�::natNT2 Ref. 30TKY1498 TKY1001mRFP-APE1:HIS Ref. 30TKY1732 BJ2168 ATG2-TAP:TRP1 atg18�::natNT2 This studyScHY-2569 BJ2168 atg18�::kanMX6 This study

Crystal Structure of KmHsv2

SEPTEMBER 14, 2012 • VOLUME 287 • NUMBER 38 JOURNAL OF BIOLOGICAL CHEMISTRY 31683

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

yeast nitrogen base (without amino acids and ammonium sul-fate) and 2% D-glucose for 4 h or treated with rapamycin (finalconcentration, 0.2 �g/ml; Sigma) for 1–3 h.Plasmid Construction for Yeast Experiments—The pRS316-

based plasmid for Atg18 with a 3�HA-enhanced GFP(EGFP) tag (hereafter referred to as Atg18-HG) was gener-ated as reported previously (30). Point mutations were intro-duced by PCR-based site-directed mutagenesis using thepRS316-based plasmid for Atg18-HG as a template. The suc-cessful introduction of the point mutations was confirmedby sequencing.Microscopy Observations—The intracellular localization of

monomeric red fluorescent protein (mRFP)- or EGFP-taggedproteins was visualized using an inverted fluorescence micro-scope (IX-71, Olympus) equipped with an EM-CCD digitalcamera (ImagEM, Hamamatsu Photonics K.K.). Images were

acquired using AquaCosmos 2.6 software (Hamamatsu Photo-nics K.K.) and processed using PhotoshopCS4 software (AdobeSystems). To observe the PAS, yeast cells were treated withrapamycin (final concentration, 0.2 �g/ml) for 1 h to induceautophagy.FM4-64 Staining—Cells at the logarithmic phase were

loaded with 2 �g/ml FM4-64 (Invitrogen) for 30 min, washed,and chased with FM4-64-free medium for 30 min.Pho8�60 Alkaline Phosphatase Assay—To quantify bulk

autophagic activity, we utilized the Pho8�60 alkaline phospha-tase assay as described previously (31).Preparation of Total Lysates and Immunoblotting—Yeast

protein extracts were prepared as reported previously (30).Immunoblotting was performed using anti-Ape1 (API-2),anti-HA (3F10), or anti-Pgk1 antibodies (Invitrogen). Chemilu-minescence detection was performed using Pierce Western

FIGURE 2. A, structurally annotated sequence alignment of Atg18 homologs. Gaps have been introduced to maximize the similarity. The conserved residues areshaded in black. The secondary structural elements of Hsv2 are shown above the sequence. Sc, S. cerevisiae; Pp, P. pastoris; Hs, H. sapiens. B, mapping of theresidues conserved among the Atg18 orthologs on the KmHsv2 structure. The conserved residues are shaded in black. The residues in parentheses are theAtg18 residues that are the structural equivalent of KmHsv2 residues. Blade 2 AB and BC loops as well as sites 1 and 2 are shown in dashed circles.

Crystal Structure of KmHsv2

31684 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 38 • SEPTEMBER 14, 2012

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

blotting substrate (Thermo Scientific) and detected using anLAS-4000 mini image analyzer (GE Healthcare).Co-immunoprecipitation—Cells were treated with 200

�g/ml Zymolyase 100T (07665-55, Nacalai Tesque) for 45 minat 30 °C in spheroplasting buffer (50mMHEPES-KOH (pH 7.2),1 M sorbitol, 1% yeast extract, 2% Bacto-peptone, 1% glucose,and 10 mM DTT). The spheroplasts were washed once withspheroplasting buffer, grown for 20 min at 30 °C, and thentreated with 0.5 �g/ml rapamycin for 1 h at 30 °C. The sphero-plasts were harvested and treatedwith 0.5%TritonX-100 for 30min on ice in lysis buffer (20mMTris-HCl (pH 8.0), 50mMKCl,5 mM MgCl2, and protease inhibitor mixture (P8340, Sigma)).The total lysate was centrifuged at 17,400� g for 20min at 4 °C,and the resulting supernatantwas incubatedwithGFP-Trap_M(gtm-20, ChromoTek) or rabbit IgG-conjugated magneticbeads (DynabeadsM-270 epoxy, Invitrogen) for 3 h at 4 °C. Thebound materials were washed three times with lysis bufferand then eluted with SDS-PAGE sample buffer for 15 min at65 °C.

RESULTS

Overall Structure of KmHsv2—We first tried to crystallizeS. cerevisiaeAtg18 but failed to obtain good diffracting crystals,so we adopted a strategy to obtain crystals from various Atg18homologs/paralogs, including S. cerevisiaeAtg21, Pichia pasto-ris Atg18, Arabidopsis thaliana Atg18b, and Homo sapiensAtg18 homologs (WIPI-1–4). However, these trials also failed.Recently, we succeeded in determining the solution structure ofAtg10 by switching the target from S. cerevisiae Atg10 toK. marxianus Atg10 (32). K. marxianus is a thermotolerantyeast, so the homologs in this yeast could be expected to havehigher stability than those in other eukaryotes. We thusattempted to crystallize K. marxianus Atg18 paralogs and suc-ceeded in obtaining good diffracting crystals of KmHsv2(referred to simply as Hsv2 hereafter), and we determined thecrystal structure of Hsv2 by the multiwavelength anomalous dis-persion method using a selenomethionine-substituted crystal(supplemental Fig. S1). The structure was refined against 2.6 Ådata to an R-factor of 0.224 and a free R-factor of 0.252 (Table 1).The asymmetric unit of the crystal contains two Hsv2 mole-

cules (Hsv2A and Hsv2B). The Hsv2 model obtained lacks 13N-terminal residues and some loop regions (residues 164–181in Hsv2A and residues 163–181 and 268–284 in Hsv2B) due toundefined electron density. Hsv2 possesses a seven-bladed�-propeller fold, in which each blade consists of a four-�-stranded antiparallel �-sheet, resembling each other (Fig. 1A).Circular permutation, which is frequently observed in �-pro-peller proteins, is not observed in this fold. Residues 263–289 ofHsv2A form a large extended loop connecting the C and D�-strands (CD loop) in blade 6, which protrudes from the�-propeller fold as far as �25 Å. The ordered conformation ofthis loop is stabilized by crystal packing. In contrast, the elec-tron density of the equivalent residues of Hsv2B is disordered,suggesting that these residues have a flexible conformation insolution.Two Binding Sites for Phosphoinositides—The phospho-

inositide-binding FRRG motif (residues 229–232) of Hsv2 islocated at the D �-strand in blade 5 and the loop connecting

blades 5 and 6. Intriguingly, the side chains of the two arginineresidues of the motif, Arg-230 and Arg-231, point in oppositedirections (Fig. 1B). In the proximity of Arg-230, there is a basicpocket composed of Ser-209, Thr-213, and Arg-216 from blade5 andHis-189, Thr-190, andAsn-191 from the loop connectingblades 4 and 5. A sulfate ion is bound to this basic pocket, sug-gesting that it has a role in accommodating phosphoinositides;

FIGURE 3. Localization of mutant Atg18-HG constructs and vacuolar mor-phology. Exponentially growing atg18� cells (TKY1001) expressing the indi-cated mutant Atg18-HG constructs were labeled with FM4-64 and subjectedto fluorescence microscopy. Scale bar � 2 �m.

Crystal Structure of KmHsv2

SEPTEMBER 14, 2012 • VOLUME 287 • NUMBER 38 JOURNAL OF BIOLOGICAL CHEMISTRY 31685

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

thus, this pocket was named site 1. The structure of site 1 is verysimilar between Hsv2A and Hsv2B (Fig. 1B). Arg-231 isinvolved in the construction of another basic pocket composedof Ser-254, Lys-256, Thr-258, and His-260 from blade 6. Thestructure of this second pocket is somewhat distinct betweenHsv2A and Hsv2B. In Hsv2A, the loop connecting blades 5 and6 has a conformation closer to blade 6, and the side chain ofAsp-234 on the loop is bound deeply in the pocket, whereas inHsv2B, the pocket is occupied by the side chain of Glu-336 ofthe crystallographically adjacent Hsv2 molecule. In both cases,the basic pocket is occupied by a negatively charged carboxylgroup, suggesting that it has a similar role to site 1 in accom-modating phosphoinositides; thus, the second pocket wasnamed site 2. The residues constituting sites 1 and 2 are highlyconserved among theAtg18 homologs/paralogs, suggesting thepossibility that Atg18 and its relatives possess two bindingpockets for phosphoinositides. During the preparation of thismanuscript, Baskaran et al. (33) and Krick et al. (34) reportedthe crystal structure of Kluyveromyces lactis Hsv2. The overallstructure of K. lactis Hsv2 is similar to that of KmHsv2, and it

possesses two basic pockets similar to sites 1 and 2 in KmHsv2.They showed that both pockets are important for recognizingphosphoinositides according to in vitro mutational analyses.We also observed that a single mutation at either Arg-230 (site1) or Arg-231 (site 2) resulted in a partial defect, and simulta-neous mutation at both residues resulted in a more severedefect in autophagy (supplemental Fig. S2), suggesting thatboth sites are important for recognizing PtdIns(3)P. Besidestwobinding pockets, Baskaran et al. (33) also showed that a longloop in blade 6 of K. lactisHsv2, which is equivalent to the longloop (residues 263–289) of KmHsv2, is important for its asso-ciation with membranes.Effect of Mutating Conserved Residues of Atg18 on Vacuolar

Morphology—Structurally annotated multiple sequence align-ment of KmHsv2 with Atg18 orthologs (S. cerevisiae, K. marx-ianus, P. pastoris, and H. sapiensWIPI-1) showed that the res-idues in blades 2, 3, 5, and 6 are highly conserved among theAtg18 orthologs (Fig. 2A). Fig. 2B shows the location of theconserved residues in the Atg18 orthologs on the structure ofKmHsv2. Conserved exposed residues are especially clustered

FIGURE 4. Mutational effect of the conserved residues of Atg18 on autophagy. A, autophagic activity was estimated using an alkaline phosphatase (ALP)assay (see “Experimental Procedures”). The white and black bars indicate alkaline phosphatase activity at 0 and 4 h after starvation, respectively. Values are themeans � S.D. of three independent experiments. B, total lysates from atg18� cells carrying the indicated plasmids were subjected to Western blotting usinganti-Ape1, anti-HA, or anti-Pgk1 (loading control) antiserum. To induce autophagy, the cells were treated with rapamycin for 3 h.

Crystal Structure of KmHsv2

31686 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 38 • SEPTEMBER 14, 2012

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

at blades 2 and 3. To evaluate the functional significance ofthese conserved residues in the regulation of vacuolarmorphol-ogy, we introduced point mutations at these sites, especiallythose conserved among Atg18 orthologs but not in Hsv2, andprepared the following six mutants: F54A/S55A and S57A/L58A (both at the AB loop in blade 2), I49K/L96K (at strand Ain blades 2 and 3), P72A/R73A (at the BC loop in blade 2),M121A/R122A/L123A (at the CD loop in blade 3), and T126R/N132R (at strand D in blade 3). As mentioned above, many ofthe residues in sites 1 and 2 are conserved among Atg18homologs/paralogs and are responsible for the recognition ofphosphoinositides. We selected His-244 from site 1 and His-315 from site 2 and prepared an Atg18 mutant with an alaninesubstitution at both histidine residues (H244A/H315A). Theseseven mutants and wild-type Atg18 were expressed as fusionproteins with a 3�HA-EGFP tag (Atg18-HG) in atg18� cellsusing the pRS316 centromeric plasmid and visualized by fluo-rescence microscopy (Fig. 3,middle panels). At the same time,the vacuoleswere visualized using FM4-64 staining (Fig. 3, rightpanels). Although wild-type Atg18-HG localized to the vacuo-lar membrane and properly regulated the vacuolar morphol-ogy, Atg18(H244A/H315A)-HG was not recruited to the vacu-olar membrane, and cells expressing Atg18(H244A/H315A)-HG

showed abnormally enlarged vacuoles. We confirmed that theH244A/H315A mutations abrogated the binding affinity ofAtg18 for PtdIns(3,5)P2 by an in vitro pulldown assay using PIPbeads (EchelonBiosciences) (data not shown). These results areconsistent with previous reports showing that the bindingactivity of Atg18 for PtdIns(3,5)P2 is required for the localiza-tion of Atg18 to the vacuolar membrane and for the mainte-nance of vacuolarmorphology (11, 16). Conversely, all of the sixmutants at blades 2 and 3 showed normal localization to thevacuolar membrane, and cells expressing these mutantsshowed normal vacuolar morphology. These data indicate thatthe mutations at blades 2 and 3 do not affect the affinity ofAtg18 for phosphoinositides or abrogate its role in the regula-tion of vacuolar morphology.Effect ofMutating Conserved Residues of Atg18 onAutophagy—

Next, we studied the significance of the conserved residues ofAtg18 in autophagy using the same panel of mutants.Autophagic activity was estimated using the Pho8�60 assay(31). This method utilizes a genetically engineered cytosolicform of Pho8 alkaline phosphatase, Pho8�60, which is deliv-ered into the vacuole exclusively by autophagy and activated.Thus, autophagic activity correlateswell with alkaline phospha-tase activity. As shown in Fig. 4A, atg18� cells expressing

FIGURE 5. The BC loop in blade 2 is essential for the PAS targeting of Atg18. atg18� cells carrying integrated mRFP-Ape1 and mutant Atg18-HG constructswere observed by microscopy after rapamycin treatment for 1 h. The arrows indicate the PAS. Scale bar � 2 �m.

Crystal Structure of KmHsv2

SEPTEMBER 14, 2012 • VOLUME 287 • NUMBER 38 JOURNAL OF BIOLOGICAL CHEMISTRY 31687

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Atg18(P72A/R73A)-HG (Atg18(P72A/R73A)-HG cells) showedalmost no autophagic activity. Atg18(F54A/S55A)-HG, Atg18(S57A/L58A)-HG, and Atg18(M121A/R122A/L123A)-HG cellsshowed mildly but significantly reduced autophagic activity,amongwhich that of Atg18(F54A/S55A)-HG cells was the lowest.Conversely, Atg18(I46K/L96K)-HG and Atg18(T126R/N132R)-HG cells showed autophagic activity comparable with that ofwild-type Atg18-HG cells. Atg18(H244A/H315A)-HG cellsshowed approximately half of the autophagic activity of wild-typeAtg18-HG cells, suggesting that this mutant retained weakPtdIns(3)P-binding ability that is sufficient for the partial pro-gression of autophagy.Next, we monitored aminopeptidase I (Ape1) maturation.

The premature form of Ape1 (prApe1) is transported to thevacuole via the cytoplasm-to-vacuole targeting pathway undernutrient-rich conditions and by autophagy in response to star-vation or rapamycin treatment (35). In the vacuole, prApe1 isprocessed into a mature form (mApe1), which can be moni-tored by Western blotting for Ape1. As shown in Fig. 4B, cellsexpressing wild-type andmutant Atg18-HG constructs, exceptfor Atg18(P72A/R73A)-HG, showed a strong mApe1 band anda weak prApe1 band in response to rapamycin treatment.Because monitoring Ape1maturation is a muchmore sensitivetechnique to detect autophagic activity than the Pho8�60 assay,the normal Ape1maturation observed in theAtg18(F54A/S55A)-HG, Atg18(S57A/L58A)-HG, and Atg18(M121A/R122A/L123A)-HG cells may be due to the high sensitivity of this method.Nevertheless, Atg18(P72A/R73A)-HG cells showed a signifi-cantly weaker mApe1 band and a stronger prApe1 band, indi-cating that these cells retained minimal autophagy activity.These data, together with the results shown in Fig. 3, suggestthat the AB loop (containing Phe-54 and Ser-55) and BC loop(containing Pro-72 and Arg-73) in blade 2 in Atg18 are impor-tant for autophagy, but they are dispensable for the regulationof vacuolar morphology and phosphoinositide recognition.TheBCLoop inBlade2 Is Essential for PASTargeting ofAtg18—

The AB and BC loops in blade 2 are not responsible for recog-nizing phosphoinositides; nevertheless, they were shown to beimportant for autophagy. To function in autophagy, Atg18 hasto localize to the PAS; therefore, we speculated that the loops inblade 2 are important for PAS localization of Atg18. To confirmthis speculation, we monitored the PAS localization ofAtg18-HG constructs upon rapamycin treatment. As a PASmarker, mRFP-Ape1 was coexpressed with Atg18-HG con-structs. As shown in Fig. 5, wild-type Atg18-HG co-localizedwith mRFP-Ape1 at a dot proximal to the vacuole, suggestingits PAS localization. Atg18(F54A/S55A)-HG and Atg18(H244A/H315A)-HG also localized to the PAS, which mayreflect their remaining autophagic activity. In contrast,Atg18(P72A/R73A)-HGdid not localize to the PAS. These datasuggest that the BC loop in blade 2 is essential for the PAStargeting of Atg18.Atg18 Recognizes Atg2 Using Blade 2—Atg18 and Atg2 local-

ize to the PAS interdependently. This observation suggests thatthe formation of the complex betweenAtg18 andAtg2 is essen-tial for their PAS targeting. Because the BC loop in blade 2 isessential for PAS localization of Atg18, it is possible that theloop is involved in the interaction with Atg2. To study this

possibility, we examined the Atg18-Atg2 interaction using co-immunoprecipitation. Wild-type Atg18-HG and mutantAtg18-HG (F54A/S55A, P72A/R73A, and H244A/H315A)were coexpressed with Atg2 fused to a tandem affinity purifica-tion (TAP) tag in atg18� cells, and Atg18-HG constructs werepulled down using GFP-Trap magnetic beads. As shown inFig. 6A, wild-type Atg18-HG, Atg18(F54A/S55A)-HG, andAtg18(H244A/H315A)-HG, but not Atg18(P72A/R73A)-HG,interacted with Atg2. Compared with wild-type Atg18-HG, theinteraction ofAtg18(F54A/S55A)-HGwithAtg2wasmildly butsignificantly attenuated, whereas that of Atg18(H244A/H315A)-HG was slightly enhanced. Similar results wereobtained when Atg2 fused to a TAP tag was pulled down usingIgG-conjugated magnetic beads (Fig. 6B). These results showthat the AB and BC loops in blade 2 are important for the inter-action of Atg18 with Atg2.

DISCUSSION

In this study, we determined the crystal structure of KmHsv2and revealed the seven-bladed �-propeller architecture con-served among the Atg18 family of proteins. Furthermore, using

FIGURE 6. Analysis of the interaction between Atg18 and Atg2. Co-immu-noprecipitation (IP) experiments were performed as described under “Exper-imental Procedures.” A, Atg18-HG constructs were pulled down using GFP-Trap magnetic beads. The protein bands for Atg2 and Pgk1 were detectedusing rabbit IgG and anti-Pgk1 antibody, respectively. The protein bands forAtg18 were detected using anti-GFP or anti-HA antibody. B, Atg2 fused to aTAP tag was pulled down using IgG-conjugated magnetic beads. The proteinbands for Atg2, Atg18, and Pgk1 were detected using anti-Atg2, anti-HA, andanti-Pgk1 antibody, respectively.

Crystal Structure of KmHsv2

31688 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 38 • SEPTEMBER 14, 2012

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

in vivo mutational analyses of Atg18, we showed that the loopregions in blade 2 play a critical role in autophagy through theirinteractionwithAtg2, whereas the two basic pockets in blades 5and 6 are responsible for phosphoinositide binding and play anessential role in the regulation of vacuolar morphology. Immu-noprecipitation experiments showed that the AB and BC loopsin blade 2, which are located on opposite surfaces of the ring-like structure of Atg18, are important for its interaction withAtg2. These observations suggest that Atg2 recognizes bothsurfaces of the ring simultaneously, whichmight be achieved bygripping the ring from the side of blade 2 (Fig. 7). It appears tobe easy forAtg2 to perform such interactions because of its verylarge size (�180 kDa). Very recently, Baskaran et al. (33) andKrick et al. (34) showed that Hsv2 could interact edge-on to themembrane such that the two binding pockets for phospho-inositides in blades 5 and 6 might contact the membrane sur-face, whereas the CD loop in blade 6might penetrate themem-brane (Fig. 7). According to this model, the membrane will notinterfere with the interaction between Atg18 and Atg2 becauseblade 2 of Atg18 will be located distally from the membrane,thus enabling the simultaneous interaction of Atg18 with bothAtg2 and themembrane. Thismodel of Atg18-Atg2 interactionappears to be conserved in their human counterparts becausethe residues responsible for the interaction with Atg2 are con-served between S. cerevisiae Atg18 and H. sapiens WIPI-1, anAtg18 ortholog. This hypothesis is supported by the report thatH. sapiens Atg2A directly interacts with S. cerevisiae Atg18(36).In addition to the autophagic function of Atg18, it also has a

role in the maintenance of vacuolar morphology. The interac-tion between Atg18 and Vac14, one of regulators ofPtdIns(3,5)P2, appears to be required for the proper regulationof PtdIns(3,5)P2 and vacuolar morphology (37). In this study,Atg18(P72A/R73A)-HG cells had normal sized vacuoles (Fig.3), indicating that Atg18(P72A/R73A)-HG could be bound to

Vac14. This observation suggests that theVac14-binding site ofAtg18 is distinct from its Atg2-binding site. Because the func-tion of Atg18 in the regulation of vacuolar morphology is notconserved among the homologs in higher eukaryotes, such asmammals, the Vac14-binding site may be located in the non-conserved region of Atg18.We recently showed thatAtg18 ismost likely to be important

for the localization of Atg2 to the PAS (30). Conversely, theinteraction of Atg18 with Atg2 was demonstrated to be essen-tial for the PAS targeting of Atg18 (Fig. 5). This kind of inter-dependence between two Atg proteins for PAS targeting is alsoobserved for Atg6-Atg14 and Atg5-Atg16 complexes (38, 39).However, the mechanisms underlying this interdependent tar-geting of Atg proteins to the PAS have yet to be elucidated.After targeting to the PAS, the Atg2-Atg18 complex plays anabsolutely critical role in autophagosome formation togetherwith other Atg groups. However, the specific roles of this com-plex also have yet to be established. A structural study of theAtg2-Atg18 complex is required to uncover these critical issuesand will create a path to understanding the molecular mecha-nisms of autophagosome formation.

Acknowledgments—We are grateful to Hiromi Kirisako and ChikaKondo-Kakuta for technical support. The synchrotron radiationexperiments were performed at beamline NW12A at KEK.

REFERENCES1. Mizushima, N., Yoshimori, T., and Ohsumi, Y. (2011) The role of Atg

proteins in autophagosome formation. Annu. Rev. Cell Dev. Biol. 27,107–132

2. Mizushima, N. (2007) Autophagy: process and function. Genes Dev. 21,2861–2873

3. Levine, B., Mizushima, N., and Virgin, H. W. (2011) Autophagy in immu-nity and inflammation. Nature 469, 323–335

4. Mizushima, N., and Komatsu, M. (2011) Autophagy: renovation of cellsand tissues. Cell 147, 728–741

5. Nakatogawa, H., Suzuki, K., Kamada, Y., andOhsumi, Y. (2009) Dynamicsand diversity in autophagymechanisms: lessons from yeast.Nat. Rev.Mol.Cell Biol. 10, 458–467

6. Suzuki, K., andOhsumi, Y. (2010) Current knowledge of the pre-autopha-gosomal structure (PAS). FEBS Lett. 584, 1280–1286

7. Noda, N. N., Ohsumi, Y., and Inagaki, F. (2009) ATG systems from theprotein structural point of view. Chem. Rev. 109, 1587–1598

8. Barth, H., Meiling-Wesse, K., Epple, U. D., and Thumm, M. (2001) Au-tophagy and the cytoplasm-to-vacuole targeting pathway both requireAut10p. FEBS Lett. 508, 23–28

9. Guan, J., Stromhaug, P. E., George, M. D., Habibzadegah-Tari, P., Bevan,A., Dunn,W. A., Jr., and Klionsky, D. J. (2001) Cvt18/Gsa12 is required forcytoplasm-to-vacuole transport, pexophagy, and autophagy in Saccharo-myces cerevisiae and Pichia pastoris.Mol. Biol. Cell 12, 3821–3838

10. Suzuki, K., Kubota, Y., Sekito, T., and Ohsumi, Y. (2007) Hierarchy of Atgproteins in pre-autophagosomal structure organization. Genes Cells 12,209–218

11. Dove, S. K., Piper, R. C., McEwen, R. K., Yu, J. W., King, M. C., Hughes,D. C., Thuring, J., Holmes, A. B., Cooke, F. T., Michell, R. H., Parker, P. J.,and Lemmon,M. A. (2004) Svp1p defines a family of phosphatidylinositol3,5-bisphosphate effectors. EMBO J. 23, 1922–1933

12. Strømhaug, P. E., Reggiori, F., Guan, J., Wang, C. W., and Klionsky, D. J.(2004) Atg21 is a phosphoinositide-binding protein required for efficientlipidation and localization of Atg8 during uptake of aminopeptidase I byselective autophagy.Mol. Biol. Cell 15, 3553–3566

13. Obara, K., Sekito, T., Niimi, K., and Ohsumi, Y. (2008) The Atg18-Atg2

FIGURE 7. Schematic representation of the Atg2-Atg18 complex on themembrane. Atg2 (yellow) grips the ring-like structure of Atg18 (green) atblade 2, whereas Atg18 interacts with membrane (dark red) at blades 5 and 6.

Crystal Structure of KmHsv2

SEPTEMBER 14, 2012 • VOLUME 287 • NUMBER 38 JOURNAL OF BIOLOGICAL CHEMISTRY 31689

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

complex is recruited to autophagic membranes via phosphatidylinositol3-phosphate and exerts an essential function. J. Biol. Chem. 283,23972–23980

14. Obara, K., Sekito, T., and Ohsumi, Y. (2006) Assortment of phosphatidyl-inositol 3-kinase complexes–Atg14p directs association of complex I tothe pre-autophagosomal structure in Saccharomyces cerevisiae.Mol. Biol.Cell 17, 1527–1539

15. Obara, K., and Ohsumi, Y. (2008) Dynamics and function of PtdIns(3)P inautophagy. Autophagy 4, 952–954

16. Efe, J. A., Botelho, R. J., and Emr, S. D. (2007) Atg18 regulates organellemorphology and Fab1 kinase activity independent of its membrane re-cruitment by phosphatidylinositol 3,5-bisphosphate. Mol. Biol. Cell 18,4232–4244

17. Krick, R., Tolstrup, J., Appelles, A., Henke, S., and Thumm,M. (2006) Therelevance of the phosphatidylinositol phosphate-binding motif FRRGT ofAtg18 and Atg21 for the Cvt pathway and autophagy. FEBS Lett. 580,4632–4638

18. Proikas-Cezanne, T., Waddell, S., Gaugel, A., Frickey, T., Lupas, A., andNordheim, A. (2004) WIPI-1� (WIPI49), a member of the novel seven-bladedWIPI protein family, is aberrantly expressed in human cancer andis linked to starvation-induced autophagy. Oncogene 23, 9314–9325

19. Meiling-Wesse, K., Barth, H., Voss, C., Eskelinen, E. L., Epple, U. D., andThumm, M. (2004) Atg21 is required for effective recruitment of Atg8 tothe pre-autophagosomal structure during the Cvt pathway. J. Biol. Chem.279, 37741–37750

20. Krick, R., Henke, S., Tolstrup, J., and Thumm, M. (2008) Dissecting thelocalization and function of Atg18, Atg21, and Ygr223c. Autophagy 4,896–910

21. Otwinowski, Z., andMinor,W. (1997) Processing of x-ray diffraction datacollected in oscillation mode.Methods Enzymol. 276, 307–326

22. Sheldrick, G. M. (2008) A short history of SHELX. Acta Crystallogr. A 64,112–122

23. Cowtan, K. (2006) The Buccaneer software for automatedmodel building.1. Tracing protein chains. Acta Crystallogr. D Biol. Crystallogr. 62,1002–1011

24. Winn,M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans,P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G., McCoy, A., McNicholas,S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read,R. J., Vagin, A., and Wilson, K. S. (2011) Overview of the CCP4 suite andcurrent developments. Acta Crystallogr. D Biol. Crystallogr. 67, 235–242

25. Emsley, P., Lohkamp, B., Scott,W. G., and Cowtan, K. (2010) Features anddevelopment of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501

26. Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gros, P.,Grosse-Kunstleve, R.W., Jiang, J. S., Kuszewski, J., Nilges,M., Pannu,N. S.,Read, R. J., Rice, L. M., Simonson, T., andWarren, G. L. (1998) Crystallog-raphy & NMR system: a new software suite for macromolecular structuredetermination. Acta Crystallogr. D Biol. Crystallogr. 54, 905–921

27. Amberg, D. C., Burke, D., and Strathern, J. N. (2005) Methods in Yeast

Genetics: A Cold Spring Harbor Laboratory Course Manual, Cold SpringHarbor Laboratory Press, Cold Spring Harbor, NY

28. Knop, M., Siegers, K., Pereira, G., Zachariae, W., Winsor, B., Nasmyth, K.,and Schiebel, E. (1999) Epitope tagging of yeast genes using a PCR-basedstrategy: more tags and improved practical routines. Yeast 15, 963–972

29. Longtine, M. S., McKenzie, A., 3rd, Demarini, D. J., Shah, N. G.,Wach, A.,Brachat, A., Philippsen, P., and Pringle, J. R. (1998) Additional modules forversatile and economical PCR-based gene deletion and modification inSaccharomyces cerevisiae. Yeast 14, 953–961

30. Kobayashi, T., Suzuki, K., and Ohsumi, Y. (2012) Autophagosome forma-tion can be achieved in the absence of Atg18 by expressing engineeredPAS-targeted Atg2. FEBS Lett. 586, 2473–2478

31. Noda, T.,Matsuura, A.,Wada, Y., andOhsumi, Y. (1995)Novel system formonitoring autophagy in the yeast Saccharomyces cerevisiae. Biochem.Biophys. Res. Commun. 210, 126–132

32. Yamaguchi, M., Noda, N. N., Yamamoto, H., Shima, T., Kumeta, H., Ko-bashigawa, Y., Akada, R., Ohsumi, Y., and Inagaki, F. (2012) Structuralinsights into Atg10-mediated formation of the autophagy-essentialAtg12-Atg5 conjugate. Structure 20, 1244–1254

33. Baskaran, S., Ragusa, M. J., Boura, E., and Hurley, J. H. (2012) Two-siterecognition of phosphatidylinositol 3-phosphate by PROPPINs in au-tophagy.Mol. Cell 47, 339–348

34. Krick, R., Busse, R. A., Scacioc, A., Stephan, M., Janshoff, A., Thumm, M.,and Kühnel, K. (2012) Structural and functional characterization of thetwo phosphoinositide-binding sites of PROPPINs, a �-propeller proteinfamily. Proc. Natl. Acad. Sci. U.S.A. 109, E2042-E2049

35. Lynch-Day, M. A., and Klionsky, D. J. (2010) The Cvt pathway as a modelfor selective autophagy. FEBS Lett. 584, 1359–1366

36. Romanyuk, D., Polak, A., Maleszewska, A., Sieńko, M., Grynberg, M., andZoładek, T. (2011) Human hAtg2A protein expressed in yeast is recruitedto pre-autophagosomal structure but does not complement autophagydefects of atg2� strain. Acta Biochim. Pol. 58, 365–374

37. Jin, N., Chow, C. Y., Liu, L., Zolov, S. N., Bronson, R., Davisson, M., Pe-tersen, J. L., Zhang, Y., Park, S., Duex, J. E., Goldowitz, D., Meisler, M. H.,and Weisman, L. S. (2008) VAC14 nucleates a protein complex essentialfor the acute interconversion of PI(3)P and PI(3,5)P2 in yeast and mouse.EMBO J. 27, 3221–3234

38. Matsushita, M., Suzuki, N. N., Obara, K., Fujioka, Y., Ohsumi, Y., andInagaki, F. (2007) Structure of Atg5-Atg16, a complex essential for au-tophagy. J. Biol. Chem. 282, 6763–6772

39. Noda, N. N., Kobayashi, T., Adachi, W., Fujioka, Y., Ohsumi, Y., and Ina-gaki, F. (2012) Structure of the novel C-terminal domain of vacuolar pro-tein sorting 30/autophagy-related protein 6 and its specific role in au-tophagy. J. Biol. Chem. 287, 16256–16266

40. Robinson, J. S., Klionsky, D. J., Banta, L. M., and Emr, S. D. (1988) Proteinsorting in Saccharomyces cerevisiae: isolation of mutants defective in thedelivery and processing of multiple vacuolar hydrolases.Mol. Cell. Biol. 8,4936–4948

Crystal Structure of KmHsv2

31690 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 287 • NUMBER 38 • SEPTEMBER 14, 2012

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Akada, Fuyuhiko Inagaki, Yoshinori Ohsumi and Nobuo N. NodaYasunori Watanabe, Takafumi Kobayashi, Hayashi Yamamoto, Hisashi Hoshida, Rinji

Phosphoinositides in Atg18Structure-based Analyses Reveal Distinct Binding Sites for Atg2 and

doi: 10.1074/jbc.M112.397570 originally published online July 31, 20122012, 287:31681-31690.J. Biol. Chem. 

  10.1074/jbc.M112.397570Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

  http://www.jbc.org/content/suppl/2012/07/31/M112.397570.DC1

  http://www.jbc.org/content/287/38/31681.full.html#ref-list-1

This article cites 39 references, 11 of which can be accessed free at

by guest on March 27, 2018

http://ww

w.jbc.org/

Dow

nloaded from