A vesicle carrier that mediates peroxisome protein
traffic from the endoplasmic reticulum
Sheung Kwan Lama,1, Naofumi Yodaa,1,2, and Randy Schekmana,b,3
aDepartment of Molecular and Cell Biology andbHoward Hughes Medical Institute, University of California, Berkeley, CA 94720
Edited by David D. Sabatini, New York University School of Medicine, New York, NY, and approved October 29, 2010 (received for review September 7, 2010)
Pex19p, a soluble cytoplasmic transport protein, is required for the
traffic of the peroxisomal membrane proteins Pex3p and Pex15p
from the endoplasmic reticulum (ER) to the peroxisome. We docu-
mentedPex15ptraffic fromtheERusinga chimericproteincontain-
in wild-type yeast cells is N-glycosylated and functions properly in
the peroxisome. In contrast, pex19Δ-mutant cells accumulate the
glycoprotein Pex15Gp in the ER. We developed a cell-free preper-
oxisomal vesicle-budding reaction in which Pex15Gp and Pex3p are
packaged into small vesiclesin the presence of cytosol, Pex19p, and
a SNARE protein (soluble N-ethylmaleimide-sensitive attachment
protein receptor) occurs in the same incubation but does not de-
pend on Pex19p. Conversely a dominant GTPase mutant Sar1p
which inhibits COPII has no effect on Pex3p packaging. Pex15Gp
and Pex3p budded vesicles sediment as low-buoyant-density mem-
branes on a Nycodenz gradient and copurify by affinity isolation
using native but not Triton X-100–treated budded vesicles. ER–
peroxisome transport vesicles appear to rely on a novel budding
mechanism requiring Pex19p and additional unknown factors.
mals (1–3). Despite decades of work on the biosynthetic origin of
the peroxisome, uncertainty remains about the source of mem-
brane precursors. Classic morphologic work by Novikoff and Shin
(4) pointed to an origin of peroxisomal membrane in the endo-
plasmic reticulum (ER). Subsequent work by Lazarow and Fujiki
(5) and many others favored an independent origin in which per-
oxisomal proteins are imported from the cytoplasm into an or-
ganelle that propagates by growth and division of a preexisting
the biogenic dependence or independence of the organelle (6–9).
The growth and division model waschallenged in genetic studies
of peroxisome biogenesis in yeast where it was possible to dem-
onstrate that the generation of new peroxisomes does not depend
on a preexisting organelle (10). Two proteins, Pex3p and Pex19p,
terminally anchored membrane protein, serves an unknown but
binds to peroxisomal membrane proteins (PMPs) and may chap-
erone their insertion into the peroxisomal membrane (10–13). In-
duction of Pex19p in a pex19Δ cell results in the creation of new
peroxisomes (14). Furthermore, at least two PMPs, Pex3p and
Pex15p, appear to be localized to the ER in pex19Δ cells (14–16).
Several other peroxisomal proteins also may originate in the ER
(15). Pex15p is a C-terminal tail-anchor protein that requires the
GET1, -2, and -3 gene products, which act at the ER, to assemble
other C-terminal tail-anchored proteins required for secretory
traffic (16). Hoepfner et al. (14) showed that newly made YFP-
Pex3p migrates from the ER to the peroxisome, suggesting a novel
of YFP-Pex3p as an intermediate in this traffic path. Indepen-
dently, using a photoactivable form ofGFP,Kim et al. (17)showed
that Pex16, a mammalian PMP, originates in the ER. However,
other than Pex19p, the requirements for the formation and tar-
eroxisomes function in selected metabolic and biosynthetic
pathways that are essential for normal development in mam-
geting of a membrane carrier from the ER to the peroxisome
We developed two independent tests ofthe origin and vesicular
carrier mechanism of peroxisomal protein transfer from the ER.
A hybrid form of Pex15p was created containing a glycosylation
sequence which we used to document that Pex15p is exposed to
the ER-localized N-glycosylation apparatus en route to the per-
oxisome. Further, we established a cell-free vesicle-budding re-
action that reproduces the formation of a distinct membrane
carrier responsible for transit of Pex15p and Pex3p from the ER.
These ER-derived membrane carriers may fuse homotypically
to create a new peroxisome or fuse with a preexisting peroxisome
to sustain their growth and division.
Pex15p as a Reporter of Peroxisome Traffic via the ER. Pex15p is a C-
terminal tail-anchored membrane protein that has been reported
to assemble in the ER membrane (16, 18). A special insertion
and in its absence Pex15p appears in mitochondria (16). Thus, as
has been documented for Pex3p traffic, we predicted that Pex15p
would assemble in the peroxisome from an ER precursor form.
We adopted a variation of an approach Schuldiner et al. (16)
used to demonstrate ER insertion of tail-anchored proteins. An N-
terminal protein A (pA) and GFP extension of Pex15p was tagged
at the C terminus with a short opsin tag containing one potential
N-glycosylation site (pA-GFP-Pex15G). This fusion gene was ex-
pressedunder controlofa methionine-repressiblepromoter inWT
cells or pex19-deletion mutant cells which lack peroxisomes.
Quantitative immunoblot assays using Pex15p serum showed the
plasmid copy and the chromosomally encoded Pex15p were ex-
pressed at levels comparable to the endogenous levels in untrans-
fected cells (Fig. S1). In WT cells, pA-GFP-Pex15Gp localized to
discrete punctae that coincide with peroxisomes as marked by
a CFP-peroxisomal target signal 1 (PTS1) (Fig. 1 A and B). In
contrast, in a pex19Δ strain, pA-GFP-Pex15Gp localized to the
nuclear envelope and cortical ER (Fig. 1A). The steady-state level
of pA-GFP-Pex15Gp was reduced in pex19Δ vs. WT cells. Perhaps
ER-localized Pex15p is subject to ER-associated degradation.
Elgersma et al. (18) used a Pex15-invertase fusion protein to
detect ER insertion of Pex15p. Overexpressed fusion protein
fragment linked to Pex15p did not interfere with Pex15 function
or localization to the peroxisome (Fig. 1B). To assess the role of
the ER in traffic of Pex15p, we examined the glycosylation of pA-
Author contributions: S.K.L., N.Y., and R.S. designed research; S.K.L. and N.Y. performed
research; S.K.L., N.Y., and R.S. analyzed data; and S.K.L., N.Y., and R.S. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1S.K.L. and N.Y. contributed equally to this work.
2Present address: Mitsubishi Chemical Group Science and Technology Research Center
Inc., Yokohama 227-8502, Japan.
3To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| December 14, 2010
| vol. 107
| no. 50
GFP-Pex15Gp in WT and pex19Δ cells. Fig. 1C shows that pA-
GFP-Pex15Gp was sensitive to endoglycosidase H (endoH), and
the deglycosylated species migrated close to the same position as
pA-GFP-Pex15p. Similarly, pA-GFP-Pex15Gp made in pex19Δ
was sensitive to endoH (Fig. 1D).
To be certain that glycosylated Pex15Gp trafficked to the per-
oxisome, we fractionated membranes from WT cells expressing
Pex15Gp (upper and lower bands, respectively) cosedimented with
peroxisomes, marked by another peroxisomal protein, Pex3-pAp,
and separated from ER membranes, marked by Sec61p, on a
buoyant density gradient. These results support the proposal that
Pex15p passes through the ER en route to the peroxisome.
Reconstitution of Pex15Gp Budding from the ER. Hoepfneretal.(14)
showedthat peroxisomes arise denovo from theER when Pex19p
expression is induced in a pex19Δ cell. They found that Pex3p
accumulates in the ER in a pex19Δ strain and collects in ER
punctae that migrate to peroxisomes when Pex19p expression
We sought a biochemical test of the role of Pex19p in Pex15p
and Pex3p traffic from the ER. For this purpose, we created
pex19Δ yeast strains harboring a methionine promoter-controlled
pA-GFP-Pex15G fusion gene or a Pex-positive functional in-
tegrated form of Pex3-pA. For experiments where pA-GFP-
Pex15Gp was measured, cells were grown in synthetic medium
with methionine (20 mM) to reduce Pex15Gp expression, and the
total Pex15p level remained the same even after Pex15Gp ex-
pression (Fig. S1). Cells were converted to spheroplasts and lysed
gently under conditions in which ER membranes sediment rap-
idly. The gentle lysate served as a donor membrane for a vesicle-
budding reaction modeled on a reconstitution that allowed us to
isolate the COPII coat responsible for secretory cargo protein
traffic from the ER (19–21).
Membranes were mixed with cytosol from WT or pex19Δ-
and GTP and were incubated at 22 °C for 30 min. Donor ER
membranes were centrifuged for 5 min at 20,000 × g, and the
medium-speed supernatant (MSS) fraction was removed and
analyzed by SDS/PAGE and immunoblot for the presence of pA-
GFP-Pex15Gp. We detected Pex15Gp (ca. 1% of total) in the
MSS fractions from incubations containing WT cytosol, but the
signal was dramatically decreased (from 1 to <0.1%) with pex19Δ-
mutant cytosol (Fig. 2A, lanes 1 and 2). Recombinant Pex19p
expressed in and purified from Escherichia coli restored budding
of Pex15Gp in the presence of mutant cytosol but not without
cytosol (Fig. 2A, lanes 3 and 4). As controls, we detected the cyto-
solic protein 3-phosphoglycerate kinase (PGK1p) in the MSS
fraction but not the ER-resident protein, Sec61p. In contrast, the
COPII membrane cargo protein, Sec22p, was detected in the
MSS fraction independent of Pex19p. Pex15Gp release into the
substitute for ATP, and GTPγs had no effect on the ATP-stimu-
lated reaction (Fig. 2B).
anNterminustagged withpAandGFPandaCterminustagged withanopsin
peptide containing a predicted N-glycosylation site was cloned under control
of a methionine-repressible promoter. The tagged Pex15 was expressed in
WT and pex19Δ-mutant cells. GFP fluorescence was localized to punctae in
WT cells and labeled the nuclear envelope and peripheral ER in pex19Δ
cells. (B) Localization of pA-GFP-Pex15Gp was compared with a PMP, CFP-
PTS1, in WT cells. (C) N-glycosylation of pA-GFP-Pex15Gp and pA-GFP-Pex15p
(without a C-terminal glycosylation site) in WT cells was analyzed by EndoH
treatment. pA-GFP-Pex15Gp displayed decreased mobility on SDS/PAGE com-
pared with pA-GFP-Pex15p (lane1 and 2). EndoH treatment increased the
mobility of pA-GFP-Pex15Gp to approximately that of pA-GFP-Pex15p (lane
3). (D) EndoH treatment of pA-GFP-Pex15Gp expressed in pex19Δ cells re-
duced the apparent size of the protein to that of pA-GFP-Pex15Gp. (E) pA-
GFP-Pex15Gp was expressed under a methionine-repressible promoter in WT
cells, and cell extracts were fractionated by centrifugation on a 10–50% step
Nycodenz gradient. pA-GFP-Pex15Gp cofractionated with a genomically
encoded Pex3-pAp fusion protein and separated from the ER marker Sec61p.
Expression of pA-GFP-Pex15Gp. (A) A fusion gene coding Pex15 with
branes isolated from pex19Δ cells expressing pA-GFP-Pex15Gp were in-
cubated with nucleotide WT cytosol (lane 1), pex19Δ cytosol (lane 2), pex19Δ
cytosol plus Pex19 recombinant protein (rPex19, 4 μg/mL) (lane 3), rPex19
alone (4 μg/mL) (lane 4), buffer only (lane 5) or 10% load of starting mi-
crosome (lane 6). After 30 min incubation, microsomes were centrifuged, and
the MSS fraction was analyzed by SDS/PAGE and immunoblot. (B) The re-
quirements for vesicle budding of pA-GFP-Pex15Gp were assessed in incuba-
tions containing WT cytosol with ATP-regenerating system (ATPr) and 100 μM
GTP (lane 1); WT cytosol with 100 μM GTP (lane 2); WT cytosol with ATPr and
100 μM GTPγS (lane 3); WT cytosol with 100 μM GTPγS (lane 4); ATPr and 100
μM GTP only (lane 5); or 10% load of starting microsome (lane 6).
Cell-free vesicle budding of pA-GFP-Pex15Gp. (A) Microsome mem-
| www.pnas.org/cgi/doi/10.1073/pnas.1013397107Lam et al.
We applied the same biochemical budding reaction to evaluate
the packaging of Pex3-pAp into small vesicles. As with pA-GFP-
Pex15Gp, Pex3-pAp was detected in the MSS fraction (ca. 10% of
total) when gently lysed cells were incubated with WT cytosol,
ATP, and ATP-regenerating components, but the signal was
strongly reduced with pex19Δ-mutant cytosol (from 10 to <1%),
and no signal was detected with buffer alone (Fig. 3 A, lanes 1–3,
and B). Additional controls showed that the COPII cargo protein,
Sec22p, was packaged in reactions containing WT or pex19Δ cy-
tosol, whereas Sec61p, the resident ER protein, was retained in
the medium-speed pellet fraction. Conversely, pure COPII pro-
teins packaged Sec22p efficiently but not Pex3-pAp (lane 4).
We also used selective inhibition of COPII budding to docu-
ment distinct requirements for Pex3-pAp and secretory protein
cargo packaging. A dominant GTPase-mutant form of Sar1p
(T54A) that blocks COPII coat disassembly inhibited the pack-
aging of Sec22p but had no effect on the incorporation of Pex3-
pAp into slowly sedimenting membrane vesicles (Fig. 3C).
Pex3-pAp and pA-GFP-Pex15Gp Packaged Together into Membrane
Vesicles. We used flotation of the MSS fraction on a Nycodenz
density gradient to determine if pA-GFP-Pex15Gp and Pex3-pAp
were associated with buoyant membranes. Samples were mixed
withNycodenz toa 35% final concentration andplaced below less
dense Nycodenz fractions. Centrifuged samples were fractionated
and examined by immunoblot to detect peroxisomal proteins,
cytosolic protein (PGK1p), and COPII vesicles (Sec22p). Fig. 4 A
and B documents that both peroxisomal proteins associated with
buoyant membranes that were slightly more buoyant than COPII
vesicles and were resolved from the cytosolic marker.
Two further experiments were conducted to examine the
membrane integration of Pex15Gp and Pex3p. The pex19Δ strain
harboring an integrated Pex3-pA was transformed with a centro-
mere(CEN) plasmid carrying a pA-GST-Pex15G construct under
control of the methionine promoter. Cells grown in methionine-
prepared for the budding reaction were incubated with WT cy-
with glutathione-agarose beads in the absence or presence of 1%
Triton X-100 to capture exposed pA-GST-Pex15Gp. pA-GST-
Pex15Gp, Pex3-pAp, and Sec22p were present in the starting
fraction, but after extensive washing with buffer only Pex15Gp
and Pex3p were retained on the glutathione beads (Fig. 5A, lanes
1 and 2). However, in the presence of detergent, only Pex15Gp
wasrecovered onbeads (Fig.5A,lanes 2and3).We conclude that
some Pex15Gp and Pex3p are packaged together in a common
membrane vesicle, although the possibility of separate vesicles for
a fraction of each protein remains possible.
In another experiment, we examined the membrane associa-
tion of Pex3-pAp in vesicles in the MSS fraction after a budding
reaction. A standard incubation of pex19Δ membranes was con-
ducted with pex19Δ cytosol and purified GST-Pex19p (to provide
a tag for Pex19p detection). We found that GST-Pex19p pro-
moted Pex3-pAp budding as efficiently as Pex19p produced by
proteolytic cleavage of the GST hybrid protein. The MSS fraction
was centrifuged at high speed to collect small vesicles in a pellet
which was resuspended in buffer (B88), 0.1 M Na2CO3(to extract
peripheral membrane proteins), or 1% Triton X-100 (to solubi-
lize membranes). After a repeat high-speed centrifugation, the
examined by immunoblot for the pA and GST-tagged peroxi-
somal proteins. GST-Pex19p was membrane associated, possibly
by direct interaction with Pex3p (22), but was readily removed by
carbonate or detergent. In contrast, Pex3-pAp was retained in
membranes except in the presence of detergent. We conclude
that Pex3p is packaged as an integral membrane protein, along
with Pex15Gp, into buoyant vesicles in a budding reaction that
requires Pex19p and other cytosolic proteins.
Other Cytoplasmic Peroxisomal Proteins Are Not Required for Pex3p
Budding. We observed that Pex19p is not sufficient to package pA-
GFP-Pex15Gp into budded vesicles (Fig. 2A, lanes 3 and 4). The
other known cytoplasmic, Pex proteins (Pex1p, Pex5p, Pex6p,
Pex7p, and Pex11p) are required for protein import into the or-
ganelle or for organelle dynamics but not for the formation of the
organelle per se (23–25). Nonetheless, we tested their role in the
budding reaction using cytosol prepared from strains harboring
pex19Δ cells expressing a chromosomally integrated form of Pex3-pA were
incubated at room temperature (22 °C) for 30 min with WT cytosol (lane 1),
pex19Δ cytosol (lane 2), no cytosol (lane 3), or COPII purified proteins (Sar1p,
Sec23p/24p, and Sec13p/31p) only. MSS fractions were obtained by centri-
fugation at 15,000 rpm for 5 min in an Eppendorf refrigerated bench top
centrifuge, and aliquots (17 μL) were sampled for SDS/PAGE. Lanes 5–8
represent a 10% load of total reactions. (B) Packaging of Pex3-pAp was
carried in the presence of both ATPr and GTP (lane 1), GTP only (lane 2), ATPr
only (lane 3), or in the absence of nucleotide (lane 4). In lane 5 there was
a 10% load of starting microsome. (C) Budding reaction of Pex3-pAp was
conducted in the presence of nucleotides with different cytosol fractions:
WT cytosol (lane 1); WT cytosol with Sar1(T54A) (lane 2); pex19Δ cytosol
(lane 3); pex19Δ cytosol with Sar1(T54A) (lane 4); or 10% load of starting
microsome fraction (lane 5).
Cell-free packaging of Pex3-pAp. (A) Microsome membranes from
(A) The MSS fraction from a cell-free incubation of Pex3-pAp microsomes
was evaluated using a density gradient flotation assay. Membrane-bound
Sec22p and Pex3-pAp centrifuged near the top of a Nycodenz gradient
(fractions 1–3), whereas soluble protein (PGK1p) remained in the densest
Nycodenz fractions (fractions 4–7). (B) The MSS fraction from a cell-free in-
cubation of pA-GFP-Pex15Gp microsomes was evaluated by density gradient
flotation. Membrane-bound Sec22p and pA-GFP-Pex15Gp centrifuged near
the top of a Nycodenz gradient (fractions 1–3), whereas soluble protein
(PGK1p) remained in the dense Nycodenz phase (fractions 4–6).
Pex3-pAp and pA-GFP-Pex15Gp are packaged into buoyant membrane.
Lam et al.PNAS
| December 14, 2010
| vol. 107
| no. 50
deletions in each of these genes and from pex3Δ cells and com-
pared the activity of these fractions with the standard pex19Δ
cytosol in a Pex3-pAp budding reaction. As seen before, pex19Δ
cytosol was defective in Pex3-pAp budding but not in Sec22p
budding (Fig. 6). Conversely, pure COPII proteins packaged
Sec22p but not Pex3-pAp. Cytosols from the other pex-mutant
cells showed no defect in Pex3-pAp packaging. The additional
soluble proteins required to produce Pex3-pAp budded vesicles
The biogenic origin of peroxisomes has been a controversial
subject for decades. In recent years, most investigators have fo-
cused on the ER as a source of membrane and at least some of
the integral membrane proteins that end up in the peroxisome.
Most of the evidence in support of a role for the ER comes from
visualization of peroxisomal precursors in the ER membrane
followed kinetically by movement to the peroxisome (14, 15).
Although it is clear that Pex3p and Pex15p accumulate in the ER
in a pex19Δ yeast mutant (14, 15), the possibility that this path is
a minor one taken when no peroxisome exists for these proteins
to enter could not be ruled out. Thus it has been difficult to test
the role of the ER as an essential station in this pathway.
We have developed two lines of evidence concerning the role of
as a glycoprotein in WT and pex19Δ-mutant cells. Fluorescence
localization and cell fractionation of PEX-positive cells confirms
an interruption of normal traffic in the pex19 mutant, Pex15Gp
shows the mark of an ER intermediate in peroxisome assembly.
Pex19p has been found in molecular complexes with newly
synthesized PMPs (12, 26). This finding was interpreted as a role
for Pex19p in chaperoning the import of membrane proteins di-
rectly from the cytoplasm into the peroxisome. What role could
Pex19p play if PMPs are first assembled in the ER? To examine
the mechanism of peroxisomal protein transit from the ER, we
developed a cell-free vesicle-budding reaction in which two dif-
ferent PMPs are incorporated into a slowly sedimenting vesicle
fraction dependent on Pex19p, ATP, and additional cytosolic
factors. Our evidence demonstrates that this budding process is
independent of the COPII machinery responsible for secretory
cargo packaging at the ER.
Pex3p is the one of the most crucial PMPs required for per-
oxisome formation. No peroxisomal structures are detected in
pex3-mutant cells (10). New peroxisomes form when Pex3p syn-
thesis is induced, and Pex19p is required for this de novo process
(10, 14). Our cell-free budding reaction recapitulates the first step
of this pathway. Indeed, the importance of Pex3p in peroxisome
biogenesis may be reflected in the efficiency with which it is
packaged in our cell-free budding reaction in relation to Pex15p
(10% vs. 1% of total). Pex19p contains a dedicated domain for
Pex3p interaction and a common domain for the other PMPs
(13), perhaps explaining why Pex3p has a higher packaging effi-
ciency than Pex15p. Another possibility is that pA-GFP in some
way sterically hinders the preperoxisomal packaging of Pex15p.
Preperoxisomal vesicle biogenesis may begin with the recruitment
of Pex19p to the membrane peroxisomal targeting signal (mPTS)
on an ER-integrated form of Pex3p. Likewise, Pex15p and other
PMPsmay display mPTSontheERthat alsoare boundby Pex19p.
In contrast, resident ER proteins or biosynthetic membrane pro-
teins destined for transport through the secretory pathway would
lack such mPTS sequences. Thus, Pex19p may constitute an
adaptor-like protein to initiate the sorting of PMPs from other
membrane proteins in the ER.
Our recombinant Pex19p was made in E. coli; thus it lacks the
C-terminal prenylation group attached to the native protein in
yeast (11, 27). Nonetheless, the recombinant proteins, as well as
a GST-Pex19p chimera, were functional in our cell-free budding
reaction. However, activity depended on additional heat-sensi-
tive cytosolic factors contained in the WT or pex19Δ-mutant cell
lysates. The unidentified factors, probably proteins, do not in-
clude known peroxisomal cytoplasmic proteins, COPII, or COPI.
One or more of the additional proteins or possibly a membrane
protein may confer ATP dependence on this budding reaction.
fivefold scaled-up vesicle-budding incubation was carried out with micro-
somes prepared from pex19Δ cells coexpressing pA-GST-Pex15Gp (CEN plas-
mid) and Pex3-pAp (chromosomally integrated). The resulting MSS fraction
(lane 1) was incubated with glutathione-agarose beads in buffer with (lane 3)
or without (lane 2) Triton X-100. Sec22p is a COPII vesicle cargo protein. (B)
The MSS fraction was prepared from a budding reaction using membranes
from the Pex3-pAp strain mixed with pex19Δ cytosol and purified GST-
Pex19p. The MSS fraction was subjected to high-speed centrifugation to
obtain vesicles enriched in an HSP. The HSP was resuspended with B88 buffer
(lanes 1 and 2), 0.1 M Na2CO3(lanes 3 and 4), and 1% Triton X-100 (lanes 5
and 6). The resuspended HSP was centrifuged again at 55,000 rpm in
TLA100.3 for 30 min and separated into pellet (P) and supernatant (S). Lane 7,
10% load of the starting microsome fraction.
pA-GFP-Pex15Gp and Pex3-pAp co-isolate in budded vesicles. (A) A
toplasmic peroxisomal proteins. Microsomes were incubated with cytosol
prepared from WT, pex19Δ, pex3Δ, pex1Δ, pex5Δ, pex6Δ, pex7Δ, pex11Δ,
without cytosol, and with COPII recombinant proteins only. MSS fractions
were obtained from each reaction and detected as described in Materials
and Methods. Pex3-pAp in the MSS fraction was examined by SDS/PAGE and
immunoblot (A) and quantified (B) by Typhoon and Image Quant Software
(Molecular Dynamics). n = 2.
Pex3-pAp packaging is dependent on Pex19p but not on other cy-
| www.pnas.org/cgi/doi/10.1073/pnas.1013397107 Lam et al.
Neither Pex3p nor Pex19p is predicted to contain an ATP-
binding domain (10–12). The unknown factors could bind ER-
associated Pex19p and convey Pex19p-bound cargo into a per-
oxisomal transport vesicle.
Peroxisomes are not essential for yeast cell growth on glucose
medium.Thus, the genetic screens for pex mutants wereconducted
initially by mutagenesis and then through a survey of the entire set
of viable deletion-mutant yeast strains (6, 8, 28). Despite this ex-
haustive search, no additional candidate genes encoding cytosolic
proteins as important as Pex19p have emerged. The unidentified
genes maybeessential foryeastcellviability(e.g.,ifpreperoxisome
vesicle budding requires proteins also engaged in an essential
vesicle traffic event), or they may encode redundant functions (29).
Additional genetic screens that incorporate these possibilities may
turn up such candidates. However, our cell-free reaction provides
an alternative to allow the purification of functional forms of the
Materials and Methods
Yeast Strains and Media. ProteinA was integrated at theC-terminal endofthe
Pex3 locusby a PCR-based one-step transformation procedure(30). Primers for
C-terminal protein A integration were 5′-ATACAGCAACTTTGGCGTCTCCA-
GCTCGTTTTCCTTCAAGCCT and 3′-CGCTATATATATATATATTCTGGTGTGAG-
TGTCAGTACTTATTCA. The PCR product was introduced into pex19Δ cells from
the Euroscarf collection (Mat α; his3Δ; leu2 Δ; lys2 Δ; ura3 Δ; pex19::kanMX4)
to obtain NYY9 (Mat α; leu2 Δ; lys2 Δ; ura3 Δ; pex19::kanMX4 Pex3-pA::His3),
which was used for microsome preparation. Pex15G was cloned from yeast
genomic DNA with an additional opsin tag sequence (ATGATGTCTAGA-
TGAT) at the 3′ end and was inserted downstream of pA and GFP in a MET25
vector (31). For cytosol preparation, all the deletion-mutant cells were from
the Euroscarf deletion collection. Cells were grown at 30 °C in yeast-peptone-
dextrose (YPD) medium (1% yeast extract, 2% peptone, 2% glucose) or syn-
thetic medium (2% glucose, Difco yeast nitrogen base without amino acid,
of microsomal and cytosolic fractions. For oleic acid selection, cultures were
grown at 30 °C on oleate plates containing 0.67% yeast nitrogen base, 0.25%
oleic acid, 0.25% Tween 40, 2% agar, with the supplement of amino acids.
EndoH Treatment. pA-GFP-Pex15p or pA-GFP-Pex15Gp was expressed under
a methionine-repressible promoter in WT cells and pex19Δ cells, and 5 mL of
cell culture (OD600= 0.8) was centrifuged at 4,000 × g for 5 min. The cell
pellet was resuspended in 1 mL of 20 mM Tris·HCl buffer. Total proteins were
extracted by grinding in liquid nitrogen and heated at 95 °C with 1% SDS for
5 min. The lysate was centrifuged at 15,000 rpm in a bench-top ultracen-
trifuge (Eppendorf 5417C with rotor F45-30-11), and the supernatant frac-
tion was collected. Aliquots (1–20 μg protein) were mixed with 1 μL 10×
denaturing buffer (New England Biolabs), and water was added to a volume
of 10 μL. The mixture was heated at 95 °C for 10 min. Once the mixture
cooled, 5 μL (5,000 units) of EndoH (New England Biolabs) and 2 μL 10× G5
reaction buffer (New England Biolabs) were added, and the volume was
adjusted to 20 μL with water. The reaction was incubated at 37 °C for 1 h.
Samples (15 μL) were heated with SDS-loading buffer and sampled for SDS/
PAGE and immunoblot analysis.
Subcellular Fractionation of Peroxisome. pA-GFP-Pex15G was expressed under
a methionine-repressible promoter in WT cells, and 5 mL of cell culture
(OD600= 0.8) was centrifuged at 4,000 × g for 5 min. The cell pellet was
resuspended in 1 mL of lyticase buffer (0.7 M sorbitol, 0.75× yeast extract
Peptone, 0.5% glucose, 10 mM Hepes/NaOH, pH 7.4) and digested with
lyticase (6,000 units/mL) at 30 °C for 30 min. Cells were centrifuged at 4,000 ×
g for 10 min at 4 °C and resuspended in 1 mL of B88 [20 mM Hepes-KOH (pH
6.8), 0.25 M Sorbitol, 0.15 M KOAc, 1 mM MgOAc2] with protease inhibitor
mixture (Roche). The suspension was homogenized with 20 strokes in
a Potter-Elvehjem homogenizer. The lysate was centrifuged at 4,000 × g for
10 min. The postnuclear supernatant (300 μL) was loaded on the top of
a Nycodenz step gradient ranging from 10 to 50% (200 μL: 50%, 40%, 30%,
20%; 150 μL: 10%) and centrifugated for 2 h at 55,000 rpm at 4 °C in
a Beckman TLS-55 ultracentrifuge. The top 150 μL was removed, and 13 frac-
tions (60 μL) were collected. An aliquot (15 μL) of each fraction was heated at
65 °C with SDS loading buffer for 10 min and sampled for SDS/PAGE and
Protein Purification. GST-Pex19 was prepared as an N-terminal GST fusion
protein from pGEX-2T (GE Healthcare) in E. coli BL21 (DE3). The GST-tagged
protein was purified by a method described previously (32, 33) with slight
modification. GST-Pex19 expression was induced by isopropyl-β-D-thio-
galactoside (0.1 mM) at 30 °C for 3 h. GST was cleaved from Pex19 by thrombin
cleavage in 20 mM Tris·HCl (pH 8) buffer with 2.5 mM CaCl2at 22 °C for 2 h.
Sar1, Sec13/31, and Sec23/24 were prepared as described (20). A dominant-
negative form of Sar1 (T54A) (34) was purified from E. coli expressing GST-
fusion proteins. Thrombin was used to cleave the GST from Sar1p, which was
purified further by MonoQ column.
Cell-Free Vesicle Budding Reaction. For the preparation of concentrated cy-
tosol, yeast was grown in 1 L YPD at 30 °C, and cells were harvested at ∼1
OD600/mL, washed in cold water, and resuspended in 2 mL of B88 with pro-
tease inhibitor mixture (Roche). A concentrated suspension was dropped into
a mortar filled with liquid nitrogen to form small beads which were ground
to a fine powder with a pestle. The frozen lysate was transferred into a 50-mL
tube, thawed on ice, and centrifuged at 3,000 rpm for 5 min, then at 15,000
rpm for 10 min (Beckman SS34), and finally at 34,000 rpm for 1 h (Beckman
SW55Ti). The supernatant was collected carefully to avoid a lipid phase at the
meniscus. Microsomal membranes for use in vesicle-budding reactions were
prepared from NYY9 cells as described (20). The conditions of incubation for
peroxisomal vesicle budding were similar to those used to form COPII vesicles
(20). Aliquots (15 μg protein) of the microsome fraction were washed twice in
2.5 M urea in B88, rinsed twice in B88, and the washed/rinsed microsome
fraction was centrifugated at 15,000 rpm in an Eppendorf refrigerated bench-
top centrifuge. The final pellet was resuspended in 800 μL of B88. An aliquot
(100 μL) of the microsomal suspension was mixed with cytosol (final con-
centration 4 mg protein/mL) in the presence of ATP (1 mM) with an ATP
regeneration system (ATPr) (35) and GTP (0.1 mM) in a final volume of 250 μL.
In one experiment, a dominant-negative form of Sar1 (T54A) was added at
the concentration of 10 μg/mL. Purified COPII proteins were added at a con-
centration of 10 μg/mL Sar1p, 20 μg/mL Sec23p/24p complex, and 20 μg/mL
Sec13p/31p complex. The reaction mixture was incubated at 22 °C for 30 min.
The MSS fraction was obtained after centrifugation at 20,000 × g for 5min. In
some experiments, the MSS fraction was analyzed further by high-speed
centrifugation at 55,000 rpm for 30 min at 4 °C in a Beckman TLA100 ultra-
centrifuge or by equilibrium density flotation (20) at 55,000 rpm for 2 h at
4 °C in a Beckman TLS-55 ultracentrifuge. Briefly, MSS fractions (150 μL) were
mixed with 70% Nycodenz (300 μL) in B88 (final Nycodenz concentration
∼35%) and placed at the bottom of a TLS-55 tube. Three solutions of dif-
ferent density (200 μL each of 35%, 25%, and 10% Nycodenz in B88) were
layered on top. Tubes were centrifuged at 55,000 rpm for 2 h (Beckman TLS-
55). Fractions (150 μL) were collected from the top, mixed with SDS loading
buffer, and heated at 55 °C for 10 min, followed by SDS/PAGE and immu-
noblot analysis. Antibodies used in this study were rabbit peroxidase anti-
peroxidase (Jackson ImmunoResearch Laboratories, Inc.) for the detection of
Pex3-pAp, pA-GFP-Pex15p, pA-GFP-Pex15Gp, and pA-GST-Pex15Gp; anti-
Sec22p, anti-Sec61p, anti-GST (Santa Cruz), and anti-Pgk1p (Molecular
Probes). ECL plus Western blotting detection reagent (GE Healthcare) was
used for detection. An alkaline sodium carbonate fractionation assay was
performed as follows: Microsomal membranes were incubated with GST-
Pex19p (20 μg/mL) with pex19 cytosol (4 mg/mL). The MSS fraction (150 μL)
was collected after the incubation, followed by high-speed centrifugation as
described above. The high-speed pellet (HSP) fraction was resuspended in 500
μL of B88 in three tubes. After a repeat centrifugation, the HSP was resus-
pended in 150 μL B88, 0.1 M Na2CO3, or 1% Triton X-100 and incubated on ice
for 30 min; then soluble and integral membrane proteins were separated by
high-speed centrifugation (Beckman TLA100) at 55,000 rpm for 30 min.
Glutathione Bead Co-Isolation. A fivefold vesicle budding was carried out with
microsomes prepared from pex19Δ cells coexpressing pA-GST-Pex15G (CEN
plasmid) and Pex3-pA (chromosomally integrated). The resulting MSS frac-
tion (1 mL) was incubated with 50 μL glutathione-agarose beads in B88 with
or without 1% Triton X-100 overnight at 4 °C. Beads were settled by gravity,
and the supernatant was removed. The beads were rinsed three times with
1.5 mL of B88 with or without 1% Triton X-100. The settled beads were
heated to 95 °C with SDS loading buffer for 5 min and sampled for SDS/PAGE
and immunoblot analysis.
ACKNOWLEDGMENTS. We thank Jonathan Weissman (Howard Hughes
Medical Institute, Department of Cellular and Molecular Pharmacology,
University of California, San Francisco, and California Institute for Quanti-
tative Biosciences, San Francisco) for the opsin tag constructs and Suresh
Subramani (Section of Molecular Biology, Division of Biological Sciences,
Lam et al.PNAS
| December 14, 2010
| vol. 107
| no. 50
University of California at San Diego, La Jolla, CA) for the Pex15 antibody.
We also thank Lazar Dimitrov for insightful discussions. S.K.L. is supported by
a Human Frontier Science Program Long-term Fellowship. R.S. is an investi-
gator of the Howard Hughes Medical Institute.
1. Kunau WH (1998) Peroxisome biogenesis: From yeast to man. Curr Opin Microbiol 1:
2. Steinberg SJ, et al. (2006) Peroxisome biogenesis disorders. Biochim Biophys Acta
3. Brosius U, Gärtner J (2002) Cellular and molecular aspects of Zellweger syndrome and
other peroxisome biogenesis disorders. Cell Mol Life Sci 59:1058–1069.
4. Novikoff AB, Shin WY (1964) The endoplasmic reticulum in the Golgi zone and its
relations to microbodies, Golgi apparatus, and autophagic vacuoles in rat liver cells. J
Microsc (Paris) 3:187–206.
5. Lazarow PB, Fujiki Y (1985) Biogenesis of peroxisomes. Annu Rev Cell Biol 1:489–530.
6. Erdmann R, Veenhuis M, Mertens D, Kunau WH (1989) Isolation of peroxisome-
deficient mutants of Saccharomyces cerevisiae. Proc Natl Acad Sci USA 86:5419–5423.
7. Tsukamoto T, Yokota S, Fujiki Y (1990) Isolation and characterization of Chinese
hamster ovary cell mutants defective in assembly of peroxisomes. J Cell Biol 110:
8. Elgersma Y, van den Berg M, Tabak HF, Distel B (1993) An efficient positive selection
procedure for the isolation of peroxisomal import and peroxisome assembly mutants
of Saccharomyces cerevisiae. Genetics 135:731–740.
9. Gould SJ, Valle D (2000) Peroxisome biogenesis disorders: Genetics and cell biology.
Trends Genet 16:340–345.
10. Höhfeld J, Veenhuis M, Kunau WH (1991) PAS3, a Saccharomyces cerevisiae gene
encoding a peroxisomal integral membrane protein essential for peroxisome
biogenesis. J Cell Biol 114:1167–1178.
11. Götte K, et al. (1998) Pex19p, a farnesylated protein essential for peroxisome
biogenesis. Mol Cell Biol 18:616–628.
12. Sacksteder KA, et al. (2000) PEX19 binds multiple peroxisomal membrane proteins, is
predominantly cytoplasmic, and is required for peroxisome membrane synthesis. J
Cell Biol 148:931–944.
13. Fang Y, Morrell JC, Jones JM, Gould SJ (2004) PEX3 functions as a PEX19 docking
factor in the import of class I peroxisomal membrane proteins. J Cell Biol 164:863–875.
14. Hoepfner D, Schildknegt D, Braakman I, Philippsen P, Tabak HF (2005) Contribution of
the endoplasmic reticulum to peroxisome formation. Cell 122:85–95.
15. van der Zand A, Braakman I, Tabak HF (2010) Peroxisomal membrane proteins insert
into the endoplasmic reticulum. Mol Biol Cell 21:2057–2065.
16. Schuldiner M, et al. (2008) The GET complex mediates insertion of tail-anchored
proteins into the ER membrane. Cell 134:634–645.
17. Kim PK, Mullen RT, Schumann U, Lippincott-Schwartz J (2006) The origin and
maintenance of mammalian peroxisomes involves a de novo PEX16-dependent
pathway from the ER. J Cell Biol 173:521–532.
18. Elgersma Y, et al. (1997) Overexpression of Pex15p, a phosphorylated peroxisomal
integral membrane protein required for peroxisome assembly in S.cerevisiae, causes
proliferation of the endoplasmic reticulum membrane. EMBO J 16:7326–7341.
19. Matsuoka K, et al. (1998) COPII-coated vesicle formation reconstituted with purified
coat proteins and chemically defined liposomes. Cell 93:263–275.
20. Barlowe C, et al. (1994) COPII: A membrane coat formed by Sec proteins that drive
vesicle budding from the endoplasmic reticulum. Cell 77:895–907.
21. Rexach MF, Schekman RW (1991) Distinct biochemical requirements for the budding,
targeting, and fusion of ER-derived transport vesicles. J Cell Biol 114:219–229.
22. Matsuzono Y, Matsuzaki T, Fujiki Y (2006) Functional domain mapping of peroxin
Pex19p: Interaction with Pex3p is essential for function and translocation. J Cell Sci
23. Titorenko VI, Rachubinski RA (2000) Peroxisomal membrane fusion requires two AAA
family ATPases, Pex1p and Pex6p. J Cell Biol 150:881–886.
24. Li X, Gould SJ (2002) PEX11 promotes peroxisome division independently of
peroxisome metabolism. J Cell Biol 156:643–651.
25. Meinecke M, et al. (2010) The peroxisomal importomer constitutes a large and highly
dynamic pore. Nat Cell Biol 12:273–277.
26. Jones JM, Morrell JC, Gould SJ (2004) PEX19 is a predominantly cytosolic chaperone
and import receptor for class 1 peroxisomal membrane proteins. J Cell Biol 164:57–67.
27. Vastiau IM, et al. (2006) Farnesylation of Pex19p is not essential for peroxisome
biogenesis in yeast and mammalian cells. Cell Mol Life Sci 63:1686–1699.
28. Lockshon D, Surface LE, Kerr EO, Kaeberlein M, Kennedy BK (2007) The sensitivity of
yeast mutants to oleic acid implicates the peroxisome and other processes in
membrane function. Genetics 175:77–91.
29. Perry RJ, Mast FD, Rachubinski RA (2009) Endoplasmic reticulum-associated secretory
proteins Sec20p, Sec39p, and Dsl1p are involved in peroxisome biogenesis. Eukaryot
30. Kragler F, Lametschwandtner G, Christmann J, Hartig A, Harada JJ (1998) Identi-
fication and analysis of the plant peroxisomal targeting signal 1 receptor NtPEX5.
Proc Natl Acad Sci USA 95:13336–13341.
31. Mumberg D, Müller R, Funk M (1994) Regulatable promoters of Saccharomyces
cerevisiae: Comparison of transcriptional activity and their use for heterologous
expression. Nucleic Acids Res 22:5767–5768.
32. Wang CW, Hamamoto S, Orci L, Schekman R (2006) Exomer: A coat complex for
transport of select membrane proteins from the trans-Golgi network to the plasma
membrane in yeast. J Cell Biol 174:973–983.
33. Wang F, et al. (2006) Development of antibody-based assays for omega-conotoxin
MVIIA. J Biochem Biophys Methods 67:49–56.
34. Saito Y, Kimura K, Oka T, Nakano A (1998) Activities of mutant Sar1 proteins in
guanine nucleotide binding, GTP hydrolysis, and cell-free transport from the
endoplasmic reticulum to the Golgi apparatus. J Biochem 124:816–823.
35. Baker D, Hicke L, Rexach M, Schleyer M, Schekman R (1988) Reconstitution of SEC
gene product-dependent intercompartmental protein transport. Cell 54:335–344.
| www.pnas.org/cgi/doi/10.1073/pnas.1013397107Lam et al.