A Complex of Pdi1p and the Mannosidase Htm1p
Initiates Clearance of Unfolded Glycoproteins
from the Endoplasmic Reticulum
Robert Gauss,1,4Kazue Kanehara,2,4,5Pedro Carvalho,3Davis T.W. Ng,2,* and Markus Aebi1,*
1Institute of Microbiology, Department of Biology, Swiss Federal Institute of Technology Zurich, 8093 Zurich, Switzerland
2Temasek Life Sciences Laboratory and Department of Biological Sciences, National University of Singapore, Singapore 117604
3Department of Cell and Developmental Biology, Centre for Genomic Regulation (CRG), 08003 Barcelona, Spain
4These authors contributed equally to this work
5Present address: Max Planck Institute for Plant Breeding Research, 50829 Cologne, Germany
*Correspondence: firstname.lastname@example.org (D.T.W.N.), email@example.com (M.A.)
Endoplasmic reticulum (ER)-resident mannosidases
generate asparagine-linked oligosaccharide signals
that trigger ER-associated protein
(ERAD) of unfolded glycoproteins. In this study, we
provide in vitro evidence that a complex of the yeast
protein disulfide isomerase Pdi1p and the mannosi-
dase Htm1p processes Man8GlcNAc2carbohydrates
bound to unfolded proteins, yielding Man7GlcNAc2.
This glycan serves as a signal for HRD ligase-medi-
ated glycoprotein disposal. We identified a point
mutation in PDI1 that prevents complex formation
of the oxidoreductase with Htm1p, diminishes
mannosidase activity, and delays degradation of
unfolded glycoproteins in vivo. Our results show
that Pdi1p is engaged in both recognition and glycan
signal processing of ERAD substrates and suggest
that protein folding and breakdown are not sepa-
rated but interconnected processes. We propose
a stochastic model for how a given glycoprotein is
partitioned into folding or degradation pathways
and how the flux through these pathways is adjusted
to stress conditions.
Endoplasmic reticulum (ER) quality control (ERQC) monitors
protein biogenesis and redirects unfolded and flawed polypep-
tides to the cytosol for destruction by the 26S proteasome
(Buchberger et al., 2010; Ma ¨a ¨tta ¨nen et al., 2010; Stolz and
Wolf, 2010). Asparagine (N)-linked oligosaccharides (NLOs)
coupled to the majority of polypeptides synthesized in the ER
are directly involved in this process (Aebi et al., 2010). Glycosyl-
hydrolases process NLOs in a time- and protein conformation-
dependent manner and produce signals that reflect the current
folding and maturation state of the attached polypeptide (Helen-
ius and Aebi, 2004). Accordingly, molecular chaperones assist
folding of nascent polypeptides presenting Glc3–0Man9GlcNAc2
glycans, while specialized ubiquitin-protein ligases of the ER-
associated protein degradation (ERAD) system clear glycopro-
teins displaying a bipartite signal— unfolded domains and a
Man5–7GlcNAc2 sugar—from the compartment (Hirsch et al.,
2009). Thus, cleavage of more than one mannosyl residue by
class 47-glycosylhydrolases (GH47s) extracts proteins from
the folding process and directs them to disposal in the cytosol
(Aebi et al., 2010; Helenius and Aebi, 2004).
The yeast Saccharomyces cerevisiae expresses two ER-local-
ized GH47-proteins with an a1,2-mannosidase homology (MH)
domain: Mns1p (ER mannosidase 1) and Htm1p (homolog to
ER mannosidase 1, also termed Mnl1p for mannosidase-like
1). In vitro, Mns1p removes the terminal mannosyl residue of
the B branch of free Man9GlcNAc2oligosaccharides (Jelinek-
Kelly et al., 1985; Jelinek-Kelly and Herscovics, 1988). In vivo,
Mns1p processes the entire pool of protein-bound Man9-
GlcNAc2present in the ER (Jakob et al., 1998). In MNS1-defi-
cient cells, however, breakdown of ERAD model proteins from
the ER is impeded (Jakob et al., 1998). Although Mns1p has no
peptide specificity, glycoproteins appear to be protected from
degradation until Mns1p processes attached glycans.
Deletion of the second GH47 gene, HTM1, also delays turn-
over of ERAD model substrates (Jakob et al., 2001; Nakatsukasa
et al., 2001). While initially considered a lectin, recent genetic
experiments provide evidence that Htm1p actively engages in
removing the outmost mannosyl moiety from the C branch of
the glycan (Clerc et al., 2009). The product Man7GlcNAc2oligo-
saccharide exposes an a1,6-linked mannosyl residue that is
the substrate of Yos9p, the lectin subunit of the HRD ligase
(Carvalho et al., 2006; Denic et al., 2006; Gauss et al., 2006;
Quan et al., 2008). Mns1p activity is a precondition for Htm1p
cleavage, suggesting that the two enzymes function in subse-
quent steps (Clerc et al., 2009). In addition to its N-terminal
MH domain, Htm1p features a large C-terminal (CT) domain
that is conserved only among fungal orthologs of Htm1p and
does not display significant homology to any known functional
domain (Sakoh-Nakatogawa et al., 2009). Truncating this
domain disturbs the function of Htm1p and prevents association
with its binding partner, the oxidoreductase Pdi1p (Clerc et al.,
Yeast Pdi1p is an essential folding catalyst that introduces
disulfide bonds into its substrates and rearranges or reduces
782 Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc.
pre-existing ones (Farquhar et al., 1991; Gilbert, 1997; Hatahet
and Ruddock, 2009). Besides its oxidoreductase activity,
Pdi1p acts as a molecular chaperone and also assists in the
folding of proteins that do not contain disulfide bonds (Cai
et al., 1994; Hatahet and Ruddock, 2009; Song and Wang,
1995; Wang and Tsou, 1993). Moreover, Pdi1p was assigned
a function in retrotranslocation of proteins into the cytosol or un-
folding of proteins prior to their export (Gillece et al., 1999; Lee
et al., 2010; Tsai et al., 2001). Pdi1p also introduces a disulfide
(Sakoh-Nakatogawa et al., 2009).
In this study, we expressed the Pdi1p-Htm1p complex in
insect cells and demonstrated that the purified complex pro-
cessed Man8GlcNAc2glycans bound to unfolded glycoproteins,
yielding Man7GlcNAc2. Moreover, we showed that Pdi1p played
a direct role in the clearance of unfolded glycoproteins from the
ER. In summary, we propose that the Pdi1p-Htm1p complex is
a functional unit in ERQC and put forward a model of how this
complex initiates breakdown of unfolded proteins.
The Pdi1p-Htm1p Complex Converts Man8GlcNAc2
to Man7GlcNAc2In Vitro
To express and purify Htm1p, we took advantage of the insect
cell-based baculovirus system. As a control protein, we included
Mns1p in our analysis. The genes of interest were fused to
N-terminal affinity tags and equipped with a C-terminal ER-
retention signal. GST-Mns1p was readily expressed and purified
a small fraction of GST-Htm1p was soluble when expressed as
a single protein (data not shown). Coexpression of its complex
partner HIS-Pdi1p, however, increased the solubility of GST-
Htm1p, and the Pdi1p-Htm1p complex was efficiently purified
(data not shown and Figure 1A). Copurification of HIS-Pdi1p
was dependent on GST-Htm1p because HIS-Pdi1p did not
bind to the affinity matrix.
The purified complex allowed us to study the activity of Htm1p
in vitro. We metabolically labeled N-linked oligosaccharides with
3H-mannose and prepared yeast whole cell extracts yielding
mutant strains to obtain glycoproteins with defined oligosaccha-
ride structures. Extracts were incubated with GSH beads that
were loaded with proteins purified from insect cells. After
removing the beads by centrifugation, N-linked glycans were
liberated from the polypeptides by PNGase F and analyzed by
HPLC to monitor glycan processing. To validate our experi-
mental approach, we first analyzed the activity of recombinant
Mns1p. Incubation of3H-mannose-labeled glycoproteins from
Dmns1 cells that predominantly display Man9GlcNAc2 NLOs
with GST did not alter the NLO profile (Figure 1B). GST-Mns1p
to Man8GlcNAc2, unless Mns1p activity was suppressed by
addition of EDTA (Figure 1B). In line with previous findings
demonstrating that Mns1p does not require the presence of
a polypeptide attached to its glycan substrate (Jakob et al.,
1998; Jelinek-Kelly et al., 1985), GST-Mns1p was active on
free oligosaccharides in our assay (Figure 1C).
Figure 1. Purification of Mns1p and Pdi1p-Htm1p Complex and
In Vitro Activity of Mns1p
(A) Insect cells were infected with viruses carrying expression copies of GST,
GST-Mns1p, GST-Htm1p, and HIS-Pdi1p. Cells were harvested and lysed and
Input (in), flowthrough (out), and bead fractions were separated by SDS-PAGE.
Proteins were visualized by staining with Coomassie Brilliant Blue. Signals
representing GST, GST-Mns1p, GST-Htm1p, and HIS-Pdi1p are indicated.
(B) Denatured glycoproteins were extracted from3H-mannose-labeled YJU90
cells (Dmns1) and incubated with GST, or Mns1p. When indicated, EDTA was
using PNGase F, purified, and separated by HPLC. Elution of radiolabeled
Man8GlcNAc2(M8) and Man9GlcNAc2(M9) were determined by comparison to
an LLO standard profile. Displayed are representative HPLC elution profiles.
(C)3H-mannose-labeled N-linked glycans from strain YJU90 were liberated
from polypeptides by PNGase F treatment and purified. Oligosaccharides
were incubated with GST or Mns1p, purified, and analyzed as described in (B).
Function of the Pdi1p-Htm1p Complex
Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc. 783
When we incubated glycoprotein extracts from wild-type cells
that predominantly carry Man8GlcNAc2NLOs with recombinant
Pdi1p-Htm1p, about 10% of Man8GlcNAc2 was trimmed to
Man7GlcNAc2(Figures 2A and 2D). Concomitantly, the Man8-
GlcNAc2population was decreased, indicating that it served
as substrate for the trimming reaction. Application of twice the
volume of Pdi1p-Htm1p-loaded beads did not increase the
amount of Man7GlcNAc2produced in the assay, suggesting
that the available substrate pool for the processing enzyme
was depleted (data not shown). Addition of EDTA abolished
lent cations to be functional (Figures 2A and 2D). Reduction and
alkylation of cell lysates was required to achieve maximal
conversion of Man8GlcNAc2NLOs by the mannosidase complex
(Figure S1). Pdi1p-Htm1p was not active on Man9GlcNAc2NLOs
(Figures 2C and 2E); only when Mns1p was included in the reac-
tion did we observe anincreased signal for Man7GlcNAc2. More-
than Mns1p (Figures 2B and 2D). From these data we concluded
that the Pdi1p-Htm1p complex is an active mannosidase that
preferentially processes a fraction of protein-bound Man8-
GlcNAc2oligosaccharides. Importantly, our in vitro results accu-
rately replicated in vivo experiments showing that Htm1p activity
depends on preceding glycan trimming by adding Mns1p and
that Htm1p only processes a portion of the total N-linked glycan
pool (Clerc et al., 2009). It is possible that the peptide linked to
the N-glycan determined whether the oligosaccharide was
a substrate for the Pdi1p-Htm1p mannosidase.
Isolation of a Viable PDI1 Point Mutation that Depends
on an Intact UPR Pathway
The in vitro experiments suggested that Pdi1p and Htm1p form
a functional unit. We wanted to corroborate this hypothesis in in-
vivo experiments. In a screen for mutant strains that depend on
an intact unfolded protein response (UPR) for viability, we iso-
lated a missense mutation in the PDI1 gene, named pdi1-1, that
resulted in a leucine to proline exchange at position 313 of the
protein (see Supplemental Experimental Procedures for details
on the screening procedure). Located near the center of the
b0domain of the enzyme (Figure 3A), the L313P mutation might
interfere with substrate binding or crosstalk between the b0and
a0domains and impair the functions of Pdi1p (Hatahet and
Ruddock, 2009; Serve et al., 2010). Importantly, pdi1-1 cells
were viable, suggesting that the essential functions of the oxido-
(Figure 3B). Only when we exposed the cells to the reducing
agent DTT did we observe a mild growth defect (Figure 3B).
Analyzing Pdi1-1p via immunoblotting, we detected a slight
hypoglycosylation of the protein and a minor increase in protein
level as compared to wild-type Pdi1p (Figure S2A). The stability
of Pdi1-1p was the same as for the wild-type protein, and UPR
was not induced in pdi1-1 cells (Figures S2B and S2C).
The pdi1-1 Mutation Does Not Affect Oxidative Protein
Folding but Impairs Disposal of Glycoproteins
To substantiate these findings at a molecular level, we analyzed
the biogenesis of two model proteins, CPY and Gas1p, in pulse-
chase experiments. While CPY/Prc1p is a soluble carboxypepti-
dase of the yeast vacuole, Gas1p is an extracellular glucanosyl
transferase bound to the plasma membrane by a GPI anchor.
The native structure of both proteins is stabilized by disulfide
bonds. Maturation of both proteins can be easily monitored
since posttranslational modifications change their electrophor-
etical mobility as they travel along the secretory pathway.
Compared to wild-type cells, maturation of neither of the two
proteins was delayed in cells expressing Pdi1-1p (Figures 3C
and 3D), demonstrating that the mutation did not interfere with
functions of Pdi1p in oxidative folding.
Next, we analyzed the breakdown of CPY*, a mutant form of
CPY that is not properly folded. Since fully glycosylated CPY*
can use both the ERAD system and the vacuolar pathway for
degradation, we used a single-glycan variant, CPY*abcD, that
can only use the Htm1p-Yos9p-dependent pathway of ERAD
(Kawaguchi et al., 2010). Compared to wild-type cells, disposal
of CPY*abcD was delayed in pdi1-1 cells (Figure 3E). We
observed the same effect on degradation of PrA* (data not
shown). Importantly, as observed for Dhtm1, the pdi1-1 allele
slightly increased the degradation of a glycan-independent
ERAD substrate, ngPrAD295-331 (Kanehara et al., 2010) (Fig-
ure 3F). These findings demonstrated that the pdi1-1 mutation
deteriorated efficient clearance of unfolded glycoproteins from
the ER but did not generally impair Hrd1p-dependent protein
Overexpression of Htm1p and Deletion of the ALG3
Gene Suppress the pdi1-1 Phenotype
Since Pdi1p forms a complex with Htm1p, we asked whether
elevated levels of Htm1p suppressed the degradation pheno-
type of pdi1-1. In wild-type cells, overexpression of HTM1 from
a PRC1 promoter accelerated degradation of CPY*abcD, but
slightly delayed breakdown of ngPrAD295-331 (Figures 4A and
S3). Increased levels of HTM1 in cells carrying the pdi1-1 muta-
tion restored breakdown of CPY*abcD to wild-type levels
(Figures 4A and 4B). In cells lacking Alg3p, Man5GlcNAc2oligo-
saccharides are transferred onto proteins. These sugars expose
an a1,6-linked mannosyl moiety and serve as signals for Yos9p-
mediated disposal. Eliminating the activity of Alg3p in cells lack-
ing the HTM1 gene restores breakdown of CPY* and CPY*abcD
(Clerc et al., 2009) (Figure 4C). Interestingly, the same effect
was observed when ALG3 was deleted in cells expressing
Pdi1-1p (Figure 4D). These results indicated that the pdi1-1
phenotype was suppressed by altering the structure of glycans
linked to proteins and suggested that Pdi1p was engaged in
manufacturing the Man7GlcNAc2glycan prior to Yos9p-medi-
The pdi1-1 Mutation Disrupts Association of Pdi1p
and Htm1p and Decreases Htm1p Stability
Next, we asked whether the pdi1-1 mutation affected the ability
of Pdi1p to associate with Htm1p and employed nondenaturing
immunoprecipitation experiments. C-terminally tagged Htm1p
(Htm1pmyc) was expressed from its chromosomal locus. Yeast
cells were lysed and Htm1pmycand Pdi1p were precipitated
with anti-Myc and anti-Pdi1p antibodies, respectively (Figures
Function of the Pdi1p-Htm1p Complex
784 Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc.
efficiently coprecipitated with Htm1pmyc, and vice versa.
Contrarily, Pdi1-1p was almost absent from the Htm1pmyc
precipitate, and Htm1p was barely detectable when Pdi1-1p
was pulled down, suggesting that the association of Htm1p
and Pdi1-1p was abrogated because of the L313P mutation.
Figure 2. The Pdi1p-Htm1p Complex Processes Man8GlcNAc2Glycans In Vitro
(A) Denatured glycoproteins from3H-mannose-labeled wild-type strain YWO1 were incubated with GST or Pdi1-Htm1p complex. When indicated, EDTA was
added. After incubation, oligosaccharides were purified and analyzed as described in Figure 1B. Displayed are representative HPLC elution profiles. M7,
Man7GlcNAc2; M8, Man8GlcNAc2.
(B)3H-labeled N-linked glycans from YWO1 were liberated from polypeptides using PNGase F and purified. Oligosaccharides were incubated with GST or
Pdi1p-Htm1p, purified, and analyzed as described in Figure 1B.
(C)Radiolabeled glycoproteins fromYJU90(Dmns1)wereisolatedas inFigure1A andincubatedwiththeindicatedpurifiedproteins.NLO profiles wererecorded,
and representative HPLC elution profiles are displayed.
(D) Relative peak areas of M7 and M8 from three independent experiments, as shown in (A) and (B), were quantified. Error bars indicate the standard deviation of
mean (SEM) values. P values were determined using paired Student’s t test: *p < 0.05; **p < 0.01.
(E) Relative peak areas M7, M8, and M9 from four independent experiments, as shown in (C), were quantified relative to the sum of all peak areas. P values:
*p < 0.01; **p < 0.001; ***p < 0.0001. See also Figure S1.
Function of the Pdi1p-Htm1p Complex
Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc. 785
Moreover, our findings indicated that only a small fraction of
Pdi1p associated with Htm1p. Although we depleted an HA-
HDEL-tagged version of Pdi1p from cell extracts, Htm1pmyc
was not codepleted in this experiment (Figure S4). It is possible
thatthe Pdi1p-Htm1pcomplex isnotstableunder theconditions
used or the addition of the tag to Pdi1p affected complex forma-
tion. Alternatively, Htm1p might have other yet unknown binding
Importantly, we also observed decreased steady-state levels
of Htm1p in pdi1-1 cells (Figure 5A, compare lanes 3 and 4)
and we, therefore, monitored the stability of Htm1p in pulse-
chase experiments. In wild-type cells, the level of labeled
Figure 3. The pdi1-1 Mutation Only Affects
Breakdown of Glycoproteins, Not Oxidative
(A) Location of the pdi1-1 (L313P) mutation in the
crystal structure of Pdi1p. The four thioredoxin-
like domains (a, b, b0, a0) and two active sites
(CxxC) are indicated.
(B) Serial 10-fold dilutions of logarithmically
growing strains W303a [WT] and KKY415 [pdi1-1]
were spotted on YPD and YPD containing 5 mM
DTT and incubated at 30?C or 37?C for 2 days.
(C) Proteins of logarithmically growing strains
W303a and KKY415 were labeled with
cysteine/methionine for 5 min. Samples were
taken at the indicated time after initiating a chase
by addition of nonradioactive amino acids. CPY
was immunoprecipitated from whole cell extracts
and analyzed by SDS-PAGE. Representative
phosphor scan images of the gels are shown.
Signal intensities of precursor and matured forms
were quantified, and the ratio was plotted against
time. The data reflect three independent experi-
ments with error bars indicating the SEM values.
(D) As in (C), but with Gas1p immunoprecipitated.
(E) As in (C), but with CPY*abcDHAwas ex-
pressed in KKY854 (WT, pWX60) and KKY853
(pdi1-1, pWX60), and cells were labeled for
10 min. CPY*abcDHAwas immunoprecipitated
and analyzed. Representative phosphor scan
images are shown. The intensities of CPY*abcDHA
signals were quantified relative to the signal at the
start of the chase. The data reflect three inde-
pendent experiments witherror bars indicating the
(F) As in (E), but with ngPrAD295-331HAwas ex-
pressed in KKY622 (WT, pKK261), KKY789
(pdi1-1, pKK261) and KKY624 (Dhrd1, pKK261).
See also Figure S2.
Htm1p was reduced by 25% in cell
extracts after a chase period of 2 hr (Fig-
ure 5C). When we introduced the pdi1-1
mutant, clearance of labeled Htm1p was
increased (40% loss after 2 hr). Because
Htm1p does not carry an ER-retention
Pdi1p that carries an HDEL sequence
might result in leakage of Htm1p into the
Golgi. In fact, Htm1p displayed an altered
electrophoretic mobility in pdi1-1 cells, arguing for Golgi modifi-
cation of attached glycans and breakdown of the protein in the
vacuole (Figures 5A and 5C). Alternatively, Htm1p can be
removed from the ER by the ERAD system. When Htm1p was
overexpressed in wild-type cells, the protein was stable (Fig-
ure 5D). However, when we introduced the pdi1-1 mutation,
Htm1p was rapidly lost from cell extracts. When HRD1, the cata-
lytic subunit of the HRD ligase, was deleted in addition, Htm1p
was stabilized. These findings indicated that a significant frac-
tion of Htm1p was removed from the ER by the HRD pathway
when not associated with Pdi1p. Inaddition, the cellular concen-
tration of Pdi1p exceeded that of Htm1p and, thus, Pdi1p was
Function of the Pdi1p-Htm1p Complex
786 Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc.
able to bind overexpressed Htm1p and prevented its degrada-
tion. The observation that Htm1p is unstable when binding to
Pdi1p is disrupted could also explain why we were not able to
express Htm1p in insect cells without its complex partner.
The pdi1-1 Mutation Diminishes Production
of Man7GlcNAc2by Htm1p In Vivo
Finally, we asked whether the pdi1-1 mutation also interfered
with the mannosidase function of Htm1p. We analyzed NLO
profiles of wild-type, pdi1-1, and Dhtm1 strains that harbored
either a control plasmid or a 2m plasmid encoding for an expres-
sion copy of HTM1. We observed a reduced Man7GlcNAc2
rable to the situation observed in Dhtm1 cells and validated our
assumption that Htm1p and Pdi1p act in the same pathway.
Overexpression of HTM1 in wild-type and Dhtm1 cells increased
the Man7GlcNAc2 signal. The pdi1-1 mutation inhibited an
increase in the Man7GlcNAc2population when HTM1 was over-
expressed. We concluded that the function of Pdi1p-Htm1p is
limiting in degradation of glycoproteins. A small, mass-action-
driven increase of mannosidase activity upon overexpression
of HTM1 in pdi1-1 cells results in increased disposal of
CPY*abcD (Figure 4B), but is not sufficient to yield elevated
levels of Man7GlcNAc2in the NLO profiles.
We evaluated the structure of N-linked glycans in cells defi-
cient for HRD1. Unfolded polypeptides accumulate in these
cells, and we expected an increase of the Man7GlcNAc2pool
when compared to wild-type. However, overexpression of
HTM1 in Dhrd1 cells did not result in the anticipated outcome
(Figures 6C and 6D). The result suggested that, under the condi-
tions employed, the majority of Man7GlcNAc2oligosaccharides
were present on folded proteins that were not disposed of by
the HRD pathway. More interestingly, these data also implied
that the activity of the Pdi1p-Htm1p complex was decreased
when ERAD was blocked, indicating that the flux through the
degradation pathway was regulated by the activity of the
In this study, we provided in vitro and in vivo evidence that the
yeast protein disulfide isomerase Pdi1p and the mannosidase
Htm1p assembled into an operative complex initiating clearance
of unfolded glycoproteins from the ER. Using recombinant
proteins, we demonstrated that the Pdi1p-Htm1p complex
acted as a peptide-dependent exomannosidase, generating
N-linked Man7GlcNAc2 oligosaccharides that can serve as
signals for degradation by the HRD ligase. In contrast to
Mns1p, which consumed the entire Man9GlcNAc2NLO pool,
Man8GlcNAc2under the experimental conditions used (Figures
1 and 2). This outcome was consistent with earlier observations
that overexpression of HTM1 in vivo results in about 15% of
Man8GlcNAc2converted to Man7GlcNAc2(Clerc et al., 2009)
and that not all glycans attached to a protein are used as a signal
for ERAD. Only the very C-terminal glycan bound to CPY* and
clearance of the protein from the ER (Kostova and Wolf, 2005;
Figure 4. Overexpression of HTM1 or Deletion of ALG3 Supresses
the pdi1-1 Mutation
(A) Proteins of strains KKY855 (wild-type [WT], pWX60, pDN251) and KKY856
(wild-type, pWX60, pKK225) were radioactively labeled for 10 min and decay
kinetics of CPY*abcDHAwere determined as described (Figures 3C and 3E).
Representative phosphor scan images are shown. The quantified data reflect
three independent experiments with error bars indicating the SEM values.
(B) As in (A), but strains KKY857 (pdi1-1, pWX60, pDN251) and KKY858
(pdi1-1, pWX60, pKK225) were used.
(C) As in (A), but strains KKY853 (wild-type [WT], pWX60), KKY899 (Dhtm1,
pWX60), and KKY893 (Dhtm1, Dalg3, pWX60) were used.
(D) As in (A), but strains KKY853 (wild-type [WT], pWX60), KKY854 (pdi1-1,
pWX60), and KKY884 (pdi1-1, Dalg3, pWX60) were used. See also Figure S3.
Function of the Pdi1p-Htm1p Complex
Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc. 787
Spear and Ng, 2005). Moreover, distinct regions within a protein
appear to be essential for effective degradation (Spear and Ng,
2005; Vashist et al., 2001; Xie et al., 2009). Taken together, these
findings imply that the mannosidase activity of the Pdi1p-Htm1p
complex is restricted to glycans that are in a compatible spatial
orientation to a distinct, presumably unfolded, polypeptide
sequence bound by the complex. Indeed, the Pdi1p-Htm1p
mannosidase did not efficiently process free Man8GlcNAc2
oligosaccharides (Figure 2B).
Pdi1p has a modular and dynamic domain structure, and the
functions of its four thioredoxin domains (a, b, b0, and a0) are
asymmetrical; while domains a and a0are involved in thiol
exchange reactions, b and b0are catalytically inactive, and
domain b0is considered the primary substrate-binding site (Ha-
tahet and Ruddock, 2009; Klappa et al., 1998). However, b0is
not only involved in client-protein capture; in addition, it also
serves as the main binding interface for partner proteins of
PDI1 or its homologs (Hatahet and Ruddock, 2009). The P
domain of calnexin binds to a highly charged region of the
b0domain of human ERp57, located opposite to the substrate
binding surface (Ellgaard and Frickel, 2003; Kozlov et al.,
2006; Russell et al., 2004). Conversely, mutations in the
substrate binding site of human PDI b0domain do not inhibit
its assembly into collagen prolyl 4-hydroxylase tetramers (Koi-
vunen et al., 2005; Pirneskoski et al., 2004). Introducing a single
point mutation into b0domain of Pdi1p disrupted its ability to
bind Htm1p, and we speculate that the Pdi1p-Htm1p complex
follows the same structural theme as described for ERp57-
CNX and the collagen prolyl 4-hydroxylase tetramers. The CT
domain of Htm1p might bind to the b0domain of Pdi1p; such
an arrangement would recruit Htm1p to Pdi1p substrates and
enable the mannosidase to process N-linked Man8GlcNAc2
of mammalian cells is more complex. It is extended by the
calnexin-calreticulin cycle, and there are multiple members of
the GH47 family found in the ER: ER a1,2-mannosidase-I (ER-
ManI); and three homologs of Htm1p, i.e., EDEM1, EDEM2,
and EDEM3 (Aebi et al., 2010; Olivari and Molinari, 2007). Inter-
estingly, EDEM1 forms a complex with BiP and ERdj5, a DnaJ-
type chaperone with six thioredoxin-like domains (Hagiwara
et al.,2011; Ushioda et al., 2008). In vitro and in vivo experiments
revealed that it can act as a reductase, cleaving disulfide bonds
in unfolded model substrates, and can accelerate the turnover of
disulfide bonds containing EDEM1 substrates (Hagiwara et al.,
2011; Ushioda et al., 2008). In view of our results, we speculate
Figure 5. The pdi1-1 Mutation Disrupts the
(A) Whole cell extracts were prepared from strains
W303a (wild-type; loaded in lanes 1, 5, 9), KKY415
(pdi1-1; lanes 2, 6, 10), YRG421 (wild-type,
HTM1myc; lanes 3, 7, 11), and YRG422 (pdi1-1,
HTM1myc; lanes 4, 8, 12), and Htm1pmycwas
immunoprecipitated. Input (in), flowthrough (ft),
and precipitated (IP) fractions were analyzed by
SDS-PAGE and western blotting (WB). Samples
for detection of the precipitated protein were
diluted 5-fold (WB anti-Myc, lanes 9–12). The
symbol ‘‘+’’ indicates the presence of Htm1pmyc,
wild-type PDI1, and pdi1-1. The symbol ‘‘*’’ indi-
cates that the secondary antibody used in the WB
also binds to the heavy chain of the antibody used
(B) As in (A), but with Pdi1p immunoprecipated
fromlysatesof YRG421(lanes 1,3,5)andYRG422
(lanes 2, 4, 6). Samples for detection of the
precipitated protein were diluted 5-fold (WB anti-
Pdi1p, lanes 5 and 6).
(C) Cells from strains W303a and KKY415 were
radioactively labeled, and degradation kinetics of
Htm1p was determined as described (Figures 3C
and 3E). Htm1p was immunoprecipitated by anti-
Htm1p antibodies. Representative phosphor scan
images are shown. The quantified data reflect
three independent experiments with error bars
indicating the SEM values.
KKY866 (wild-type [WT], pKK207), KKY868 (pdi1-
1,pKK207),and KKY917 (pdi1-1,Dhrd1, pKK207).
See also Figure S4.
Function of the Pdi1p-Htm1p Complex
788 Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc.
that, comparably to Pdi1p in yeast, ERdj5 acts as a chaperone
adaptor of EDEM1 to capture substrate proteins.
Based on our results, we suggest a stochastic model of how
the Pdip-Htm1p complex initiates breakdown of unfolded
proteins and contributes to ER protein homeostasis (Figure 7).
Only a fraction of the whole Pdi1p pool associates with Htm1p,
and we speculate that unfolded polypeptides bind to free
Figure 6. The pdi1-1 Mutation Abrogates Generation of Man7GlcNAc2NLOs In Vivo
(A)Cellsfrom strainsW303a (wild-type[WT]),KKY415 (pdi1-1),and ESY679 (Dhtm1),carrying either an empty vector(YEp352; [e.v.]) or a2mplasmid encoding for
Htm1p (pHTM1u-1; [HTM1]) were labeled with3H-mannose, and whole cell lysates were prepared. Glycans were released from polypeptides by PNGase F
treatment, purified, and separated by HPLC. Elution profiles of radiolabeled oligosaccharides were recorded by flow scintillation detection. Retention times of
Man5GlcNAc2(M5), Man7GlcNAc2(M7), and Man8GlcNAc2(M8) are indicated. Displayed are representative HPLC elution profiles. Note that the Man5GlcNAc2
glycan is not a product of the Pdi1p-Htm1p mannosidase (data not shown); its origin and function will be addressed elsewhere.
(B) Relative peak areas of M7 and M8 from three independent experiments, as shown in (A), were quantified. Error bars indicate the standard deviation of mean
(SEM) values. P value was determined using paired Student’s t test: *p < 0.005.
(C) Asin (A),but with strains W303a and ESY279 (Dhrd1) carrying either an empty vector (YEp352;[e.v.]) or a2m plasmid encoding for Htm1p (pHTM1u-1; [HTM1])
(D) As in (B), but quantifying peak areas from (C). P value was determined using paired Student’s t test: *p < 0.005.
Function of the Pdi1p-Htm1p Complex
Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc. 789
to oxidoreductase and chaperone functions, the Pdi1p-Htm1p
complex possesses exomannosidase activity. If present at
a compatible position relative to the Pdi1p-binding site, Htm1p
trims the Man8GlcNAc2 glycan to produce Man7GlcNAc2.
Once released from Pdi1p, the polypeptide can fold, reassociate
with Pdi1p, or bind to the HRD ligase. Binding of a Man7GlcNAc2
oligosaccharide to the lectin subunit Yos9p either increases the
binding of polypeptides or triggers removal of the polypeptide
from the compartment (Carvalho et al., 2006, 2010; Denic
et al., 2006; Gauss et al., 2006; Quan et al., 2008).
It isimportant to note that, within the framework of our hypoth-
esis, Pdi1p does not distinguish between folding intermediates
and polypeptides trapped in an unfolded conformation; Htm1p
can process both specimens. The decisive variable determining
the net output of the system is the time a given polypeptide
displays features recognized by Pdi1p. The slower it folds—
i.e., the longer it exposes hydrophobic patches—the higher the
probability of encountering the Pdi1p-Htm1p complex that trims
Man8GlcNAc2glycans. Mutant proteins like CPY* are trapped in
an unfolded state and, thus, the majority of these proteins is
removed by the ERAD pathway. Mns1p provides directionality
and attenuates the system by producing the glycan detected
by Pdi1p-Htm1p. Since free and Htm1p-bound Pdi1p compete
for the same substrates, the relative levels of Pdi1p modulate
the net output of the system. The higher the relative concentra-
tion of Pdi1p-Htm1p, the higher the fraction of proteins flagged
with a Man7GlcNAc2 glycan. It is noteworthy that in yeast,
PDI1—but not HTM1—is a target of the UPR (Travers et al.,
2000). ER stress condition will, therefore, trigger upregulation
of the oxidoreductase, shift the ratio toward monomeric Pdi1p,
and favor folding over degradation. Thus, turnover of secretory
glycoproteins in the ER is regulated by adjusting the relative
level of the Pdi1p-Htm1p mannosidase to ensure protein
Materials and Reagents
The monoclonal 9E10 anti-Myc antibody was purchased from Sigma-Aldrich
Research Products (Princeton, NJ). Polyclonal anti-Htm1p antibodies were
raised in rabbits against an Htm1p-glutathione S-transferase fusion protein
purified from E. Coli BL21 cells, using the pGEX-4T-3 expression system
(GE Healthcare Life Sciences, Uppsala, Sweden). The polyclonal anti-Gas1p
was produced as described (Spear and Ng, 2003). Anti-Kar2p and anti-
Sec61p antibodies were provided by Peter Walter (University of California
San Francisco, San Francisco, CA). Anti-Pdi1p antibodies were kind gifts
from Karin Ro ¨mish (University of Cambridge, Cambridge, UK) and Jakob
Winther (University of Copenhagen, Copenhagen, Denmark).
Yeast Strains and Plasmids
Standard yeast media and genetic techniques were used as described
(Guthrie and Fink, 1991). Yeast strains used in this study are listed in Table
S1. PCR-based methods were implemented to introduce C-terminal tags
and deletions into the yeast genome (Janke et al., 2004; Knop et al., 1999).
Plasmids and primers used in this study are listed in Tables S2 and S3. Plas-
mids were constructed using standard cloning protocols. See Supplemental
Experimental Procedures for further details on the plasmids. Open reading
frames of all constructs were confirmed by nucleotide sequencing.
Metabolic Labeling with35S and Immunoprecipitation
Cell labeling and pulse-chase analysis were performed as described (Kane-
hara et al., 2010). Cells were grown to log phase and 3 3 107cells were har-
vested and resuspended in 0.9 ml synthetic complete media lacking
methionine and cysteine. Cells were incubated at the appropriate temperature
for 30 min. Then, cells were pulse-labeled in the presence of 82.5 mCi
35S-methionine/cysteine (1 mCi/mmol) (EasyTag EXPRESS35S, Perkin Elmer)
for 5 min (maturation assays) or 10 min (degradation assays). Nonradioactively
labeled methionine/cysteine was added to a final concentration of 2 mM to
initiate the chase. To terminate label incorporation at each time point, TCA
was added to a final concentration of 10%, and cells were subsequently dis-
rupted with the glass bead method. The homogenate was cleared by centrifu-
gation at 18,000 g for 15 min. Pellets were resuspended in 120 ml of TCA
suspension buffer (100 mM Tris-HCl, 3% SDS, and 3 mM DTT) and heated
at 100?C for 15 min. Insoluble material was removed by centrifugation, and
40 ml of the lysate was added to 560 ml of IPS II (13.3 mM Tris-HCl, 150 mM
NaCl, 1% Triton X-100, 0.02% NaN3, 1 mM PMSF, and 1 ml protease inhibitor
cocktail [Sigma-Aldrich]) and the appropriate antiserum. Samples were
Figure 7. A Stochastic Model for the Function of the Pdi1p-Htm1p
Complex in ER Protein Homeostasis
(1–9) Aschematicillustration of thefoldingand glycan-processing pathwaysof
a soluble glycoprotein in the ER is shown. After translocation into the lumen
and covalent attachment of a Glc3Man9GlcNAc2oligosaccharide, initial trim-
mingstepsby glucosidase I and IIgenerateaMan9GlcNAc2(M9)glycan on the
unfolded polypeptide (1). Mns1p trims glycans of unfolded (2) or folded
proteins (4). Pdi1p assists oxidative protein folding irrespective of the glycan
status of the protein (3). Folded proteins leave the ER (5). Free Pdi1p and
Htm1p-bound Pdi1p bind unfolded polypeptides with equal affinities (3) and
(6), but only the Pdi1p-Htm1p can process the Man8GlcNAc2(M8) oligosac-
charide to yield Man7GlcNAc2(M7) (7). The HRD ligase captures unfolded
proteins flagged with M7 (8) and initiates clearance from the compartment (9).
Function of the Pdi1p-Htm1p Complex
790 Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc.
incubated at 4?C for 2 hr and then centrifuged at 18,000 g for 15 min. The
supernatant was recovered and protein A-Sepharose beads were added.
Samples were incubated for 1 hr and washed three times with IPS I (0.2%
SDS, 1% Triton X-100, 20 mM Tris-HCl [pH 7.4], 150 mM NaCl, and 0.02%
NaN3) and once with TBS (20 mM Tris-HCl [pH 7.4] and 150 mM NaCl). Immu-
noprecipitated proteins were eluted with SDS gelsample buffer and separated
by SDS-PAGE. Gels were exposed to phosphor screens for 24–48 hr and the
exposed screens were scanned using a phosphorimager (Typhoon, GE
Healthcare). Signals were quantified using ImageQuant TL software (GE
Healthcare). All data plots reflect three independent experiments with SEM
indicated with error bars.
3H-Mannose Labeling and NLO Profiling
To analyze NLO profiles of yeast cells, a modified version of the protocol
described was implemented (Clerc et al., 2009; Jakob et al., 1998). Yeast cells
were grown in appropriate medium at 30?C to early logarithmic phase corre-
sponding to an OD600of 0.8–1.2. An equivalent of 50 OD cells (5 3 108cells)
was harvested and washed once with 25 ml of YP0.1D (1% yeast extract,
2% peptone, and 0.1% glucose). Cells were resuspended in 200 ml YP0.1D
and incubated at 30?C with shaking (1000 rpm) for 5 min. Metabolic labeling
was started by adding 100 ml YP0.1D containing 100 mCi D-2-3H-mannose
(23.7 Ci/mmol, Perkin Elmer) and performed at 30?C for 20 min. To stop the
bated on ice for 5 min. After centrifugation at 20,000 g for 5 min, the precipitate
was washed twice with ice-cold acetone and dried at 50?C for 10 min. The
pellet was resuspended in 200 ml buffer S2 (50 mM sodium buffer [pH 7.5],
0.4% SDS, and 40 mM DTT) and vortexed with glass beads at 50?C for
60 min. The slurry was extracted twice with 250 ml S2 buffer and once with
200 ml S2 buffer. Supernatants were collected, cleared by centrifugation,
and 500 ml of the lysate was transferred to a new tube. Fifty five ml of
a 450 mM iodoacetamide was added, the mixture incubated at 37?C for
15 min, and 62 ml of NP-40 and 69 ml of 500 mM sodium buffer pH 7.5 were
added. This mixture was used as substrate for the in vitro mannosidase assay.
Finally, the sugars were cleaved from the polypeptides with 1 ml PNGase (NEB)
at 37?C for 4 hr or 16 hr. Clean-up and HPLC analysis of oligosaccharides and
data evaluation was performed as described (Clerc et al., 2009; Jakob et al.,
Yeast cells weregrown at30?C inYPDto early logarithmic phase, correspond-
ing to an OD600of 0.8 to 1.2. An equivalent of 50 OD cells (5 3 108cells) was
harvested and washed once with ice-cold 1 mM PMSF in H20. Cells were
washed with 1 ml ice-cold IP33 (50 mM HEPES-NaOH [pH 7.2], 50 mM
NaCl, 125 mM KOAc, 2 mM MgCl2, 1 mM EDTA, 3% glycerol, 0.5% Nonidet
P-40, and 1 mM PMSF), resuspended in 400 ml IP33, and disrupted with the
glass beads method. The slurry was extracted twice with 750 ml IP33, and
the supernatants were collected. Cellular debris was removed by high speed
centrifugation (20,000 g at 4?C for 12 min). Htm1pmycand Pdi1p were precip-
itated from the supernatant by addition of 1 ml anti-Myc (Sigma-Aldrich,
beads (GE Healthcare,17-52800-1) and incubation at 4?C for 3 hr. Beads were
recovered by centrifugation and washed three times with 1 ml of IP33. Bound
proteins were eluted by adding 50 ml of SDS sample buffer and incubation at
65?C for 15 min. When indicated, samples were diluted 1:5 in sample buffer.
Proteins were separated by SDS-PAGE and immunoblotting, using the indi-
Protein Expression and Purification in Insect Cells
or GST were produced according to the manufacturer’s instructions (Invitro-
gen). Following three rounds of amplification, the virus stocks were used for
protein expression. For small scale analysis of protein expression, 2 3 106
Spodoptera frugiperda (Sf9) cells were infected with 150 ml baculovirus solu-
tion and incubated at 27?C for 72 hr. Cells were harvested, washed once
with ice-cold PBS, and lysed in 250 ml PBS-Tx (PBS + 1% Triton X-100) sup-
plemented with 1 mM E-64 (Roche Applied Science) at 4?C for 20 min. Insol-
uble material was removed by centrifugation at 20,000 g for 5 min, and the
supernatant was used for further experiments. For protein purification,
16.7 3 106cells were infected with 1.3 ml virus suspension. Cells were lysed
in 1.5 ml PBS-Tx containing E-64, and recombinant proteins were bound to
100 ml GSH-Sepharose beads (50% suspension, GE Healthcare) at 4?C for
4 hr. Sepharose beads were washed three times with ice-cold PBS and stored
as a 50% suspension in PBS at 4?C.
In Vitro Mannosidase Assay
Whole-cell lysates containing denatured, reduced, and alkylated glycopro-
teins with3H-mannose labeled NLOs were prepared, implementing the NLO
extraction protocol described above. Prior to addition of PNGase F, 20 ml
GSH-Sepharose beads loaded with purified Mns1p or 100 ml of GSH-beads
loaded with Pdi1p-Htm1p was added and incubated for 16 hr. Beads were
removed by centrifugation, and NLOs were cleaved by PNGase F. Glycans
were isolated and analyzed as described above.
Supplemental Information includes four figures, three tables, Supplemental
Experimental Procedures, and Supplemental References and can be found
with this article online at doi:10.1016/j.molcel.2011.04.027.
We thank Karelia Velez de Joerg, Jeremy Brodhead, Sandy Toh, Songyu
Wang, and Jun Yang for excellent technical assistance; Anna Mohr and Chris-
tine Neupert for critical comments on experiments and the manuscript; Karin
Ro ¨misch, Peter Walter, Jakob Winther, and Thomas Sommer for providing
materials. The identification of the pdi1-1 allele was part of an expanded
UPR synthetic lethal screen carried out by Woong Kim and isolation of the
complementing PDI1 clone by Nurzian Ismail. Funding for this work was
provided by funds from the Swiss National Science Foundation (grant
3100317.127098 for M.A.), the Swiss Federal Institute of Technology, and
the Singapore Millennium Foundation. R.G. received an EMBO long-term
fellowship. K.K. received a grant from the Japan Society for the Promotion
Received: December 20, 2010
Revised: March 18, 2011
Accepted: April 18, 2011
Published: June 23, 2011
Aebi, M., Bernasconi, R., Clerc, S., and Molinari, M. (2010). N-glycan
structures: recognition and processing in the ER. Trends Biochem. Sci. 35,
Buchberger, A., Bukau, B., and Sommer, T. (2010). Protein quality control in
the cytosol and the endoplasmic reticulum: brothers in arms. Mol. Cell 40,
Cai, H., Wang, C.C., and Tsou, C.L. (1994). Chaperone-like activity of protein
disulfide isomerase in the refolding of a protein with no disulfide bonds. J. Biol.
Chem. 269, 24550–24552.
Carvalho, P., Goder, V., and Rapoport, T.A. (2006). Distinct ubiquitin-ligase
complexes define convergent pathways for the degradation of ER proteins.
Cell 126, 361–373.
Carvalho, P., Stanley, A.M., and Rapoport, T.A. (2010). Retrotranslocation of
a misfolded luminal ER protein by the ubiquitin-ligase Hrd1p. Cell 143,
Clerc, S., Hirsch, C., Oggier, D.M., Deprez, P., Jakob, C., Sommer, T., and
Aebi, M. (2009). Htm1 protein generates the N-glycan signal for glycoprotein
degradation in the endoplasmic reticulum. J. Cell Biol. 184, 159–172.
Denic, V., Quan, E.M., and Weissman, J.S. (2006). A luminal surveillance
complex that selects misfolded glycoproteins for ER-associated degradation.
Cell 126, 349–359.
Function of the Pdi1p-Htm1p Complex
Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc. 791
Ellgaard, L., and Frickel, E.-M. (2003).Calnexin, calreticulin, and ERp57: team-
mates in glycoprotein folding. Cell Biochem. Biophys. 39, 223–247.
Ellis,R.W., Freedman, R.B., and Tuite, M.F. (1991). Protein disulfide isomerase
is essential for viability in Saccharomyces cerevisiae. Gene 108, 81–89.
Gauss, R.,Jarosch, E.,Sommer, T.,and Hirsch, C.(2006). A complex of Yos9p
and the HRD ligase integrates endoplasmic reticulum quality control into the
degradation machinery. Nat. Cell Biol. 8, 849–854.
Gilbert, H.F. (1997). Protein disulfide isomerase and assisted protein folding.
J. Biol. Chem. 272, 29399–29402.
Gillece, P., Luz, J.M., Lennarz, W.J., de La Cruz, F.J., and Ro ¨misch, K. (1999).
Export of a cysteine-free misfolded secretory protein from the endoplasmic
reticulum for degradation requires interaction with protein disulfide isomerase.
J. Cell Biol. 147, 1443–1456.
Guthrie, C., and Fink, G.R. (1991). Guide to Yeast Genetics and Molecular
Biology. Methods in Enzymology, Volume 194 (San Diego, CA: Academic
Hagiwara,M.,Maegawa, K., Suzuki, M.,Ushioda, R.,Araki, K.,Matsumoto,Y.,
Hoseki, J., Nagata, K., and Inaba, K. (2011). Structural basis of an ERAD
pathway mediated by the ER-resident protein disulfide reductase ERdj5.
Mol. Cell 41, 432–444.
Hatahet, F., and Ruddock, L.W. (2009). Protein disulfide isomerase: a critical
evaluation of its function in disulfide bond formation. Antioxid. Redox Signal.
Helenius,A.,andAebi, M.(2004).RolesofN-linked glycans intheendoplasmic
reticulum. Annu. Rev. Biochem. 73, 1019–1049.
Hirsch, C., Gauss, R., Horn, S.C., Neuber, O., and Sommer, T. (2009). The
ubiquitylation machinery of the endoplasmic reticulum. Nature 458, 453–460.
Jakob, C.A., Burda, P., Roth, J., and Aebi, M. (1998). Degradation of misfolded
endoplasmic reticulum glycoproteins in Saccharomyces cerevisiae is deter-
mined by a specific oligosaccharide structure. J. Cell Biol. 142, 1223–1233.
Jakob, C.A., Bodmer, D., Spirig, U., Battig, P., Marcil, A., Dignard, D.,
Bergeron, J.J., Thomas, D.Y., and Aebi, M. (2001). Htm1p, a mannosidase-
like protein, is involved in glycoprotein degradation in yeast. EMBO Rep. 2,
Janke, C., Magiera, M.M., Rathfelder, N., Taxis, C., Reber, S., Maekawa, H.,
Moreno-Borchart, A., Doenges, G., Schwob, E., Schiebel, E., and Knop, M.
(2004). A versatile toolbox for PCR-based tagging of yeast genes: new fluores-
cent proteins, more markers and promoter substitution cassettes. Yeast 21,
Jelinek-Kelly, S., and Herscovics, A. (1988). Glycoprotein biosynthesis in
Saccharomyces cerevisiae. Purification of the alpha-mannosidase which re-
moves one specific mannose residue from Man9GlcNAc. J. Biol. Chem.
Jelinek-Kelly, S., Akiyama, T., Saunier, B., Tkacz, J.S., and Herscovics, A.
(1985). Characterization of a specific alpha-mannosidase involved in oligosac-
charide processing in Saccharomyces cerevisiae. J. Biol. Chem. 260, 2253–
Kanehara, K., Xie, W., and Ng, D.T.W. (2010). Modularity of the Hrd1 ERAD
complex underlies its diverse client range. J. Cell Biol. 188, 707–716.
Kawaguchi, S., Hsu, C.-L., and Ng, D.T.W. (2010). Interplay of substrate reten-
tion and export signals in endoplasmic reticulum quality control. PLoS ONE 5,
Klappa, P., Ruddock, L.W., Darby, N.J., and Freedman, R.B. (1998). The
b0domain provides the principal peptide-binding site of protein disulfide isom-
erase but all domains contribute to binding of misfolded proteins. EMBO J. 17,
Knop, M., Siegers, K., Pereira, G., Zachariae, W., Winsor, B., Nasmyth, K.,
and Schiebel, E. (1999). Epitope tagging of yeast genes using a PCR-based
strategy: more tags and improved practical routines. Yeast 15 (10B),
Koivunen, P., Salo, K.E., Myllyharju, J., and Ruddock, L.W. (2005). Three
binding sites in protein-disulfide isomerase cooperate in collagen prolyl
4-hydroxylase tetramer assembly. J. Biol. Chem. 280, 5227–5235.
Kostova, Z., and Wolf, D.H. (2005). Importance of carbohydrate positioning in
the recognition of mutated CPY for ER-associated degradation. J. Cell Sci.
Kozlov, G., Maattanen, P., Schrag, J.D., Pollock, S., Cygler, M., Nagar, B.,
Thomas, D.Y., and Gehring, K. (2006). Crystal structure of the bb0domains
of the protein disulfide isomerase ERp57. Structure 14, 1331–1339.
Lee, S.-O., Cho, K., Cho, S., Kim, I., Oh, C., and Ahn, K. (2010). Protein disul-
phide isomerase is required for signal peptide peptidase-mediated protein
degradation. EMBO J. 29, 363–375.
Ma ¨a ¨tta ¨nen, P., Gehring, K., Bergeron, J.J., and Thomas, D.Y. (2010). Protein
quality control in the ER: the recognition of misfolded proteins. Semin. Cell
Dev. Biol. 21, 500–511.
Nakatsukasa, K., Nishikawa, S., Hosokawa, N., Nagata, K., and Endo, T.
(2001). Mnl1p, an alpha -mannosidase-like protein in yeast Saccharomyces
cerevisiae, is required for endoplasmic reticulum-associated degradation of
glycoproteins. J. Biol. Chem. 276, 8635–8638.
Olivari,S.,andMolinari, M.(2007).Glycoproteinfolding andtheroleofEDEM1,
EDEM2 and EDEM3 in degradation of folding-defective glycoproteins. FEBS
Lett. 581, 3658–3664.
Pirneskoski, A.,Klappa, P.,Lobell,M.,Williamson, R.A.,Byrne,L.,Alanen,H.I.,
Salo, K.E.H., Kivirikko, K.I., Freedman, R.B., and Ruddock, L.W. (2004).
Molecular characterization of the principal substrate binding site of the ubiqui-
tous folding catalyst protein disulfide isomerase. J. Biol. Chem. 279, 10374–
Quan, E.M., Kamiya, Y., Kamiya, D., Denic, V., Weibezahn, J., Kato, K., and
Weissman, J.S. (2008). Defining the glycan destruction signal for endoplasmic
reticulum-associated degradation. Mol. Cell 32, 870–877.
Russell, S.J., Ruddock, L.W., Salo, K.E.H., Oliver, J.D., Roebuck, Q.P.,
Llewellyn, D.H., Roderick, H.L., Koivunen, P., Myllyharju, J., and High, S.
for endoplasmic reticulum lectin association. J. Biol. Chem. 279, 18861–
Sakoh-Nakatogawa, M., Nishikawa, S.-I., and Endo, T. (2009). Roles of
protein-disulfide isomerase-mediated disulfide bond formation of yeast
Mnl1p in endoplasmic reticulum-associated degradation. J. Biol. Chem. 284,
Serve, O., Kamiya, Y., Maeno, A., Nakano, M., Murakami, C., Sasakawa, H.,
Yamaguchi, Y., Harada, T., Kurimoto, E., Yagi-Utsumi, M., et al. (2010).
Redox-dependent domain rearrangement of protein disulfide isomerase
coupled with exposure of its substrate-binding hydrophobic surface. J. Mol.
Biol. 396, 361–374.
Song, J.L., and Wang, C.C. (1995). Chaperone-like activity of protein disulfide-
isomerase in the refolding of rhodanese. Eur. J. Biochem. 231, 312–316.
Spear, E.D., and Ng, D.T. (2003). Stress tolerance of misfolded carboxypepti-
dase Y requires maintenance of protein trafficking and degradative pathways.
Mol. Biol. Cell 14, 2756–2767.
Spear, E.D., and Ng, D.T.W. (2005). Single, context-specific glycans can
target misfolded glycoproteins for ER-associated degradation. J. Cell Biol.
Stolz, A., and Wolf, D.H. (2010). Endoplasmic reticulum associated protein
degradation: a chaperone assisted journey to hell. Biochim. Biophys. Acta
Travers, K.J., Patil, C.K., Wodicka, L., Lockhart, D.J., Weissman, J.S., and
Walter, P. (2000). Functional and genomic analyses reveal an essential coordi-
nation between the unfolded protein response and ER-associated degrada-
tion. Cell 101, 249–258.
Tsai, B., Rodighiero, C., Lencer, W.I., and Rapoport, T.A. (2001). Protein disul-
fide isomerase acts as a redox-dependent chaperone to unfold cholera toxin.
Cell 104, 937–948.
Function of the Pdi1p-Htm1p Complex
792 Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc.
Ushioda, R., Hoseki, J., Araki, K., Jansen, G., Thomas, D.Y., and Nagata, K. Download full-text
(2008). ERdj5 is required as a disulfide reductase for degradation of misfolded
proteins in the ER. Science 321, 569–572.
Vashist, S., Kim, W., Belden, W.J., Spear, E.D., Barlowe, C., and Ng, D.T.
(2001). Distinct retrieval and retention mechanisms are required for the quality
control of endoplasmic reticulum protein folding. J. Cell Biol. 155, 355–368.
Wang, C.C., and Tsou, C.L. (1993). Protein disulfide isomerase is both an
enzyme and a chaperone. FASEB J. 7, 1515–1517.
Xie, W., Kanehara, K., Sayeed, A., and Ng, D.T. (2009). Intrinsic conforma-
tional determinants signal protein misfolding to the Hrd1/Htm1 endo-
plasmic reticulum-associated degradation system. Mol. Biol. Cell 20,
Function of the Pdi1p-Htm1p Complex
Molecular Cell 42, 782–793, June 24, 2011 ª2011 Elsevier Inc. 793