the role of the endoplasmic reticulum in...

The Role of the Endoplasmic Reticulumin Peroxisome Biogenesis

Lazar Dimitrov1,2, Sheung Kwan Lam1, and Randy Schekman1,2

1Department of Molecular and Cell Biology, University of California, Berkeley, California 947202Howard Hughes Medical Institute, University of California, Berkeley, California 94720

Correspondence: [email protected]

Peroxisomes are essential cellular organelles involved in lipid metabolism. Patients affectedby severe peroxisome biogenesis disorders rarely survive their first year. Genetic screens inseveral model organisms have identified more than 30 PEX genes that are required for theformation of functional peroxisomes. Despite significant work on the PEX genes, the bio-genic origin of peroxisomes remains controversial. For at least two decades, the prevailingmodel postulated that peroxisomes propagate by growth and fission of preexisting peroxi-somes. In this review, we focus on the recent evidence supporting a new, semiautonomousmodel of peroxisomal biogenesis. According to this model, peroxisomal membrane proteins(PMPs) traffic from the endoplasmic reticulum (ER) to the peroxisome bya vesicular budding,targeting, and fusion process while peroxisomal matrix proteins are imported into the or-ganelle by an autonomous, posttranslational mechanism. We highlight the contradictoryconclusions reached to answer the question of how PMPs are inserted into the ER. Wethen review what we know and what still remains to be elucidated about the mechanismof PMP exit from the ER and the contribution of preperoxisomal vesicles to mature peroxi-somes. Finally, we discuss discrepancies in our understanding of de novo peroxisome bio-genesis in wild-type cells. We anticipate that resolving these key issues will lead to a morecomplete picture of peroxisome biogenesis.

The separation of metabolic and biosyntheticfunctions within discrete organelles is a hall-

mark of eukaryotic cells. Peroxisomes are sin-gle-membrane-bound organelles whose mostconserved function is the compartmentaliza-tion of b-oxidation of fatty acids and the break-down of the hydrogen peroxide generated bythis process. The importance of peroxisomesfor normal human development is underscoredby a spectrum of peroxisome biogenesis dis-orders (PBDs) (Kunau 1998; Brosius and Gart-

ner 2002; Steinberg et al. 2006). Patients affect-ed by the most severe of the PBDs—Zellwegersyndrome, lack functional peroxisomes, sufferfrom developmental delay, low muscle tone,hearing impairment, and rarely survive theirfirst year (reviewed by Sacksteder and Gould2000).

The search for genes defective in PBD pa-tients has been greatly aided by genetic screensfor peroxisome biogenesis (pex) mutants inseveral divergent yeasts (Erdmann et al. 1989;

Editors: Susan Ferro-Novick, Tom A. Rapoport, and Randy Schekman

Additional Perspectives on The Endoplasmic Reticulum available at www.cshperspectives.org

Copyright # 2013 Cold Spring Harbor Laboratory Press; all rights reserved; doi: 10.1101/cshperspect.a013243

Cite this article as Cold Spring Harb Perspect Biol 2013;5:a013243

1

on March 12, 2020 - Published by Cold Spring Harbor Laboratory Press http://cshperspectives.cshlp.org/Downloaded from

http://cshperspectives.cshlp.org/

Gould et al. 1992; Liu et al. 1992; Tan et al.1995). Yeast serves as an excellent model systembecause the process of peroxisome biogenesisis evolutionarily very well conserved. Thus, themajority of human genes defective in PBDs werediscovered by searching for the human homo-logs of the PEX genes first cloned in yeast (Sack-steder and Gould 2000).

Despite decades of research on peroxisomebiogenesis, the origin of the organelle remainscontroversial. Classic morphologic work point-ed to an origin of peroxisomal membranes fromthe endoplasmic reticulum (ER) (Novikoff andNovikoff 1972). Subsequently, it was discoveredthat peroxisomal matrix proteins are translatedon free ribosomes and imported posttransla-tionally directly from the cytosol (Goldman andBlobel 1978) (Fig. 1). These proteins contain twovarieties of peroxisomal targeting signals (PTS1or PTS2) that are necessary and sufficient fordirecting soluble proteins to the peroxisomal

matrix (Gould et al. 1987; Swinkels et al. 1991).The analogy with mitochondrial and chloroplastprotein import led to a shift in the scientificthinking toward a “growth and fission” modelin which peroxisomes were considered autono-mous organelles that derive from preexistingones (Lazarow and Fujiki 1985) (Fig. 1). In theirinfluential review, Lazarow and Fujiki’s main ar-gument against the ER origin of peroxisomeswas that classical electron micrographs do notconvincingly show direct lumenal connectionsbetween peroxisomes and the ER (Lazarow andFujiki 1985). Nonetheless, peroxisomal contentcould flow from the ER by a vesicular budding,targeting, and fusion process that would not de-pend on lumenal continuity (Fig. 2).

Another reason why the origin of the per-oxisomal membrane remains difficult to pindown is that peroxisomes or remnants thereofare not detectable in only two, pex19 and pex3,of the more than 30 pex mutants identified thus

Peroxisomal matrix protein traffic

Pex5p/Pex7p

Matrix proteins

Importomer(RING and docking subcomplexes)

Ribosome

ER

PMPPMP

Pex19p

Pex3p

Peroxisome

Sec61p GET complex

Current model for PMP trafficOld model

Figure 1. Protein traffic to the peroxisome. Under the old model (Lazarow and Fujiki 1985), peroxisomalmembrane proteins (PMPs) target to the organelle after translation on cytosolic ribosomes. Recent evidence,however, argues that PMPs are first inserted into the endoplasmic reticulum (ER) via either the Sec61 transloconor the GET (Get1p/Get2p/Get3p) complex. Subsequently, the PMPs exit the ER to reach the peroxisome. Thereis consensus in the literature that peroxisomal matrix proteins target to the organelle after translation oncytosolic ribosomes.

L. Dimitrov et al.

2 Cite this article as Cold Spring Harb Perspect Biol 2013;5:a013243



far in the yeast Saccharomyces cerevisiae (Hoh-feld et al. 1991; Gotte et al. 1998). In mam-malian cells but not in yeast, an additionalgene, PEX16, is required for peroxisomal mem-brane biogenesis (Honsho et al. 1998; Southand Gould 1999). The precise molecular func-tions of the Pex19, Pex3, and Pex16 proteins arenot known. Pex19p is a predominantly cytosolicprotein that binds to the membrane protein tar-geting signal (mPTS) motifs of peroxisomalmembrane proteins (PMPs) (Jones et al. 2004;Rottensteiner et al. 2004) (Fig. 1). Within thecontext of the “growth and fission” model ofperoxisome biogenesis, Pex19p was thought tofacilitate PMP insertion directly into the perox-

isomal membrane as either a PMP receptor orchaperone (Hettema et al. 2000; Snyder et al.2000; Jones et al. 2004). Pex3p is an integralmembrane protein that binds to Pex19p (Gotteet al. 1998) and is proposed to function as aPex19p docking factor in the direct import ofPMPs into the peroxisome (Fang et al. 2004)(Fig. 1). In mammalian cells, PEX16 is proposedto function as a receptor for Pex3p-Pex19p com-plexes at the peroxisomal membrane (Matsu-zaki and Fujiki 2008).

In this review we focus on the recent litera-ture that supports the resurgent model of per-oxisome biogenesis from the ER. For a moredetailed historical account of the vicissitudes

Peroxisome

Peroxisome Fission

Growth

PPV (preperoxisomal vesicle)

Budding (requires ATP, Pex19p,and unknown factors)

PMP

Peroxisome

ER

Figure 2. Contribution of the endoplasmic reticulum to peroxisome biogenesis. Peroxisomal membrane pro-teins (PMPs) exit the endoplasmic reticulum (ER) in preperoxisomal vesicles (PPVs), which target to and fusewith mature peroxisomes. Apart from Pex19p, the cytosolic factors required for PPV budding are unknown.PPVs contribute to peroxisomal growth by delivering new lipids and PMPs. Subsequent fission of matureperoxisomes maintains organelle numbers during each cell division. The existence of a retrograde traffickingpathway from peroxisomes to the ER is an open question. Moreover, the extent to which PPVs can fuse with eachother to generate a new peroxisome in wild-type cells appears to differ between yeast and mammalian cells (Kimet al. 2006; Motley and Hettema 2007; van der Zand et al. 2012).

The ER and Peroxisome Biogenesis

Cite this article as Cold Spring Harb Perspect Biol 2013;5:a013243 3



in models about peroxisome biogenesis, includ-ing less recent findings suggesting an ER originof the organelle, we refer the reader to excellentreviews elsewhere (Kunau and Erdmann 1998;Titorenko and Rachubinski 1998; Tabak et al.2008). The import of peroxisomal matrix pro-teins and the inheritance of the organelle havebeen reviewed elsewhere as well (Fagarasanu etal. 2010; Schliebs et al. 2010).

THE ER ORIGIN OF PEROXISOMES

The “growth and fission” model prevailed forthe two decades since Lazarow and Fujiki’s re-view (Lazarow and Fujiki 1985) despite accu-mulating evidence that peroxisomes can arisede novo. In mutants that lack detectable per-oxisomes, the organelle is regenerated on rein-troduction of the wild-type version of the mu-tated gene (reviewed by Subramani 1998). Suchde novo biogenesis has not been observed forautonomous organelles such as mitochondriaand chloroplasts.

The ER origin of peroxisomes received itsmost direct support in a paper from the Tabakgroup (Hoepfner et al. 2005). The authors de-veloped a pulse-chase assay in which PEX3-YFPis expressed from the galactose-inducible pro-moter as the sole Pex3p source in cells other-wise devoid of peroxisomes. A 30-min galactosepulse followed by a glucose chase results inthe one-time expression of Pex3p-YFP to levelscomparable to those achieved from the endog-enous PEX3 promoter. These physiological lev-els of Pex3p are sufficient to regenerate import-competent peroxisomes. Moreover, time-lapsemicroscopy reveals that Pex3p-YFP first con-centrates into foci on the ER membrane andthen migrates in a Pex19p-dependent mannerto peroxisomes. This evidence argues stronglythat the intracellular traffic of at least one PMP,Pex3p, proceeds from the ER to peroxisomes.Two independent groups came to similar con-clusions about the trafficking of Pex3p to per-oxisomes (Kragt et al. 2005; Tam et al. 2005). In2010, the Tabak group extended their fluores-cence pulse-chase analysis to many other PMPsand concluded that most, if not all, S. cerevisiaePMPs traffic to peroxisomes via the ER in a

Pex3p- and Pex19p-dependent manner (vander Zand et al. 2010).

Evidence in favor of the ER to peroxisomepathway in mammalian cells focused on thePex16 protein (Kim et al. 2006). The authorsfound that Pex16p localizes to peroxisomes inCOS-7 cells at low expression levels. However,on Pex16p overexpression, the predominant lo-calization of the protein is to the ER. Presum-ably, the machinery for Pex16p traffic from theER to peroxisomes is saturated at higher expres-sion levels. Using a photoactivatable version ofGFP to tag PEX16, the authors show convinc-ingly that the ER localization of Pex16p repre-sents an intermediate in the delivery of Pex16pto peroxisomes. This delivery is dependent onthe mPTS of Pex16p located in the protein’samino terminus. Moreover, the authors appendto Pex16p an amino-terminal type I signal an-chor sequence to force Pex16p to be cotransla-tionally inserted into the ER. In this case, theprotein still targets to peroxisomes and comple-ments PEX16-deficient human cells.

HOW ARE PEROXISOMAL MEMBRANEPROTEINS INSERTED INTO THE ER?

The majority of proteins targeted to the ER co-or posttranslationally are substrates of the Sec61translocon (Osborne et al. 2005). If PMPs traf-fic to peroxisomes via the ER, then what is themechanism for their insertion into the ER mem-brane? The literature on this subject has beensurprisingly controversial as well. The first pub-lished study on the requirement of the Sec61translocon for peroxisome biogenesis was a neg-ative result but appeared during the reign of the“growth and fission” model (South et al. 2001).The authors used a cold-sensitive sec61-11 allelethat blocks protein translocation into the ERonly 15 min after a shift to 178C. They followedde novo peroxisome biogenesis by expressingPex3p under the galactose-inducible promoter.However, because they did not perform a glu-cose chase, a caveat to their study is that Pex3p isoverexpressed. Unlike the more recent Pex3ppulse-chase experiments (Hoepfner et al. 2005;van der Zand et al. 2010; Thoms et al. 2012),South et al. induced Pex3p expression for about

L. Dimitrov et al.




20 h before they saw import-competent peroxi-somes in about 30% of wild-type cells. Theyattributed this long period of peroxisome for-mation to the inhibitory effects of Pex3p over-expression. Despite these shortcomings, de novoperoxisome biogenesis occurred at the same rateand to the same extent in wild-type and sec61-11 cells at the restrictive temperature (Southet al. 2001).

A much more recent report revisited therequirement of the Sec61 translocon for per-oxisome biogenesis and came to exactly the op-posite conclusion (Thoms et al. 2012). The ex-perimental setup, however, differs in two keyaspects. First, the authors used a different con-ditional allele—sec61-2, which is temperaturesensitive but unfortunately phenotypic fortranslocation into the ER even at the permissivetemperature, raising the possibility that thestrain may have accumulated second-site mu-tations. Second, the Pex3p expression level ismore physiologically relevant because Pex3p isinduced in galactose for only 45 min and sub-sequently chased with growth on glucose. Theauthors follow peroxisome regeneration for thefirst 4 h after galactose induction. At these phys-iological levels of Pex3p reexpression, about80% of wild-type cells regenerate peroxisomes.In contrast, less than 5% of sec61-2 cells, grownat the permissive temperature, regenerate per-oxisomes in the same time window. Despite thisdefect in de novo peroxisome biogenesis, how-ever, at steady state sec61-2 cells grown at thepermissive temperature contain similar num-bers of import-competent peroxisomes as wild-type cells. This result suggests that the sec61-2allele does not block but instead delays de novoperoxisome biogenesis.

Because the experimental setups of thesetwo studies differ in several aspects, it is difficultto determine which one is responsible for thedrastic difference in their conclusions. Addi-tional experiments, which use several SEC61alleles at their restrictive temperatures in thecontext of the more physiologically relevantpulse-chase assay, should resolve the controversyabout the requirement of the Sec61 transloconin peroxisomes biogenesis. In the meantime,the results from a third independent study are

consistent with a requirement for the Sec61translocon in peroxisome biogenesis (van derZand et al. 2010). These authors drove the ex-pression of Sec62p and Sec63p from the methi-onine-repressible MET3 promoter. Cells withthese alleles were grown in methionine-repletemedium for 7 h to achieve partial depletionof Sec62p and Sec63p but not Sec61p. Micro-scopic analysis revealed the accumulation ofcytosolic Pex13p-YFP even by 6 h of promoterrepression. These results were confirmed bio-chemically by cellular fractionation of pex3Dcells to avoid the problem of leaky peroxisomes.Before SEC62 and SEC63 repression, Pex13pand Pex14p were exclusively found in the mem-brane fraction. Seven hours after the simultane-ous repression of Sec62p and Sec63p, the threePMPs were also detectable in the cytosolic frac-tion. Thus, when we take all the evidence intoaccount, the scales are tipping toward the viewthat the Sec61 translocon is required for PMPinsertion into the ER (Fig. 1).

In contrast to the conflicting results aboutthe Sec61 translocon, there is consensus in theliterature about the trafficking of the only tail-anchored PMP, Pex15p. Tail-anchored (TA)proteins targeted to the ER are recognized inthe cytosol by Get3p and the Get3p-TA complexis recruited to the ER membrane by the hetero-meric Get1p/Get2p receptor (Schuldiner et al.2008) (Fig. 1). In their seminal report on thediscovery of the GET complex, the authorsshowed that Get3p physically interacts withPex15p in a yeast two-hybrid assay (Schuldineret al. 2008). This interaction depends on thecarboxy-terminal transmembrane domain ofPex15p. Moreover, in get1D or get2D mutantcells, Pex15p initially forms cytosolic aggre-gates. Extended overexpression leads to Pex15pmislocalization to mitochondria. The lag peri-od to mitochondrial mislocalization of Pex15pis dramatically shortened in get3D cells. Ina later report, the Tabak group performed afluorescent pulse-chase experiment to followthe traffic of YFP-Pex15p (van der Zand et al.2010). They also found that Pex15p traffic isimpaired in get3D mutant cells. Thus, the GETcomplex is required for targeting Pex15p to theperoxisome via the ER.





HOW DO PEROXISOMAL MEMBRANEPROTEINS LEAVE THE ER?

The observation that newly made Pex3p-GFPmigrates from ER foci to peroxisomes (Hoepf-ner et al. 2005) raises the intriguing possibilitythat PMPs leave the ER in membrane-boundvesicles. Using S. cerevisiae as a model system,our laboratory was the first to establish a cell-free budding reaction that reproduces the for-mation of preperoxisomal vesicles (PPVs) (Fig.2) responsible for the transit of at least twoPMPs—Pex3p and Pex15p, from the ER (Lamet al. 2010). When cytosol from wild-type cellsis incubated with microsomal membranes frompex19D cells, which lack detectable peroxi-somes, Pex3p and Pex15p are incorporatedinto a slowly sedimenting vesicle fraction. Thisprocess is ATP-dependent but independent ofthe small GTPase Sar1, which initiates the as-sembly of the COPII machinery responsible forsecretory cargo packaging at the ER (Box 1)(Zanetti et al. 2012). The PPV budding reactionalso requires Pex19p in the cytosol fraction butrecombinant Pex19p is not sufficient. Thus,PPV production requires additional cytosolicfactors, which are likely to be proteins becausethe cytosolic activity is heat- and trypsin-sensi-tive (Fig. 2). However, other known Pex proteinswith cytosolic localization (Pex1p, Pex5p,Pex6p, and Pex7p) are not required in the cyto-sol for PPV budding. An independent studyfrom the Subramani group in permeabilizedPichia pastoris cells reached remarkably similarconclusions despite the different model organ-ism, source of donor membranes, PMPs inves-tigated and reaction conditions (Agrawal et al.

2011). They also found that Pex11p-containingvesicles bud even when pex3D permeabilizedcells are used as donor membranes. The resul-tant vesicles, however, lack most other PMPstested.

If PMPs bud from the ER in membrane-bound vesicles, then the next pressing ques-tion concerns the identity of the other cytosolicproteins responsible for PPV formation (Fig. 2).The in vitro budding reactions rule out the in-volvement of the canonical COPII coat (Box 1).Moreover, two independent studies in humancell lines concluded that the COPII and COPIcoats (Box 1) are not required for peroxisomalbiogenesis (South et al. 2000; Voorn-Brouweret al. 2001). Both studies used dominant nega-tive alleles of Sar1p and brefeldin A (BFA) treat-ment (Box 1) to rule out the requirement of thecanonical COPII and COPI coats, respectively,for PMP targeting to peroxisomes. On the otherhand, a study in the yeast Hansenula polymorphafound that BFA treatment leads to the reversibleaccumulation of newly synthesized PMPs andperoxisomal matrix proteins in the ER, but notto a complete block in peroxisome biogenesis(Salomons et al. 1997). Thus, there could besome species-specific differences in the involve-ment of COPI in peroxisome biogenesis. In thisregard, COPI has also been proposed to play arole in peroxisome fission because purified ratliver peroxisomes recruit COPI in GTPgS-de-pendent manner and a Chinese hamster ova-ry cell line with a temperature-sensitive alleleof 1-COP develops elongated, tubular peroxi-somes at the restrictive temperature (Passreiteret al. 1998). In summary, the identity of theadditional soluble proteins and the putative

BOX 1. DEFINITIONS

COPII Coat protein complex II responsible for the formation of vesicles involved in anterogradecargo transport from the endoplasmic reticulum (ER) to the Golgi apparatus.

COPI Coat protein complex I responsible for the formation of vesicles involved in retrogradetransport of resident proteins among Golgi cisternae and in the retrieval of ERproteins from the cis-Golgi cisterna.

Sar1 A small Arf-like GTPase that initiates the assembly of the COPII coat.Brefeldin A

(BFA)Inhibitor of a GTPexchange factor responsible for the activation of Arf1, a small GTPase

that initiates the assembly of the COPI coat.

L. Dimitrov et al.




ATPase required in the PPV budding reactionremain unknown. These proteins may havebeen missed in prior genetic screens because ofredundancy or their essential function in othercellular processes. The discovery and character-ization of these additional cytosolic factors re-mains an active area of research.

In mammalian cells, a clue to the identity ofthe machinery responsible for Pex16p exit fromthe ER comes from a recent study that impli-cates the Sec16B isoform in peroxisome bio-genesis (Yonekawa et al. 2011). Knock-downof Sec16B but not Sec16A leads to ER localiza-tion for Pex16p. Moreover, a fluorescent pulse-chase analysis revealed that Sec16B is requiredfor the delivery of Pex16p from the ER to per-oxisomes. Both Sec16 isoforms are involvedin ER exit site formation (Bhattacharyya andGlick 2007). Truncated Sec16B mutants, how-ever, showed that ER exit site localization is sep-arable from the protein’s role in Pex16p traffic.S. cerevisiae has a single SEC16 gene that is theortholog of the longer Sec16A mammalian iso-form. Thus, it is unclear if yeast contains a func-tional ortholog of the SEC16B gene. Identifyingproteins that bind specifically to Sec16B mayreveal other players in mammalian peroxisomebiogenesis from the ER.

We finish this section by mentioning twoadditional outstanding questions related to ve-sicular traffic between the ER and peroxisomes.First, how are PMPs sorted away from ER-resi-dent and classical secretory pathway proteinsduring their traffic to the peroxisome? The cur-rent literature provides no clues to answer thisquestion. Thus, genetic screens for mutants inwhich an ER-resident or secretory protein mis-localizes to peroxisomes may be necessary. Sec-ond, is there a retrograde pathway from per-oxisomes back to the ER, given the precedentfor such a pathway from the Golgi apparatus tothe ER (Fig. 2)? One proposal for a retrogradepathway from peroxisomes to the ER mediatedby COPI-coated vesicles is based on results thathave not been followed up since 1998 (Passreiteret al. 1998; Titorenko and Rachubinski 1998).Testing this hypothesis as well as the existence ofa retrograde pathway should be another focus offuture research.

HOW DO PREPEROXISOMAL VESICLESCONTRIBUTE TO MATURE PEROXISOMES?

With the exception of Salomons et al.’s work onHansenula polymorpha cells treated with BFA(Salomons et al. 1997), we are not aware of anyother report in which soluble peroxisomal ma-trix enzymes are localized to the ER. Thus, afunctional peroxisomal importomer—the trans-locon responsible for import of soluble PTS1-or PTS2-containing proteins into the peroxi-somal matrix, must not assemble in the ERmembrane. The importomer consists of twohalves—the docking subcomplex, which con-tains the PMPs Pex13p and Pex14p, and theRING subcomplex, which contains the PMPsPex2p, Pex10p, and Pex12p (Agne et al. 2003)(Fig. 1). Given our thinking that all PMPs trafficto the peroxisome via the ER, how is importo-mer assembly prevented in the ER membrane?

Progress toward answering this questioncomes from the recent demonstration thatmature, import-competent peroxisomes areformed by the heterotypic fusion of at least twomolecularly distinct PPVs (van der Zand et al.2012). The authors used a split-GFP comple-mentation assay (Hu et al. 2002) to follow thegeneration of newly synthesized importomercomplexes. The nonfluorescent amino- and car-boxy-terminal halves of Venus fluorescent pro-tein (VFP) were each fused to different PMPsin haploid cells that were subsequently mated.Fluorescence complementation in the resultingdiploid zygote indicates physical interaction be-tween the two PMPs. When wild-type haploidcells were mated, fluorescence complementationwas achieved between tagged members with-in the docking subcomplex (Pex13p-Pex14p),within the RING subcomplex (Pex2p-Pex10p),and across the twosubcomplexes (Pex2p-Pex14pand Pex10p-Pex13p). Remarkably, when pex3Dor pex19D haploid cells were mated, fluores-cence complementation was seen only betweenproteins within a subcomplex (Pex13p-Pex14pand Pex2p-Pex10p) but not for proteins acrossthe subcomplexes (Pex2p-Pex14p and Pex10p-Pex13p). The authors argue that the reconsti-tuted docking and RING subcomplexes are inthe ER membrane based on their prior pulse-





chase studies (van der Zand et al. 2010). How-ever, direct evidence for subcomplex formationin the ER membrane requires the demonstra-tion that the reconstituted VFP signal colocal-izes with an ER fluorescent marker of differentcolor. Despite this shortcoming, the evidencesuggests that the docking and RING subcom-plexes can assemble in the ER membrane with-out reconstituting a functional importomer inpex3D or pex19D cells.

Another remarkable finding is that a similarlack of VFP complementation across subcom-plex members occurs in mating pex1D or pex6Dcells. Prior studies in Yarrowia lipolytica im-plicate the NSF-like (AAAþ) proteins Pex1pand Pex6p in the heterotypic fusion of purifiedPPVs leading to the proposal that differentPPVs fuse to form mature peroxisomes (Titor-enko and Rachubinski 2000; Titorenko et al.2000). This model receives further support inthe more recent study (van der Zand et al. 2012)in which the authors provide fluorescence lo-calization evidence that the docking and RINGsubcomplexes are kept in distinct PPVs. Com-partments harboring the subcomplexes haveto fuse to reconstitute a functional importomerand generate new mature peroxisomes. Hetero-typic fusion between PPVs and preexisting per-oxisomes and between mature peroxisomes wasnot detected (van der Zand et al. 2012).

Whereas the intriguing finding that thedocking and RING subcomplexes leave the ERin molecularly distinct PPVs is a major step to-ward understanding why peroxisomal matrixproteins are not observed in the ER lumen, theoriginal question remains: What is the mecha-nism for preventing importomer reconstitutionin the ER membrane? One possibility is that thedocking and RING subcomplexes need to pair ina trans rather than cis orientation to reconstitutea functional importomer by analogy to the func-tional association of SNARE proteins in trans topromote membrane fusion. An alternative pos-sibility is that an inhibitory factor localized tothe ER membrane prevents importomer recon-stitution. Yet another possibility is that differentER subdomains are required for the budding ofthe distinct PPVs. In this case, one must ask howthe different ER subdomains are segregated, how

PMPs are sorted into the different PPVs, andwhich SNARE proteins, if any, specify the het-erotypic fusion event.

DOES DE NOVO PEROXISOME BIOGENESISOCCUR IN WILD-TYPE CELLS?

The latest finding that PPVs fuse to form newperoxisomes that add to the preexisting pool oforganelles (van der Zand et al. 2012) is hardto reconcile with the prior evidence that denovo peroxisome biogenesis occurs only in cellsdevoid of peroxisomes (Motley and Hettema2007). Both studies were conducted in S. cere-visiae but use slightly different experimental set-ups that are described in detail in Table 1. Theearlier report convincingly showed that the ma-jority of peroxisomes in wild-type S. cerevisiaecells are generated from preexisting peroxi-somes (Motley and Hettema 2007), althougha low level of de novo biogenesis may have es-caped their detection. Moreover, they show thatPex3p-GFP traffics from the ER to preexistingperoxisomes in newly formed zygotes hetero-zygous for pex19D (Table 1, fourth column).This result contradicts the finding that Pex3pis on both types of PPVs, which do not fusewith preexisting peroxisomes but instead fuseto form new peroxisomes (van der Zand et al.2012). One possibility for these discrepant re-sults is that the latest study follows new per-oxisome formation, which is assayed by recon-stitution of the split-VFP signal, every 24 hfor 72 h. This time frame is much longer thanthe five hours necessary for de novo peroxisomebiogenesis in cells devoid of peroxisomes in allprior pulse-chase studies from the same group(Hoepfner et al. 2005; van der Zand et al. 2010).

In contrast to S. cerevisiae, mammalian cellsappear to prefer de novo formation of peroxi-somes (Kim et al. 2006). The authors develop anassay based on a photoactivatable GFP-PTS1fusion that allows them to distinguish betweenpreexisting and newly generated peroxisomes.In their assay, de novo peroxisome biogenesiscontributes the majority of new cellular perox-isomes (73%), while fission of preexisting per-oxisomes accounts for the remainder. Thus, itis possible that there are species-specific differ-

L. Dimitrov et al.




ences in the regulation of the mix between denovo biogenesis and fission of peroxisomes inwild-type cells. Yeast appears to have dialed therheostat so that fission of preexisting organellesdominates peroxisome maintenance at least un-der the growth conditions tested (Motley andHettema 2007). Mammalian cells, on the otherhand, appear to favor de novo peroxisome for-mation from the ER.

CONCLUDING REMARKS

We conclude that a model of peroxisome bio-genesis, consistent with most of the existingdata, calls for a semiautonomous origin of theorganelle. In this model, peroxisomal mem-brane proteins are delivered by vesicular traffic

from the ER and matrix proteins are translo-cated posttranslationally directly from the cyto-plasm (Figs. 1 and 2). This model is also oneof “growth and fission” (Fig. 2). However, the“growth” component has a dual origin—an en-domembrane (PMPs and membrane lipids)and an autonomous one (peroxisomal matrixproteins). This model is surprisingly similar tothe one discussed by Gunter Blobel as far backas 1978 (Goldman and Blobel 1978). Given thependulum swings in scientific opinion betweenan endomembrane versus autonomous originof the peroxisome, it seems possible that thefinal model could change again.

We now briefly draw attention to some ev-idence in the literature that is hard to reconcilewith the ER origin of peroxisomal membrane

Table 1. Summary of experiments assessing the contribution of de novo peroxisome biogenesisin wild-type cells

References

(Motley and

Hettema 2007)

(Motley and

Hettema 2007)

(Motley and

Hettema 2007)

(van der Zand

et al. 2012)

Figure 4B 4C 5B 6BUse of mating

assay?No Yes Yes Yes

Strain genotype Wild type (WT) WT � pex3D WT � pex19D WT � WTLabel for

preexistingperoxisomes

3 h galactose pulseof GFP-PTS1 and2 h glucose chase

3 h galactose pulse ofGFP-PTS1 and 5 hglucose chase inWTmating partner

3 h galactose pulse ofHcRed-PTS1 and 2h glucose chase inWTmating partner

1.5 h galactose pulseof CFP-PTS1 and3 h glucose chase

Label for newperoxisomes

Constitutive HcRed-PTS1 in the samestrain

HcRed-PTS1 inpex3D matingpartner

3 h galactose pulse ofPex3p-GFP and 2 hglucose chase inpex19D matingpartner

Reconstitution ofsplit-VFP signalfrom interactingPMPs

Time allowedfor de novobiogenesis

6–8 h in glucosebeyond the 2 hchase

5 h of glucose chase 2 h of glucose chaseand 2 h aftermating

Up to 72 h; figureshows 48 h timepoint

Results No red-onlyperoxisomesdetected

No red-onlyperoxisomesdetected

Most of the punctatePex3p-GFP signalco-localizes withthe redperoxisomes

Reconstituted split-VFP signal doesnot localize withCFP-PTS1 signal

Conclusions No peroxisomesform de novo inWT cells

No peroxisomes formde novo in WT �pex3Dheterozygousdiploid cells

ER-localized Pex3p-GFP can betransported topreexistingperoxisomes

PPVs fuse andmature to newperoxisomes





proteins. First, PMPs accumulate in mitochon-dria in fibroblasts from PBD patients withPEX3, PEX19, or PEX16 mutations (Sackstederet al. 2000; South et al. 2000). Moreover, denovo peroxisome biogenesis can occur in S.cerevisiae when the sole copy of Pex3p is tar-geted to the mitochondrial outer membrane(Rucktaschel et al. 2010). Is there a mitochon-dria-to-peroxisome pathway or are these obser-vations an artifact of protein overexpressionor other off-pathway effects? A mitochondria-to-peroxisome pathway has been suggestedby studies in mammalian cells (Neuspiel et al.2008; Braschi et al. 2010). Second, how do wereconcile the ER-to-peroxisome pathway withthe old reports of in vitro PMP import directlyinto purified peroxisomes (Diestelkotter andJust 1993; Imanaka et al. 1996). Are those resultsplagued by the fragility of isolated peroxisomes(reviewed by Tabak et al. 2008), by contamina-tion from other organelles or is there an alter-native cytosol-to-peroxisome pathway? If PMPscan traffic directly to peroxisomes and there isno need to invoke a vesicular carrier for lipiddelivery from the ER to the peroxisome (Ray-chaudhuri and Prinz 2008) then how do wereenvision the role of PPVs? Every field has its“dirty little secrets” (Yewdell 2005) and ours isno exception. Therefore, much remains to beclarified before a consistent and unified modelof peroxisome biogenesis can be advanced.

ACKNOWLEDGMENTS

We apologize to the authors whose work couldnot be incorporated in this review because ofspace limitations. We thank Jennie Dormanfor critical reading of the manuscript and dis-cussions. L.D. is supported as a research associ-ate by the Howard Hughes Medical Institute.S.K.L. is a Human Frontier Science Programpostdoctoral fellow. R.S. is a Senior Fellow ofthe UC Berkeley Miller Institute and is sup-ported as an Investigator of the Howard HughesMedical Institute.

REFERENCES

Agne B, Meindl NM, Niederhoff K, Einwachter H, Rehling P,Sickmann A, Meyer HE, Girzalsky W, Kunau WH. 2003.

Pex8p: An intraperoxisomal organizer of the peroxisomalimport machinery. Mol Cell 11: 635–646.

Agrawal G, Joshi S, Subramani S. 2011. Cell-free sorting ofperoxisomal membrane proteins from the endoplasmicreticulum. Proc Natl Acad Sci 108: 9113–9118.

Bhattacharyya D, Glick BS. 2007. Two mammalian Sec16homologues have nonredundant functions in endoplas-mic reticulum (ER) export and transitional ER organiza-tion. Mol Biol Cell 18: 839–849.

Braschi E, Goyon V, Zunino R, Mohanty A, Xu L,McBride HM. 2010. Vps35 mediates vesicle transportbetween the mitochondria and peroxisomes. Curr Biol20: 1310–1315.

Brosius U, Gartner J. 2002. Cellular and molecular aspects ofZellweger syndrome and other peroxisome biogenesisdisorders. Cell Mol Life Sci 59: 1058–1069.

Diestelkotter P, Just WW. 1993. In vitro insertion of the 22-kD peroxisomal membrane protein into isolated rat liverperoxisomes. J Cell Biol 123: 1717–1725.

Erdmann R, Veenhuis M, Mertens D, Kunau WH. 1989.Isolation of peroxisome-deficient mutants of Saccharo-myces cerevisiae. Proc Natl Acad Sci 86: 5419–5423.

Fagarasanu A, Mast FD, Knoblach B, Rachubinski RA. 2010.Molecular mechanisms of organelle inheritance: Lessonsfrom peroxisomes in yeast. Nat Rev Mol Cell Biol 11:644–654.

Fang Y, Morrell JC, Jones JM, Gould SJ. 2004. PEX3 func-tions as a PEX19 docking factor in the import of class Iperoxisomal membrane proteins. J Cell Biol 164: 863–875.

Goldman BM, Blobel G. 1978. Biogenesis of peroxisomes:Intracellular site of synthesis of catalase and uricase. ProcNatl Acad Sci 75: 5066–5070.

Gotte K, Girzalsky W, Linkert M, Baumgart E, Kammerer S,Kunau WH, Erdmann R. 1998. Pex19p, a farnesylatedprotein essential for peroxisome biogenesis. Mol CellBiol 18: 616–628.

Gould SG, Keller GA, Subramani S. 1987. Identification of aperoxisomal targeting signal at the carboxy terminus offirefly luciferase. J Cell Biol 105: 2923–2931.

Gould SJ, McCollum D, Spong AP, Heyman JA, Sub-ramani S. 1992. Development of the yeast Pichia pastorisas a model organism for a genetic and molecular analysisof peroxisome assembly. Yeast 8: 613–628.

Hettema EH, Girzalsky W, van den Berg M, Erdmann R,Distel B. 2000. Saccharomyces cerevisiae pex3p andpex19p are required for proper localization and stabilityof peroxisomal membrane proteins. EMBO J 19: 223–233.

Hoepfner D, Schildknegt D, Braakman I, Philippsen P,Tabak HF. 2005. Contribution of the endoplasmic retic-ulum to peroxisome formation. Cell 122: 85–95.

Hohfeld J, Veenhuis M, Kunau WH. 1991. PAS3, a Saccha-romyces cerevisiae gene encoding a peroxisomal integralmembrane protein essential for peroxisome biogenesis.J Cell Biol 114: 1167–1178.

Honsho M, Tamura S, Shimozawa N, Suzuki Y, Kondo N,Fujiki Y. 1998. Mutation in PEX16 is causal in the perox-isome-deficient Zellweger syndrome of complementa-tion group D. Am J Hum Genet 63: 1622–1630.

L. Dimitrov et al.




Hu C.-D, Chinenov Y, Kerppola TK. 2002. Visualization ofinteractions among bZIPand Rel family proteins in livingcells using bimolecular fluorescence complementation.Mol Cell 9: 789–798.

Imanaka T, Shiina Y, Takano T, Hashimoto T, Osumi T. 1996.Insertion of the 70-kDa peroxisomal membrane proteininto peroxisomal membranes in vivo and in vitro. J BiolChem 271: 3706–3713.

Jones JM, Morrell JC, Gould SJ. 2004. PEX19 is a predom-inantly cytosolic chaperone and import receptor for class1 peroxisomal membrane proteins. J Cell Biol 164: 57–67.

Kim PK, Mullen RT, Schumann U, Lippincott-Schwartz J.2006. The origin and maintenance of mammalian perox-isomes involves a de novo PEX16-dependent pathwayfrom the ER. J Cell Biol 173: 521–532.

Kragt A, Voorn-Brouwer T, van den Berg M, Distel B. 2005.Endoplasmic reticulum-directed Pex3p routes to perox-isomes and restores peroxisome formation in a Saccha-romyces cerevisiae pex3Delta strain. J Biol Chem 280:34350–34357.

Kunau WH. 1998. Peroxisome biogenesis: From yeast toman. Curr Opin Microbiol 1: 232–237.

Kunau WH, Erdmann R. 1998. Peroxisome biogenesis: Backto the endoplasmic reticulum? Curr Biol 8: R299–302.

Lam SK, Yoda N, Schekman R. 2010. A vesicle carrier thatmediates peroxisome protein traffic from the endoplas-mic reticulum. Proc Natl Acad Sci 107: 21523–21528.

Lazarow PB, Fujiki Y. 1985. Biogenesis of peroxisomes. AnnuRev Cell Biol 1: 489–530.

Liu H, Tan X, Veenhuis M, McCollum D, Cregg JM. 1992.An efficient screen for peroxisome-deficient mutants ofPichia pastoris. J Bacteriol 174: 4943–4951.

Matsuzaki T, Fujiki Y. 2008. The peroxisomal membraneprotein import receptor Pex3p is directly transported toperoxisomes by a novel Pex19p- and Pex16p-dependentpathway. J Cell Biol 183: 1275–1286.

Motley AM, Hettema EH. 2007. Yeast peroxisomes multiplyby growth and division. J Cell Biol 178: 399–410.

Neuspiel M, Schauss AC, Braschi E, Zunino R, Rippstein P,Rachubinski RA, Andrade-Navarro MA, McBride HM.2008. Cargo-selected transport from the mitochondria toperoxisomes is mediated by vesicular carriers. Curr Biol18: 102–108.

Novikoff PM, Novikoff AB. 1972. Peroxisomes in absorptivecells of mammalian small intestine. J Cell Biol 53: 532–560.

Osborne AR, Rapoport TA, van den Berg B. 2005. Proteintranslocation by the Sec61/SecY channel. Annu Rev CellDev Biol 21: 529–550.

Passreiter M, Anton M, Lay D, Frank R, Harter C,Wieland FT, Gorgas K, Just W. 1998. Peroxisome biogen-esis: Involvement of ARF and coatomer. J Cell Biol 141:373–383.

Raychaudhuri S, Prinz WA. 2008. Nonvesicular phospholip-id transfer between peroxisomes and the endoplasmicreticulum. Proc Natl Acad Sci 105: 15785–15790.

Rottensteiner H, Kramer A, Lorenzen S, Stein K, Landgraf C,Volkmer-Engert R, Erdmann R. 2004. Peroxisomal mem-brane proteins contain common Pex19p-binding sitesthat are an integral part of their targeting signals. MolBiol Cell 15: 3406–3417.

Rucktaschel R, Halbach A, Girzalsky W, Rottensteiner H,Erdmann R. 2010. De novo synthesis of peroxisomesupon mitochondrial targeting of Pex3p. Eur J Cell Biol89: 947–954.

Sacksteder KA, Gould SJ. 2000. The genetics of peroxisomebiogenesis. Annu Rev Genet 34: 623–652.

Sacksteder KA, Jones JM, South ST, Li X, Liu Y, Gould SJ.2000. PEX19 binds multiple peroxisomal membrane pro-teins, is predominantly cytoplasmic, and is required forperoxisome membrane synthesis. J Cell Biol 148: 931–944.

Salomons FA, van der Klei IJ, Kram AM, Harder W,Veenhuis M. 1997. Brefeldin A interferes with peroxisom-al protein sorting in the yeast Hansenula polymorpha.FEBS Lett 411: 133–139.

Schliebs W, Girzalsky W, Erdmann R. 2010. Peroxisomalprotein import and ERAD: Variations on a commontheme. Nat Rev Mol Cell Biol 11: 885–890.

Schuldiner M, Metz J, Schmid V, Denic V, Rakwalska M,Schmitt HD, Schwappach B, Weissman JS. 2008. TheGET complex mediates insertion of tail-anchored pro-teins into the ER membrane. Cell 134: 634–645.

Snyder WB, Koller A, Choy AJ, Subramani S. 2000. Theperoxin Pex19p interacts with multiple, integral mem-brane proteins at the peroxisomal membrane. J Cell Biol149: 1171–1178.

South ST, Gould SJ. 1999. Peroxisome synthesis in the ab-sence of preexisting peroxisomes. J Cell Biol 144: 255–266.

South ST, Sacksteder KA, Li X, Liu Y, Gould SJ. 2000. Inhib-itors of COPI and COPII do not block PEX3-mediatedperoxisome synthesis. J Cell Biol 149: 1345–1360.

South ST, Baumgart E, Gould SJ. 2001. Inactivation ofthe endoplasmic reticulum protein translocation factor,Sec61p, or its homolog, Ssh1p, does not affect peroxi-some biogenesis. Proc Natl Acad Sci 98: 12027–12031.

Steinberg SJ, Dodt G, Raymond GV, Braverman NE,Moser AB, Moser HW. 2006. Peroxisome biogenesis dis-orders. Biochim Biophys Acta 1763: 1733–1748.

Subramani S. 1998. Components involved in peroxisomeimport, biogenesis, proliferation, turnover, and move-ment. Physiol Rev 78: 171–188.

Swinkels BW, Gould SJ, Bodnar AG, Rachubinski RA,Subramani S. 1991. A novel, cleavable peroxisomal tar-geting signal at the amino-terminus of the rat 3-ketoacyl-CoA thiolase. EMBO J 10: 3255–3262.

Tabak HF, van der Zand A, Braakman I. 2008. Peroxisomes:Minted by the ER. Curr Opin Cell Biol 20: 393–400.

Tam YYC, Fagarasanu A, Fagarasanu M, Rachubinski RA.2005. Pex3p initiates the formation of a preperoxisomalcompartment from a subdomain of the endoplasmic re-ticulum in Saccharomyces cerevisiae. J Biol Chem 280:34933–34939.

Tan X, Titorenko VI, van der Klei IJ, Sulter GJ, Haima P,Waterham HR, Eyers M, Harder W, Veenhuis M,Cregg JM. 1995. Characterization of peroxisome-defi-cient mutants of Hansenula polymorpha. Curr Genet 28:248–257.

Thoms S, Harms I, Kalies K-U, Gartner J. 2012. Peroxisomeformation requires the endoplasmic reticulum channelprotein Sec61. Traffic 13: 599–609.





Titorenko VI, Rachubinski RA. 1998. The endoplasmic re-ticulum plays an essential role in peroxisome biogenesis.Trends Biochem Sci 23: 231–233.

Titorenko VI, Rachubinski RA. 2000. Peroxisomal mem-brane fusion requires two AAA family ATPases, Pex1pand Pex6p. J Cell Biol 150: 881–886.

Titorenko VI, Chan H, Rachubinski RA. 2000. Fusion ofsmall peroxisomal vesicles in vitro reconstructs an earlystep in the in vivo multistep peroxisome assembly path-way of Yarrowia lipolytica. J Cell Biol 148: 29–44.

van der Zand A, Braakman I, Tabak HF. 2010. Peroxisomalmembrane proteins insert into the endoplasmic reticu-lum. Mol Biol Cell 21: 2057–2065.

van der Zand A, Gent J, Braakman I, Tabak HF. 2012. Bio-chemically distinct vesicles from the endoplasmic retic-ulum fuse to form peroxisomes. Cell 149: 397–409.

Voorn-Brouwer T, Kragt A, Tabak HF, Distel B. 2001. Perox-isomal membrane proteins are properly targeted to per-oxisomes in the absence of COPI- and COPII-mediatedvesicular transport. J Cell Sci 114: 2199–2204.

Yewdell JW. 2005. The seven dirty little secrets of majorhistocompatibility complex class I antigen processing.Immunol Rev 207: 8–18.

Yonekawa S, Furuno A, Baba T, Fujiki Y, Ogasawara Y,Yamamoto A, Tagaya M, Tani K. 2011. Sec16B is involvedin the endoplasmic reticulum export of the peroxisomalmembrane biogenesis factor peroxin 16 (Pex16) in mam-malian cells. Proc Natl Acad Sci 108: 12746–12751.

Zanetti G, Pahuja KB, Studer S, Shim S, Schekman R. 2012.COPII and the regulation of protein sorting in mammals.Nat Cell Biol 14: 20–28.

L. Dimitrov et al.




2013; doi: 10.1101/cshperspect.a013243Cold Spring Harb Perspect Biol Lazar Dimitrov, Sheung Kwan Lam and Randy Schekman The Role of the Endoplasmic Reticulum in Peroxisome Biogenesis

Subject Collection The Endoplasmic Reticulum

the Golgi Apparatus through ProteomicsCell Biology of the Endoplasmic Reticulum and

Jeffrey Smirle, Catherine E. Au, Michael Jain, et al.Entry, Replication, and AssemblyHow Viruses Use the Endoplasmic Reticulum for

Takamasa Inoue and Billy Tsai

Reticulum-Associated Degradation SystemThe Mammalian Endoplasmic

ChristiansonJames A. Olzmann, Ron R. Kopito and John C.

NetworksExpanding Proteostasis by Membrane Trafficking

Darren M. Hutt and William E. Balch

of the Nuclear Pore and Coat Protein ComplexesFunctional Insights from Studies on the Structure

Thomas SchwartzEndoplasmic Reticulum

-Linked Protein Glycosylation in theN

Jörg Breitling and Markus Aebi

PathwayTail-Anchored Membrane Proteins by the GET Endoplasmic Reticulum Targeting and Insertion of

Vladimir Denic, Volker Dötsch and Irmgard Sinning

Endoplasmic ReticulumRetrograde Traffic from the Golgi to the

Anne Spang

Reticulum and MitochondriaLipid Transport between the Endoplasmic

Vid V. Flis and Günther DaumPeroxisome BiogenesisThe Role of the Endoplasmic Reticulum in

SchekmanLazar Dimitrov, Sheung Kwan Lam and Randy

Protein Folding in the Endoplasmic ReticulumIneke Braakman and Daniel N. Hebert Reticulum and Beyond

Sphingolipid Homeostasis in the Endoplasmic

David K. Breslow

Interconnections with Other OrganellesEndoplasmic Reticulum Structure and

Amber R. English and Gia K. Voeltz Endoplasmic ReticulumCharacterizing the Proteins and Functions of the The Contribution of Systematic Approaches to

Maya Schuldiner and Jonathan S. Weissman

Unfolded Protein ResponseEndoplasmic Reticulum Stress Sensing in the

et al.Brooke M. Gardner, David Pincus, Katja Gotthardt,

Endoplasmic ReticulumProtein Translocation across the Rough

GilmoreElisabet C. Mandon, Steven F. Trueman and Reid

http://cshperspectives.cshlp.org/cgi/collection/ For additional articles in this collection, see

Copyright © 2013 Cold Spring Harbor Laboratory Press; all rights reserved


http://cshperspectives.cshlp.org/cgi/collection/

http://cshperspectives.cshlp.org/cgi/collection/


the role of the endoplasmic reticulum in...

Documents