2705
Journal of Cell Science 107, 2705-2717 (1994)
Printed in Great Britain © The Company of Biologists Limited 1994
Retention and retrieval: both mechanisms cooperate to maintain calreticulin
in the endoplasmic reticulum
Birte Sönnichsen1,*, Joachim Füllekrug1, Phuc Nguyen Van1, Wilfried Diekmann2, David G. Robinson2 and
Gottfried Mieskes1,†
1Abt.Klinische Biochemie, University of Göttingen, Robert-Koch-Str.40, 37070 Göttingen, FRG
2Pflanzenphysiologisches Institut, Abt. Cytologie, University of Göttingen, Untere Karspüle 2, 37073
Göttingen, FRG
*Present address: Cell Biology Laboratory, ICRF, PO Box 123, Lincoln’s Inn Fields, London WC2A 3PX, UK
†Author for correspondence
SUMMARY
Many soluble resident proteins of the endoplasmic
reticulum share a COOH-terminal Lys-Asp-Glu-Leu
(KDEL) sequence. Current opinion favours a model in
which these proteins can escape from the endoplasmic
reticulum (ER) by bulk flow and are recognized and sorted
in the Golgi apparatus by binding to a specific KDELreceptor, which returns them to the ER. Through biochemical, morphological and mutational analysis we have
studied the mechanisms that determine the localization of
calreticulin, a soluble 60 kDa KDEL-protein of the ER.
Immunogold labelling established the ER localization of
calreticulin in transfected and nontransfected COS cells.
Although the ER cisternae in transfected cells were enormously dilated and heavily labelled by gold particles we
found no significant label in any other compartment. In
vivo pulse chase experiments with [35S]methionine followed
by biochemical fractionation of calreticulin overexpressing
COS cells (50- to 100-fold) revealed that only a minor part
of labelled calreticulin leaves the ER. Retrieval from the
Golgi was confirmed by a partial redistribution of the
endogenous KDEL-receptor as shown by double immunofluorescence. These data suggest a KDEL-independent
retention of calreticulin in the ER. Further supporting
evidence has come from morphological in vivo studies
using calreticulin-transfected and vesicular stomatitis virus
(ts045)-infected COS cells. Stimulation of vesicular
transport from the ER by releasing the temperaturedependent transport block for the viral G-protein resulted
in a small but significant appearance of calreticulin in a
post-ER compartment. In contrast a calreticulin mutant,
which lacked the Ca2+-binding domain but included the
KDEL sequence, could escape from the ER to a much
higher extent. Secretion of the nonmutated calreticulin was
very low (1-2% of total calreticulin in 3 hours) compared
to the mutated form (18% in 3 hours). Deletion of the
KDEL sequence led to an increase in secretion to 29% over
a 3 hour period, which is much less than expected for a
secretory protein. Taken together these results strongly
support the hypothesis of two independently operating
retention/retrieval mechanisms for calreticulin: one
providing for direct retention in the ER with a very high
capacity and having Ca2+-dependent properties; the other
a KDEL-based retrieval system for escaped calreticulin
present in the Golgi apparatus.
INTRODUCTION
subsequently retrieved from a post ER compartment (Palade,
1975; Pfeffer and Rothman, 1987; Klausner, 1989).
A recognition motif has been described for soluble ER
proteins, the COOH-terminal Lys-Asp-Glu-Leu (KDEL)
sequence, which can vary to some extent (KEEL, HVEL etc.;
Haugejorden et al., 1991; Andres et al., 1990; Munro and
Pelham, 1987). By contrast resident membrane proteins have
a double Lys-motif (Nilsson et al., 1989; Jackson et al., 1990)
as a targeting signal. Evidence is emerging for the existence of
a retrieval mechanism for soluble ER proteins, which involves
recognition of the KDEL signal by a specific KDEL-receptor
in a post ER compartment, segregation from the bulk of the
transported proteins and the targeting to a retrograde transport
pathway back to the ER.
Resident proteins of the endoplasmic reticulum (ER) need to
be sorted from other proteins that follow the secretory pathway
to the trans-Golgi network where they are distributed to their
target compartments, e.g. cell surface, lysosomes or secretory
vesicles. Current opinion favours the idea that transport of
proteins from the ER to the Golgi and further on to the plasma
membrane occurs by default. Consequently a highly selective
and specific mechanism is required that recognizes ER-resident
proteins and causes their localization in this organelle.
Basically the correct localization of ER proteins can be
achieved in one of two ways: either they are retained by
exclusion from export or they enter transport vesicles and are
Key words: endoplasmic reticulum, Golgi, protein transport, KDELreceptor, calreticulin, electron microscopy
2706 B. Sönnichsen and others
The KDEL-receptor was first described in yeast as the
product of the ERD2 gene (Semenza et al., 1990). This gene
encodes an integral membrane protein of 26 kDa, whose
abundance determines the capacity of the HDEL (analogous to
the mammalian KDEL) retrieval system. Subsequently it was
shown that a receptor similar in size and sequence is expressed
in mammalian cells (Lewis and Pelham, 1990, 1992; Hsu et
al., 1992; Tang et al. 1993). This receptor which was localized
in the Golgi apparatus, was found ubiquitously in all cell types
examined and binds the KDEL sequence in vitro (Tang et al.,
1993; Wilson et al., 1993; Townsley et al., 1993). In a recent
model the formation of the receptor-ligand complex in the
Golgi initiates retrograde transport to the ER. The ligand
protein is then released into the ER lumen and the unbound
KDEL-receptor is recycled back to the Golgi (Pelham, 1991).
It remains an open question as to the identity of the Golgi compartment from which proteins can be retrieved to the ER, nor
is it clear what the requirements for this retrieval are.
Experimental evidence for a two-way traffic between ER
and Golgi derives in part from studies with the fungal metabolite brefeldin A (BFA; Lippincott-Schwartz et al., 1990; see
also review by Klausner et al., 1992) and through the morphological analysis of some marker proteins in a putative intermediate compartment (Lippincott-Schwartz, 1989; Schweizer
et al., 1990; Saraste and Svensson, 1991; Hsu et al., 1992;
Hauri and Schweizer, 1992). One effect of BFA is the inhibition of ADP-ribosylation factor (ARF) binding, a low
molecular mass GTP-binding protein, to the donor membrane
(Kahn and Gilman, 1986; Weiss et al., 1989; Serafini et al.,
1991; Donaldson et al., 1992a; Helms and Rothman, 1992;
Tsai et al., 1993) thus preventing the formation of β-COPcoated vesicles (Donaldson et al., 1991, 1992b; Peter et al.,
1993; Orci et al., 1993a,b; Palmer et al., 1993; Ostermann et
al., 1993). The consequence is a blockage in the anterograde
transport and the formation of uncoated tubular structures
derived from the Golgi apparatus, which mediate the retrograde transport of Golgi material to the ER (Orci et al., 1991;
Ladinsky and Howell, 1993; Tang et al., 1993). A similar BFAlike phenotype is induced by the overexpression of a human
KDEL-receptor, the ERD2-like protein ELP-1 (Hsu et al.,
1992; Tang et al., 1993). Together with the redistribution of
the KDEL-receptor from the Golgi to the ER by the overexpression of a KDEL-protein (Lewis and Pelham, 1992; Hsu et
al., 1992; Tang et al., 1993) these results point to a role for the
KDEL-receptor in establishing a balance between membrane
traffic to and from the Golgi apparatus.
According to the bulk flow theory if there is no limitation to
KDEL-proteins exiting the ER and being transported to the
Golgi apparatus, then the concentration of these proteins within
the ER, their rate of transport to the Golgi and back to the ER
should be important factors in maintaining the dynamic
balance between these two compartments. To address the
mechanism by which a resident soluble ER protein is maintained in the ER and how its overexpression can influence
membrane equilibrium between ER and Golgi, we have studied
the localization, transport and sorting of calreticulin (CR)
under different in vivo and in vitro conditions.
Calreticulin has been reported to be a luminal resident
protein of the ER with a molecular mass of 60-63 kDa. It is
characterized by a high Ca2+-binding capacity (Nguyen Van et
al., 1989; for a review see Michalak et al., 1992). It bears a
COOH-terminal KDEL sequence and has been reported to be
N-glycosylated with a terminal galactose residue in rat hepatocytes (Peter et al., 1992). This modification is characteristic
for the trans-Golgi (Kornfeld and Kornfeld, 1985), hence CR
has to be transported to and must be recycled back from this
compartment. The proposed retrieval system from different
stacks of the Golgi apparatus indicates a quality control system
for the removal of KDEL or related proteins at successive
points in the secretory pathway. We isolated calreticulin cDNA
from rat liver and used it for expression studies in COS and
CHO cells. We now report that the overexpressed calreticulin
is efficiently retained in the ER, and provide both biochemical
and morphological evidence that only a minor portion of newly
synthesized calreticulin leaves the ER and reaches the Golgi
where it can induce the partial redistribution of the KDELreceptor. In addition we demonstrate that high overexpression
of calreticulin induces massive dilation of ER cisternae similar
to cells that have been stressed by tunicamycin.
MATERIALS AND METHODS
Chemicals
All reagents were of analytical grade and purchased from Merck
(Darmstadt, FRG), Serva (Heidelberg, FRG), Sigma (München,
FRG), Biomol (Ilvesheim, FRG), and Boehringer (Mannheim, FRG),
unless otherwise stated.
Radioactive uridine-5′-bisphospho-D-[3H]galactose was purchased from Amersham-Buchler (Braunschweig, FRG), Express
[35S]methionine/cysteine labelling mix was from NEN Dupont (Bad
Homburg, FRG).
For transient expression of calreticulin the pCMV2 vector of
Anderson et al. (1989) was used, which was kindly provided by Dr
G. Thiel (Inst. for Genetics, University of Köln, FRG). The h-ERD2
cDNA was expressed using the pBEH vector (Artelt et al., 1988),
which was supplied by Dr C. Peters (Dept of Biochemistry II, University of Göttingen, FRG).
Antibodies
Antibodies against rat liver calreticulin, a peptide-specific antibody
against the C-terminal domain of the KDEL-receptor
(CLYITKVLKGKKLSLPA; for other studies the N-terminal cysteine
was added to the peptide) and BiP were produced in rabbits according
to standard procedures. For double immunofluorescence studies a goat
antibody against rabbit calreticulin was supplied by Dr M. Michalak
(University of Alberta, Edmonton, Canada). Unless otherwise stated
the previously mentioned rabbit antibody against calreticulin was
used. Mouse monoclonal antibodies prepared against p53 (mAb
G1/93; Schweizer et al., 1988) and p63 (mAb G1/296) (Schweizer et
al., 1993) were kindly provided by Dr H. P. Hauri (Dept of Pharmacology, Basel, Switzerland). Monoclonal antibody P5D4 against the
C-terminal domain of VSV G-protein was provided by Dr W. E. Balch
(Scripps Research Inst., La Jolla, California, USA).
Cloning and sequencing of calreticulin
Calreticulin was cloned from a rat liver cDNA library in the λZAPII
vector by immunoscreening and conventional southern hybridization.
A 1.5 kb EcoRI fragment was subcloned into pCMV2 for expression
and M13 mp18 for sequencing according to Sambrook et al. (1989).
Generation of calreticulin mutants
Calreticulin mutants were generated by standard PCR techniques
using the above mentioned cDNA as template and the following
oligonucleotides: sense primer for all mutants (5′-GC AAG CTT
GAA TTC CCT CGG CCC GCC ATG CTC CTT TCG GTG CCG
Retention and retrieval of calreticulin 2707
CTC-3′), antisense for the CR∆C mutant (5′-GC GAA TTC AAG CTT
CTA CAG CTC ATC CTT CTG CTT GTC CTT CAT CTG CTT
CTC-3′), antisense for the CR∆KDEL mutant (5′-GC GAA TTC AAG
CTT CTA GGC TTG GCC AGT GGC ATC CTC CTC ATC TTC3′) and antisense for the CRHVEL mutant (5′-GC GAA TTC AAG CTT
CTA CAG CTC CAC GTG GGC TTG GCC AGT GGC ATC-3′).
PCR products were verified by sequencing, and subcloned into the
expression vector pCMV2.
Cell culture and transfection
COS cells (ATCC CRL-1650) were grown in Dulbecco’s minimal
essential medium (DMEM, Biochrom, Berlin, FRG) with 10% FCS.
Cells were incubated under standard tissue culture conditions. Transfections were carried out by the calcium-phosphate coprecipitation
method according to Kaufman (1990). After 48 hours, cells were
processed for immunofluorescence or subcellular fractionation.
Subcellular fractionation
Gradients were preformed by freezing and thawing of a step gradient
containing 2 ml of a 5% and 2 ml of a 35% (w/v) Nycodenz solution
(Nycomed, Oslo, Norway) in 25 mM Tris-HCl, pH 7.5. All steps of
the fractionation procedure were performed at 4°C. Transfected COS
cells were grown for 48 hours on two 145 cm2 Petri dishes and washed
with PBS (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM
KH2PO4, pH 7.5). Cells were collected in 5 ml homogenization buffer
(130 mM KCl, 25 mM Tris/HCl, pH 7.5) with a rubber policeman
and sedimented for 10 minutes at 500 g. The pellet was resuspended
in 500 µl homogenization buffer and homogenized with a 1 ml pipette
tip (10×) and subsequently by passing through a syringe (10× 22 G/0.7
mm). The homogenate was centrifuged for 10 minutes at 1000 g. The
resulting postnuclear supernatant was layered onto the formed
Nycodenz gradient. After centrifugation for 25 minutes at 100,000 g,
nine fractions were collected from top to bottom.
Gradient characterization
Rotenone-insensitive NADH cytochrome c reductase (Sottocasa et al.,
1967) for the ER and UDP-galactosytransferase (Verdon and Berger,
1983) for Golgi were measured according to standard procedures.
For western blot analysis, membranes of the diluted fractions were
collected by centrifugation at 100,000 g for 1 hour. Proteins of the
resulting pellets were separated on 10% SDS gels (Laemmli, 1970),
and blotted onto nitrocellulose (Sartorius, Göttingen, FRG). The blots
were blocked with PBS-5% dry milk, and incubated with primary
antibodies (1:1000) followed by detection with peroxidase-coupled
secondary antibodies using 4-chloro-1-naphtol and N,N-diethyl-pphenylenediamine as substrates.
Labelling of transfected cells and immunoprecipitation
Cells were grown on 60 mm Petri dishes, preincubated for 60 minutes
with α-MEM without methionine (Gibco BRL, Berlin, FRG), and
labelled with 200 µCi [35S]methionine in 1 ml of the same medium
adjusted with 25 mM HEPES/KOH, pH 7.2. After 15 minutes
labelling medium was exchanged for 1.5 ml of DMEM supplemented
with 2 mM cold methionine and 1 mM cycloheximide for different
chase times. Supernatants and cells were collected separately, and
cells were sedimented at 500 g for 10 minutes. The sediments were
resuspended in 500 µl TIN buffer (50 mM imidazole, pH 7.5, 150
mM NaCl, 0.5% Triton X-100). For immunoprecipitation 40 µl calreticulin antiserum and 50 µl Staphylococcus aureus membranes
(Pansorbin, Calbiochem, Bad Soden, FRG) were added to chase
supernatants and cell homogenates, respectively, and incubated for
one hour. Precipitates were sedimented at 10,000 g for 5 minutes, and
washed three times with 1% Triton X-100, 0.5% sodium desoxycholate in PBS, and twice with PBS. Samples were analysed on 10%
SDS gels, and radioactive calreticulin was quantified using a Bio-Rad
GS-250 molecular imager.
For pulse chase experiments with subsequent subcellular fraction-
ation, cells were grown on 145 cm2 Petri dishes, and labelled for 8
minutes with 1.25 mCi [35S]methionine per dish. Labelled material
was chased for the indicated time intervals. Postnuclear supernatents
were fractionated on Nycodenz gradients, and calreticulin was
immuno-precipitated from ER- and Golgi-enriched fractions sedimented at 100,000 g for 1 hour.
Immunofluorescence microscopy
Cells grown and transfected on coverslips were fixed with 3%
paraformaldehyde in PBS at room temperature for 20 minutes, and
permeabilized with 0.3% Triton X-100 in PBS for 10 minutes. Antibodies were incubated for 1 hour at room temperature in 0.2%
gelatine/PBS. Secondary antibodies were coupled to fluorescein or
tetramethylrhodamine (Dianova, Hamburg, FRG). Coverslips were
mounted in Mowiol and viewed using a Zeiss Axioskop microscope
equipped with a Plan-Neofluar ×100/1.30 objective.
Immunogold labelling
Cells were fixed with 1% formaldehyde, 0.1% glutaraldehyde in PBS
for 1 hour at room temperature. Postfixation was performed for 1 hour
with 0.1% OsO4 in 50 mM potassium phosphate buffer, pH 7. Cells
were stained with 2% aqueous uranyl acetate overnight, dehydrated
through an ethanol series, and embedded in LR-White (London Resin
White (hard grade), Plano, Marburg, FRG) overnight. Embedding
medium was polymerized at 50°C for 16 hours. 90 nm thin sections
were cut by standard procedures. After blocking the samples with 3%
BSA plus 0.1% BSA-C (acetylated BSA, Aurion, Wageningen,
Netherlands) in PBS for 30 minutes, primary antibodies were
incubated in PBS containing 1% BSA for 1 hour. After washing in
PBS containing 1% BSA (5× 5 minutes) sections were exposed to 10
nm gold-conjugated secondary antibodies (diluted 1:100) in PBS plus
0.01% BSA-C for 1 hour. Antigen-conjugated complexes were stabilized with 1% glutaraldehyde for 10 minutes then washed in H2O and
poststained with 3% uranyl acetate and 0.3% aqueous lead citrate for
10 minutes, respectively, and observed in a Philips CM 10 electron
microscope operating at 80 kV.
Vesicular stomatitis virus infection
COS cells were grown on coverslips, and infected with vesicular
stomatitis virus (VSV) strain VSV tsO45 (kindly provided by Dr K.
Simons, EMBL, Heidelberg, FRG) at a multiplicity of 20-50 plaqueforming units per cell for 45 minutes at 32°C α-MEM, 20 mM HEPES
pH 7.2. After incubation for 3 hours at 39.5°C in α-MEM/5% FCS,
COS cells were incubated at 32°C for different time intervals.
RESULTS
Localization of calreticulin is not altered by its
overexpression
We have cloned and sequenced calreticulin (CR) from rat liver
(sequence submitted to EMBL Database Library). Based on
blotting experiments and transfection efficiency, transient
expression in COS cells using the high expression vector
pCMV2, which contained the cytomegalovirus promotor
resulted in a 50- to 100-fold overexpression of CR per cell.
Western blotting revealed a doublet at 62 and 63 kDa for the
expressed protein. A similar doublet is seen in highly purified
CR from rat liver, although the intensity of the slower
migrating band is less prominent. Probing with different anticalreticulin antibodies raised in goat, sheep and rabbit
confirmed that both proteins refer to calreticulin. A COOHterminal KDEL-specific antibody recognizes both bands in
western blots. As both bands showed positive general carbo-
2708 B. Sönnichsen and others
hydrate staining, we assume that a different, yet unknown glycosylation status is responsible for the expression of the two
forms (data not shown).
Biochemical fractionation of transiently transfected COS
cells on a Nycodenz-gradient, followed by a 100,000 g sedimentation to separate soluble (generated during homogenization) from vesicular CR revealed a clear colocalization with
the typical ER markers immunoglobulin heavy chain binding
protein (BiP) and rotenone-insensitive NADH cytochrome c
reductase. No CR could be detected in those fractions bearing
the Golgi marker UDP-galactosyltransferase (Fig. 1A and B).
This result indicates a retention system for CR with a very high
capability, either by retaining CR already in the ER or by a
very efficient and fast KDEL-based retrieval system.
Overexpressed CR leaves the ER only to a small
extent
Since subcellular fractionation simply reflects the steady state
distribution of CR we performed pulse chase experiments as
well as indirect immunofluorescence to see whether newly synthesized CR is able to reach the Golgi. In pulse chase experi-
Fig. 1. Subcellular distribution of calreticulin in transfected and nontransfected COS cells. Postnuclear supernatants were fractionated on
5% to 35% Nycodenz rate zonal gradients by centrifugation for 25
minutes at 100,000 g. Nine fractions were collected starting at the
top. Activities (A) of rotenone-insensitive NADH cytochrome c
reductase (Rot. ins. Cyt. c R.) for ER and UDP-galactosyltransferase
(GalTrase) for Golgi membranes were calculated as % of total
activity in the gradient. Steady state distribution of calreticulin and
BiP in transfected (COS × CR-WT) and non-transfected cells was
determined by western blotting (B). Pulse chase distribution (8
minutes pulse, chase times as indicated) in transfected COS cells (C)
was determined by immunoprecipitation of calreticulin from Golgi
(fraction 4) and ER-enriched fractions (fraction 7). For details see
Materials and Methods.
ments we labelled CR-transfected COS cells for 8 minutes with
[35S]methionine and chased the cells for the times indicated
with 2 mM methionine (Fig. 1C). At each timepoint the cells
were collected, homogenized and fractionated on a Nycodenz
gradient. Membranes of each fraction were sedimented and CR
immunprecipitated using polyclonal rabbit anti-CR antibodies.
Comparison of the relative distribution of CR between Golgi
and ER fractions reveals a small, hardly-detectable part of the
newly synthesized CR in the Golgi fractions after 5 minutes,
which disappears again after 15-30 minutes. We never found
any indication for an increase of BiP in CR-transfected COS
cells.
High overexpression of a KDEL-protein results in the redistribution of the KDEL-receptor to the ER (Lewis and Pelham,
1992; Hsu et al., 1992; Tang et al., 1993) and can be used as
indirect evidence for the sorting of a KDEL-protein in the
Golgi. We therefore performed indirect immunofluorescence
using polyclonal antibodies against CR and the cytoplasmic
tail of the KDEL-receptor (Erd2). In addition we used monoclonal antibodies against two markers of the intermediate compartment p53 and p63. Fig. 2 demonstrates that even with the
highly overexpressed CR only a partial redistribution of the
endogenous KDEL-receptor was observed. In the CR-transfected cells (two transfected cells and one nontransfected cell
are seen in Fig. 2A and B) the KDEL-receptor appeared in
more peripheral dotted structures, which only partially show
the reticulate pattern typical for the ER. A similar observation
was made for the intermediate compartment marker p53, which
in nontransfected cells was located in Golgi-like structures
(perinuclear cap, Fig. 2C and D). Only in the binuclear transfected cell p53 shifted to peripheral dotted and ER-like structures (Fig. 2D). The distribution of p63 did not change (Fig.
2E and F). In all our experiments p63 had an ER localization,
regardless of whether or not the cell was transfected.
Stimulation of vesicular transport does not
significantly increase the exit of overexpressed CR
from the ER
Two assumptions are inherent in the current view of the
KDEL-based retrieval system (Lewis and Pelham, 1992):
(1) diffusion of the KDEL-proteins within the ER is not
limited; and (2) export of these proteins is not selective.
Therefore, by inducing the exit of large amounts of secretory
proteins, we wondered whether stimulation of vesicular
transport would result in an increased leakage of CR from the
ER. To test this hypothesis we transiently transfected COS
cells with CR and infected these cells with the VSV ts045. The
G-protein of this virus accumulates in the ER at the restrictive
temperature of 39.5°C and its vesicular transport can be
induced by lowering the temperature to 32°C (permissive temperature) (Doms et al., 1993). Double immunofluorescence
revealed that at the restrictive temperature both proteins colocalize in the ER (Fig. 3A and B). At 8 minutes after shifting
to the permissive temperature VSV G-protein can be observed
in a mixture of ER, Golgi, and the typical dotted structures
believed to be the intermediate compartment (Fig. 3C;
Schwaninger et al., 1992; Plutner et al., 1992). The CR staining
is restricted to ER and some punctate structures, which colocalize with VSV G-protein (Fig. 3D). After 60 minutes the Gprotein is detectable in the Golgi, whereas CR resumed a pure
ER distribution (Fig. 3E and F). That fraction of CR that tran-
Retention and retrieval of calreticulin 2709
siently reached the dotted structures became more apparent
after transported proteins were allowed to accumulate for 60
minutes at 15°C (Fig. 3G and H). From these experiments we
conclude that only a small part of CR is accessible to transport
vesicles, even under conditions when massive vesicular
transport is induced.
High overexpression of CR leads to dilation of ER
cisternae
We wondered whether there is a morphological basis for the
observation that hardly any CR leaves the ER, despite its high
level of overexpression in COS cells (50- to 100-fold). For this
reason we performed postembedding immunogold labelling on
transiently transfected COS cells. In order to obtain high specificity of labelling the binding of the highly specific polyclonal
CR antibodies was visualized by using a gold-conjugated antirabbit IgG secondary antibody. In nontransfected cells CR was
localized clearly and exclusively to cisternal and tubular rough
ER (arrows in Fig. 4A and C). The high CR overexpressing
COS cells showed by comparison an enormous dilation of the
ER (Fig. 4B), which is filled with CR (arrows in Fig. 4B and
D). Neither in the control cells nor in the transfected cells was
immunogold labelling seen concentrated at the membrane.
Instead it appeared to be spread evenly across the lumen of the
Fig. 2. Double immunofluorescence of
calreticulin (A,C and E), Erd2 (B), and the
intermediate compartment marker proteins p53
(D) and p63 (F) in COS cells transfected with
calreticulin. Note the partial redistribution of
Erd2 and p53 in transfected cells as indicated
by arrows. (Detection of calreticulin with goat
antibody, see Materials and Methods.)
Bar, 10 µm.
ER. No other structures, including the Golgi-like membranes
were found to be labelled by the gold particles (Fig. 4E). These
dilated ER cisternae correlate well with the massive overproduction of CR. Whether these structures represent a parallel
increase of ER membrane synthesis or whether the ER
cisternae simply expanded by the accumulation of CR cannot
be decided on the basis of these experiments.
Taken together the electron microscopic, the immunofluorescence and the biochemical data are seen to support the
hypothesis of a very efficient and highly capable retention
mechanism for CR in the ER.
The C-domain participates in retention of CR
As a consequence of the above results we decided to look for
an additional domain in the CR sequence that might be involved
in ER retention. In order to address this question we constructed
three CR mutants. We considered that removal of the KDEL
sequence should lead to a significant but relatively slow
secretion of the mutated CR (CR∆KDEL) because of the presence
of the putative, additional retention domain. We argued that this
domain might be the high capacity Ca2+-binding domain of CR
(C-domain, Michalak et al., 1992). A hint in this direction was
an earlier report (Booth and Koch, 1989) that addition of a Ca2+chelator to murine fibroblasts resulted in the secretion of protein
2710 B. Sönnichsen and others
Fig. 3. Double immunofluorescence of VSV Gprotein and calreticulin in VSV ts045-infected COS
cells overexpressing calreticulin. G-protein
accumulated in the ER at the restrictive
temperature of 39.5°C for 3 hours. Transport was
monitored in vivo after lowering the temperature to
32°C for the indicated time intervals (A to F).
Colocalization of VSV G-protein and calreticulin
could be observed in the ER at 0 minutes (A and
B), and in vesicular intermediates (arrows, C and
D) as well as in the Golgi region after 8 minutes.
Transport intermediates are more pronounced after
incubation at 15°C for 60 minutes (G and H). Bar,
10 µm.
disulfide isomerase (PDI). We removed this domain whilst
saving the KDEL sequence (CR∆C). In addition we used a
KDEL to HVEL mutation (CRHVEL) to see whether its retention
would be as efficient as the CR wild type (Robbi and Beaufay,
1991). Double immunofluorescence studies on the distribution
of the KDEL-receptor in cells expressing the mutant forms
confirmed our assumptions (Fig. 5). The change of KDEL for
HVEL had no effect and resulted in a very similar pattern to
the nonmutated CR. CRHVEL was located in the typical ER
network (Fig. 5A, one transfected and one nontransfected cell
is seen) and the KDEL-receptor staining showed a shift to more
peripheral elements (compare Figs 5B and 2B). In contrast the
CR∆C mutation (Fig. 5C and D, one transfected and one nontransfected cell is seen) causes a nearly complete redistribution
of the KDEL-receptor to ER-like structures, indicating that
retention of CR∆C in the ER is reduced, and has to be compensated by increased retrieval. The receptor in the nontransfected
cell (small arrow, Fig. 5D) was unaffected. Even the CR∆KDEL
mutant showed an ER pattern (Fig. 5E, one transfected and one
nontransfected cell is seen) although it seems that perinuclear
staining is more pronounced. As expected the KDEL-receptor
distribution remained unaltered (Fig. 5F).
Subcellular fractionation of COS cells transfected with the
mutants CRHVEL, CR∆KDEL and CR∆C confirmed the results
obtained by immunofluorescence analysis. Cells were harvested
48 hours after transfection, the 1000 g supernatant separated on
a Nycodenz gradient and analyzed for the distribution of BiP,
CR and the KDEL-receptor (Erd2) by western blotting. All
gradients were prepared under identical conditions and the
maximum activities of the marker enzymes were consistently
Fig. 4. Immunoelectron microscopy of COS cells overexpressing
calreticulin. In control cells (A and C) calreticulin is localized in
tubular-cisternal ER structures (arrows). In transfected cells (B and
D) calreticulin is highly enriched in extremely dilated ER vesicles
(arrows). No labelling was found in the Golgi apparatus (E). N,
nucleus. Bars: A, 0.5 µm; B, 1 µm; C, 0.2 µm; D, 0.2 µm; E, 0.5 µm.
Retention and retrieval of calreticulin 2711
2712 B. Sönnichsen and others
found in fraction 6 for rotenone-insensitive cytochrome c
reductase (ER) and fraction 4 for UDP-galactosyltransferase
(Golgi) (Fig. 6). The intensity and localization of BiP in fraction
6 (ER) was the same in all experiments. The distribution of CR
or its mutants colocalizes perfectly with BiP regardless of
whether the KDEL sequence or the Ca2+-binding domain was
deleted. In agreement with the results of the morphological
studies the Erd2 distribution depends on the expression of
different CR forms. Both CR and CRHVEL overexpression
caused a shifting of the receptor to fractions 5 and 6 compared
to fractions 3 to 6 in nontransfected COS cells, reflecting a
partial shift from Golgi to ER compartments. As expected,
expression of CR∆KDEL was without significant influence on the
distribution of Erd2 (maximum in fraction 4 and 5). In contrast,
removing the Ca2+-binding domain caused a redistribution of
Erd2 to the ER fractions 5 and 6, with a clear maximum in
fraction 6.
CR∆C mutant shows increased exit from the ER
upon stimulation of vesicular transport
Both the immunofluorescence and the subcellular fractionation
studies gave indirect evidence for the participation of the Cdomain in the retention of CR. To show exit of the CR∆C
mutant from the ER more directly we used the same experimental setup as described in Fig. 3. In this experiment we
showed that only a minor fraction of nonmutated CR leaves
the ER upon stimulation of vesicular transport. Consequently,
removal of an important retention signal should lead to
increased amounts of escaped CR∆C in post-ER compartments.
Conversely loss of the KDEL-signal should not influence ERretention, but escaped protein cannot be rescued by retrieval.
COS cells, transiently transfected with the mutants CR∆KDEL
and CR∆C, were infected with VSV ts045. At 39.5°C the distribution of the VSV G-protein and the mutated CR proteins
(data not shown) was very similar to those shown in Fig. 3A
and B. At 15 minutes after shifting the temperature to 32°C the
CR∆C was located to the same Golgi-like structures as the viral
G-protein (Fig. 7C and D). At 45 minutes the mutated CR
appeared in a mixture of ER-like structures and structures in
which the G-protein was localized (Fig. 7E and F). Only a part
of the CR∆C was redistributed to the ER. The KDEL-based
retrieval system might easily be saturated by this massive
appearance of KDEL-proteins in the Golgi, especially when
considering the finding that expression of CR∆C without additional stimulation of vesicular transport, already led to a substantial redistribution of the KDEL-receptor (Fig. 5C and D).
The situation for CR∆KDEL was as predicted. At 15 minutes
after shifting to the permissive temperature the G-protein
localized clearly to Golgi-like structures (Fig. 8C), whereas
CR∆KDEL remains partially in the ER (right cell in Fig. 8D),
and is partially concentrated in perinuclear structures (left cell
in Fig. 8D). In contrast to the experiments with CR∆C,
CR∆KDEL did not redistribute after 45 minutes to the ER but
moved on with the G-protein to the trans-Golgi network
Fig. 5. Double immunofluorescence of
calreticulin (A,C and E) and Erd2 (B,D and F) in
COS cells overexpressing 3 different calreticulin
mutants: exchange of the KDEL sequence for
HVEL (A) did not alter the distribution of
calreticulin and Erd2 (B) compared to
overexpression results of the wild-type protein
(Fig. 2A and B). Deletion of the C-terminal
calcium-binding domain, but preservation of the
KDEL sequence (C) led to a stronger
redistribution of Erd2 to the endoplasmic
reticulum (D, note the Golgi localization of Erd2
in the non-transfected cell, arrow). After removal
of the KDEL sequence of calreticulin (E)
redistribution of Erd2 could no longer be
observed. Localization of the receptor was
indistinguishable from localization in nontransfected cells (F). (Detection of calreticulin
with goat antibody, see Materials and Methods).
Bars, 10 µm.
Retention and retrieval of calreticulin 2713
(TGN) or to the plasma membrane (Fig. 8E and F; three cells
are depicted, one is transfected).
Deletion of either the retention or the retrieval signal
results in secretion of CR mutants
Both the immunofluorescence and the subcellular fractionation
studies indicate that CR and its mutated forms should be
secreted at different rates. Overexpression of CR should lead,
if at all, only to a very low secretion due to its efficient
retention in the ER. By contrast removal of one of the two
proposed retention/retrieval signals should result in a substantial secretion of CR. Retention of CRKDEL depends only upon
interactions in the ER without the possibility of retrieving
escaped protein, and expression of CR∆C should saturate the
Fig. 6. Subcellular distribution of calreticulin and Erd2 in nontransfected COS cells, and COS cells expressing wild-type
calreticulin (COS × CR-WT), calreticulin with KDEL substituted by
HVEL (COS × CR-HVEL), calreticulin without KDEL (COS ×
CR∆KDEL), and without the C-terminal calcium-binding domain
(COS × CR∆C). For details of fractionation see Fig. 1 and Materials
and Methods.
KDEL-receptor-based retrieval system because it cannot be
retained in the ER. The results of a pulse chase experiment
were as expected (Fig. 9). We labelled transfected COS cells
for 15 minutes with [35S]methionine, chased the cells for 3
hours in the presence of 2 mM unlabelled methionine and
1 mM cycloheximide and immunprecipitated the different CR
forms from the medium and from the collected cells. Overexpression of CR resulted in only a marginal secretion of 1-2%
of total CR, representing normal cell death during the chase
time. The overexpressed mutated forms CR∆C and CR∆KDEL
were secreted to the amounts of 18% and 29%, respectively.
DISCUSSION
CR is a soluble, resident protein of the ER. According to its
Ca2+-binding properties it has been suggested to serve as a
Ca2+-storage protein in non-muscle cells (Nguyen Van et al.,
1989; for review see Michalak et al., 1992). It bears a Cterminal KDEL retrieval sequence and has been reported to be
sorted in the trans-Golgi from other proteins transported along
the secretory pathway (Peter et al., 1992). This led to the
assumption that different soluble ER proteins might be
retrieved from different parts of the secretory pathway,
depending on the conditions required for their binding to the
KDEL-receptor (ionic conditions, pH). Therefore it seemed
plausible to postulate a progressive extraction of ER proteins
from a series of compartments, which should increase the efficiency of the sorting process (Rothman, 1981).
To investigate the localization and sorting of CR in more
detail we have cloned the cDNA of rat liver CR and overexpressed it in COS cells. We expected that high level overexpression would lead to a shift in the steady state distribution of
CR from a clear ER to an ER/Golgi distribution and a subsequential secretion of CR due to saturation of the KDELreceptor-based retrieval system. To our surprise neither
immunofluorescence nor subcellular fractionation gave any
indication for a post-ER localization of expressed CR. In
general this can be easily explained by assuming a slow exit
from the ER and a rapid retrieval from a salvage compartment.
Moreover, considering the high level of CR per overexpressing cell (about 100-fold), one must infer that the retrieval
system should combine both a high capacity and a high efficiency in order to maintain the original steady state distribution.
It is known that the KDEL receptor can easily be saturated
by overexpression of a KDEL-tagged secretory protein
(lysozyme-KDEL), and this leads to a complete redistribution
of a c-myc-Erd2 construct from the Golgi to the ER (Lewis
and Pelham, 1992). This seemed to be in contrast to our experiments: in transiently transfected COS cells a high level of CR
expression resulted in only a partial redistribution of the
endogenous KDEL receptor. Thus we concluded that CR is
retained in the ER, whereas the lysozyme-KDEL construct,
lacking a special ER-retention motif, can easily escape and
induce the redistribution of the KDEL-receptor.
Pulse chase experiments with transiently transfected COS
cells followed by subcellular fractionation confirmed the
results of our immunofluorescence studies. During the chase
period only minimal amounts of labelled CR were transiently
observed in the Golgi-containing fractions of the Nycodenz
2714 B. Sönnichsen and others
Fig. 7. Double immunofluorescence of VSV
G-protein and calreticulin in VSV ts045infected COS cells overexpressing calreticulin
and lacking the C-terminal calcium-binding
domain. G-protein was accumulated in the ER
at the restrictive temperature of 39.5°C for 3
hours. Transport was monitored in vivo after
lowering the temperature to 32°C for the time
intervals indicated. Colocalization of the virus
protein and the calreticulin mutant could be
observed in transport intermediates (A and B)
and in the Golgi apparatus (C and D). The
calreticulin mutant was partly redistributed to
the ER after 45 minutes (F). Bars, 10 µm.
gradient, although misfolding could be a reason for retention
of CR in the ER. However, we found neither an increase of
BiP nor observed a degradation of newly synthesized CR
within 3 hours in any of our experiments. This indicates the
correct folding of overexpressed CR. The experiments demonstrate that CR is able to exit the ER at a very low rate and
becomes sorted from secretory proteins in the compartment to
which the KDEL-receptor is localized, most likely the Golgi
apparatus (Pelham et al., 1988; Klausner et al., 1992; Hsu et
al., 1992). These data are therefore in contrast to the report by
Peter et al. (1992). We cannot rule out the possibility that continuous leakage out of the ER leads to a significant galactosylation of CR in rat liver cells. But we point out that whereas
the latter authors relied mostly on the equivocal specificity of
quantitative precipitation of newly synthesized CR from crude
microsomal fractions with the lectin Jacalin (Sastry et al.,
1986), our fractionation data clearly show that most CR stays
in the ER directly after synthesis.
Secretion experiments using labelled CR or its mutated
forms confirmed our results. Despite its high overexpression in
COS cells CR was almost quantitatively retained in the ER.
Only 1-2% of total CR was detected in the medium within 3
hours in the presence of cycloheximide. This represents the
normal rates of cell death in the cultures during our experiments. Under the same conditions the percentage of secreted
protein was 29% for the CR∆KDEL mutant. This corresponds to
other luminal proteins of the ER, like BiP or PDI. Character-
istic for CR are its Ca2+-binding properties. For this reason we
have speculated that the C-domain (low affinity, high capacity
Ca2+-binding domain) might participate in the retention of CR.
The P-domain (Michalak et al., 1992) of CR has also been
reported to bind Ca2+ with high affinity and low capacity and
might, in addition, be involved in retention of the protein.
However, in analogy to the chaperone calnexin this domain
could instead be associated with the function of the protein
(Helenius, 1994). Indeed, deletion of the C-domain increased
the secretion of overexpressed CR∆C about 15-fold above
background. Furthermore, we could directly follow increased
exit of this mutant from the ER after stimulation of vesicular
transport in VSV-infected cells. As expected CR∆C was partly
transported back to the ER, because it still had the retrieval
sequence whereas the escaped CR∆KDEL mutant lacking this
sequence moved further on along the secretory pathway (Figs
7 and 8).
The immunofluorescence and the biochemical data indicate
that: (a) overexpressed mutated CR forms are able to leave the
ER, showing that they are correctly folded; and (b) the Cdomain participates in the retention of CR in the ER. A similar
situation was recently reported for the retention of interleukin6 in the ER (Rose-John et al., 1993). Addition of the KDEL
sequence alone to the COOH terminus of interleukin-6 was not
sufficient to accomplish full retention in the ER of HepG2
cells. Complete retention was, however, achieved when the 14
COOH-terminal amino acids (MEEDDDQKAVKDEL) of PDI
Retention and retrieval of calreticulin 2715
Fig. 8. Double immunofluorescence of VSV
G-protein and calreticulin in VSV ts045infected COS cells overexpressing
calreticulin and lacking the KDEL sequence
at 5, 15 and 45 minutes after shifting to the
permissive temperature. Virus infection was
as described in Fig. 7. Only partial
colocalization of the G-protein and the
calreticulin mutant in transport intermediates
(A and B) and the Golgi apparatus (C and D)
was seen. Note that there is no redistribution
of calreticulin lacking KDEL after 45
minutes (F). Bars, 10 µm.
including the KDEL signal were added. Interestingly this
contains a cluster of 5 acidic residues that resembles the acidic
clusters of CR in its C-domain.
The electron microscopic studies provide an insight into the
morphological consequences of the high overexpression of a
resident ER protein. First, they unequivocally demonstrate the
localization of endogenous CR to the ER, which is recognized
by its attached ribosomes. Second they demonstrate the
impressive changes in ER morphology in transfected COS
cells from the typical cisternal-tubular structure to large dilated
vesicles, heavily labelled with immunogold particles. Neither
in the nontransfected nor in the transfected cells could any CR
be found in the Golgi, nucleus or other compartments. This is
not necessarily in contradiction to the reports of Michalak et
al. (1992), who partially localized CR to the nucleus and Burns
et al. (1994) as well as Dedhar et al. (1994) who suggested a
function for CR in the nucleus. The nuclear localization is
mainly based on immunfluorescence studies of proliferating rat
L6 myoblasts and the functional studies rely mainly on in vitro
binding characteristics of CR to a synthetic peptide
(KXGFFKR), found in the cytoplasmic domains of all integrin
α-subunits. Our results do not rule out the possibility that this
nuclear localization may be observed in special cell types or
under specific cell conditions, like proliferation or differentiation. However, except when one assumes a differential
spliced protein lacking the ER signal sequence is it difficult to
conceive how an ER resident protein might enter the nucleus.
But even then, an import from the cytosol remains to be
demonstrated. The other possibility, a direct import of CR from
the lumen of the ER into the nucleus, is even more speculative. The dilated ER in the CR overexpressing COS cells
resemble those structures observed in tunicamycin-stressed H35 Reuber hepatoma cells (Nguyen Van et al., 1993). Although
it is not yet clear whether the dilation of the ER cisternae represents a de novo synthesis of ER membrane or whether the
Fig. 9. Secretion of calreticulin and calreticulin mutants expressed in
COS cells. Cells were labelled for 15 minutes with [35S]methionine
and chased for 3 hours in the presence of unlabelled methionine and
1 mM cycloheximide. Calreticulin was immunoprecipitated from the
culture medium and whole cell extracts. For details see Materials and
Methods.
2716 B. Sönnichsen and others
cisternae simply expand, an overproduction of soluble ER
proteins is certainly the cause of this dilation.
There are two possibilities for the retention of a soluble ER
protein. Either it does not reach the exit sites of the ER-Golgi
transport vesicles or if it does it is excluded from these vesicles.
In the case of overexpressed CR we are not able to discriminate between these two possibilities. The fact that CR is in
principle able to leave the ER might argue against an exclusion
from the exit sites. Our experiments point to a retention
mechanism in which CR forms matrix-like interactions with
itself or other components of the ER (Booth and Koch, 1989;
Suzuki et al., 1991). This idea is supported by experiments in
which removal of the KDEL-sequence of ER proteins like
UDP-glucuronosyltransferase (Jackson et al., 1993), BiP
(Munro and Pelham, 1987) and ERp59/PDI (Mazzarella et al.,
1990) results in slow if any transport of the truncated proteins
out of the ER. In terms of efficiency it would make sense if
such prominent proteins like CR, BiP or PDI are prevented
from leaving the ER. The combination of high capacity
retention with a KDEL-based retrieval system would ensure a
minimum of leakage, especially during stress situations.
In contrast to the receptor-based retrieval system such an ER
retention mechanism should in principle not be saturable. This
could explain the enormous dilation of the ER cisternae
observed in CR-transfected COS cells. A high capacity does
not necessarily implicate a high efficiency. Our experiments
neither give answers as to how much CR is able to leave the
ER under normal conditions nor how it can be sorted from
different stacks of the Golgi. However, if only minor amounts
of CR leave the ER under normal conditions it is not absolutely
necessary to postulate a retrieval from late compartments like
the trans-Golgi. Our results show that the efficiency of
retention in the ER varies with different proteins. CaBP1,
another Ca2+-binding KDEL-protein of the ER, is also not
secreted after high overexpression in COS cells. In contrast to
CR, removal of its KDEL sequence resulted in the secretion of
49% of total CaBP1 in 3 hours (G. Mieskes et al., unpublished
data). However, even this rate is significantly lower than the
values reported for lysozyme, a secretory protein (about 85%
within 4 hours; Hsu et al., 1992) and a glycosylated tripeptide
(about 10 minutes for the half-time for secretion; Wieland et
al., 1987). The possible formation of a matrix composed of the
major ER proteins could be of great importance for their
function. For PDI and BiP it is known that they are part of the
folding machinery for nascent proteins (Nigam et al., 1994). It
will be very interesting to see whether CR plays a similar role
in these processes.
We are grateful to Dr W. E. Balch for the introduction to the morphological in vitro ER-Golgi transport system and for the monoclonal
anti-VSV G-protein antibodies; to Dr H.-P. Hauri and Dr M. Michalak
for providing us with anti-p53 and anti-CR antibodies; to Dr G. Thiel
for the pCMV2 vector and the COS cells; and to Dr K. Simons for
the VSV ts045, the VSV wt and the CHO 15B cells. This work was
supported by the Deutsche Forschungsgemeinschaft (SFB 236,
project B18, and Ro 440/10-1) and by a fellowship from the
Boehringer Ingelheim Foundation to B. Sönnichsen. The nucleotide
sequence data reported in this paper will appear in the EMBL,
GenBank and DDBJ Nucleotide Sequence Databases under the
accession number X79327.
REFERENCES
Anderson, S., Davis, D. L., Dahlbäck, H., Jörnvall, H. and Russell, D. W.
(1989). Cloning, structure, and expression of the mitochondrial cytochrome
P-450 sterol 26-hydroxylase, a bile acid biosynthetic enzyme. J. Biol. Chem.
264, 8222-8229.
Andres, D. A., Dickerson, I. M. and Dixon, J. E. (1990). Variants of the
carboxyl-terminal KDEL sequence direct intracellular retention. J. Biol.
Chem. 265, 5952-5955.
Artelt, P., Morelle, C., Ausmeier, M., Fitzek, M. and Hauser, H. (1988).
Vectors for efficient expression in mammalian fibroblastoid, myeloid and
lymphoid cells via transfection or infection. Gene 68, 213-219.
Booth, C. and Koch, G. L. E. (1989). Perturbation of cellular calcium induces
secretion of luminal ER proteins. Cell 59, 729-737.
Burns, K., Duggan, B., Atkinson, E. A., Famulski, K. S., Nemer, M.,
Bleackley, R. C. and Michalak, M. (1994). Modulation of gene expression
by calreticulin binding to the glucocorticoid receptor. Nature 367, 476-480.
Dedhar, S., Rennie, P. S., Shago, M., Hagesteijn, C.-Y. L., Yang, H.,
Filmus, J., Hawley, R. G., Bruchovsky, N., Cheng, H., Matusik, R., J. and
Giguère, V. (1994). Inhibition of nuclear hormone receptor activity by
calreticulin. Nature 367, 480-483.
Doms, R. W., Lamb, R. A., Rose, J. K. and Helenius, A. (1993). Folding and
assembly of viral membrane proteins. Virology 193, 545-562.
Donaldson, J. G., Kahn, R. A., Lippincott, S. J. and Klausner, R. D. (1991).
Binding of ARF and β-COP to Golgi membranes: possible regulation by a
trimeric G protein. Science 254, 1197-1199.
Donaldson, J. G., Cassel, D., Kahn, R. A. and Klausner, R. D. (1992a).
ADP-ribosylation factor, a small GTP-binding protein, is required for
binding of the coatomer protein β-COP to Golgi membranes. Nat. Acad. Sci.
USA 89, 6408-6412.
Donaldson, J. G., Finazzi, D. and Klausner, R. D. (1992b). Brefeldin A
inhibits Golgi membrane-catalyzed exchange of guanine nucleotide onto
ARF protein. Nature 360, 350-352.
Haugejorden, S. M., Srinivasan, M. and Green, M. (1991). Analysis of the
retention signals of two resident luminal endoplasmic reticulum proteins by
in vitro mutagenesis. J. Biol. Chem. 266, 6015-6018.
Hauri, H.-P. and Schweizer, A. (1992). The endoplasmic reticulum-Golgi
intermediate compartment. Curr. Opin. Cell Biol. 4, 600-608.
Helenius, A. (1994). How N-linked oligosaccharides affect glycoprotein
folding in the endoplasmic reticulum. Mol. Biol. Cell (in press).
Helms, J. B. and Rothman, J. E. (1992). Inhibition by brefeldin A of a Golgi
membrane enzyme that catalyzes exchange of guanine nucleotide bound to
ARF. Nature 360, 352-354.
Hsu, V. W., Shah, N. and Klausner, R. D. (1992). A brefeldin A-like
phenotype is induced by the overexpression of a human ERD-2-like protein,
ELP-1. Cell 69, 625-635.
Jackson, M. R., Nilsson, T. and Peterson, P. A. (1990). Identification of a
consensus motif for retention of transmembrane proteins in the endoplasmic
reticulum. EMBO J. 9, 3153-3162.
Jackson, M. R., Nilsson, T. and Peterson, P. A. (1993). Retrieval of
transmembrane proteins to the endoplasmic reticulum. J. Cell Biol. 121, 317333.
Kahn, R. A. and Gilman, A. G. (1986). The protein cofactor necessary for
ADP-ribosylation of Gs by cholera toxin is itself a GTP binding protein. J.
Biol. Chem. 261, 7906-7911.
Kaufman, R. J. (1990). Selection and coamplification of heterologous genes in
mammalian cells. Meth. Enzymol. 185, 537-566.
Klausner, R. D. (1989). Sorting and traffic in the central vacuolar system. Cell
57, 703-706.
Klausner, R. D., Donaldson, J. G. and Lippincott-Schwartz, J. (1992).
Brefeldin A: insights into the control of membrane traffic and organelle
structure. J. Cell. Biol. 116, 1071-1080.
Kornfeld, R. and Kornfeld, S. (1985). Assembly of asparagine-linked
oligosaccharides. Annu. Rev. Biochem. 54, 631-664.
Ladinsky, M. S. and Howell, K. E. (1993). The trans-Golgi network can be
dissected structurally and functionally from the cisternae of the Golgi
complex by brefeldin A. Eur. J. Cell Biol. 59, 92-105.
Laemmli, U. K. (1970). Cleavage of structural proteins during assembly of the
head of bacteriophage T4. Nature 227, 680-685.
Lewis, M. J. and Pelham, H. R. B. (1990). A human homologue of the yeast
HDEL receptor. Nature 348, 162-163.
Lewis, M. J. and Pelham, H. R. B. (1992). Ligand-induced redistribution of a
human KDEL-receptor from the Golgi complex to the endoplasmic
reticulum. Cell 68, 353-364.
Retention and retrieval of calreticulin 2717
Lippincott-Schwartz, J., Yuan, L. C., Bonifacino, J. S. and Klausner, R. D.
(1989). Rapid redistribution of Golgi proteins into the ER in cells treated with
brefeldin A: evidence for membrane recycling from Golgi to ER. Cell 56,
801-813.
Lippincott-Schwartz, J., Donaldson, J. G., Schweizer, A., Berger, E. G.,
Hauri, H.-P., Yuan, L. C. and Klausner, R. D. (1990). Microtubuledependent retrograde transport of proteins into the ER in the presence of
brefeldin A suggests an ER recycling pathway. Cell 60, 821-836.
Mazzarella, R. A., Srinivasan, M., Haugejorden, S. M. and Green, M.
(1990). Erp72, an abundant luminal endoplasmic reticulum protein, contains
three copies of the active site sequence of protein disulfide isomerase. J. Biol.
Chem. 265, 1094-1101.
Michalak, M., Milner, R. E., Burns, K. and Opas, M. (1992). Calreticulin.
Biochem. J. 285, 681-692.
Munro, S. and Pelham, H. R. B. (1987). A C-terminal signal prevents
secretion of luminal ER proteins. Cell 48, 899-907.
Nguyen Van, P., Peter, F. and Söling, H.-D. (1989). Four intracisternal
calcium-binding glycoproteins from rat liver microsomes with high affinity
for calcium. No indication for calsequestrin-like proteins in inositol 1,4,5trisphosphate sensitive calcium sequestering rat liver vesicles. J. Biol. Chem.
264, 17494-17501.
Nguyen Van, P., Rupp, K., Lampen, A. and Söling, H.-D. (1993). CaBP2 is a
rat homolog of ERp72 with proteindisulfide isomerase activity. Eur. J.
Biochem. 213, 789-795.
Nigam, S. K., Goldberg, A. L., Ho, S., Rhode, M. F., Bush, K. T. and
Sherman, M. Y. (1994). A set of endoplasmic reticulum proteins possessing
properties of molecular chaperons includes Ca2+-binding proteins and
members of the thioredoxin superfamily. J. Biol. Chem. 269, 1744-1749.
Nilsson, T., Jackson, M. R. and Peterson, P. A. (1989). Short cytoplasmic
sequences serve as retention signals for transmembrane proteins in the
endoplasmic reticulum. Cell 58, 707-718.
Orci, L., Tagaya, M., Amherdt, M., Perrelet, A., Donaldson, J. G.,
Lippincott-Schwartz, J., Klausner, R. D. and Rothman, J. E. (1991).
Brefeldin A, a drug that blocks secretion, prevents the assembly of nonclathrin-coated buds on Golgi cisternae. Cell 64, 1183-1195.
Orci, L., Palmer, D. J., Ravazzola, M., Perrelet, A., Amherdt, M. and
Rothman, J. E. (1993a). Budding from Golgi membranes requires coatomer
complex of non-clathrin coat proteins. Nature 362, 648-652.
Orci, L., Palmer, D. J., Amherdt, M. and Rothman, J. E. (1993b). De novo
assembly of coated vesicles from Golgi cisternae requires only coatomer and
ARF proteins from the cytosol. Nature 364, 732-734.
Ostermann, J., Orci, L., Tani, K., Amherdt, M., Ravazzola, M., Elazar, Z.
and Rothman, J. E. (1993). Stepwise assembly of functionally active
transport vesicles. Cell 75, 1015-1025.
Palade, G. (1975). Intracellular aspects of the processing of protein synthesis.
Science 189, 347-358.
Palmer, D. J., Helms, J. B., Beckers, C. J. M., Orci, L. and Rothman, J. E.
(1993). Binding of coatomer to Golgi membranes requires ADP-ribosylation
factor. J. Biol. Chem. 268, 12083-12089.
Pelham, H. R. B., Hardwick, K. G. and Lewis, M. J. (1988). Sorting of
soluble ER proteins in yeast. EMBO J. 7, 1757-1762.
Pelham, H. R. B. (1991). Recycling of proteins between the endoplasmic
reticulum and Golgi complex. Curr. Opin. Cell Biol. 3, 585-591.
Peter, F., Nguyen Van, P. and Söling, H.-D. (1992). Different sorting of lysasp-glu-leu proteins in rat liver. J. Biol. Chem. 267, 10631-10637.
Peter, F., Plutner, H., Zhu, H., Kreis, T. E. and Balch, W. E. (1993). β-COP
is essential for transport of protein from the endoplasmic reticulum to the
Golgi in vitro. J. Cell Biol. 122, 1155-1167.
Pfeffer, S. R. and Rothman, J. E. (1987). Biosynthetic protein transport and
sorting in the endoplasmic reticulum and Golgi. Annu. Rev. Biochem. 56,
829-852.
Plutner, H., Davidson, H. W., Saraste, J. and Balch, W. E. (1992).
Morphological analysis of protein transport from the ER to Golgi membranes
in digitonin-permeabilized cells: role of the p58 containing compartment. J.
Cell Biol. 119, 1097-1116.
Robbi, M. and Beaufay, H. (1991). The COOH terminus of several liver
carboxylesterases targets these enzymes to the lumen of the endoplasmic
reticulum. J. Biol. Chem. 266, 20498-20503.
Rose-John, S., Schooltink, H., Schmitz-Van de Leur, H., Müllberg, J.,
Heinrich, P. C. and Graeve, L. (1993). Intracellular retention of
interleukin-6 abrogates signalling. J. Biol. Chem. 268, 22084-22091.
Rothman, J. E. (1981). The Golgi apparatus: two organelles in tandem.
Science 213, 1212-1219.
Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning: a
Laboratory Manual. Second edition. Cold Spring Harbor Laboratory Press,
New York.
Saraste, J. and Svensson, K. (1991). Distribution of the intermediate elements
operating in ER to Golgi transport. J. Cell Sci. 100, 415-430.
Sastry, M. V. K., Banarjee, P., Patanjali, S. R., Swamy, M. J., Swarnalatha,
G. V. and Surolia, A. (1986). Analysis of saccharide binding to Artocarpus
integrifolia lectin reveals specific recognition of T-antigen (β-D-Gal(1→3)DGalNAc). J. Biol. Chem. 261, 11726-11733.
Schwaninger, R., Plutner, H., Davidson, H. W., Pind, S. and Balch, W. E.
(1992). Transport of protein between endoplasmic reticulum and Golgi
compartments in semiintact cells. Meth. Enzymol. 219, 111-124.
Schweizer, A., Fransen, J. A. M., Bächi, T., Ginsel, L. and Hauri, H. P.
(1988). Identification, by a monoclonal antibody, of a 53-kD protein
associated with a tubulo-vesicular compartment at the cis-side of the Golgi
apparatus. J. Cell Biol. 107, 1643-1653.
Schweizer, A., Fransen, J. A. M., Matter, K., Kreis, T. E. and Hauri, H.-P.
(1990). Identification of an intermediate compartment involved in protein
transport from the endoplasmic reticulum to the Golgi apparatus. Eur. J. Cell
Biol. 53, 185-196.
Schweizer, A., Ericsson, M., Bächi, T., Griffiths, G. and Hauri, H. P.
(1993). Characterization of a novel 63 kDa membrane protein: implications
for the organization of the ER-to-Golgi pathway. J. Cell Sci. 104, 671-683.
Semenza, J. C., Hardwick, K. G., Dean, N. and Pelham, H. R. B. (1990).
ERD2, a yeast gene required for the receptor-mediated retrieval of luminal
ER proteins from the secretory pathway. Cell 61, 1349-1357.
Serafini, T., Stenbeck, G., Brecht, A., Lottspeich, F., Orci, L., Rothman, J.
E. and Wieland, F. T. (1991). A coat subunit of Golgi-derived non-clathrincoated vesicles with homology to the clathrin-coated vesicle coat protein
β-adaptin. Nature 349, 215-220.
Sottocasa, G. L., Kuylenstierna, B., Ernster, L. and Bergstrand, A. (1967).
An electron-transport system associated with the outer membrane of liver
mitochondria: a biochemical and morphological study. J. Cell Biol. 32, 415438.
Suzuki, C. K., Bonifacino, J. S., Lin, A. Y., Davis, M. M. and Klausner, R.
D. (1991). Regulating the retention of T-cell receptor α chain variants within
the endoplasmic reticulum: Ca2+-dependent association with BiP. J. Cell
Biol. 114, 189-205.
Tang, B. L., Wong, S. H., Qi, X. L., Low, S. H. and Hong, W. (1993).
Molecular cloning, characterization, subcellular localization and dynamics
of p23, the mammalian KDEL-receptor. J. Cell Biol. 120, 325-338.
Townsley, F. M., Wilson, D. W. and Pelham, H. R. (1993). Mutational
analysis of the human KDEL-receptor: distinct structural requirements for
Golgi retention, ligand binding and retrograde transport. EMBO J. 12, 28212829.
Tsai, S.-C., Adamik, R., Haun, R. S., Moss, J. and Vaughan, M. (1993).
Effects of brefeldin A and accessory proteins on association of ADPribosylation factor 1, 3, and 5 with Golgi. J. Biol. Chem. 268, 10820-10825.
Verdon, B. and Berger, E. G. (1983). Galactosyltransferase. Meth. Enzymol.
Anal. 3, 374-381.
Weiss, O., Holden, J., Rulka, C. and Kahn, R. A. (1989). Nucleotide binding
and cofactor activities of purified bovine brain and bacterially expressed
ADP-ribosylation factor. J. Biol. Chem. 264, 21066-21072.
Wieland, F. T., Glaeson, M. L., Serafini, T. A. and Rothman, J. E. (1987).
The rate of bulk flow from the endoplasmic reticulum to the cell surface. Cell
50, 289-300.
Wilson, D. W., Lewis, M. J. and Pelham, H. R. B. (1993). pH-dependent
binding of KDEL to its receptor in vitro. J. Biol. Chem. 268, 7465-7468.
(Received 30 March 1994 - Accepted 2 June 1994)