QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Vol. 16, Issue 10, 4941-4953, October 2005

This Article Abstract Full Text (PDF) All Versions of this Article: E05-02-0143v1 16/10/4941    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chang, T. Articles by Dunn, W. A. Search for Related Content PubMed PubMed Citation Articles by Chang, T. Articles by Dunn, W. A., Jr.

PpATG9 Encodes a Novel Membrane Protein That Traffics to Vacuolar Membranes, Which Sequester Peroxisomes during Pexophagy in Pichia pastoris

Tina Chang ^*, Laura A. Schroder ^*, J. Michael Thomson ^*, Amy S. Klocman ^*, Amber J. Tomasini ^*, Per E. Strømhaug , and William A. Dunn, Jr. ^*

^* Department of Anatomy and Cell Biology, University of Florida College of Medicine, Gainesville, FL 32610-0235; Division of Biological Sciences, University of Missouri, Columbia, MO 65201

Submitted February 22, 2005; Revised July 21, 2005; Accepted July 22, 2005
Monitoring Editor: Suresh Subramani

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
When Pichia pastoris adapts from methanol to glucose growth, peroxisomes are rapidly sequestered and degraded within the vacuole by micropexophagy. During micropexophagy, sequestering membranes arise from the vacuole to engulf the peroxisomes. Fusion of the sequestering membranes and incorporation of the peroxisomes into the vacuole is mediated by the micropexophagy-specific membrane apparatus (MIPA). In this study, we show the P. pastoris ortholog of Atg9, a novel membrane protein is essential for the formation of the sequestering membranes and assembly of MIPA. During methanol growth, GFP-PpAtg9 localizes to multiple structures situated near the plasma membrane referred as the peripheral compartment (Atg9-PC). On glucose-induced micropexophagy, PpAtg9 traffics from the Atg9-PC to unique perivacuolar structures (PVS) that contain PpAtg11, but lack PpAtg2 and PpAtg8. Afterward, PpAtg9 distributes to the vacuole surface and sequestering membranes. Movement of the PpAtg9 from the Atg9-PC to the PVS requires PpAtg11 and PpVps15. PpAtg2 and PpAtg7 are essential for PpAtg9 trafficking from the PVS to the vacuole and sequestering membranes, whereas trafficking of PpAtg9 proceeds independent of PpAtg1, PpAtg18, and PpVac8. In summary, our data suggest that PpAtg9 transits from the Atg9-PC to the PVS and then to the sequestering membranes that engulf the peroxisomes for degradation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Cellular activities are regulated and maintained as a well-controlled balance between protein synthesis and degradation of essential proteins and enzymes. Protein lifetimes can last from minutes to days, with a majority of the cellular proteins being degraded in a lytic compartment such as the lysosome or vacuole. Autophagy is the primary pathway for the delivery of endogenous proteins and organelles to this compartment (Levine and Klionsky, 2004; Klionsky, 2005). There exist a number of autophagic pathways in yeast: micro- and macro-autophagy, micro- and macro-pexophagy, cytoplasm-to-vacuole targeting pathway (CVT), and vacuole import and degradation pathway. The CVT pathway is constitutive, whereas the other pathways are regulated by environmental signals. For example, autophagy is enhanced when cells are deprived of nutrients such as amino acids and glucose, whereas pexophagy is activated when methylotrophic yeasts adapt from growth on methanol medium to growth on glucose or ethanol medium. Autophagy sequesters cytosolic proteins and organelles nonselectively, whereas pexophagy is responsible for the selective degradation of peroxisomes. Despite their differences, these pathways share a number of common molecular events. For example, Atg7/Gsa7/Apg7, Atg1/Gsa10/Apg1/Aut3, Atg2/Gsa11/Apg2, Atg18/Gsa12/Cvt18, and Vps15 are required for pexophagy, autophagy, and CVT pathways, whereas Atg11/Gsa9/Cvt9 and Vac8 are essential for pexophagy and CVT pathways, but not autophagy (Stasyk et al., 1999; Yuan et al., 1999; Scott et al., 2000; Guan et al., 2001; Kim et al., 2001; Strømhaug et al., 2001; Strømhaug and Klionsky, 2001; Wang et al., 2001; Huang and Klionsky, 2002; Abeliovich et al., 2003). Because many autophagy genes and their orthologues have been identified using different yeast models, the nomenclature for these genes had become confusing. Therefore, we will utilize the ATG nomenclature for those genes uniquely essential for autophagy (We will conform to the ATG nomenclature; all the autophagy-related [ATG] genes will contain a genus and species prefix, e.g., PpAtg for the P. pastoris and ScAtg for the S. cerevisiae homologues; Klionsky et al., 2003).

Pichia pastoris is able to synthesize enzymes to assimilate methanol for energy and growth. These enzymes such as alcohol oxidase (AOX) are housed primarily in the peroxisomes (Tuttle et al., 1993). When these cells adapt from methanol to ethanol or glucose, the now superfluous peroxisomes are rapidly degraded by autophagic events. During ethanol adaptation, peroxisomes are individually and selectively sequestered into autophagosomes that then fuse with the vacuole. This process is called macropexophagy. A second venue for selective peroxisome degradation occurs during glucose adaptation. Micropexophagy proceeds by a mechanism whereby the vacuolar membrane invaginates, resulting in protrusions that surround and engulf the entire cluster of peroxisomes (Tuttle et al., 1993; Sakai et al., 1998). Both pathways result in the degradation of peroxisomes within the vacuole by proteolytic enzymes.

Pichia pastoris is a particularly good genetic model for the study of pexophagy because of its ability to rapidly and selectively degrade peroxisomes (i.e., AOX) when adapting from methanol to glucose medium. The sequestration and degradation of peroxisomes during glucose adaptation is completed within 6 h (Tuttle and Dunn, 1995). Based on the movements of the vacuole that can be easily observed by fluorescence and electron microscopy, glucose-induced micropexophagy has been described as a multistage process that includes: signaling events (vacuole round), early sequestration events (vacuole indented), intermediate sequestration events (vacuole indented with short armlike processes), late sequestration events (vacuole indented with long armlike processes), and vacuole degradation (vacuole with autophagic bodies; Sakai et al., 1998; Strømhaug et al., 2001; Mukaiyama et al., 2002). We and others have utilized genetic screens to identify a number of unique proteins required for peroxisome degradation in P. pastoris (Tuttle and Dunn, 1995; Sakai et al., 1998; Strømhaug et al., 2001). To better understand the functions of these proteins, we examined the vacuole morphology upon glucose-induced pexophagy. For example, PpVps15 (Paz13), PpAtg11 (Gsa9), and PpAtg18 (Gsa12) appear to act at early sequestration events (Stasyk et al., 1999; Guan et al., 2001; Mukaiyama et al., 2002), PpAtg2 (Gsa11) and PpAtg7 (Gsa7/Paz12) at intermediate sequestration events (Yuan et al., 1999; Strømhaug et al., 2001), and PpAtg1 (Paz1/Gsa10) at late sequestration events (Mukaiyama et al., 2002). Many of these proteins are structurally and functionally homologous to those proteins essential for autophagy and CVT pathways in Saccharomyces cerevisiae (Habibzadegah-Tari and Dunn, 2003; Klionsky et al., 2003). Furthermore, structural homologues of these proteins can be found in various invertebrates and vertebrates including humans (Klionsky et al., 2003).

In this study, we have sequenced and characterized a unique membrane protein that is required for both pexophagy and autophagy in P. pastoris. Based on vacuole morphology during glucose adaptation, we project that PpAtg9 is required for an early sequestration event. In growing cells, this protein localizes to unique foci near the cell periphery. During glucose-induced pexophagy, PpAtg9 traffics from these vesicles to perivacuolar structures that appear to adjoin the vacuole and then to the sequestering membranes that arise from the vacuole. The trafficking of PpAtg9 requires PpAtg7, PpAtg11, PpAtg2, and PpVps15, but not PpAtg1, PpAtg18, or PpVac8. Furthermore, our data suggest that the transfer of PpAtg9 to the vacuole is a prerequisite for the formation of those sequestering membranes that engulf the peroxisome for incorporation into the vacuole for degradation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Yeast Strains and Media
The yeast strains used in this study are listed in Table 1 and were routinely cultured at 30°C in YPD (1% Bacto yeast extract, 2% Bacto peptone, and 2% dextrose). P. pastoris was grown in YNM (0.67% yeast nitrogen base, 0.4 mg/l biotin, and 0.5% methanol) to induce peroxisome biogenesis. The degradation of peroxisomes was induced when cells grown in YNM were transferred to YND (0.67% yeast nitrogen base, 0.4 mg/l biotin, and 2% glucose) or YNE (0.67% yeast nitrogen base, 0.4 mg/l biotin, and 0.5% ethanol). Nitrogen starvation medium contained 0.17% yeast nitrogen base (without amino acids and NH₄SO₄) and 2% glucose. All media contained 2% agar when made as plates. Histidine or arginine or both were added at 40 µg/ml when needed. Vector amplification was done in Escherichia coli (DH5) cultured at 37°C in LB (0.5% Bacto yeast extract, 1% Bacto tryptone, and 1% NaCl) with ampicillin (100 µg/ml). Zeocin was added at 25 µg/ml when culturing DH5 and 100 µg/ml when culturing P. pastoris.

Yeast Transformation
Cells grown overnight in YPD to an optical density (OD₆₀₀) of 1.0 were harvested and treated with 10 mM dithiothreitol (DTT) in YPD containing 25 mM HEPES, pH 8, for 15 min at 30°C. The cells were washed twice in ice-cold water and once in 1 M sorbitol and then resuspended into 1 M sorbitol. Cells (40 µl) were mixed with 0.2-1 µg of linearized vector and transferred to a 0.2-cm gap cuvette (Bio-Rad Laboratories, Hercules, CA), and the DNA was introduced by electroporation at 1.5 kV, 25 µF, 400 (Gene Pulser, Bio-Rad Laboratories). The cells transformed with vectors containing the His4 gene were transferred to plates containing 0.67% yeast nitrogen base without amino acids, 2% glucose, 1 M sorbitol, 0.4 mg/L biotin, and 2% agar and incubated at 30°C for 3-5 d before colonies appeared. Cells transformed with vectors containing the zeocin resistance gene (zeo^R) were grown on YPD with 100 µg/ml zeocin.

Isolation of gsa Mutants and Cloning of GSA Genes by Restriction Enzyme-mediated Integration Mutagenesis
Mutagenesis was performed by randomly inserting the pREMI-Z vector (restriction enzyme-mediated integration [REMI]; provided by Dr. Ben Glick, University of Chicago) that contained the zeocin resistance gene into the genome of P. pastoris as previously described (Strømhaug et al., 2001). Those gsa mutants caused by disruption of gene expression were identified by direct colony assays (Strømhaug et al., 2001). The site of insertion of the pREMI-Z and identification of the disrupted gene was done as described (Strømhaug et al., 2001). Based on the genomic sequences around the pREMI-Z insertion site that was isolated along with the pREMI-Z vector from the R19 mutant, GSA14 was cloned and sequenced from genomic DNA using a linker-mediated PCR method previously described (van Der Wel et al., 2001). We were able to completely assemble the GSA14 gene (National Center for Biotechnology Information [NCBI] accession number AY075105 [GenBank] ) and show it encodes a protein homologous to ScAtg9 of S. cerevisiae, using the adapted ATG nomenclature for those genes uniquely essential for autophagy (Klionsky et al., 2003). A search of the NCBI database revealed structural homologues of PpAtg9 and ScAtg9 in Schizosaccharomyces pombe (NP_596247 [GenBank] ), Arabidopsis thaliana (NP_180684), Neurospora crassa (XP_331198 [GenBank] ), Drosophila melanogaster (NP_611114 [GenBank] ), Caenorhabditis elegans (NP_503178 [GenBank] ), and Homo sapiens (NP_076990 [GenBank] ; Yamada et al., 2005). The alignment of these proteins reveals a large central region (190-679 residues) of homology containing at least five putative transmembrane domains. The pREMI-Z disruption of PpAtg9 occurred after the first putative transmembrane domain at aspartic acid 245. Using this approach, we have isolated R2, R12, R13, and R22 mutants. The R2 (his4 Ppatg18-1::zeo^R) and R13 (his4 Ppatg11-1::zeo^R) mutants have been described elsewhere (Guan et al., 2001; Kim et al., 2001; Strømhaug et al., 2001). The R12 mutants had the pREMI-Z inserted into the PpATG1 gene loci. The R22 (his4 Ppatg2-2::zeo^R) and WDK011 (his4 Ppatg2::zeo^R) mutants have been characterized previously (Strømhaug et al., 2001). The Ppatg7/gsa7 mutants have been previously described (Yuan et al., 1999). The null mutant of PpVps15 was provided by Dr. J. Cregg (Keck Graduate Institute; Stasyk et al., 1999). PpVAC8 was cloned from genomic DNA using degenerate primers combined with linker-mediated PCR (van Der Wel et al., 2001). The entire gene was sequenced and assembled (NCBI accession number AY886543 [GenBank] ). A Ppvac8 null mutant was constructed by replacing the entire gene (-7 base pairs through 1669 base pairs) with the zeo^R gene driven by the TEF1 promoter. The null mutants were selected on YPD plates containing 100 µg/ml zeocin and replica-plated to YNM plates. Null mutants were identified by direct colony assay (see below) and verified by PCR using primers flanking PpVAC8 and within the zeo^R gene (unpublished data).

Measurements of Alcohol Oxidase (AOX) and Endogenous Protein Degradation
The direct colony and liquid medium assays to detect and measure the degradation of peroxisomal AOX was performed as previously described (Strømhaug et al., 2001). Briefly, cells were grown in YNM for 40 h. At which time, glucose (2%) or ethanol (0.5%) was added. Aliquots of cells (2 ml of OD₆₀₀ = 1.4) at 0 and 6 h of glucose or ethanol adaptation were pelleted and resuspended in 1 ml 20 mM Tris, pH 7.5, containing 50 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 µg/ml pepstatin A, and 0.5 µg/ml leupeptin. The cells were then lysed by vortexing in the presence of glass beads (425-600 µm). The glass beads and cellular debris were removed by centrifugation and AOX measured (Tuttle and Dunn, 1995; Yuan et al., 1999). The degradation of endogenous proteins during nitrogen starvation was performed as described previously (Strømhaug et al., 2001). Briefly, cellular proteins were radiolabeled with ¹⁴C-valine for 16 h, and the cells switched to nitrogen starvation medium containing 0.17% yeast nitrogen base (without amino acids and NH₄SO₄) and 2% glucose and supplemented with 10 mM valine. The rates of protein degradation were calculated from the slopes of the linear plots of TCA-soluble radioactivity over 2-24 h of chase.

Western Blot Analysis
Cells (2 ml) from cultures grown to an optical density (OD₆₀₀) of 1.4 were collected by centrifugation and prepared for SDS-PAGE as previously described (Tuttle and Dunn, 1995). The cells were lysed in 67 mM Tris, pH 6.8, 2% SDS, 10% glycerol, 0.1% bromophenol blue, 1.5% DTT solution, and 2 µl of a proteinase inhibitor cocktail (200 mM PMSF, 14.5 mM pepstatin A, 10.5 mM leupeptin in dimethyl sulfoxide) by vortexing with glass beads. The proteins were separated by SDS-PAGE and then transferred to nitrocellulose by Trans-Blot SemiDry Transfer Cell (Bio-Rad Laboratories) for 1 h. The blots were blocked in 5% nonfat dried milk in phosphate-buffered saline and then incubated with mouse anti-AOX or rabbit anti-HA antibody (Covance, Princeton, NJ). After incubation with secondary goat anti-mouse antibody conjugated with HRP (Covance), the blots were washed and HRP was detected using ECL-plus (Amersham, Piscataway, NJ) and quantified using the Typhoon 9400 laser scanner (Molecular Dynamics, Sunnyvale, CA).

Construction of PpAtg Expression Vectors
The gene for the green fluorescent protein (GFP) was inserted behind the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) promoter into the EcoRI site of pIB2 (Sears et al., 1998). The resulting expression vector, pPS55, was then used to construct the GFP fusion protein of PpAtg9 being expressed by the constitutive and glucose-inducible GAPDH promoter. PpATG9 was amplified from genomic DNA by PCR with EnzyPlus polymerase (Enzypol, Boulder, CO) using a forward primer of 5'-CTCTCACATTGTCGGTACCATGCATAAGAATAACACGAC-3' that contained a KpnI site. The reverse primer 5'-GTTTTGGACTCGAGGGTACTAATGCTTCATT-3' contained an XhoI site. PpATG9 was inserted behind the GFP gene in pPS55. A second expression vector called pTC2 was created by inserting PpATG9 with its endogenous promoter in front of GFP in pWD3 whereby GFP had been inserted into the SphI site of pIB1 (Sears et al., 1998). PpATG9 was amplified from genomic DNA by PCR using a forward primer 5'-GCAGGCTAGGGTACCGGTACTGGCACATT-3' that contained KpnI site and a reverse primer 5'-CAAGATCTATGCTCGAGAACAAATAATGCCTTATGCTGTTGACTAA-3' with an XhoI site. This product was then inserted into the KpnI and XhoI sites of pWD3, the resulting fusion construct verified by sequencing, and the vector used to transform R19 (his4 atg9::zeo^R) cells. pTC3 was made using PpATG9 with a GAPDH promoter in place of the PpATG9 endogenous promoter by amplifying it from genomic DNA using a forward primer 5'-CTCTCACATTGTCGGTACCATGCATAAGAATAACACGAC-3' containing a KpnI site and a reverse primer 5'-CAAGATCTATGCTCGAGAACAAATAATGCCTTATGCTGTTGACTAA-3' with an XhoI site. It was inserted in front of the GFP gene of pWD4, which had been inserted into the SphI site of pIB2 (Sears et al., 1998). Two additional expression vectors were made by inserting the PpATG9 gene behind the GFP and mRFP genes that were inserted into the EcoRI site of pGAPZ (Invitrogen, San Diego, CA). The gene for mRFP was kindly provided by Dr. R. Y. Tsein (University of California at San Diego; Campbell et al., 2002). PpATG9 was amplified from genomic DNA using a forward primer 5'-CTCTCACATTGTCGGTACCATGCATAAGAATAACACGAC-3' containing a KpnI site and a reverse primer 5'-CATTATTATTCAAGACCGCGGAATGAAAA-3' with a SacII site. The PCR product was then cut with KpnI and SacII and inserted into pGAPz-GFP and pGAPz-RFP resulting in pWD17 and pAJM6 vectors, respectively. GFP-Ppatg9(N) lacking the N-terminus upstream of the first transmembrane region and GFP-Ppatg9(L3) lacking a highly conserved third loop between the 3rd and 4th transmembrane regions was constructed by PCR. A Ppatg9 gene product lacking M1 through Y211 was amplified from genomic DNA by PCR using a forward primer 5'-CGACTATGGTACCGGAAATGGATTCAA-3' with a KpnI site and a reverse primer 5'-GTTTTGGACTCGAGGGTACTAATGCTTCATT-3' with an XhoI site. The resulting gene was inserted into the KpnI and XhoI sites of pPS55. A second Ppatg9 gene product lacking T471 through A523 was constructed by independently amplifying the gene fragments upstream and downstream of the deletion. The upstream region of the PpATG9 gene was amplified using a forward primer of 5'-CTCTCACATTGTCGGTACCATGCATAAGAATAACACGAC-3' that contained a KpnI site and a reverse primer of 5'-CCGCAGGATTAAGCTTAAAATCGTAAAAAA-3' containing a unique HindIII site. The downstream region of the PpATG9 gene was amplified using a forward primer of 5'-CTACAACGAAGCTTCTGAAGTTCATCATGT-3' containing an HindIII site and a reverse primer 5'-GTTTTGGACTCGAGGGTACTAATGCTTCATT-3' containing an XhoI site. These two fragments were digested with the appropriate restriction enzymes and ligated into the KpnI and XhoI sites of pPS55. The expression vector, pWD21, contained PpAtg8 with a GFP at its N-terminus. PpAtg8 was amplified from genomic DNA by PCR using a forward primer of 5'-CCATGAATCCATGCGATCGCAATTTAAAGACGAACA-3' containing an EcoRI site and a reverse primer of 5'-GGCACTACTCGAGTTATTATTCAATCTCCTCAACACCTGGAA-3' containing an XhoI site. PpAtg8 was inserted behind the GFP gene in pPS55. The final expression vector, pASK1, contained PpAtg9 with an HA tag at its N-terminus driven by the endogenous promoter of PpATG9. This construct was made by PCR of the PpATG9 promoter and open reading frame separately, which was then followed by a PCR amplification of the entire gene including promoter. The promoter was amplified using a forward primer of 5'-GCAGGCTAGGGTACCGGTACTGGCACATT-3' containing a KpnI site and a reverse primer of 5'-GCGTAATCTGGAACATCGTATGGATACATTCAATCGACAATGTGAGAGATTCAGTGAAG-3' containing the HA tag. The open reading frame was amplified using a forward primer of 5'-TGTATCCATACGATGTTCCAGATTACGCGATGCATAAGAATAACACGACATTTTTATCC-3' containing an HA tag (MYPYDVPDYA) insertion in front of the PpAtg9 start codon and a reverse primer of 5'-GTCAGACTCCAAACTCGAGTTCATTTTCAA-3' containing an XhoI site. Finally, the entire gene with promoter was amplified from the above products by PCR using forward 5'-GCAGGCTAGGGTACCGGTACTGGCACATT-3' and reverse 5'-GTCAGACTCCAAACTCGAGTTCATTTTCAA-3' primers. The resulting product was cut with KpnI and XhoI and inserted into pIB1. The CoxIV signal sequence plus the upstream ADH1 promoter was excised from pCC4 by cutting with EcoRV and XbaI (Campbell and Thorsness, 1998). The fragment was then inserted into the SmaI and SpeI sites upstream of the GFP gene of pWD3, which had been constructed by inserting the GFP gene into the SphI site of pIB3 (Sears et al., 1998). The resulting expression vector, pAJM3, produced a GFP protein that was targeted to the mitochondria. The construction of pPS55-G12, pPS64, and pPS69 vectors have been previously described (Guan et al., 2001; Kim et al., 2001; Strømhaug et al., 2001) and pPOP-S7-GFPX3, pPOP-SEC13-GFP, and pIB2-DsRED-HDEL were provided by Dr. Ben Glick (University of Chicago; Bevis et al., 2002). For transformation, these vectors were first linearized by cutting either within the HIS4 gene with StuI or SalI or within the PpATG9 gene with SacI.

Fluorescence microscopy and FM 4-64 Labeling
Cells expressing GFP or RFP fusion proteins were grown in either YPD for 24 h or YNM for 20 h. Cells grown on YNM medium were then transferred to YND or YNE for 1-4 h. FM 4-64 (Molecular Probes, Eugene, OR) was added to a final concentration of 20 µg/ml and the cells incubated for 2-12 h. The cells were washed of unbound FM 4-64 and examined immediately using a Zeiss Axiophot fluorescence microscope (Thornwood, NY). Image capture was done using SPOT camera (Diagnostics Instruments, Sterling Heights, MI) interfaced with IP Lab software.

Electron Microscopy
The cellular ultrastructure of Ppatg9 mutants was examined as previously described (Tuttle and Dunn, 1995). Briefly, cells grown in YNM or grown in YNM and adapted to YND or YNE were harvested by centrifugation, washed in water, and fixed in 1.5% KMnO₄ in veronal-acetate buffer (28 mM sodium acetate, 28 mM sodium barbital, pH 7.6) for 20 min at 22°C (Veenhuis et al., 1983). The specimens were dehydrated with increasing concentrations of ethanol and infiltrated with POLY/BED 812 (Polysciences, Warrington, PA) with accelerator 2,4,6-tri(dimethylaminomethyl) phenol (DMP-30, Polysciences) for 2 d at 22°C under vacuum. After polymerization of the resin, the samples were sectioned and examined using a JEOL 100CX transmission electron microscope (Peabody, MA).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Ppatg9 Mutants Are Defective in Glucose-induced Pexophagy and Starvation-induced Autophagy
P. pastoris can assimilate methanol for growth by synthesizing peroxisomal enzymes such as alcohol oxidase (AOX). When these cells are switched from a medium containing methanol to one containing glucose, the peroxisomes are selectively sequestered by the vacuole for degradation by a process called micropexophagy (Tuttle and Dunn, 1995). To identify the molecular components of this degradative process, we developed a novel approach whereby genes can be randomly mutagenized in vivo by the integration of a vector containing a zeo^R gene, pREMI-Z. Those zeocin-resistant cells that were unable to degrade AOX during glucose adaptation were identified by direct colony assay and verified by liquid medium assay (see Materials and Methods). Using this approach, we have isolated a number of mutant strains based on their poor ability to degrade AOX during glucose adaptation. The gene mutated by the insertion of the pREMI-Z was sequenced and identified in each of our mutants (see Table 1). We then compared the inability of these mutants to degrade peroxisomes to parental GS115 and to vacuole defective mutants that lack proteinases A (pep4) and B (prb1) (Figure 1A). Within 6 h of glucose adaptation, more than 90% of the AOX was degraded by the GS115 cells. In comparison, only 10% of the AOX was degraded in mutants lacking proteinases A and B. Less than 60% of the AOX was degraded in R19 cells that contained a pREMI-Z insertion within the open reading frame of the PpATG9 gene. AOX degradation in these R19 cells was comparable to that observed for Ppatg1, Ppatg2, and Ppatg11 mutants, but higher than the 10-30% degraded by Ppatg7, Ppatg18, Ppvps15, or Ppvac8 mutants. Next, we examined the ability of R19 cells to degrade AOX during ethanol adaptation (Figure 1B). We observed on Western blots a substantial loss of AOX protein when GS115 cells adapt from methanol to ethanol medium. However, no loss of AOX was detected in R19 cells over 24 h, suggesting that macropexophagy was suppressed in the cells lacking PpAtg9.

View larger version (24K): [in this window] [in a new window]   Figure 1. Pexophagy and autophagy are defective in Ppatg9 mutants. (A) Wild-type GS115, R12 (Ppatg1), R22 (Ppatg2-2), WDY7 (Ppatg7), R19 (Ppatg9), R13 (Ppatg11-1), R9 (Ppatg18), Ppvps15 (Ppvps15), WDY53 (Ppvac8), and SMD1163 (pep4, prb1) cells were grown in YNM for 36 h. At that time, cells were switched to a medium containing 2% glucose. Aliquots were removed at 0 and 6 h of adaptation, the cells were lysed, and AOX activities were measured as describe in Materials and Methods. The data are expressed as a percentage of AOX remaining at 6 h relative to 0 h and represents the mean ± SE of 3-6 trials. (B) Wild-type GS115 and R19 (Ppatg9) cells were grown in YNM for 36 h and then switched to YNE medium. Aliquots were removed, the cells were lysed, and AOX protein was visualized by Western blotting as describe in Materials and Methods. (C) Wild-type GS115, R12 (Ppatg1), R22 (Ppatg2-2), WDY7 (Ppatg7), R19 (Ppatg9), R13 (Ppatg11-1), R9 (Ppatg18), Ppvps15 (Ppvps15), WDY53 (Ppvac8), and SMD1163 (pep4, prb1) cells were grown in minimal medium containing 14C-valine for 18 h. The cells were pelleted and resuspended in medium lacking amino acids and nitrogen and containing 10 mM valine. The production of TCA-soluble radioactivity was measured at 2, 5, 8, and 24 h of chase, and the rates calculated by linear regression of the slope of the line. The rates represent the mean ± SE of 5-7 trials.

PpAtg9 Is Essential for a Sequestration Event in Pexophagy
We have shown that during glucose-induced micropexophagy the vacuole indents while sequestering membranes encircle multiple peroxisomes (Tuttle and Dunn, 1995). The sequestering membranes are labeled with FM 4-64, a dye that selectively labels vacuole membranes, suggesting that they likely form from the vacuole. A micropexophagic membrane apparatus (MIPA) containing PpAtg8 forms near the tips of the sequestering membranes. The MIPA is thought to assist in the fusion of the sequestering membranes, thereby incorporating the peroxisomes into the vacuole for degradation. During ethanol-induced macropexophagy, individual peroxisomes are selectively incorporated into pexophagosomes whose membranes contain PpAtg8. The pexophagosome then fuses with the vacuole and its contents degraded. To better assess the site of blockage of pexophagy in the R19 cells, we examined the formation of the sequestering membranes and the MIPA during micropexophagy and of the pexophagosome during macropexophagy. GS115 (Figure 2, A and C) and R19 (Figure 2, E, G, I, and K) cells were grown in methanol-enriched YNM medium and then adapted to glucose-enriched YND (Figure 2, E and I) or ethanol-enriched YNE (Figure 2, G and K) medium. Cells were harvested at 3 h, fixed in potassium permanganate, and prepared for electron microscopy. During glucose adaptation, profiles of peroxisomes engulfed by the vacuole were routinely observed in GS115 cells (Figure 2A). However, when R19 cells were adapted to glucose, the sequestering membranes appear to be absent, but instead the vacuole appeared only slightly indented (Figure 2E), with an occasional short armlike projection (Figure 2I). Similar vacuole morphology has been reported for Ppatg11, Ppatg18, and Ppvps15 mutants (Stasyk et al., 1999; Guan et al., 2001; Kim et al., 2001). Next, we examined whether the MIPA was forming in R19 cells. WDY70 cells expressing BFP-SKL and GFP-PpAtg8 were adapted from YNM to YND for 2 h. At this time, PpAtg8 was localized to the MIPA and foci adjacent to the peroxisomes (Figure 2B). In the absence of PpAtg9, the MIPA was absent and PpAtg8 localized solely to foci (Figure 2, F and J). During ethanol adaptation, peroxisomes within pexophagosomes were observed by electron (Figure 2C) and fluorescence (Figure 2D) microscopy. In WDY70 cells, the PpAtg8 was localized almost exclusively to pexophagosomes (Figure 2D). Pexophagosomes were absent in cells lacking PpAtg9 (Figure 2, G, H, K, and L), and PpAtg8 was localized to multiple foci that appear within and about the peroxisome cluster (Figure 2, H and L). The data demonstrate that PpAtg9 is essential for the formation of sequestering membranes that arise from the vacuole and the assembly of the MIPA during micropexophagy and of the pexophagosome during macropexophagy.

View larger version (85K): [in this window] [in a new window]   Figure 2. Pexophagy is blocked at an early sequestration event in cells lacking PpAtg9. Wild-type cells, GS115 (A and C) and WDY70 (B and D) and cells lacking PpAtg9, R19 (E, G, I, and K) and WDY71 (F, H, J, and L) were grown in YNM for 24-36 h. At that time, cells were switched to medium containing 2% glucose (A, B, E, F, I, and J) or 0.5% ethanol (C, D, G, H, K, and L) for 3 h. Cells were fixed with potassium permanganate and prepared for viewing on a JEOL 100CX transmission electron microscope (A, C, E, G, I, and K) or viewed in situ by fluorescence microscopy (B, D, F, H, J, and L). When GS115 and WDY70 cells were adapted to glucose for 3 h, many of the cells lacked peroxisomes, whereas in others the vacuole was found to virtually surround the peroxisome cluster (A and B). In R19 and WDY71 cells, peroxisome clusters were evident in virtually all the cells. The vacuole was either oblong with a cuplike depression at the surface adjacent to the peroxisomes (E and F) or round with short armlike segments (I and J). By fluorescence microscopy, the vacuole was observed by staining with FM 4-64, the peroxisomes by expressing BFP-SKL, and the MIPA and pexophagosomes by expressing GFP-PpAtg8. GFP-PpAtg8 was observed at the MIPA in WDY70 cells (B, arrow), but appeared as foci in WDY71 cells lacking PpAtg9 (F and J, arrowheads). During ethanol adaptation, individual peroxisomes were sequestered into pexophagosomes (C, arrow) that contained PpAtg8 (D, arrow). No pexophagosomes were evident in R19 and WDY71 cells lacking PpAtg9 (G, H, K, and L). In the absence of macropexophagy, PpAtg8 resided in numerous foci (H and L, arrowheads). N, nucleus; P, peroxisome; V, vacuole.

View larger version (60K): [in this window] [in a new window]   Figure 8. PpAtg9 domains required for function and trafficking. In A, wild-type GS115, R19 (Ppatg9), and R19 cells expressing mRFP-PpAtg9, GFP-PpAtg9, GFP-Ppatg9N (M1-Y221), or GFP-Ppatg9L3 (T471-A523) were grown in YNM medium for 36 h. All proteins were expressed behind the GAPDH promoter. Cells were then switched to glucose medium. Aliquots were removed at 0 and 6 h of adaptation, the cells were lysed, and AOX activities were measured as described in Materials and Methods. The data are expressed as a percentage of AOX remaining at 6 h relative to 0 h and represent the mean ± SE of 4-6 trials. ASK5 (B) and ASK4 (C) cells were grown in YNM for 20 h in the presence of FM 4-64. The cells were then transferred to YND medium for 3 h and then visualized by fluorescence microscopy. GFP-Ppatg9N was found at the Atg9-PC (arrowheads), PVS (arrows), and the vacuolar membranes. GFP-Ppatg9L3 was absent from the Atg9-PC and PVS, but found in a compartment that resembled the endoplasmic reticulum (large arrows).

View larger version (73K): [in this window] [in a new window]   Figure 3. Cellular localization of PpAtg9. TC14 expressing PpAtg9-GFP behind the endogenous PpAtg9 promoter (A and B), AJM27 expressing GFP-PpAtg9 behind the GAPDH promoter and dsRFP-HDEL (C), AJM44 expressing mRFP-Atg9 behind the GAPDH promoter and CoxIV-GFP (D), AJM19 expressing mRFP-PpAtg9 behind the GAPDH promoter and Sec13-GFP (E), and AJM18 expressing mRFP-PpAtg9 behind the GAPDH promoter and Sec7-GFP (F) were grown in YPD and the cellular distribution of the GFP- and mRFP-tagged proteins were visualized by in situ fluorescence microscopy. In TC14 cells, PpAtg9-GFP was present in foci situated at the cell periphery (arrows). The GFP-PpAtg9 structures appeared to be juxtaposed to the endoplasmic reticulum containing dsRFP-HDEL (C) and mitochondria containing CoxIV-GFP (D), but distinct from intermediate vesicles containing Sec13-GFP (E) and Golgi apparatus containing Sec7-GFP (F).

Cellular Trafficking of PpAtg9 during Glucose-induced Pexophagy
The trafficking of PpAtg9 during glucose-induced pexophagy was visualized in TC3 cells expressing BFP-SKL behind the AOX promoter and GFP-PpAtg9 behind the GAPDH promoter (Figure 4). When TC3 cells were grown in YNM, GFP-PpAtg9 localized to one or more structures, described above as the Atg9-PC (Figure 4A). On adapting these cells from YNM to YND for 3 h, GFP-PpAtg9 localized to multiple structures of differing sizes and shapes that were positioned at the vacuole surface labeled with FM 4-64 and to the sequestering membranes that could be seen flanking the peroxisomes labeled with BFP-SKL (Figure 4). We refer these structures that appear as dots and patches as perivacuolar structures (PVS). The PVS were routinely observed near those sites where the sequestering membranes joined the vacuole, suggesting they may function in the formation of these membranes from the vacuole. Indeed, the sequestering membranes contained both PpAtg11 and PpAtg18 present at the vacuole membrane and stain with FM 4-64, suggesting that they originated from the vacuole (Guan et al., 2001; Kim et al., 2001). These results were not due to overexpression of GFP-PpAtg9 that may be caused by the GAPDH promoter, because PpAtg9-GFP whose expression is regulated by the endogenous PpAtg9 promoter in TC14 cells showed a similar distribution but weaker signal (unpublished data). In a given population of cells, the onset of micropexophagy is variable resulting in cells at differing stages of micropexophagy. Nevertheless, we have presented a time course based on images obtained from 0-2 h of glucose adaptation that best illustrates our understanding of PpAtg9 trafficking (Figure 4C). We show that during glucose-induced pexophagy, PpAtg9 traffics from the Atg9-PC (arrowheads) to the PVS (arrows) and then to the vacuole membrane and those sequestering membranes (double arrowheads) that engulf the peroxisomes for degradation within the vacuole.

View larger version (67K): [in this window] [in a new window]   Figure 4. PpAtg9 relocates from peripheral structures to the sequestering membranes during glucose-induced pexophagy. TC3 cells expressing both BFP-SKL behind the AOX1 promoter and GFP-PpAtg9 behind the GAPDH promoter were grown in YNM medium in the presence of FM 4-64, adapted to YND medium for 0 h (A) and 3 h (B), and the cellular distribution of GFP-PpAtg9 visualized by in situ fluorescence microscopy. Peroxisomes were identified by the presence of BFP, which was targeted by its SKL signal and the vacuole by the red dye FM 4-64. At 0 h, many cells had BFP-containing peroxisomes and a single round vacuole. GFP-PpAtg9 was localized to foci near the cell periphery (arrowheads). At 3 h of glucose adaptation, numerous profiles of vacuoles with armlike extensions surrounding the BFP-containing peroxisomes could be observed. GFP-PpAtg9 was present at perivacuolar structures (arrows) and at the vacuole membrane (double arrowhead). In C, images from 0 to 2 h of glucose adaptation depict the movements of GFP-PpAtg9 from peripheral structures (arrowheads) to the perivacuolar structures (arrows) and sequestering and vacuole membranes (double arrowhead). In D, ASK2 cells expressing HA-PpAtg9 behind the endogenous PpAtg9 promoter were grown in methanol medium and then adapted to glucose medium. At 0-8 h, equivalent numbers of cells based on the optical density (OD600) of the cultures were solubilized in SDS, and the proteins were separated by SDS-PAGE. HA-PpAtg9 was then visualized by Western blotting using anti-HA antibodies and quantified using a Typhoon laser scanner. The values were normalized to 0 h and presented as the mean ± SE (n = 6).

Cellular Trafficking of PpAtg9 during Glucose-induced Pexophagy Requires other PpAtg Proteins
Many Atg proteins interact with each other, thereby influencing function and trafficking (Wang and Klionsky, 2003; Reggiori et al., 2004). In fact, the trafficking of ScAtg9 appears to be regulated by a number of ScAtg proteins, including ScAtg1 and ScAtg18 (Reggiori et al., 2004). Therefore, we investigated whether other PpAtg proteins were required for the trafficking of PpAtg9 from the Atg9-PC to the PVS and sequestering membranes. We examined the trafficking of GFP-PpAtg9 during glucose-induced pexophagy in mutants defective in early (Ppatg11, Ppatg18, and Ppvps15), intermediate (Ppatg2 and Ppatg7), and late (Ppatg1 and Ppvac8) sequestration events. These mutants were transformed by electroporation with pTC1 and GFP-PpAtg9 expression verified by Western blotting and fluorescence microscopy. When the resulting transformants were grown in YNM, GFP-PpAtg9 was found almost exclusively at the Atg9-PC (unpublished data). After growth in YNM, the mutants expressing GFP-PpAtg9 were then adapted to glucose for 3 h. In Ppatg11, Ppatg18, and Ppvps15 mutants, the vacuoles were predominantly round with a slight indentation and no armlike extensions were evident (Figure 5). GFP-PpAtg9 expressed in Ppatg11 and Ppvps15 cells was found predominantly at peripheral foci (Figure 5). In addition, the vacuole membrane was void of GFP-PpAtg9 in these mutants. In Ppatg18 cells, GFP-PpAtg9 localized to the PVS and to the vacuole membrane (Figure 5). The vacuoles in Ppatg7 and Ppatg2 mutants were indented with short extensions (Figure 6). In these cells, GFP-PpAtg9 was found in peripheral structures (arrowheads) and the PVS (arrows) located either at the vacuole or at the site where the sequestering membranes join the vacuole (Figure 6). Although the location of the PVS appeared unaltered, the PVS morphology appeared to be smaller and less diffuse in these mutants than that observed in control TC3 cells (see Figure 4). Furthermore, GFP-PpAtg9 was not detected at the sequestering or vacuole membranes. During micropexophagy, the vacuoles in Ppatg1 and Ppvac8 mutants contained an extensive segmented array of sequestering membranes that extended from the vacuole to almost completely surround the peroxisome cluster (Figure 7). GFP-PpAtg9 was found at the PVS and sequestering membranes (Figure 7).

View larger version (113K): [in this window] [in a new window]   Figure 5. Trafficking of PpAtg9 in mutants defective in early sequestration events. TC6 (Ppatg11), TC9 (Ppatg18), and TC11 (Ppvps15) mutants expressing GFP-PpAtg9 behind the GAPDH promoter were grown in YNM medium in the presence of FM 4-64 then adapted to YND medium for 3 h and the cellular distribution of GFP-PpAtg9 visualized by in situ fluorescence microscopy. The vacuoles in the Ppvps15, Ppatg11, and Ppatg18 mutants were either round, flattened, or slightly indented. GFP-PpAtg9 was present in peripheral structures (arrowheads) in the Ppvps15 and Ppatg11 mutants. However, in Ppatg18 mutants, GFP-PpAtg9 was visualized at the large sometimes diffuse perivacuolar structures (arrows) and the vacuole membrane (double arrowheads).

View larger version (83K): [in this window] [in a new window]   Figure 6. Trafficking of PpAtg9 in mutants defective in intermediate sequestration events. TC19 (Ppatg2) and TC5 (Ppatg7) mutants expressing GFP-PpAtg9 behind the GAPDH promoter were grown in YNM medium in the presence of FM 4-64 then adapted to YND medium for 3 h, and the cellular distribution of GFP-PpAtg9 was visualized by in situ fluorescence microscopy. The vacuoles in Ppatg2 and Ppatg7 mutants were indented with an occasional short segmented armlike extension. In these mutants, GFP-PpAtg9 localized to peripheral structures (arrowheads) and the PVS at the vacuole surface (arrows), but was absent from the vacuolar membrane. Similar results were obtained in TC8 (Ppatg2) and TC4 (Ppatg7) cells.

View larger version (81K): [in this window] [in a new window]   Figure 7. Trafficking of PpAtg9 in mutants defective in late sequestration events. TC7 (Ppatg1) and TC12 (Ppvac8) mutants expressing GFP-PpAtg9 behind the GAPDH promoter were grown in YNM medium in the presence of FM 4-64 and then adapted to YND medium for 3 h, and the cellular distribution of GFP-PpAtg9 was visualized by in situ fluorescence microscopy. The vacuoles in Ppatg1 and Ppvac8 mutants were indented with multiple "segmented" armlike extensions that surrounded the peroxisome cluster. GFP-PpAtg9 was visualized at the vacuole membrane but appeared to be more concentrated at the sequestering membrane extensions of the vacuole (double arrowheads). In addition, GFP-PpAtg9 was found at the PVS situated at sites of the vacuole surface adjacent to the armlike extensions (arrows).

Perivacuolar Structures Contain PpAtg9 and PpAtg11, but not PpAtg2
We have previously shown that Atg2 becomes associated with peripheral structures during pexophagy and that this association requires PpAtg9 (Strømhaug et al., 2001). In addition, we have reported that PpAtg11 localizes to the vacuole membrane and to one or more structures associated with either the vacuole or the armlike extensions that surround the peroxisomes (Kim et al., 2001). In this study, we have shown that the trafficking of PpAtg9 requires PpAtg2 and PpAtg11. Therefore, we compared the localization of mRFP-PpAtg9 with PpAtg2-GFP and PpAtg11-GFP in cells adapting from YNM to YND (Figure 9). We first compared the localization of mRFP-PpAtg9 with PpAtg18-GFP, which we have previously shown to be present at the vacuole and sequestering membranes. In cells undergoing micropexophagy, mRFP-PpAtg9 was present in the Atg9-PC and at the PVC localized at sites where the sequestering membranes appear to extend from the vacuole (Figure 9, arrows). Unlike GFP-PpAtg9, mRFP-PpAtg9 was not present at the vacuole membrane, but instead was present within the vacuole that was delineated by PpAtg18-GFP and PpAtg11-GFP. One possible explanation is that the mRFP at the N-terminus of PpAtg9 is exposed to vacuolar enzymes and proteolytically removed from PpAtg9. However, the accumulation of mRFP within the vacuole persisted in cells lacking Pep4 or Prb1 (Tomasini and Dunn, unpublished observations). The data reveal that PpAtg2 does not colocalize with the Atg9-PC nor the PVS. However, some of the PpAtg2 vesicles can be found in close proximity to the PVS (see Figure 9, arrows). PpAtg11 localized to the vacuole membrane and the PVS containing PpAtg9 (Figure 9, arrows), but did not colocalize with the peripheral structures of PpAtg9 (Figure 9, arrowhead). The data suggest that the Atg9-PC does not contain PpAtg2 and that the peripheral compartment of PpAtg2 may be unique. Furthermore, we have shown that the PVS contains PpAtg9 and PpAtg11 but not PpAtg2 or PpAtg18.

View larger version (118K): [in this window] [in a new window]   Figure 9. Perivacuolar structures contain PpAtg9 and PpAtg11. AJM33 expressing GFP-PpAtg18 and mRFP-PpAtg9, AJM25 expressing GFP-PpAtg2 and mRFP-PpAtg9 and AJM32 expressing GFP-PpAtg11 and mRFP-PpAtg9 were grown in YNM medium and then adapted to YND medium for 2 h, and the cellular distribution of GFP- and RFP-tagged proteins was visualized by in situ fluorescence microscopy. PpAtg9 was found at peripheral structures (arrowheads) and PVS (arrows) and within the vacuole delineated by PpAtg18 and PpAtg11. The Atg9-PC did not contain PpAtg2, PpAtg11, or PpAtg18. GFP-PpAtg2 resided in distinct vesicles of which a few were in close association with the PVS. Meanwhile, virtually all the PVS contained both GFP-PpAtg11 and mRFP-PpAtg9 (arrows).

PpAtg9 Does Not Localize to MIPA or the Pexophagosome
Finally, we examined whether mRFP-PpAtg9 colocalized with GFP-PpAtg8 at the MIPA, at the pexophagosomes, or at the perivacuolar foci referred to as preautophagosome structures (PAS) in S. cerevisiae. During glucose-induced micropexophagy, PpAtg8 was present at perivacuolar foci and at the MIPA (arrowheads), which did not contain PpAtg9 (Figure 10A). During ethanol-induced macropexophagy, PpAtg8 was found at perivacuolar foci and at the pexophagosome (arrowheads), whereas PpAtg9 localized to the Atg9-PC (small arrow) and the PVS (large arrow; Figure 10B). PpAtg9 did not colocalize with PpAtg8 and was absent from the pexophagosome. To the contrary, PpAtg8 was not present at the Atg9-PC (Figure 10B, small arrow) or at the PVS (Figure 10, large arrows). The data suggest that PpAtg9 is not a component of the MIPA or the pexophagosome and that the PVS lacks PpAtg8.

View larger version (73K): [in this window] [in a new window]   Figure 10. PpAtg9 does not colocalize with PpAtg8 structures. WDY75 expressing GFP-PpAtg8 and mRFP-PpAtg9 were grown in YNM and then adapted to glucose or ethanol medium for 2 h, and the cellular distribution of GFP- and RFP-tagged proteins was visualized by in situ fluorescence microscopy. During glucose adaptation (A), mRFP-PpAtg9 was found within the vacuole and at the PVS (large arrows), but was absent from PpAtg8 foci and MIPA (arrowheads). When cells were transferred from methanol to ethanol medium (B), mRFP-PpAtg9 was present within the vacuole and at the peripheral compartment (small arrow) and the PVS (large arrow), but not at the PpAtg8 foci or pexophagosome (arrowheads).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

In S. cerevisiae, ScAtg9 resides in two populations of vesicles, one at the cell periphery and a second found adjacent to the vacuole. These vesicles distribute as a single peak on sucrose gradients, which can be resolved from the vacuole (Pho8), plasma membrane (Pma1), Golgi apparatus (Kex2), endoplasmic reticulum (Sec12), and endosomes (Pep12; Noda et al., 2000). In addition, our results suggest that those vesicles adjacent to the vacuole are not the prevacuolar compartment (Noda et al., 2000). The preautophagosome structure (PAS) appears as a single structure located at the vacuolar surface and is composed of a number of proteins such as ScAtg2 (Apg2), ScAtg8 (Aut7), ScAtg9 (Apg9), ScAtg11 (Cvt9), ScAtg17 (Apg17), ScAtg18 (Cvt18), ScAtg20 (Cvt20), and ScAtg24 (Cvt13) (Suzuki et al., 2001; Huang and Klionsky, 2002; Kim et al., 2002; Nice et al., 2002; Reggiori et al., 2004). The PAS is thought to be responsible for organizing the formation of the autophagosome and transferring ScAtg8 to the autophagosome membrane. However, the appearance of these structures does not change when cells are starved for amino acids, and other components of the PAS such as ScAtg9 do not appear to associate with autophagosomes (Noda et al., 2000; Kim et al., 2002). Those ScAtg9 vesicles not adjacent to the vacuole but more at the cell periphery differ from the PAS in that they do not contain ScAtg8 (Kim et al., 2002; Tucker et al., 2003). In S. cerevisiae, ScAtg9 appears to recycle between the peripheral vesicles and the PAS. Furthermore, the movements of ScAtg9 from the PAS to the peripheral vesicles require ScAtg1 independent of its kinase activity, ScAtg2, ScAtg18, and the PtdIns 3-kinase complex I, which includes ScVps34, ScVps15, and ScAtg14 (Reggiori et al., 2004).

In Figure 11, we have presented a working model for the trafficking of PpAtg9 during glucose-induced pexophagy. In growing cells, PpAtg9 resides in the Atg9 peripheral compartment (Atg9-PC) that consists of one or more structures distributed near the cell periphery. These structures are neither intermediate vesicles nor the Golgi apparatus because they lack PpSec13 and PpSec7, respectively. However, they occasionally contain the endoplasmic reticulum marker, mRFP-HDEL, whereas others are in close association with the endoplasmic reticulum, suggesting these structures may be a subcompartment of the endoplasmic reticulum. Furthermore, Ppatg9L3 is not present at the Atg9-PC, but appears to be associated with the endoplasmic reticulum. However, additional studies are needed to substantiate the relationship between the endoplasmic reticulum and the Atg9-PC. ScAtg9 appears to recycle between the peripheral compartment and the PAS (Reggiori et al., 2004). Such recycling was not detected in P. pastoris, although we have not analyzed this under the same conditions defined in S. cerevisiae. We have also shown that the PpAtg9-PC is sometimes associated with mitochondria. However, the functional significance of this location remains unclear. Nevertheless, based on our findings and those from S. cerevisiae, we suggest that the peripheral vesicles containing PpAtg9 are a unique compartment. On the onset of glucose-induced pexophagy, PpAtg9 is recruited from the Atg9-PC to two or more perivacuolar structures (PVS) and then to those membranes that sequester the peroxisomes. The PVS is in many ways similar to the PAS defined in S. cerevisiae. For example, both contain PpAtg9 and PpAtg11 (see Figure 10). Ano et al. (2005) have shown that similar perivacuolar spots contain PpAtg24. Both PVS and PAS are located at the vacuole and appear to be essential for organizing the formation of sequestering membranes. However, the PVS does not contain PpAtg2 or PpAtg8 and appears to be structurally different from the PAS. There exist multiple foci of PVS of differing sizes situated about the vacuole and the sequestering membranes, whereas the PAS usually consists of a single, well-defined structure at the vacuole. PpAtg17, a component of the PAS in S. cerevisiae, resides in a single structure juxtaposed to the vacuole (Ano et al., 2005). PpAtg24 colocalizes with PpAtg17 but is also found at multiple other perivacuolar structures (Ano et al., 2005). PAS has been implicated in the formation of the autophagosome from membranes of unknown origin, but its function remains unclear (Suzuki et al., 2001). We have demonstrated that the PVS is situated at sites where the sequestering membranes appear to form from the vacuole and that PpAtg9 is transported from the PVS to those sequestering membranes that engulf the peroxisomes. Therefore, we project that PVS may act as the site of assembly in the formation of the sequestering membranes that appear to originate from the vacuole because they stain with FM 4-64. It is unclear how these membranes are formed. Mukaiyama et al. (2002) suggest these membranes form by a "septation" of the vacuole. Our data suggest that the formation of these membranes requires protein synthesis and both PpAtg9 and PpAtg18. We have previously shown that these membranes do not form in the presence of cycloheximide (Tuttle and Dunn, 1995). In this study, we show that the cellular levels of PpAtg9 are increased during pexophagy, suggesting a role for protein synthesis. In addition, we have demonstrated that when PpAtg9 is absent or does not traffic to the PVS, the vacuole arms are either not present or short. Conversely, when PpAtg9 localizes to the PVS and the vacuole membrane but PpAtg18 is missing, the vacuole arms are absent. Furthermore, we have shown that PpAtg9 lacking its N-terminus is sorted to the PVS, but the peroxisomes are not sequestered or degraded. This suggests that there exists a region within the N-terminus of PpAtg9 that is essential for its function once it arrives at the PVS. This domain and its function have yet to be defined and are currently under investigation.

View larger version (15K): [in this window] [in a new window]   Figure 11. Model of PpAtg9 trafficking during pexophagy. During growth in YNM, PpAtg9 (green) is in the peripheral compartment (Atg9-PC), PpAtg11 (red) at the vacuole membrane, and PpAtg2 (brown) cytosolic. Upon glucose-induced pexophagy, PpAtg9 is trafficked from the Atg9-PC, to the PVS juxtaposed to the vacuole which also contains PpAtg11, and to the sequestering membranes (SM) that surround the peroxisomes (blue). Also at this time, PpAtg2 becomes associated with unknown structures situated near the PVS. The trafficking of PpAtg9 from the Atg9-PC to the PVS requires PpAtg11 and PpVps15. The movements of PpAtg9 from the PVS to the vacuolar and sequestering membranes require PpAtg7 and PpAtg2. The formation of the sequestering membranes from the vacuole requires PpAtg18, but the movements of PpAtg9 to the vacuolar membrane proceeds normally in the absence of this protein. Finally, the fusion of the sequestering membranes that enable the incorporation of the peroxisomes into the vacuole is guided by the micropexophagy-specific membrane apparatus (MIPA) and requires PpVac8 and PpAtg1.

We have shown that PpAtg11 and PpVps15, but not PpAtg18, PpAtg2, PpAtg7, PpAtg1, or PpVac8, are essential to recruit PpAtg9 to the PVS. In S. cerevisiae, the overexpression of ScAtg11 appears to enhance the recruitment of ScAtg9 to the PAS (Kim et al., 2002). In cells lacking ScAtg1, ScAtg2, ScAtg14, or ScAtg18, ScAtg9 is found predominantly at the PAS either because trafficking to the PAS is enhanced or recycling to the peripheral vesicles is suppressed (Reggiori et al., 2004). ScAtg11 and ScAtg18 are present at the PAS and vacuole surface, but the localization of ScVps15 has not yet been determined (Guan et al., 2001; Kim et al., 2001). From the PVS, PpAtg9 is transferred to the membranes of the vacuole and those sequestering membranes that extend from the vacuole. The apparent movement of PpAtg9 from the PVS to the sequestering membranes requires PpAtg2 and PpAtg7, but not PpAtg18, PpAtg1, or PpVac8. We have previously shown that during pexophagy PpAtg2 becomes associated with foci of unknown origin (Strømhaug et al., 2001). In this study, we show that these structures are situated juxtaposed to the PVS. The homotypic fusion of the sequestering membranes and formation of the autophagic body within the vacuole completes the sequestration of the peroxisomes. A membrane sac called MIPA (micropexophagy-specific membrane apparatus), which contains PpAtg8 and PpAtg26, forms between the tips of the sequestering membranes to presumably direct membrane fusion (Oku et al., 2003; Mukaiyama et al., 2004). In addition to PpAtg8 and PpAtg26, the assembly of MIPA requires the lipidation of PpAtg8, which is mediated by PpAtg3, PpAtg4, and PpAtg7. We show here that the assembly of MIPA also requires PpAtg9. However, it is unclear whether PpAtg9 has a direct role in MIPA formation or if the assembly of MIPA requires the presence of the sequestering membranes. Our data suggest that late sequestration or fusion events require PpAtg1, a serine-threonine protein kinase, and PpVac8, an Armadillo-repeat protein structurally homologous to ScVac8 that when anchored to the vacuole membrane by palmitoylation will promote homotypic vacuole fusion (Wang et al., 1998). PpAtg24 is also essential for homotypic fusion of the sequestering membranes (Ano et al., 2005). PpAtg24 is not required for MIPA formation (Ano et al., 2005), but it is unclear whether PpAtg1 and PpVac8 are essential for the formation of the MIPA.

In summary, we have characterized the functional role of PpAtg9 in glucose-induced micropexophagy. We have shown that PpAtg9 is essential for the formation of the sequestering membranes that engulf the peroxisomes for degradation within the vacuole. During micropexophagy, PpAtg9 traffics from a unique compartment of the Atg9-PC to the PVS juxtaposed to the vacuole where it then becomes associated with the vacuole and those sequestering membranes that engulf the peroxisomes for degradation (see Figure 11). We have also demonstrated that the trafficking of PpAtg9 requires PpAtg11, PpVps15, PpAtg2, and PpAtg7, but not PpAtg1, PpAtg18, or PpVac8. We propose that upon the onset of micropexophagy, PpAtg11 recruits PpAtg9 to the PVS, which act as sites of formation of the sequestering membranes presumably by causing segmentation of the vacuole. These membranes then engulf the peroxisomes and eventually fuse with the assistance of PpAtg1 and PpVac8 to incorporate the peroxisomes into the vacuole for degradation.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
We thank Dr. B. S. Glick (University of Chicago) for generously providing the pREMI-Z, pIB1, and pIB2 vectors. Finally, we would like to thank Mr. Todd Barnash for his help in assembling the figures, and Dr. Michelle Fry and Debbie Akin for helpful discussions and editing. This work was supported by grants from the National Science Foundation Grant MCB-9817002 and the National Cancer Institute, National Institutes of Health Grant CA95552 to W.A.D. and a grant from The Norwegian Cancer Society to P.E.S.

	Footnotes

 
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E05-02-0143) on August 3, 2005.

Abbreviations used: AOX, alcohol oxidase; FM 4-64, N-(triethylammoniumpropyl)-4-(p-diethylaminophenylhexatrienyl) pyridinium dibromide; TCA, trichloroacetic acid; GFP, green fluorescent protein; BFP, blue fluorescent protein; mRFP, monomeric red fluorescent protein; GSA, glucose-induced selective autophagy; REMI, restriction enzyme-mediated integration; CVT, cytoplasm-to-vacuole targeting; Atg9-PC, Atg9 peripheral compartment; PVS, perivacuolar structure; MIPA, micropexophagy-specific membrane apparatus; PAS, pre-autophagosome structure.

Address correspondence to: William A. Dunn Jr. (dunn{at}ufl.edu).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Abeliovich, H., Zhang, C., Dunn, W. A., Jr., Shokat, K. M., and Klionsky, D. J. ((2003). ). Chemical genetic analysis of Apg1 reveals a non-kinase role in the induction of autophagy. Mol. Biol. Cell 14, , 477-490.[Abstract/Free Full Text]

Ano, Y., Hattori, T., Oku, M., Mukaiyama, H., Baba, M., Ohsumi, Y., Kato, N., and Sakai, Y. ((2005). ). A sorting nexin PpAtg24 regulates vacuolar membrane dynamics during pexophagy via binding to phosphatidylinositol-3-phosphate. Mol. Biol. Cell 16, , 446-457.[Abstract/Free Full Text]

Bevis, B. J., Hammond, A. T., Reinke, C. A., and Glick, B. S. ((2002). ). De novo formation of transitional ER sites and Golgi structures in Pichia pastoris. Nat. Cell Biol. 4, , 750-756.

Campbell, C. L., and Thorsness, P. E. ((1998). ). Escape of mitochondrial DNA to the nucleus in yme1 yeast is mediated by vacuolar-dependent turnover of abnormal mitochondrial compartments. J. Cell Sci. 111, (Pt 16), 2455-2464.[Abstract]

Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. ((2002). ). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99, , 7877-7882.[Abstract/Free Full Text]

Cregg, J. M., Barringer, K. J., Hessler, A. Y., and Madden, K. R. ((1985). ). Pichia pastoris as a host system for transformations. Mol. Cell. Biol. 5, , 3376-3385.[Abstract/Free Full Text]

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

Habibzadegah-Tari, P., and Dunn, W. A., Jr. ((2003). ). Glucose-induced pexophagy in Pichia pastoris. In: Autophagy, ed. D. J. Klionsky, Georgetown, TX: Landes Bioscience.

Hanaoka, H., Noda, T., Shirano, Y., Kato, T., Hayashi, H., Shibata, D., Tabata, S., and Ohsumi, Y. ((2002). ). Leaf senescence and starvation-induced chlorosis are accelerated by the disruption of an Arabidopsis autophagy gene. Plant Physiol. 129, , 1181-1193.[Abstract/Free Full Text]

Huang, W. P., and Klionsky, D. J. ((2002). ). Autophagy in yeast: a review of the molecular machinery. Cell Struct. Funct. 27, , 409-420.[CrossRef][Medline]

Kim, J., Huang, W. P., Strømhaug, P. E., and Klionsky, D. J. ((2002). ). Convergence of multiple autophagy and cytoplasm to vacuole targeting components to a perivacuolar membrane compartment prior to de novo vesicle formation. J. Biol. Chem. 277, , 763-773.[Abstract/Free Full Text]

Kim, J., Kamada, Y., Strømhaug, P. E., Guan, J., Hefner-Gravink, A., Baba, M., Scott, S. V., Ohsumi, Y., Dunn, W. A., and Klionsky, D. J. ((2001). ). Cvt9/gsa9 functions in sequestering selective cytosolic cargo destined for the vacuole. J. Cell Biol. 153, , 381-396.[Abstract/Free Full Text]

Klionsky, D. J. ((2005). ). The molecular machinery of autophagy: unanswered questions. J. Cell Sci. 118, , 7-18.[Abstract/Free Full Text]

Klionsky, D. J., Cregg, J. M., Dunn, W. A., Jr., Emr, S. D., Sakai, Y., Sandoval, I. V., Sibirny, A., Subramani, S., Thumm, M., Veenhuis, M., and Ohsumi, Y. ((2003). ). A unified nomenclature for yeast autophagy-related genes. Dev. Cell 5, , 539-545.[CrossRef][Medline]

Levine, B., and Klionsky, D. J. ((2004). ). Development by self-digestion: molecular mechanisms and biological functions of autophagy. Dev. Cell 6, , 463-477.[CrossRef][Medline]

Mukaiyama, H., Baba, M., Osumi, M., Aoyagi, S., Kato, N., Ohsumi, Y., and Sakai, Y. ((2004). ). Modification of a ubiquitin-like protein Paz2 conducted micropexophagy through formation of a novel membrane structure. Mol. Biol. Cell 15, , 58-70.[Abstract/Free Full Text]

Mukaiyama, H., Oku, M., Baba, M., Samizo, T., Hammond, A. T., Glick, B. S., Kato, N., and Sakai, Y. ((2002). ). Paz2 and 13 other PAZ gene products regulate vacuolar engulfment of peroxisomes during micropexophagy. Genes Cells 7, , 75-90.[Abstract]

Nice, D. C., Sato, T. K., Strømhaug, P. E., Emr, S. D., and Klionsky, D. J. ((2002). ). Cooperative binding of the cytoplasm to vacuole targeting pathway proteins, Cvt13 and Cvt20, to phosphatidylinositol 3-phosphate at the pre-autophagosomal structure is required for selective autophagy. J. Biol. Chem. 277, , 30198-30207.[Abstract/Free Full Text]

Noda, T., Kim, J., Huang, W. P., Baba, M., Tokunaga, C., Ohsumi, Y., and Klionsky, D. J. ((2000). ). Apg9p/Cvt7p is an integral membrane protein required for transport vesicle formation in the Cvt and autophagy pathways. J. Cell Biol. 148, , 465-480.[Abstract/Free Full Text]

Oku, M., Warnecke, D., Noda, T., Muller, F., Heinz, E., Mukaiyama, H., Kato, N., and Sakai, Y. ((2003). ). Peroxisome degradation requires catalytically active sterol glucosyltransferase with a GRAM domain. EMBO J. 22, , 3231-3241.[CrossRef][Medline]

Reggiori, F., Tucker, K. A., Strømhaug, P. E., and Klionsky, D. J. ((2004). ). The Atg1-Atg13 complex regulates Atg9 and Atg23 retrieval transport from the pre-autophagosomal structure. Dev. Cell 6, , 79-90.[CrossRef][Medline]

Sakai, Y., Koller, A., Rangell, L. K., Keller, G. A., and Subramani, S. ((1998). ). Peroxisome degradation by microautophagy in Pichia pastoris: identification of specific steps and morphological intermediates. J. Cell Biol. 141, , 625-636.[Abstract/Free Full Text]

Scott, S. V., Nice, D. C., 3rd, Nau, J. J., Weisman, L. S., Kamada, Y., Keizer-Gunnink, I., Funakoshi, T., Veenhuis, M., Ohsumi, Y., and Klionsky, D. J. ((2000). ). Apg13p and Vac8p are part of a complex of phosphoproteins that are required for cytoplasm to vacuole targeting. J. Biol. Chem. 275, , 25840-25849.[Abstract/Free Full Text]

Sears, I. B., O'Connor, J., Rossanese, O. W., and Glick, B. S. ((1998). ). A versatile set of vectors of constitutive and regulated gene expression in Pichia pastoris. Yeast 14, , 783-790.[CrossRef][Medline]

Stasyk, O. V., van der Klei, I. J., Bellu, A. R., Shen, S., Kiel, J. A., Cregg, J. M., and Veenhuis, M. ((1999). ). A Pichia pastoris VPS15 homologue is required in selective peroxisome autophagy. Curr. Genet. 36, , 262-269.[CrossRef][Medline]

Strømhaug, P. E., Bevan, A., and Dunn, W. A., Jr. ((2001). ). GSA11 encodes a unique 208-kDa protein required for pexophagy and autophagy in Pichia pastoris. J. Biol. Chem. 276, , 42422-42435.

Strømhaug, P. E., and Klionsky, D. J. ((2001). ). Approaching the molecular mechanism of autophagy. Traffic 2, , 524-531.[CrossRef][Medline]

Suzuki, K., Kirisako, T., Kamada, Y., Mizushima, N., Noda, T., and Ohsumi, Y. ((2001). ). The pre-autophagosomal structure organized by concerted functions of APG genes is essential for autophagosome formation. EMBO J 20, , 5971-5981.[CrossRef][Medline]

Tucker, K. A., Reggiori, F., Dunn, W. A., Jr., and Klionsky, D. J. ((2003). ). Atg23 is essential for the Cvt pathway and efficient autophagy but not pexophagy. J. Biol. Chem. 278, , 48445-48452.[Abstract/Free Full Text]

Tuttle, D. L., and Dunn, W. A., Jr. ((1995). ). Divergent modes of autophagy in the methylotrophic yeast Pichia pastoris. J. Cell Sci. 108, , 25-35.[Abstract]

Tuttle, D. L., Lewin, A. S., and Dunn, W. A., Jr. ((1993). ). Selective autophagy of peroxisomes in methylotrophic yeasts. Eur. J. Cell Biol. 60, , 283-290.[Medline]

van Der Wel, H., Morris, H. R., Panico, M., Paxton, T., North, S. J., Dell, A., Thomson, J. M., and West, C. M. ((2001). ). A non-Golgi alpha1,2-fucosyltransferase that modifies Skp1 in the cytoplasm of Dictyostelium. J. Biol. Chem. 276, , 33952-33963.

Veenhuis, M., Douma, A., Harder, W., and Osumi, M. ((1983). ). Degradation and turnover of peroxisomes in the yeast Hansenula polymorpha induced by selective inactivation of peroxisomal enzymes. Arch. Microbiol. 134, , 193-203.[CrossRef][Medline]

Wang, C. W., Kim, J., Huang, W. P., Abeliovich, H., Stromhaug, P. E., Dunn, W. A., Jr., and Klionsky, D. J. ((2001). ). Apg2 is a novel protein required for the cytoplasm to vacuole targeting, autophagy, and pexophagy pathways. J. Biol. Chem. 276, , 30442-30451.[Abstract/Free Full Text]

Wang, C. W., and Klionsky, D. J. ((2003). ). The molecular mechanism of autophagy. Mol. Med. 9, , 65-76.[Medline]

Wang, Y. X., Catlett, N. L., and Weisman, L. S. ((1998). ). Vac8p, a vacuolar protein with armadillo repeats, functions in both vacuole inheritance and protein targeting from the cytoplasm to vacuole. J. Cell Biol. 140, , 1063-1074.[Abstract/Free Full Text]

Yamada, T., Carson, A. R., Caniggia, I., Umebayashi, K., Yoshimori, T., Nakabayashi, K., and Scherer, S. W. ((2005). ). Endothelial nitric-oxide synthase antisense (NOS3AS) gene encodes an autophagy-related protein (APG9-like2) highly expressed in trophoblast. J. Biol. Chem. 280, , 18283-18290.[Abstract/Free Full Text]

Yuan, W., Strømhaug, P. E., and Dunn, W. A., Jr. ((1999). ). Glucose-induced autophagy of peroxisomes in Pichia pastoris requires a unique E1-like protein. Mol. Biol. Cell 10, , 1353-1366.[Abstract/Free Full Text]

Yuan, W., Tuttle, D. L., Shi, Y. J., Ralph, G. S., and Dunn, W. A., Jr. ((1997). ). Glucose-induced microautophagy in Pichia pastoris requires the alpha-subunit of phosphofructokinase. J. Cell Sci. 110, , 1935-1945.[Abstract]

This article has been cited by other articles:

				L. A. Schroder, M. V. Ortiz, and W. A. Dunn Jr. The Membrane Dynamics of Pexophagy Are Influenced by Sar1p in Pichia pastoris Mol. Biol. Cell, November 1, 2008; 19(11): 4888 - 4899. [Abstract] [Full Text] [PDF]

				R. Krick, Y. Muehe, T. Prick, S. Bremer, P. Schlotterhose, E.-L. Eskelinen, J. Millen, D. S. Goldfarb, and M. Thumm Piecemeal Microautophagy of the Nucleus Requires the Core Macroautophagy Genes Mol. Biol. Cell, October 1, 2008; 19(10): 4492 - 4505. [Abstract] [Full Text] [PDF]

				T. Y. Nazarko, A. S. Polupanov, R. R. Manjithaya, S. Subramani, and A. A. Sibirny The Requirement of Sterol Glucoside for Pexophagy in Yeast Is Dependent on the Species and Nature of Peroxisome Inducers Mol. Biol. Cell, January 1, 2007; 18(1): 106 - 118. [Abstract] [Full Text] [PDF]

				A. R. J. Young, E. Y. W. Chan, X. W. Hu, R. Kochl, S. G. Crawshaw, S. High, D. W. Hailey, J. Lippincott-Schwartz, and S. A. Tooze Starvation and ULK1-dependent cycling of mammalian Atg9 between the TGN and endosomes J. Cell Sci., September 15, 2006; 119(18): 3888 - 3900. [Abstract] [Full Text] [PDF]

				F. Reggiori and D. J. Klionsky Atg9 sorting from mitochondria is impaired in early secretion and VFT-complex mutants in Saccharomyces cerevisiae J. Cell Sci., July 15, 2006; 119(14): 2903 - 2911. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: E05-02-0143v1 16/10/4941    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chang, T. Articles by Dunn, W. A. Search for Related Content PubMed PubMed Citation Articles by Chang, T. Articles by Dunn, W. A., Jr.